DISS. ETH No. 15930 The functional roles of PlexinD1 in chicken nervous system during development A dissertation submitted to the SWISS FEDERAL INSTITUTE OF TECHNOLOGY ZURICH for the degree of Doctor of Sciences presented by JOËLLE GEMAYEL DEA “Metabolism, Endocrinologie et Nutrition” Université Claude Bernard-Lyon born May 27, 1974 citizen of Lebanon accepted on the recommendation of Prof. Martin Schwab, examiner Prof. Lukas Sommer, co-examiner Dr. Matthias Gesemann, external supervisor 2005 DISS. ETH No. 15930 The functional roles of PlexinD1 in chicken nervous system during development A dissertation submitted to the SWISS FEDERAL INSTITUTE OF TECHNOLOGY ZURICH for the degree of Doctor of Sciences presented by JOËLLE GEMAYEL DEA “Metabolism, Endocrinologie et Nutrition” Université Claude Bernard-Lyon born May 27, 1974 citizen of Lebanon accepted on the recommendation of Prof. Martin Schwab, examiner Prof. Lukas Sommer, co-examiner Dr. Matthias Gesemann, external supervisor 2005 Summary A functional nervous system results from the coordinated generation and assembly of billions of neural cells into highly structured and well organized networks. Building blocks for these networks are neurons, which arise from neural progenitors that are deployed from specialized neuroepithelia. These neuronal precursors migrate along specific pathways populating different areas within the developing brain, spinal cord and peripheral nervous system. Once correctly positioned, differentiated neurons send out axons along highly stereotypical pathways, specifically linking neurons within different parts of the nervous system, generating a high number of functional neuronal networks. The use of appropriate migratory routes as well as directed axonal outgrowth along specific predetermined pathways is achieved by specific receptor-ligand interactions that are characteristic for each subpopulation of migratory cells or growing axons. Several families of cell surface receptors and axon guidance cues involved in directing cell migration and/or axonal outgrowth have been described over the last decade. One of the largest families of axon guidance receptors is the plexin family. Plexins and their semaphorin ligands have been shown to be involved in several aspects of axonal targeting and cell migration. However, numerous studies describe also functional roles for plexins in the development outside the nervous system, notably in the development of the cardiovascular system. In this respect, PlexinD1 (PD1) has been extensively studied in vasculogenesis, and several groups document its implication in heart development as well as in vessel patterning. However, PD1 is also expressed in several regions of the developing brain, but no available data describe its potential role in the nervous system. Experiments in the first part of this thesis demonstrate that PD1 transcripts are not only found in mouse brain but also in chicken spinal motor neurons during the period motor axons sort in the limb plexus. PD1 knock down, using in ovo RNAi, results in motor axon misguidance in the dorsal as well as the ventral crural nerve trunk, suggesting PD1 involvement in motor axon guidance and/or sorting of the crural nerve. Interestingly, PD1 loss of function experiments showed also unexpected defects in dorsal sensory root formation and abnormalities in the area of motoneuron exit points. In PD1 knock down animals, dorsal roots are less compact, and several axon bundles grow erroneously into neighboring segments whereas motor exit points display a severely altered morphology appearing much broader than usual, and the ventral motor roots show axons with aberrant courses. i ii Summary While the defects in the crural nerve could be easily explained by the lack of PD1 expression in motoneurons, phenotypes observed in the dorsal and ventral roots support a role for PD1 in either neural crest migration or the existence of a tight relation between intersomitic vessel formation and dorsal root development. In the first case, PD1 loss of function in a subpopulation of neural crest cells would lead to erroneous positioning of Schwann cells and /or boundary cap cells resulting in outgrowth and defasciculation defects within the ventral and dorsal roots. Alternatively, PD1 knock down might affect endothelial cells surrounding the spinal cord, in particular the intersomitic vessels, leading to aberrant formation of these vessels. Intersomitic vessels could form in close contact with sensory growing axons and could play a role for correct axonal outgrowth. Their aberrant formation could result, therefore, in axon guidance defects observed in the present study. Experiments described in the second and third chapter of this thesis analyze the presence and expression of semaphorins as well as plexins in the developing chicken spinal cord. Chicken plexins and semaphorins display very complex, often complementary, but also overlapping expression patterns that are highly regulated developmentally. These results suggest multifunctional roles for these proteins, most likely exerting their activities in large complexes containing intricate combinations of several members. Résumé Le bon fonctionnement du système nerveux dépend du correct assemblage de milliards de cellules neurales générées pendant le développement embryonnaire afin de pouvoir former des réseaux complexes et structurés. Les précurseurs des cellules neurales se forment au niveau d’une structure bien spécialisée appelée l’épithélium neuronal. Une fois générées, ces cellules entament leur migration tout le long de parcours bien définis afin de former les différentes structures du système nerveux central et périphérique. Par la suite les cellules neuronales projettent leurs axones tout le long de voies bien définies, souvent à de très longues distances de leur corps cellulaire pour finalement de se connecter à leur cible finale. L’achèvement d’une migration cellulaire correcte ainsi que la navigation précise des axones jusqu’à leur destination finale dans un système embryonnaire nécessite l’intégrité de familles de protéines parmi lesquelles les plexines. Les plexines ont été récemment identifiées et caractérisées; elles constituent une très grande famille de protéines contenant environ 9 membres et exercent leur action en se liant à leur ligands, les semaphorines. Plusieurs études décrivent l’implication de ces protéines dans le développement du système nerveux central mais aussi dans la formation d’autre tissues embryonnaires. PlexinD1 appartient à la famille des plexines et a été intensivement décrite comme un facteur majeur dans le développement du système vasculaire in vitro et in vivo. Bien que l’expression de cette protéine ait été identifiée dans le cerveau de souris pendant le développement embryonnaire, aucune étude ne décrit le rôle de plexinD1 dans le développement du système nerveux central. Pendant ce travail de thèse, nous avons souhaité investir le rôle potentiel de plexinD1 dans le développement de la moelle épinière dans l’embryon de poulet. Nous avons établi que cette protéine est bien exprimée pendant le développement par les neurones moteurs de la moelle épinière et cette expression correspond au moment où ces neurones s’arrangent dans le plexus à la base du pied. Les outils moléculaires visant à bloquer l’expression de plexinD1 en utilisant une technique appelée in ovo RNAi, ont pu démontrer que l’absence de plexinD1 dans les motoneurones induit la malformation de certaines branches du nerf crural. De plus, nous avons observé des anomalies au sein des fibres sensorielles résultant en la fusion des fibres entre elles, bien que plexinD1 ne soit pas exprimée par les neurones sensorielles mais par les cellules endothéliales iii iv Résumé entourant la moelle épinière et les ganglions de la racine dorsale. Afin de pouvoir expliquer ces aberrations dans la formation des fibres sensorielles, nous avons proposé deux modèles. Notre première hypothèse suppose que plexinD1 n’est pas exclusivement exprimée par les cellules endothéliales mais aussi dans les cellules de Schwann et les boundary cap cells. Une migration erronée de ces cellules menant à leur mauvais positionnement peut expliquer les anomalies observées. Une seconde explication est que la malformation des fibres sensorielles peut résulter d’une anomalie dans le développement des vaisseaux sanguins exprimant plexinD1, en particulier les vaisseaux intersomitiques, qui normalement assistent et guident les fibres sensorielles pendant leur élongation. Nous avons également montré dans une deuxième partie de notre travail, l’expression de toutes les semaphorines et plexines dans la moelle épinière du poulet au cours des différents stades du développement. Nos résultats démontrent que les ARNm étudiés sont exprimés d’une façon très complexe et dynamique suggérant l’existence de multiples fonctions exercées par ces protéines qui très souvent doivent agir en synergie plutôt que séparément. Acknowledgements I would like to thank Dr. Matthias Gesemann for giving me the opportunity to do my PhD thesis in his laboratory and introducing me to the world of axon guidance molecules. I am very grateful to Professor Martin Schwab, not only for supervising my thesis but also for always taking time to discuss my data and giving me inputs about the future directions to adopt in my projects and experiments. I thank also Professor Lukas Sommer, for being always available for supervising my thesis work and investing time to discuss my data and results. This work would not have been possible without the collaboration as well as the precious help of Professor Esther Stoeckli and all her team. I would like to thank her for giving me the chance to learn a lot about chicken embryos and helping me all through the in vivo experiments. All my gratitude goes also toward all the Stoeckli team, especially Rejina Sadhu for teaching me all I know about chicken and Olivier Mauti, for their help and cooperation and for making the work at their laboratory very enjoyable. Many thanks go to my colleagues in the Gesemann groups (Regis, Esther, Pascal, Connie, Daniele and Peter), for making hard moments possible to survive and for helping me at any moment with all the energy they have. Despite the hard moments, a researcher can pass through, the wonderful and international environment the Hifo offers, make science more efficient and the work in the lab much easier. Without the great and continuous help of all the neighboring labs, I could not have achieved my thesis. I would like to thank specially the Schwab group for rendering the institute as a big family and for helping me always as if I was a member of their team. My special thanks to Franziska Christ, Lisa Schnell, Regula Schneider, Dana Dodd, Carri Duncan, Irin Maier, Elisabeth Aloy, Barbara Niederoest and Florence Bareyre. Special thanks to Roland Schoeb for his continuous support in arranging and printing my pictures especially during the time I was writing my thesis. v vi Acknowledgements Many thanks also for the Neuhauss Group for their kind support and cooperation, especially, Ronja Bahadori ,Oliver Rinner, Oliver Biehlmaier, David Belet and Yury Makhankov. How cheap words can be when confronted to endless love and unlimited support. This thesis is dedicated with all my love to my family and friends. To my dad, for filling gaps life made, for teaching me how to be strong and fight for my dreams, for giving me the great chance of flying away to build a life but mainly for giving me endless love that one need at any time and any age. To my brothers Gino and Tony, and my sister Jessy, for believing in me in all circumstances, for all the laughers and the tears we shared and we will still share, for being four parts of one entity and finally for your unlimited love that gives me always strength to advance. This thesis is dedicated with all my love to Christian. I could have not achieved this work without your daily support and continuous love. For standing by me no matter if the in situ and cloning worked or failed, despite my long working hours in the lab even during week ends, for pushing me to advance always forward and to push further my limits. A mes meilleures amies, Chantal, Rozlaine et Rania pour avoir su être toujours présentes malgré les distances et les frontières, pour m’avoir connu mieux que je ne me connais et pour m’avoir aimé sincèrement pour ce que je suis. How can I ever thank all of you in the Hifo and neighboring institutes who are for me more than colleagues and for standing by my side during the hardest moment? For Zeina, Ronja, Elisabeth, Carri, Rejina, and Florence, “thank you” seems meaningless!!! For wiping my tears and making me laugh over endless coffee breaks, champagne drinking sessions and great dinners, you are the best souvenir I will take with me from Zurich. Je n’oublie pas de mentionner aussi tous mes amis bien loin de Zurich qui n’ont cessé de me soutenir malgré la distance et le temps. Mes plus sincères pensées et remerciements vont vers: Mme Jomain, Dr. Alain Geloen, Dr. Geneviève Barret et toute l’équipe du Laboratoire de physiologie Lyon Nord pour avoir été comme une famille pour moi en France, avoir cru en mes capacités et m’avoir aide à parvenir jusqu’aux bouts de mes ambitions. Contents Summary .................................................................................................................................... i Résumé .....................................................................................................................................iii Acknowledgements................................................................................................................... v Contents................................................................................................................................... vii Part I: Introduction.................................................................................................................. 1 1.1 Nervous system development........................................................................................... 2 1.1.1 Neural cell induction and migration.......................................................................... 2 1.1.2 Neural cell differentiation ......................................................................................... 4 1.1.3 Axon guidance......................................................................................................... 12 1.2 Axon guidance forces..................................................................................................... 17 1.3 Families of axon guidance molecules ............................................................................ 18 1.4 Plexins and Semaphorins ............................................................................................... 20 1.4.1 Semaphorin family .................................................................................................. 20 1.4.2 Semaphorin receptors and receptor complexes....................................................... 26 1.4.3 Semaphorin and Plexins beyond axon guidance ..................................................... 30 1.5 RNAi in chicken spinal cord .......................................................................................... 31 1.6 Goal of the present thesis: Role of PlexinD1 in nervous system development.............. 32 Part II: Paper 1....................................................................................................................... 35 2.1 Introduction .................................................................................................................... 37 2.2 Material and methods ..................................................................................................... 40 2.2.1 in ovo RNAi ............................................................................................................ 40 2.2.2 In situ hybridization ................................................................................................ 41 2.3 Results ............................................................................................................................ 43 2.3.1 Expression of PlexinD1 in chicken spinal cord during motor axon sorting in the limb plexus ....................................................................................................................... 43 2.3.2 PD1 knock down specifically in chicken embryo spinal cord leads to motor axon pathfinding errors at the hindlimb level ........................................................................... 44 2.3.3 PD1 knock down animals show fusion of the dorsal root entry zones ................... 46 2.3.4 PD1 knock down animals exhibit defects at the motor exit points of the ventral roots.................................................................................................................................. 47 vii viii Contents 2.4 Discussion ...................................................................................................................... 49 2.4.1 Chicken spinal motor neurons express PD1 mRNA ............................................... 49 2.4.2 PD1 involvement in the navigation of the crural nerve toward the limb ................ 49 2.4.3 PD1 knock down affects the dorsal roots entry zones and the motor exit points ... 51 2.4.4 PD1 potential binding partner(s) in chicken embryos during development............ 53 Part III: Paper 2 ..................................................................................................................... 55 3.1 Introduction .................................................................................................................... 57 3.2 Material and Methods..................................................................................................... 60 3.2.1 Assembly of chicken semaphorin cDNAs .............................................................. 60 3.2.2 Phylogenetic tree analysis ....................................................................................... 61 3.2.3 Cloning of semaphorin cDNA fragments................................................................ 61 3.2.4 In situ hybridization ................................................................................................ 62 3.3 Results ............................................................................................................................ 63 3.3.1 Identification of Sema3G, a novel member of the class III semaphorins ............... 63 3.3.2 The chicken genome has a reduced number of semaphorin genes ......................... 64 3.3.3 Class III semaphorins are highly expressed in developing motoneurons ............... 66 3.3.4 Floor plate cells expresses high levels of semaphorin V transcripts....................... 66 3.3.5 Class VI semaphorins are highly expressed in boundary cap cells..................Error! Bookmark not defined. 3.3.6 Semaphorin7A is expressed in endothelial cells and motoneurons ........................ 71 3.4 Discussion ...................................................................................................................... 72 3.4.1 The chicken genome has fewer semaphorin genes that the mammalian genome ... 72 3.4.2 Expression patterns of semaphorins are dynamically regulated during spinal cord development ..................................................................................................................... 73 Part IV: Paper 3 ..................................................................................................................... 77 4.1 Introduction .................................................................................................................... 79 4.2 Material and methods ..................................................................................................... 81 4.2.1 Assembly of chicken plexin cDNAs ....................................................................... 81 4.2.2 Phylogenetic tree and domain identity analysis ...................................................... 82 4.2.3 In situ hybridization ................................................................................................ 82 4.3 Results ............................................................................................................................ 84 4.3.1 Plexin and Neuropilin genes in chicken.................................................................. 84 4.4 Discussion ...................................................................................................................... 93 Contents ix 4.4.1 Mouse plexin genes located on the X sex chromosome are absent in chicken....... 93 4.4.2 Expression patterns of plexins and neuropilins are dynamically regulated during spinal cord development................................................................................................... 94 4.4.3 PlexinBs and PD1 are only transiently expressed in neurons ................................. 95 4.4.4 PlexinC1 is not expressed in early stages of neuronal development....................... 96 Part V: Conclusion and Outlook........................................................................................... 97 Conclusion and Outlook ........................................................................................................ 98 References ............................................................................................................................. 103 List of Publications............................................................................................................... 117 Curriculum Vitae ................................................................................................................. 119 Part I: Introduction 1 1 Introduction 1.1 Nervous system development The nervous system of vertebrates is a very complex structure that coordinates a variety of functions ranging from the integration of simple sensory and motor information to the processing of very complex behavioral tasks such as learning. The proper functioning of an adult nervous system depends mainly on the appropriate generation of different types and numbers of neural cells, the well-defined migration of these cells to their specific final position and the correct establishment of highly precise neuronal connections between the very large numbers of generated neurons. The first stage in nervous system development is the induction of the neuroectoderm to form a columnar epithelium. This so-called neural plate is underlaid by axial along with paraxial mesodermal cells and flanked by epidermal ectoderm. During a process called neurulation, the neural plate first buckles at its midline to form the neural folds and the floor plate resulting later, once the dorsal tips of the neural folds fuse, in a tube like structure called the neural tube. At the dorsal midline of the neural tube, roof plate cells along with neural crest cells are generated. All neural crest cells subsequently initiate migrations to populate different area of the developing embryo. As the neural tube folds, cell division starts at its luminal side. Dividing cells are influenced by several inductive or repressive factors secreted from the floor plate, the roof plate and the surrounding tissue that contribute to the acquisition of diverse neuronal cell fates. Neural induction and early regional fate of neural cells appear to be linked, as it is impossible to separate the induction of neural properties from the acquisition of anteroposterior regional identity. 1.1.1 Neural cell induction and migration During early neural development, axial mesodermal cells provide inductive signals to initiate neural tissue formation. The major pathway of neural induction is mediated by the inhibition of bone morphogenic protein 4 (BMP4) signaling, which when blocked lead to the formation of the anterior neural tube. Several structurally unrelated xenopus proteins (follistatin, noggin, chordin) were shown to inhibit BMP4 activity (Godsave and Slack, 1989). The neural tissue induced by follistatin, noggin and chordin exhibit an anterior character (Hemmati-Brivanlou and Melton, 1994; Lamb et al., 1993), suggesting that distinct signaling pathways may be required for inducing posterior neural tissue. Exposure of the ectoderm to fibroblast growth 2 1.1 Nervous system development 3 factor members (FGFs), under conditions in which BMP4 signaling is reduced or eliminated, leads to the generation of posterior neural plate tissue. Moreover, neural tissue characteristics of intermediate levels of the neuraxis midbrain and hindbrain can be induced by exposure of the ectoderm to both noggin and FGF (Lamb and Harland, 1995). Additionally, retinoids belonging to another class of molecules appear to be involved in the generation of posterior neural tissue: treatment of embryos with retinoic acid leads to an anterior-to-posterior transformation in the regional character of the neural tube (Durston et al., 1989; Hill et al., 1995). However, in vivo studies demonstrate that BMP4 and follistatin are not the only factors responsible for neural induction. Mutant animals lacking BMP4, follistatin or other factors from the Hensen’s node do not exhibit any obvious defect in neural induction, implying that these factors are not the only players required for neural plate formation (Matzuk et al., 1995; Winnier et al., 1995). Furthermore, it now appears that neural induction begins prior to the formation of the organizer region or the node. Thus, different factors expressed potentially in other regions than the organizer or node such as the mesoderm and endoderm might act on the neural induction of the ectoderm. These findings suggest that the suppression of BMP signaling may maintain rather than initiate the process of neural differentiation. Neuronal progenitors are generated from the rapid division of neural stem cells at the germinal neuroepithelium of the newly formed neural tube. While symmetric cell divisions along the luminal side of the neural tube give rise to two similar stem cells, asymmetric cell divisions lead to the generation of differentiated neuronal and glia cell types migrating toward the pial surface of the neural tube (Hollyday, 2001). One hallmark of vertebrate nervous system development is long-range cell migration of neuronal precursor cells. During the formation of the brain, the spinal cord and the peripheral nervous system, neuronal precursor cells migrate extensively and undergo significant rearrangements prior to differentiation into either neurons or glia. Two major patterns of migratory movements can be distinguished during nervous system development. While laminated structures such as the cerebellum, the hippocampus and the cerebral cortex are formed largely due to cell migration along radial glia, other brain areas such as the hindbrain and the thalamus are also formed by tangentially migrating neurons (Park et al., 2002; Hatten, 1999). Laminated structures built by radial migration exhibit a specific pattern of inside-out formation where newborn neurons by pass their older siblings to populate more cortical layers. This type of migration is based on cell-cell interactions between the moving neurons and the radial glial cells, of which the latter form a migratory scaffold that extends from the 4 Part I: Introduction ventricular zone to the pial surface. This process implies the existence of signals on or near the glia that can promote migration of neuroblasts in appropriate directions and arrest the movement at appropriate locations. Identification of a key molecule in radial migration came from the isolation of a new allele of a mouse mutant, reeler (Rice and Curran, 1999). In these mutants, late migrating neurons fail to pass their older siblings, leading to a scrambling of the normal inside-out relationship between birthdates and laminar position. Reelin is concentrated on the superficial cortical laminae and seems to promote dissociation of neuroblasts from the radial glia surface. While glia guided migration is certainly a key component for laminated structure formation, it has been shown lately that many cortical interneurons arise in subcortical areas rather than in the cortical ventricular zone and migrate tangentially into the cortex (Hatten, 1999). Migration of these neurons into the cortex occurs via a tangential pathway and is glia independent. Tangentially migrating neurons develop a specialized transient structure that is called the “leading process”. Although a lot remain to be discovered regarding structures and molecules implicated in tangential cell migration, it is well understood that migrating cells use their leading process to sense and probe environmental cues in the surrounding (Marin and Rubenstein, 2003). 1.1.2 Neural cell differentiation The allocation of cell fate in the central nervous system depends on two signaling systems that are activated together with the more basic program of neural induction. These two signaling systems intersect along the rostrocaudal and dorsoventral axes of the neural tube establishing a grid-like set of positional cues. The initial position of the neuronal progenitor cells along these two axes determines their exposure to different types and concentrations of inductive signals and therefore influences directly their fate (Jessell, 2000). Signaling along the rostrocaudal axis gives rise to the different subdivisions of the nervous system: the forebrain, midbrain, hindbrain and spinal cord (Lumsden and Krumlauf, 1996). Some of the factors playing a role in neural induction seem also to determine rostrocaudal differentiation. Relatively few signaling factors (retinoid, TGF β signaling, sonic hedgehog (Shh), FGFs and Wnts) account for many features of regional cell specialization within the anterior neural tube (Tanabe and Jessell, 1996). All these factors are proposed to function in different locations or developmental windows and their combinatorial actions in a single region or cell can establish regional patterning and neuronal diversity. Moreover, some 1.1 Nervous system development 5 inductive signals such as Shh can act as a gradient signal to induce different subtypes of neurons at different concentration (Poh et al., 2002). While general cues along the dorsoventral axis lead to the diversity of cell types within each rostrocaudal subdivision (Pituello, 1997), two primary signaling factors appear to induce “ventralization” or “dorsalization” of the neural tube. As neural tube folding occurs, the underlying mesodermal cells of the notochord initially provide a signal identified as Sonic Hedgehog (Shh) that “ventralize” the developing neural tube. Shh initiates the differentiation of a special set of neuroepithelial cells located immediately adjacent to the notochord called “floor plate”. Floor plate cells are the first cells in the neural plate to show signs of overt differentiation (Patten et al., 2003). Once induced, the floor plate cells themselves secrete Shh and thereby provide patterning information for neural cell types. Shh seems to function as a morphogene which can induce differentiation of distinct types of ventral cells at different concentration thresholds (Ericson et al., 1997). The graded activity of Shh subsequently activates different transcription factors in progenitor cells placed at various locations along the ventro-dorsal axis of the neural tube and subdividing a previously uniform territory into distinct domains. These transcription factors are expressed in a combinatorial manner in the dividing multipotential progenitors of the ventricular zone. Importantly, these transcription factors can repress each other in specific combinations leading to the translation of a transient graded response to Shh into establishment of sharp boundary between two cell populations (see Figure 1) (Briscoe et al., 2000). Fig. 1: Three phases of Shh-mediated ventral neural patterning. a) Shh mediates the repression or the induction of different transcription factors at variable threshold. Shh signaling defines five progenitor domains in the ventral neural tube. c) The relation between neural progenitors (p) domains and the positions at which postmitotic neurons are generated along the dorsoventral axis of the ventral spinal cord (adapted from(Jessell, 2000)). 6 Part I: Introduction Concomitant with the ventral polarization of neural tissue, signals from the epidermal ectoderm adjacent to the lateral edge of the neural plate seem to impose the dorsal pattern. BMP4, BMP7 and Dorsalin-1 are expressed in the dorsal aspect of the neural tube and are implicated in dorsalizing activity (Nguyen et al., 2000). The region of the ventral neural tube that gives rise to motoneurons subsequently generates also oligodendrocytes. Thus, the effect of Shh and BMPs extend beyond determining ventral neurons identity as it also contributes to the location of the founder cells of the oligodendrocyte lineage. Spinal cord oligodendrocytes arise from cells in the motor neuron domain (Richardson et al., 1997; Miller, 2002), and this localization is dependent on the local expression of Shh that is antagonized by BMPs signaling. Interestingly, most of the neuronal subtypes generated within the spinal cord are represented at all segmental levels, raising the issue of whether rostrocaudal positional information contributes significantly to the establishment of neuronal subtype identity at the spinal cord. 1.1.2.1 Motor neuron differentiation and migration While all spinal motoneurons neurons derive from a single ventral progenitor domain, they acquire many distinct subtype identities. The identity of a motoneuron is defined by its location in the CNS and its final synaptic connectivity in the PNS. Motor neurons differentiate exclusively in the ventral spinal cord, but once they become post mitotic, they may migrate to occupy more dorsal positions on the ipsilateral side of the spinal cord (Hollyday, 2001). In higher vertebrates, motor neurons with common target projections are aligned into longitudinally oriented columns (see Figure 2). These columns occupy distinct and discontinuous domain along the rostrocaudal axis of the spinal cord. This columnar organization is closely linked to specific formation of axon tracks and neuronal connections (Jessell, 2000). Motor neurons in the medial motor column (MMC), innervating hypaxial (back muscles) and epaxial (body wall muscles) muscles, are present along the entire rostrocaudal axis. Whereas, neurons of the lateral motor column (LMC), which innervate the limbs, are found only at brachial (forelimb) and lumbar (hind limb) levels and column of Terni (CT) neurons project to the sympathetic ganglia (Landmesser, 1978a). At a second level of organization, neurons within the same motor column are segregated into medial and lateral division and project axons along different trajectories. Within the LMC, motoneurons in the medial and the lateral divisions project to ventral and dorsal limb muscles respectively (Landmesser, 1978b), whereas motor neurons from the medial division of MMC, innervate the epaxial muscle and the lateral one target the body wall muscles. A final level of 1.1 Nervous system development 7 organization is reached when motor neurons within each division of the LMC segregate into discrete pools to innervate specific muscles in the limb. Anatomically defined motor neuron subclasses are also molecularly distinct, as defined by the restricted expression pattern of specific transcription factors. The main different columnar subclasses of motor neurons can be distinguished by the combinatorial expression of LIM homeodomain (LIM-HD) proteins (Tsuchida et al., 1994), whereas individual motoneurons pools within the LMC can be classified based on their expression of members of ETS proteins (Lin et al., 1998). Fig. 2: Columnar organization of motor neuron subtypes in the chick spinal cord. The target specificity of motor neuron subtypes are defined by distinct combinations of LIM-HD transcription factors. Axonal pathways of motor neurons subtypes are represented in transverse sections at the brachial and thoracic levels. The expression of LIM-HD transcription factors in individual motor neurons subtypes is shown by color-coding. bw, body wall; dlb, dorsal limb bud musculature: dm, dermamytome; sg, neurons of the sympathetic ganglia; vlb, ventral limb bud musculature. LMCL (blue), lateral half of lateral motor column; LMCM (green), median half of lateral motor column; MMCL (orange), lateral half of medial motor column; MMCM (red), median half of medial motor column (adapted from (Shirasaki and Pfaff, 2002)). 8 Part I: Introduction 1.1.2.2 Interneurons differentiation and migration Spinal interneurons constitute a large number of functionally distinct classes of cells present at the dorsal as well as the ventral part of the neural tube. They appear as small groups of neurons showing no particular distinguishing features or spatial organizations. However, they comprise a large number of separate neuronal populations with widely varying functions greatly outnumbering the amount of motor neurons. While some spinal interneurons process sensory information, others modulate motoneuron activity or coordinate activity at different spinal levels, and some relay sensory or proprioceptive information to the brain. The distinguishing markers for dorsal neural tube progenitor cells seem to be bHLH transcription factors rather than homeodomain transcription factors seen in the ventral spinal domains. Interestingly at the dorsal spinal domain a combination of bHLH and HD domain are likely to define progenitor domains. Interneurons are induced by BMPs signaling factors secreted by the roof plate and/or epidermal ectoderm. The down stream cascade is not very well understood but wnt1 and wnt3a are likely to be candidates because wnt1 at least is induced by BMP signaling (Panchision et al., 2001). Some transcription factors induced by BMP signaling seem to modulate the timing of cell cycle exit of progenitor cells thus controlling the cell number. Although BMPs have been considered the key signaling molecules originating from the dorsal midline, other signaling molecules are expressed in this region during generation of dorsal interneurons, such as members of FGF and Wnt families. It is still not clear if FGF play a role in dorsal interneurons formation, however Wnts are implicated in patterning, proliferation and cell determination (Wodarz and Nusse, 1998). Dorsally located interneurons can be grouped into eight distinct subtypes defined by either the expression of specific transcription factors, initial migration, dependence on roof plate signals, or date of birth (Gross et al., 2002; Muller et al., 2002). In mice, there are six early-born post mitotic dorsal interneuron populations called dI1-dI6, which are generated between E10-E12.5 and two later-born post mitotic populations called dILA and dILB produced between E11-E13. All populations are defined by the expression of specific homeodomain transcription factors. These neurons can be further classified by their dependence on roof plate signaling for formation: class A (dI1-dI3) are dependent on, and class B (dI4-dI6, dILA/B) independent of, roof plate signals (Lee et al., 2000). While class A neurons end up in deep dorsal horn layers to form relay interneurons that project contralaterally to transmit sensory information to higher brain regions (Bermingham et al., 2001), dILA/B neurons migrate to more superficial laminae of the dorsal horn and develop into 1.1 Nervous system development 9 association neurons, which serve to integrate sensory input and characteristically project ipsilaterally (see Figure 3). Fig. 3: Patterning of dorsal spinal cord showing the different classes of dorsal interneurons, the time they are generated and the transcription factors implicated in their differentiation in addition to their migratory paths (adapted from (Helms and Johnson, 2003)). In contrast to dorsal interneurons, ventral interneurons are differentiated by graded Shh induction and can be subdivided into four domains expressing differential transcription factors. The V3 domain is located close to the floor plate below the differentiating motor neurons whereas the V0, V1 and V2 are located dorsally to motor neurons and are separated by sharp boundary depending on combinatorial expression of various transcription factors (see Figure 1). Interestingly, the generation of certain sets of interneurons in the dorsal-most region of the ventral neural tube is Shh independent. These interneurons subtypes can be induced by a parallel signaling pathway that is mediated by retinoids derived from the paraxial mesoderm and possibly from the neural plate cells (Pierani et al., 1999). 1.1.2.3 Neural crest differentiation and migration A very special population of neural precursor cells, called neural crest cells, is found at the dorsal site of the early neural tube. Neural crest cells form at the border between the neural plate and the future epidermis and delaminate from the neuroepithelium in a rostro-caudal wave. Neural crest cell precursors are multi-potent progenitors and can differentiate into a very broad range of cell types. They form most of the peripheral nervous system including 10 Part I: Introduction sensory, sympathetic and enteric neurons and glia. Additionally they give rise to melanocytes, smooth muscle, dermis, connective tissue, cartilage and bone (Le Douarin and Dupin, 2003). The embryological origin of crest cells is little understood. While it has been known for some time that inductive interactions between the neural plate and the non-neural ectoderm underlie the initial specification of neural crest cells at the neural plate border, the molecular nature of these signals has been less clear. Earlier in vitro experiments in chicken suggested a role for bone morphogenic proteins (BMPs) as endogenous neural crest inducers (Basler et al., 1993), (Liem et al., 1995). Indeed, many BMPs can mimic the ability of epidermis to induce the formation of crest cells from neural plate explants. However, recent studies in frog and chicken provide strong evidence that Wnts are rather used as endogenous neural crest inducers (Garcia-Castro et al., 2002). Both gain of function as well as loss of function experiments in chicken show that Wnt6 plays a primary role as a neural crest cell inducer. Furthermore, a novel extracellular glycoprotein called Noelin-1 has been implicated in neural crest development. Noelin-1 is expressed at the lateral edges of the neural plate where it appears to maintain the competence of neural epithelial cells to form neural crest cells. Neural crest cells migrate along specific pathways to form a very diverse range of cells, ranging from melanocytes in the skin to neurons in the sensory ganglia. Although very little is known about the signals governing the differentiation of melanocytes, in vivo and in vitro analysis demonstrates that the decision to differentiate into pigment cells or neural derivatives is made early during migration. In zebrafish, medially located neural crest cells are induced by local wnt-1 and wnt-3a signals to form pigment cells, whereas lateral cells form neurons (Dorsky et al., 1998). In contrary, the autonomic nervous system induction seems to be BMP4 and BMP-7 dependent. These two factors are expressed in the dorsal aorta that is adjacent to the sympathetic ganglia and seem to induce sympathetic neurogenesis (Reissmann et al., 1996). Whereas sensory neurons are specified by the bHLH proteins Neurogenin-1 (ngn-1) and -2 that are expressed in the neural crest precursor of sensory neurons (Ma et al., 1999), several in vitro and in vivo studies implicate neuregulin, ErbB2 and ErbB3 as well as a transient effect of the notch signaling in the specification of glial cells (Morrison et al., 2000). Following their induction in the dorsolateral neural tube, neural crest cells undergo an epithelial to mesenchymal transition and begin to migrate. Recent evidence suggests that in addition to being implicated in neural crest induction, BMP signaling is also involved in downstream aspects such as the onset of neural crest migration (Sela-Donenfeld and Kalcheim, 1999). The earliest known response to this induction is the expression of two transcription factors slug and snail (Nieto et al., 1994; Sefton et al., 1998). Following 1.1 Nervous system development 11 expression of snail in epithelial cell lines, the cells become mesenchymal and migratory (Cano et al., 2000). A number of molecules that either serve as substrates for migrating crest cells or delimit migratory pathways by forming repulsive boundaries have been identified (Robinson et al., 1997; Bronner-Fraser, 1993; Perris and Perissinotto, 2000). In avian trunk, neural crest cells travel along two distinct pathways. Some cells emerging from the dorsal neural tube, adopt a ventromedial path through the rostral but not caudal somitic sclerotome, whereas other crest cells travels dorsolaterally in a uniform manner between the somites and overlying ectoderm. While cells of the ventrolateral path aggregate bilaterally along the developing spinal cord to form DRGs and sympathetic ganglia, migratory cells of the dorsolateral path contribute mainly to the formation of pigment cells (see Figure 4) (Krull, 2001). Fig. 4: Avian trunk neural crest cells travel on two distinct pathways after emigration from the neural tube. Schematic diagram representing a longitudinal view at the trunk region of an avian embryo. Some trunk crest cells (red) emerge from the dorsal neural tube and travel ventromedially, through the rostral but not caudal, somitic sclerotome. Other neural crest cells (black) migrate dorsolaterally in a uniform manner between the somites and overlying ectoderm. DM, dermamyotome; Scl, sclerotome; No, notochord; Ao, aorta, Ec, ectoderm; NT, neural tube; R, rostral; C, caudal (adapted from (Krull, 2001)). 12 Part I: Introduction 1.1.3 Axon guidance While cell migration is a prerequisite for the correct positioning of neurons within the developing embryo, directed axon outgrowth is essential for the accurate wiring of the nervous system during development. To form functional contacts with appropriate targets, axons grow in a highly stereospecific manner over considerable distances. The precise wiring of the nervous system occurs mainly by two types of mechanisms: early acting mechanisms independent of neural activity (molecular mechanism) and later-acting activity based mechanisms. Studies performed over the last two decades in an attempt to understand the early acting mechanisms have provided a detailed understanding of the cellular interactions between axonal growth cones and their surroundings that direct pathfinding. 1.1.3.1 Pathfinding of motor axons During development, motor neurons acquire distinct identities that are reflected in their choice of specific axon pathways and their synaptic targets. Motor axon pathfinding occurs in a stepwise manner and is dependent on the differential action of guidance cues, which are serially deployed at discrete locations along the axonal pathway. Thus, a motor neuron journey is divided into several stages: axonal exit from the CNS, growth along a shared common pathway and navigation to and away from different choice points (Schneider and Granato, 2003). The first step in a motor axon’s pathway is to correctly exit the CNS and project its axon into one of the segmental nerves, connecting the CNS to the periphery. Motor axons grow initially away from the floor plate and penetrate the neuroepithelium at specific exit points. Co-culture experiments in vitro demonstrate that all classes of motor axons are repelled when placed adjacent to floor plate cells (Guthrie and Pini, 1995). Additionally, all spinal motor axons choose a single ventral root within each somitic hemi-segment to leave the spinal cord. A specific type of neural crest-derived cells, called the boundary cap cells, is located at these exit points (Niederlander and Lumsden, 1996). These cells serve, most likely, as gatekeeper between the CNS and the PNS and may play a role in inducing segmental nerve patterning (Golding and Cohen, 1997). Support for this speculation comes from a recent studies which demonstrates that boundary cap cells are not implicated in attracting motor axons to their exit point but rather in keeping their cell body confined in the CNS (Vermeren et al., 2003). The mesoderm adjacent to the spinal cord is divided into a series of segmented blocks, the somites, which become partitioned into sclerotome and dermamyotome components. After 1.1 Nervous system development 13 motor axons emerge from the spinal cord at exit points, they traverse the sclerotome component of the somite only within its rostral half. Repulsive and attractive activities, derived from the caudal and rostral halves of the sclerotome respectively, impose the periodic arrangement of motor nerves exiting from the spinal cord (Keynes and Stern, 1984). Despite emerging evidence from in vitro studies proposing the involvement of diverse proteins in spinal motor axon segmentation, to date no in vivo data exist that demonstrate the implication of any axon guidance or cell adhesion molecules in this process. Thus, while caudal sclerotome cells play an essential role in guiding the segmental exit of motor axons, the identity of the cues by which they do so remains unclear. Once motor axons have left the CNS, different classes of motor neurons innervate variable muscle targets following predefined pathways. For example, motor neurons of the MMCL send axons to innervate epaxial muscle and grow towards the dermamyotome, whereas other motor axons of the MMCM avoid the dermamyotome and navigate ventrolaterally to innervate hypaxial muscles (see Figure 2). At the hindlimb level, spinal motor axons appear to grow along well-defined highways making pathway changes at specific choice point regions. In chicken, multiple spinal nerves converge to form the crural (segment L1-L3) and sciatic nerve trunk (segment L4-L8) (see Figure 5). Motor axons in each of these nerve trunks grow to the base of the limb, called the plexus area, where they pause for 24 hours before entering the limb bud. Apparently, axons wait in the plexus for limb maturation to occur (Varela-Echavarria et al., 1997). Within the plexus region, axon trajectories are highly individualistic with many abrupt turns, perhaps reflecting a process of active sorting (Tosney and Landmesser, 1985a). This process appears to be at least in part dependent on the differential expression of a set of cell adhesion molecules by different motor axons. Motor axons of the LMCM and LMCL further subdivide at the limb base to form distinct ventral and dorsal nerve trunks. This step seems to be dependent on target-derived chemoattractants such as HGF and guidance cues produced in the developing limb. Additionally, some tissues act as barriers to axons as they navigate to the hindlimb. Motor axons seem to avoid the perinotochordal mesenchyme and the pelvic girdle precursor tissue (Tosney and Oakley, 1990; Oakley and Tosney, 1991). Once motor axons are near their target muscle, they have to recognize and form synapses with the appropriate muscle fiber. Motor neurons that innervate muscles are matched with their target in such a way to generate a precise topographic map (Laskowski and Sanes, 1987). 14 Part I: Introduction Fig. 5: In avian at the hindlimb level, multiple spinal nerves converge to form the crural (segment L1-L3) and sciatic nerve trunk (segment L4-L8). Each nerve divides further into dorsal or ventral trunk targeting different muscles of the hindlimb (adapted from (Landmesser, 2001)). 1.1.3.2 Interneurons and midline crossing In the CNS of a wide variety of bilaterally symmetric organisms, different interneurons project axons along specific trajectories, which are parallel or perpendicular to the midline. The best-studied interneurons are the commissural interneurons, whose cell bodies are located in the dorsolateral spinal cord extending axons in a circumferential path toward the floor plate. The earliest commissural axons or “pioneer” axons travel along the lateral edges of the spinal cord until they reach the floor plate whereas the later projecting axons or the “followers” extend along the same pathway. After reaching the midline, these commissural axons cross 1.1 Nervous system development 15 through the ventral most third of the floor plate, subsequently turning orthogonally at the contralateral side of the floor plate (Bovolenta and Dodd, 1990). While another subtype of the dorsally located commissural interneurons, known as association interneurons execute right angle turns and extend parallel to the floor plate along the ipsilaterally-projecting lateral funiculus (Colamarino and Tessier-Lavigne, 1995b). A more ventrally positioned population of interneurons that develop in a region between the floor plate and motor neurons extend their axons along the ipsilateral longitudinal pathway within the ventral funiculus (Yaginuma et al., 1990). Regardless of the fact whether specific interneuron populations cross or do not cross the midline, all axonal interneurons are directed by the floor plate that affects noticeably the behavior of a growth cone in its vicinity. 1.1.3.3 Projections of dorsal root ganglia A subpopulation of multipotent neural crest cells migrates along stereotypic pathways and coalesces at specific locations to form the spinal sensory ganglia also called the dorsal root ganglia (DRG). Spinal sensory neurons comprise a morphologically and functionally heterogeneous group of neurons, specialized in the transfer of different sensory modalities (Farinas et al., 2002). Each DRG innervates a full array of targets in the periphery, including skin, muscle, and viscera. Individual DRG neurons connect to specific types of sensory receptors, conveying information about position in space (proprioception), pain (nociception), distension, or touch (mechanoception) to the CNS. Neurotrophins play an essential role in the maintenance of a normal complement of neurons since all sensory neurons require the presence of at least one neurotrophin during development . Although the neurotrophic hypothesis postulates that neurons become dependent on a particular neurotrophin when their axons encounter their final targets there is evidence demonstrating that neurotrophins are expressed during early development before axon-target recognition and are therefore also implicated in gangliogenesis (Buchman and Davies, 1993; Farinas et al., 1996). Anatomical and physiological data, document very well the early peripheral projection of sensory neurons. Sensory fibers innervating the hindlimb are established in a precise orderly manner (Honig, 1982). During normal development, sensory axons appear to grow on the axons of adjacent motoneurons and always project to the same muscles as the neighboring motor neurons (Tosney and Landmesser, 1985b). Therefore, the absence of motor neurons causes also severe missprojections of sensory neurons. Interestingly sensory neurons innervating skin or muscle in the periphery appear less rigidly specified than motoneurons 16 Part I: Introduction and have more flexibility in their pathway and target choices. At the stages when innervations are being established, cutaneous as well as muscle afferents, unlike motoneurons, may not yet have acquired specified identities and the ability to recognize and respond selectively to their appropriate targets (Adams and Scott, 1998). Additionally the central projections of sensory neurons follow a strict spatio-temporal pattern with different DRG neurons having central arborisations in the spinal cord that are specific for the sensory modality. In chicken, cutaneous and muscle axons of sensory afferents reach the spinal cord by stage 23, stalling there for 24 hours in the primordium of the dorsal funiculus before extending axons rostrally as well as caudally. At around stage 28 central projections begin to enter the gray matter of the spinal cord. While cutaneous afferents branch frequently remaining in the dorsal horn (Mendelson et al., 1992), proprioceptive axons reach the vicinity of motor neuron dendrites without branching and form functional contacts around stage 32 (Davis et al., 1989). The segregation of afferent inputs into laminar-specific projections is dependent on diffusible factors, integral proteins and/or extra cellular matrix proteins (Ozaki and Snider, 1997). Both types of sensory neuron projections (peripheral and central) are established precisely and correctly from the outset, and neither cell death nor retraction of axons plays a role in the development of appropriate connectivity. During the initial stage of DRG axonal growth, surrounding “non target” tissues such as dermamyotome, the notochord, and the ventral spinal cord release strong chemorepulsive signals that inhibit DRG axons (Masuda et al., 2003) in vitro. However, despite many advances made in the identification of axon guidance cues affecting sensory neurons outgrowth in vitro, our understanding about the molecular mechanisms required in vivo for the axonal pathfinding of these neurons remain fragmentary and incomplete. 1.1.3.4 Molecular mechanism of cell migration and axon guidance Neurons extend axons to appropriate targets with the help of a very specialized structure at the tip of the advancing axon called the “growth cone”. On the molecular level, our understanding about how a growth cone recognizes which path it has to take and how it reaches its precise synaptic target while encountering numerous potential, other targets on its way, is still fragmentary. Nevertheless, many advances have been made toward a better understanding of the molecular mechanisms of growth cone based on axon guidance. The identification and characterization of different families of axon guidance molecules and the use of functional tests to discover the biological roles of these guidance cues in vivo, has greatly advanced our knowledge. 1.2 Axon guidance forces 17 To accomplish the complex task of reaching a distant target, the initial axon trajectory seems to be broken into short segments that terminate at specialized cells forming intermediate targets also called “choice points”. These intermediate targets provide guidance information enabling axons to select and initiate growth along a particular segment. Growth cones that approach an intermediate target reduce their speed and assume a more complex morphology with more filopodia, presumably to better sample the environment. Therefore, two types of cellular behaviors can characterize axon growth: a simple linear growth along “highways”, punctuated by a more complex decision-making behaviors at choice points as axons switch from one highway to another. In insects, some intermediate targets are made up of cluster of “guidepost cells”, whose ablation results in misrouting of axons that normally contact them (Raper et al., 1984; Bastiani et al., 1984). Additionally, the wiring process of the nervous system occurs in a stepwise manner. While pioneer axons navigate through an axon free environment when the embryo is still relatively small, later developing axons benefit from the scaffold provided by earlier projecting axons allowing the later developing ones to grow along pre-existing tracts or fascicules for at least some of their trajectory only switching from one fascicule to another at specific choice points. This “selective fasciculation” strategy simplifies the assembly of large nervous system network by allowing follower axons to use predefined paths (Kolodkin et al., 1992). Despite their intrinsic characteristics, growth cones are not endowed with a predestined and autonomous ability to find their correct targets. For this, they must respond to and depend on cues produced by their surroundings (O'Connor, 1992). Therefore, a growth cone must express a specific set of receptor on its leading edge that reacts appropriately to the surrounding molecules (Tessier-Lavigne and Goodman, 1996). Many proteins playing a role in axon guidance have been identified in invertebrates and vertebrates and it appears that both the mechanisms and the molecules are conserved among species. In fact, evolutionary conservation of guidance molecule is so great that insights gained in invertebrates can be immediately relevant to vertebrates, and vice-versa (Tessier-Lavigne and Goodman, 1996). 1.2 Axon guidance forces Over a century ago, Cajal already proposed long-range chemoattraction as a mean to guide axons to specific targets. Only in the late 1970s, Gunderson and Barrett showed that sensory neuron growth cones can indeed respond to a gradient of a protein called nerve growth factor (NGF). Although NGF is unlikely to be a long-range chemoattractant in the developing organism, this study initiated a resurgence of interest in chemotropism. Later, experiments 18 Part I: Introduction performed by Lumdsen and colleges (Lumsden and Davies, 1983), demonstrated that cocultures of neurons and their target areas, embedded in a collagen matrix produce stable molecular gradients originating from these target tissues that are capable of attracting extending axons. More recently, the finding that axons could also be repelled in vitro by tissues that these axons normally avoid, provided strong evidence that guidance information is not only attractive but also repulsive (Fitzgerald et al., 1993; Colamarino and TessierLavigne, 1995a). Axons can also be guided at short-range by contact-mediated mechanisms involving non-diffusible cell surface or extracellular matrix (EMC) molecules. While the process of contact attraction has been implicated in “selective fasciculation”, in which growth cones confronted with several pre-existing axons fascicules select a specific pathway (Raper et al., 1984), contact repulsion of axons has been extensively documented in the retinotectal system. Thus, the growth cone appears to be guided by four general mechanisms: contact and chemoattraction as well as contact and chemorepulsion (Tessier-Lavigne and Goodman, 1996). The presence of discrete classes of diffusible and non-diffusible factors, some attractive while others repulsive are however not as strict as one might expect. Instead, axon guidance molecules appear to act through conserved mechanisms activating or inhibiting oftenoverlapping signal transduction pathways. The same molecule may function as growth inhibitor in one assay but as an attractant in another, suggesting that most of guidance molecules are not exclusively attractive or repulsive but rather bifunctional, playing variable roles in guiding different growth cones (Tessier-Lavigne and Goodman, 1996). Additionally it was recently shown that a growth cone can exhibit attractive as well as repulsive responses to the same guidance cue depending on the level of cytoplasmic cyclic AMP (cAMP) (Ming et al., 1997; Song et al., 1997). Much of the current studies of axon guidance are directed toward defining the precise complement of forces orchestrating particular guidance decisions. An individual axon might be “pushed” from behind by a chemorepellent, “pulled” from afar by a chemoattractant and “hemmed” in by attractive or repulsive local cues. These forces appear to play in concert all together to ensure accurate and specific axon guidance. 1.3 Families of axon guidance molecules In the early 1990s, the introduction of powerful in vitro assays to detect guidance activities in the developing vertebrate nervous system and the growing interest in invertebrate axon guidance led to the discovery of several conserved families of axon guidance molecules. The 1.3 Families of axon guidance molecules 19 most prominent axon guidance cue families are the netrins, slits, semaphorins, and ephrins. These molecules are not the only known guidance molecules but are by far the best studied ones. Netrins are highly conserved proteins that while exerting a chemoattractant activity on commissural axons (Culotti and Merz, 1998) are responsible for a chemorepellent effect (Colamarino and Tessier-Lavigne, 1995a) on growth cones of trochlear motor axons. In nematodes, Unc6/Netrin is known to mediate its activity through binding to unc-5 and unc-40 receptors (Hedgecock et al., 1990). Biochemical and genetic studies have confirmed that Unc40/DCC and Unc-5 receptors function as netrins receptors in several different species (Culotti and Merz, 1998; Keleman and Dickson, 2001), guiding a variety of axons and cells in vivo as well as in vitro (Winberg et al., 1998a; Yee et al., 1999). Slit proteins represent another family of conserved axon guidance molecules Slits are large secreted proteins acting as midline repellent whose function appears to be conserved along evolution. Interestingly, slit was also purified as a factor that stimulates sensory axon branching and elongation (Wang et al., 1999), suggesting that Slits like netrins may be multifunctional. However, the best-studied functions of slits are its involvement in midline guidance. In the Drosophila embryo, slit acts as a short-range repellent to prevent ipsilateral projecting axons from crossing the midline and commissural projecting axons from recrossing the midline (Battye et al., 1999). In vertebrates, slit has been shown to play a crucial role in the formation of the optic chiasm and in axon guidance at the ventral midline guidance through binding to its receptor Roundabout (Robo) preventing inaccurate crossing of ipsilaterally as well as contralaterally projecting axons (Kidd et al., 1998; Plump et al., 2002; Erskine et al., 2000). One of the largest families of mainly repulsive axon guidance molecules are the ephrins. Ephrins were initially characterized as membrane bound proteins, involved in guiding vertebrate retinal axons to appropriate topographic locations in the tectum. Ephrins and Eph receptors fall into two classes; ephrin-As, which are anchored to the membrane by a GPI linkage that bind to EphA receptors and ephrins-Bs, which are transmembrane proteins binding to EphB receptors. In the visual system, topographic mapping of retinal axons along the anterior-posterior axis depends on ephrin-A mediated repulsion. Ephrin-A are expressed in a gradient in the tectum and their receptors EphA are expressed in a complementary gradient in the retina. Thus, retinal axons with successively higher EphA levels map to successively lower points along the ephrin-A gradient (Wilkinson, 2001). 20 Part I: Introduction However, emerging evidence nowadays suggest that ephrins and their receptors control axon guidance in many other places in addition to their ability to signal bi-directionally and mediate both attraction and repulsion. 1.4 Plexins and Semaphorins 1.4.1 Semaphorin family Semaphorins belong to a large family of cell surface or secreted proteins and represent one of the best-studied families of axon guidance molecules. Initially all semaphorin family members were believed to mediate inhibitory actions on axon pathfinding, branching or targeting, but nowadays there is increasing evidence that some semaphorin proteins also play a role in chemoattraction and cell migration (Kolodkin et al., 1992; Luo et al., 1993; Matthes et al., 1995; Messersmith et al., 1995). Semaphorins are defined by a highly conserved 500 amino-acid extracellular domain, the Sema domain, which contains 14 to 16 cysteines and some conserved N-linked glycosylation sites. More than 20 different semaphorin members have been identified up to now. Based on their degree of sequence similarity and domain organization semaphorins can be grouped into seven different subclasses. While semaphorins belonging to subclass I and II are found exclusively in invertebrates, semaphorins belonging to the subclasses III-VII are specific for vertebrates. While in general the majority of semaphorin subtypes are transmembrane proteins, semaphorin members belonging to the subclasses II, III, and VII, represent exceptions, which are secreted or GPI anchored molecules respectively (see Figure 6) (Tamagnone and Comoglio, 2000). While the cytoplasmic domains of transmembrane semaphorins share no significant homology to any known protein and have very little similarity to each other, the structural conservation of the Sema domain implies that this part is essential for semaphoring binding and function. Interestingly, all semaphorins form disulfide-linked homodimers, whose oligomerization is crucial for semaphorin-signaling, and their biological activity seems to be determined by a relatively short stretch of amino acids within the Sema domain (Koppel and Raper, 1998; Klostermann et al., 1998; Koppel et al., 1997). 1.4.1.1 Semaphorin Class I & II SemaI, the first semaphorin member to be identified, is a transmembrane protein expressed on subsets of fasciculating axons and on stripes of epithelial cells in the grasshopper limb bud (Kolodkin et al., 1992). Antibody perturbation experiments in grasshopper demonstrated that 1.4 Plexins and Semaphorins 21 SemaI plays a crucial role in steering a pair of sensory neurons to the limb bud by regulating axon defasciculation and branching. The other invertebrate semaphorin, SemaII is expressed transiently in a subset of motor neurons and a single thoracic muscle during motor neuron outgrowth and synapse formation. Gain-of-function analysis in Drosophila, show that ectopic SemaII expression in muscle cells inhibits normal synaptic terminal arborisation of two different motor neurons subtypes without affecting the growth cones of other motor axons (Matthes et al., 1995). The fact that SemaII ectopic expression does not affect the oriented growth of axons toward these muscles suggests that SemaII rather serves as a selective target-derived signal inhibiting the formation of specific synaptic terminal arbores than influencing early aspects of axon guidance (Matthes et al., 1995). Fig. 6: Semaphorins are phylogenetically related proteins, sharing sema domains similar to the cystein rich Metrelated sequences (MRSs). Semaphorins are divided into 7 subgroups, Class 1 & 2 existent only in invertebrates whereas Class 3-7 are exclusive for vertebrates. Class 2&3 are secreted semaphorins while all the other members are membrane bound; Class 1, 4, 5 and 6 are transmembrane and Class 7 contains GPI linked members. Abbreviations: G-P/ IPT motif, glycine-proline repeat/ immunoglobulin-like fold; GPI, glycosyl phosphatidylinositol,; MRS, Met-related sequence (adapted from (Tamagnone and Comoglio, 2000). 1.4.1.2 Semaphorin Class III Class III Semaphorin members share three structural motifs, the 500 amino acid sema domain, a C-2 type immunoglobulin (Ig) domain, and a positively charged carboxy terminal tail. Secreted semaphorins act as diffusible signals, although their diffusion distance might be limited because they are tethered to the cell surface and extracellular matrix in vivo by the charged sequence at the C-terminus of the protein (Bagnard et al., 2000). So far, Sema3A is the most extensively studied semaphorin among all class III members. It has been identified by two independent approaches. While Kolodkin and colleges identified vertebrate Sema3A based on its homology to SemaI (Kolodkin et al., 1993), Raper’s group 22 Part I: Introduction identified Sema3A as the major growth inhibitory protein for sensory axons(Luo et al., 1993). Sema3A is widely expressed in the developing central and peripheral nervous system as well as in several tissues surrounding the spinal cord where it is responsible for collapsing a broad range of different axons. Sema3A is expressed, in the perinotochordal mesenchyme as well as mesenchymal cells around the spinal cord and DRG with the exception of nerve exit points (Puschel et al., 1995). Additionally Sema3A transcripts have been found in the dorsal aorta, the connective tissue separating the developing muscles and in the caudal half of the sclerotomes suggesting a potential involvement of Sema3A as a repulsive guidance molecule in axon pathfinding of spinal neurons (Shepherd et al., 1996; Giger et al., 1996). Despite its relatively broad expression in the developing embryo the effect of Sema3A is highly specific as in dorsal root ganglia only NGF sensitive sensory axons are affected (Messersmith et al., 1995). Nevertheless, it also collapses axons of sympathetic ganglia neurons (Puschel et al., 1995), motor neurons (Shepherd et al., 1996; Varela-Echavarria et al., 1997), sensory neurons from the trigeminal, facial and vagal cranial ganglia, in addition to axons from olfactory (Kobayashi et al., 1997), pontine (Rabacchi et al., 1999), cortical (Bagnard et al., 1998) and hippocampal neurons (Chedotal et al., 1998; Steup et al., 1999). Interestingly, Sema3A cannot induce collapse of retinal ganglion growth cones, demonstrating that Sema3A is specific to a subset of neuronal growth cones. Recent studies show that Sema3A can also suppress the migration of avian trunk neural crest cells in vitro (Eickholt et al., 1999). The analyses of Sema3A function in vivo using gene disruption in mice gave rather surprising results. While the nervous system appears normal with the exception of some minor defects in sensory axon projections, mutant mice exhibit sever malformation of skeletal structures in addition to pronounced and selective hypertrophy of the right ventricle of the heart resulting in a high degree of mortality around birth (Behar et al., 1996). The fact that Sema3A is highly expressed in rat heart, sclerotomes, ribs and pelvic girdle during development (Wright et al., 1995) might account to explain such defects after Sema3A gene disruption. Intriguingly, Sema3A mutant mice from a different laboratory are viable until adulthood, do not show any heart defects and have normal dorsal root projections. However, they appear to have disturbances in some peripheral nerve projections that are affected by Sema3A in vitro. Many cranial nerves (trigeminal, facial, vagal, accessory and glossopharyngeal) as well as spinal nerves are highly defasciculated and some nerves overshoot their targets (Taniguchi et al., 1997). These data imply that Sema3A expressed in the surrounding tissues may drive axons into fascicules and that its absence might provoke axons in the periphery of mutant animals to 1.4 Plexins and Semaphorins 23 enter regions that are normally strongly repulsive. Surprisingly, despite the wider aberrant paths adopted by some nerves, the overall target recognition seems to be well achieved in the knock out mice. The moderate defects of the nervous development in vivo observed in Sema3A might be explained by the presence of six additional class III members with overlapping distributions and potentially overlapping activities. To answer the question whether Sema3A and other Class III semaphorin play combinatorial roles in nervous system development, it will be essential to generate not only knock out animals for other Class III members but also to disrupt the expression of several semaphorin members at the same time. Other members of semaphorin class III have also been analyzed to a lesser extend than Sema3A. To date, only Sema3C and Sema3F have been reasonably well characterized in vitro and in vivo. In vitro Sema3C repels neurites from CA1 and medial septum but has no effect on CA3, dentate gyrus and entorhinal axons. Interestingly Sema3C appear to exert a dual role in vitro depending on the neuronal population studied. It acts as a repellent on sympathetic axons, has no effect on DRG neurons but in contrary attracts axons of cortical explants (Bagnard et al., 1998). As for Sema3A, Sema3C mutant mice show no obvious defects in the development of the nervous system. The terminations of septal fibers in the inner and outer molecular layers of the dentate gyrus are normal. However, mutant mice display severe congenital cardiovascular defects and die soon after birth because of the interruption of the aortic arch and improper septation of the cardiac outflow tract (Feiner et al., 2001). This phenotype is consistent with the expression of Sema3C in the mesenchyme surrounding the branchial arch arteries and in the myocardial cuff as well as the cardiac outflow tracts suggesting a role for Sema3C in guiding migratory cardiac crest cells. Sema 3F is another semaphorin member that appears to govern the pathfinding of certain nerves in the CNS and PNS in vitro and in vivo. It is expressed in embryonic hippocampal regions in mice at the time of axonal outgrowth (E15 to E17) and shows repulsive activity on CA1, CA3 and dentate gyrus axons in vitro (Chedotal et al., 1998). In addition, it is also capable of collapsing trochlear motor axons (Giger et al., 2000). Sema3F is also present in the vomeronasal organ and in cells that flank the path of vomeronasal sensory neurons as well as in the accessory olfactory bulb and the main olfactory bulb. Interestingly in vivo studies, based on the generation of Sema3F mutant mice, demonstrate that Sema3F is crucial for axon fasciculation and segregation but not for target recognition in the olfactory system (Cloutier et al., 2004). Additional analyses of the CNS of Sema3F mutant mice reveal that this protein is essential in the ventral forebrain for anterior commissure axons to fasciculate and decussate normally at the CNS midline as well for the formation of the infrapyramidal tract (Sahay et 24 Part I: Introduction al., 2003). Furthermore, Sema3F seems to be important for the proper organization of specific cranial nerve projections in the PNS, as Sema3F mutants show severe defects in the trochlear nerve where only few axons exit the hindbrain-midbrain junction. Moreover, the occulomotor nerve is largely defasciculated but maintains its peripheral trajectory (Sahay et al., 2003). The other members of Semaphorins class III have been less well studied and understood. In vitro experiments demonstrate that Sema3B can repel sympathetic axons (Takahashi et al., 1998) but no data were reported about the role of Sema3D and Sema3E in vitro or in vivo (Koppel et al., 1997; Raper, 2000). 1.4.1.3 Membrane attached Semaphorins In contrast to Semaphorins in subclasses III and IV, Sema5A and Sema5B belonging to the subclass V lack the IgG domain and have instead seven carboxy-terminal thrombospondine repeats followed by a short intracellular C-terminus that is unique for this subclass. The thrombospondine repeats, which account for more than half of the protein sequence, are components of different extracellular matrix proteins and have been shown to promote potentially neurite outgrowth. Both semaphorins are expressed during early murine embryogenesis and adult tissues in mutually exclusive domains. Sema5A is present in axial and paraxial mesodermal tissues, limb bud, optic disc and nerve whereas Sema5B expression is exclusively restricted to the neuroepithelium along the entire anterio-posterior axis (Adams et al., 1996 117). While there is no available information concerning in vitro as well as in vivo functions of Sema5B, data on Sema5A clearly suggest its functional involvement in axon guidance. The fact that Sema5A is present in the paraxial mesoderm prior to the arrival of growing axons to the limb bud, suggests a potential role for this semaphorin not only in directing crest cell migration and in sensory or motor axonal outgrowth but also in somitogenesis (Adams et al., 1996 117). Additionally, Sema5A seems to be implicated in vitro in collapsing retinal axon growth of rodents (Goldberg et al., 2004), which has been confirmed in vivo by the use of function blocking antibody. Perturbation of Sema5A function leads to retinal axons straying out of the optic nerve bundle indicating that Sema5A normally help ensheathing the retinal pathway (Oster et al., 2003). This repulsion activity is mediated through the Sema domain of the protein. The thrombospondine repeats do no show any repulsive or attractive activity in vitro on axonal growth of retinal ganglion cells (Goldberg et al., 2004). Semaphorins grouped into the subclass VI, display a relatively simple extracellular part, in which only the highly conserved Sema domain is present. However, their intracellular stretch 1.4 Plexins and Semaphorins 25 is quite long comparing to other Semaphorin members, suggesting that this part in Class VI semaphorin might have unique functions. Four different Class VI family members have been identified so far (6A-6D), with Sema6A, being the best studied one. In mice, several tissues express Sema6A during embryonic development; however, this expression is strongly down-regulated perinataly. In the nervous system, Sema6A mRNA is present in the spinal cord and the DRGs and in different regions of the Brain. While early embryonic expression is restricted to the ventral spinal cord, at later stages expression is also observed in the dorsal spinal cord in areas of lamina I and II. Although Sema6A is, absent from all cervical and thoracic sympathetic ganglia, it is expressed in skeletal muscles near the tissue encircling the ganglia as well as in glossopharyngeal and cochlear ganglia (Zhou et al., 1997). Interestingly, in vitro Sema6A acts as a repellent on E8 chicken growth cone of sympathetic ganglia and sensory NT-3 and NGF sensitive DRG neurons. Surprisingly, abolishing Sema6A transcripts in vivo by gene trapping, gives mutant mice that are viable and fertile without any behavioral or morphological defects. Although most of the tracts in the CNS appear normal, the thalamocortical tract shows pathfinding defects at the caudal level. While most rostral projections are similar to wild type, the caudal part fails to turn through the internal capsule and instead projects towards the amygdala (Leighton et al., 2001). Recent studies have demonstrated the existence a Sema6A isoform that contains a longer cytoplasmic tail. This semaphorins 6 variant is capable of directly linking Ena/VASP proteins, which are known to play a crucial role in actin filament dynamics (Klostermann et al., 2000). Sema6B is the least understood semaphorin member of all Class VI Semaphorins. Sema6B expression appears early in development of dorsal root ganglia, somites and brain of rodents (Kikuchi et al., 1997). In contrast to other semaphorins, Sema6B appear to be homogenously expressed throughout the entire spinal cord. Additionally, Sema6B expression persists in adulthood in many tissues such as brain, heart, and lungs. Interestingly, Sema6B in vitro binds specifically the SH3 domain of the proto-oncogene c-src suggesting that it can trigger intracellular signaling and act as a receptor (Eckhardt et al., 1997). Sema6C is expressed in rat spinal cord as well as in dermamyotome, DRGs and the notochord during development. Later on, Sema6C message can be found also in cranial motor ganglia, the olfactory epithelium and the cerebellar plate. Postnataly, Sema6C expression is present in different cerebellar layers, pontine and inferior olive nuclei as well as in adult skeletal muscle tissue and many CNS structures (Kikuchi et al., 1999). Thus, temporal expression of Sema6C in neurons and in their target areas during development suggests a potential role for this 26 Part I: Introduction protein in axon guidance of motor and sensory neurons as well in directing commissural or cerebellar neurons. Sema6C shows, growth cone collapse activity on DRGs in vitro and the target regions of DRG neurons express Sema6C during development. Two isoforms of Sema6C, derived from alternative splicing, were identified and their expression pattern is regulated in tissue and age dependent manner (Kikuchi et al., 1999). Sema6D, the last member of Class VI semaphorin to be characterized, exists in five different splice variants whose expression patterns are tissue specific. In vitro Sema6D has been shown to induce growth cone collapse of DRG and hippocampal neurons but had no effect on cortical neurons (Qu et al., 2002). Sema6D knock down or over expression in mice or chicken causes morphological abnormalities of the cardiac tube as well as of the neural tube, suggesting that Sema6D is involved in cardiac morphogenesis and in the neural tube formation (Toyofuku et al., 2004b; Toyofuku et al., 2004a). Such phenotypes are in accordance with the observed expression pattern of Sema6D in normal mice where high levels of Sema6D expression are observed in the developing heart and neural folds. Semaphorins belonging to subclass IV and VII have been studied mainly in the immune system where they exert immuno-modulatory effects. However increasing evidence, also point to a potential implication of these proteins in the nervous system development. Sema4D, the only semaphorin Class IV member studied in the nervous system development, has been shown to stimulate outgrowth of embryonic DRG sensory neurons in vitro (Masuda et al., 2004). Similarly, Sema7A is expressed in several structures of the rat embryonic brain and promotes in vitro the growth of numerous central axons such as the vomeronasal epithelium, the olfactory epithelium, the olfactory bulb and the cortex as well as the peripheral axons of the dorsal root ganglia. However, the disruption of Sema7A gene in mice, leads only to minor defects in the lateral olfactory tract whose axons fail to branch or to project to the most caudal region of the olfactory cortex (Pasterkamp et al., 2003). 1.4.2 Semaphorin receptors and receptor complexes Over the last couple of years, several Semaphorin-binding proteins potentially involved in semaphorin signaling have been identified. The most prominent members can be grouped into two different subclasses: the Neuropilins and the Plexins. While neuropilins seem to bind different members of Class III semaphorin, plexins have been shown to interact directly with all different subclasses of semaphorins. 1.4 Plexins and Semaphorins 27 Fig. 7: Plexins are subdivided into four subfamilies (plexins A to D). Their extracellular domains contain a sema domain and MRS repeats. The extracellular domain of plexin B contains potential cleavage sites for furin-like convertases. Neuropilin-1 and Neuropilin-2 are transmembrane proteins containing a very short cytoplasmic tail. Abbreviations: G-P/ IPT motif, glycine-proline repeat/ immunoglobulin-like fold; SP domain, sex-plexin domain; CUB domain, complement-homology domain; MAM domain, meprin/A5/mu-phosphatase homology domain, MRS, Met-related sequence (adapted from (Tamagnone and Comoglio, 2000)). 1.4.2.1 Neuropilins Neuropilins (NP) are transmembrane proteins lacking a signaling-competent cytoplasmic domain. Two neuropilins genes have been identified in the genome of birds and mammalians (NP-1 and NP-2); however, no neuropilin gene has been identified in invertebrates. The extracellular domain of neuropilins contains two repeated complement-binding domains (CUB domains a1/a2 domains), two coagulation-factor-homology domains (b1/b2 domains) and a juxtamembrane meprin/A5/mu-phosphatase (MAM) homology domain (see Figure 7). While the CUB a1/a2 and b1/b2 domains seem to be essential to define the profile of semaphorin specificity, the MAM domain seems to be crucial for the functionally required NP non-covalent oligomerization on the cell surface. Neuropilins bind to the members of semaphorin class III with different affinities. While NP-1 binds with high affinity to Sema3A (He and Tessier-Lavigne, 1997; Kolodkin et al., 1997) but not to Sema3F, NP-2 binds to Sema3F with higher affinity than Sema3A (Chen et al., 1997). The fact that NPs aggregate into dimers and NP-1 with NP-2 can for heterodimers when co- 28 Part I: Introduction expressed suggests a model whereby NP-1 homodimers confer responsivity to Sema3A; NP-2 homodimers are responsible for responsivity to Sema3F and a cooperation of both neuropilins perhaps by heterodimerization confers responsivity to Sema3C (Chen et al., 1997; Takahashi et al., 1999; Renzi et al., 1999). NP-1 is expressed in particular classes of neurons, including most peripheral sensory neurons, autonomic neurons of the sympathetic ganglia, motor neurons in the spinal cord and the medulla, neurons in the hippocampal formation, retinal ganglion cells and olfactory receptors and their target neurons in the olfactory bulb (Kawakami et al., 1996). The expression of NP1 in the nervous system is developmentally regulated in both peripheral and central nervous systems. NP-1 appears first in newly differentiated neurons and persists throughout the period of active axonal growth disappearing only after the frameworks of neuronal circuits have been established. Mice, carrying a null mutation for the NP-1 gene, are embryonic lethal and exhibit analogous but somewhat stronger axon guidance defects to those observed in Sema3A knock out mice. In addition to the abnormal defasciculation of cranial nerves, peripheral nerves in the trunk are also defasciculated, DRG packages appear to be loose and sympathetic neurons are displaced. Interestingly, sympathetic neuronal precursors are not accumulated at their initial target sites around the dorsal aorta in NP-1 mutants (Kitsukawa et al., 1997), a defect also observed in Sema3A knock out mice (Taniguchi et al., 1997). However, the migratory pathways of sympathetic neuron progenitors within the rostral sclerotomes are normal (Kawasaki et al., 2002), suggesting that NP-1-mediated Sema3A activity may play a functional role in prohibiting incorrect migration of neural crest cells of sympathetic neuron lineage, promote aggregation of sympathetic neurons and influence sympathetic neurons fasciculation. NP-1 deficiency is also associated with altered vascularization in the brain and a variety of defects in the large vessels of the heart outflow (Kawasaki et al., 1999). The observed high degree of lethality may or may not be a result of interfering with semaphorin signaling, since NP-1 also interacts with VEGF by increasing its affinity to its receptor. The strongest phenotype observed in NP-1 mutants in comparison to Sema3A mice might reflect the functional loss of more than one class III semaphorin and/or the loss of semaphorinindependent functions of NP-1. NP-2 has an overall structure, which is similar to NP-1. The protein exists in six different isoforms that are generated by alternative splicing (Chen et al., 1997). NP-2 is expressed in multiple areas in the developing CNS and PNS as well as many non-neural tissues. The expression pattern partially overlaps with NP-1 but is mostly complementary. In contrast to 1.4 Plexins and Semaphorins 29 NP-1, NP-2 expression is not detected in the heart or in capillaries but is only found in the dorsal aorta. Two independent groups have recently reported the phenotype of knockout mice for NP-2. These mice are generally viable until adulthood and exhibit defects of axon fasciculation and targeting of selected cranial nerves and central projections (Chen et al., 2000; Giger et al., 2000). The mutants also lack the trochlear nerve and showed irregular trajectories of the occulomotor nerve but have no clear abnormalities in the projections and trajectories of spinal nerves. While neuropilins are clearly required for axon guidance of several neuronal populations, the absence of an obvious intracellular domain to propagate signaling suggests the existence of additional proteins to form functional signal-transducing complexes (Nakamura et al., 1998). In addition, the absence of neuropilin genes in invertebrates and the lack of binding of nonClass III semaphorins to neuropilins support the existence of other semaphorin binding proteins. 1.4.2.2 Plexin family First evidences that plexins are Semaphorin-binding proteins were obtained not by studies in the nervous system but by experiments carried out to identify a receptor for virally encoded semaphorins. Using this approach a protein called VESPR was identified which was later renamed PlexinC1. Subsequently nine different plexins have been identified in the mammalian genome, which can be sub-grouped into four different classes (Tamagnone et al., 1999). In contrast to neuropilins, plexins are also found in invertebrates, suggesting that they represent functional receptors for Class I and II semaphorins (Winberg et al., 1998b). All plexins are large integral membrane proteins with a highly conserved cytoplasmic tail. Interestingly at their amino-terminus, they contain, as their putative ligands, a highly conserved Sema domain, suggesting that an interaction between both proteins is mediated by Sema-Sema interactions. In addition to the Sema domain, the plexin extracellular domain is characterized by two or three Met-related sequence repeats (MRS). The large cytoplasmic moiety of plexins contains a strikingly conserved sex-plexins (SP) domain, which is likely to trigger novel signal-transduction pathways (Raper, 2000; Tamagnone and Comoglio, 2000) (see Figure 7). The SP domain of plexins is unrelated to any other domain found so far, though its primary sequence is highly conserved among family members and across evolution, suggesting that plexins share common biochemical functions and signaltransduction pathways. 30 Part I: Introduction Initially plexins were studied in Drosophila where PlexinA was found to bind directly to the transmembrane Class I Semaphorins and controls important aspects of axon guidance (Winberg et al., 1998b). In contrast to plexin-semaphorin interactions in invertebrates, biochemical and cellular studies in vertebrates, suggest that members of the Plexin-A subfamily form stable complexes with NP-1 or NP-2 rather than interacting directly with semaphorins. Such complex formation does not depend on the presence of the ligand (Raper, 2000; Takahashi et al., 1999; Tamagnone et al., 1999; Rohm et al., 2000) nevertheless, Plexins do not simply provide a signaling moiety to NPs but actively influence their binding efficiency for the different subsets of secreted semaphorins (Takahashi et al., 1999; Rohm et al., 2000). Moreover Plexins, have been shown to also interact directly with transmembrane or GPI anchored semaphorins, influencing axon guidance and other developmental processes in a neuropilin independent way (Tamagnone et al., 1999; Comeau et al., 1998; Winberg et al., 1998a). In mice, PlexinA subfamily members are widely expressed in the central and peripheral nervous system and are spatio-temporally regulated. PA-1 and PA-2 mRNA expression is limited to some neurons in the DRG and absent from sympathetic ganglia (Murakami et al., 2001) whereas PA-3 appears to be expressed in all peripheral ganglia, including trigeminal and vagal ganglia, in addition to DRG and sympathetic ganglia. Additionally, PlexinA3 is strongly expressed in the whole spinal cord and PlexinA2 is expressed selectively in the dorsal spinal cord (Suto et al., 2003) whereas PlexinA4 is the most abundant plexins in DRG. Targeted disruption of PlexinA3 gene demonstrates that this plexin plays a role in fasciculating the ophthalmic branch of the trigeminal nerve and regulates the development of hippocampal projections in vivo (Cheng et al., 2001). However, PlexinD1 knock out animals display cardiac defects but unrelated to cardiac crest migration. PlexinD1 seems unnecessary for cardiac crest migration but appear implicated in outflow tract septation, development of aortic arch artery and intersomitic vessels sprouting (Gitler et al., 2004). Despite the generation of a large number of mutants for different semaphorin and plexin genes, our understanding about the biological functions in vivo of most of these proteins families in the nervous system development remain limited. 1.4.3 Semaphorin and Plexins beyond axon guidance Semaphorins and plexins are highly expressed in several tissues outside the nervous system where they play very important roles in normal and pathological situations, including cardiovascular development, immune system formation, and tumor genesis. The 1.5 RNAi in chicken spinal cord 31 transmembrane Sema4D was shown to modulate the functions of T and B-lymphocytes (Hall et al., 1996; Delaire et al., 1998), and two non-neurotropic viruses encode semaphorin like molecules might interfere with the immune system of the host contributing to the immune evasion, protecting the virus from being destroyed. Interestingly, Sema3C and 3E are also over-expressed in invasive and metastasizing tumor cells, possibly mediating cell dissociation and/or protection from apoptosis (Yamada et al., 1997; Christensen et al., 1998). In contrast to this, Sema3A was reported to induce apoptosis of selected sensory and sympathetic neurons (Gagliardini and Fankhauser, 1999; Shirvan et al., 1999). Another function for Sema3A seems the inhibition of endothelial cell motility through competition between Sema3A and vascular endothelial growth factor (VEGF) for NP-1 binding . Emerging evidence suggests that in pathological situation as diverse as nerve injury and tumor progression, the specific expression of semaphorins is modulated. Semaphorins seem to be a perfect example of a cell-cell communication code, exploited by a variety of cells and in different instances from embryo development to adult pathology . Despite all the advances over the last years, biological functions for the majority of semaphorins remain incomplete representing important challenge for the future. 1.5 RNAi in chicken spinal cord Plexins and semaphorins are broadly expressed in several different tissues outside the nervous system making the investigation of their potential roles in the nervous system in vivo very hard to tackle. So far, the generation of mutant mice for different semaphorin and plexin genes gave little insights on the implication of these genes in axon guidance, cell migration, or synapse formation during nervous development. Nevertheless, the broad and very often complementary expression of semaphorins and their receptors in the central and peripheral nervous system strongly argues for multiple yet undetected functional roles for these molecules. To understand the potential roles of some of these proteins in spinal cord development, the chicken embryo has proven to be a powerful model system (Pekarik et al., 2003). The chicken embryo has served as a classic model system for developmental studies due to its easy access for surgical manipulations and a wealth of data about chicken embryogenesis. Recently, the method of in ovo RNAi has led to a revival of the chicken system. Nevertheless, due to the lack of appropriate genetic knock out possibilities, functional studies in chicken embryos have been limited to antibody perturbation and dominant negative expression assays. With the sequencing of the chicken genome near completion, this approach provides a 32 Part I: Introduction powerful opportunity to examine the function of chicken genes. While in ovo electroporation has been effectively used earlier for ectopic or over expression analyses, the injection of long double stranded RNA (ds RNA) into the spinal central canal followed by electroporation represents an excellent tool to study loss of function phenotypes. Interestingly, this method allows the targeting of specific subpopulations of neural cells within the developing spinal cord depending on embryonic age at the injection time. Ds RNA injected- around stage 14 (E2) targets, in addition to spinal neurons, also crest cell derivatives, whereas injections after stage 16 (E2.5) specifically targets spinal cells, and injections later than stage 18 (E3) fails to target motor neurons (Pekarik et al., 2003). Thus, it seems appealing to use the RNAi technique in chicken embryo to knock down different plexins or semaphorins in crest cells and/or spinal cells with the intention to understand better their function in spinal cord development in vivo. 1.6 Goal of the present thesis: Role of PlexinD1 in nervous system development Despite the numerous studies describing the role and function of PlexinD1 in the vascular system development in vivo, very little is known about the involvement of this protein in axon guidance and/or cell migration in the nervous system formation (Gitler et al., 2004; TorresVazquez et al., 2004). However, several lines of evidence show the expression of PlexinD1 in the central nervous system of mice and rats (van der Zwaag et al., 2002), (unpublished data). Our expression analysis using two chicken EST clones matching the PlexinD1 sequence revealed that PlexinD1 is not only expressed in endothelial cells but also in motor neurons. Expression in chicken motor neurons is highly regulated developmentally, and the highest expression levels correspond well with the timing motoneurons sort in the limb plexus and project into different nerve trunks. PlexinD1 knock down experiments lead to several defects in motor axon guidance and surprisingly also to aberrant growth of sensory neurons at the dorsal root entry zones. Our data strongly suggest that PlexinD1 besides influencing directly motor axon outgrowth also affects directly or indirectly the sensory dorsal root afferents as well as the ventral motor roots. In an attempt to investigate whereas any member of the semaphorin family might correspond to a potential binding partner(s) mediating PD1 signaling in vivo, we performed a detailed spatio-temporal expression study to analyze the distribution of all semaphorins in the 1.6 Goal of the present thesis: Role of PlexinD1 in nervous system development 33 developing chicken spinal cord. Different plexin and semaphorin members display complementary as well as overlapping expression patterns in chicken spinal cord, suggesting the presence of large functional complexes between different compounds that are responsible for mediating their biological activity in vivo. 34 Part II: Paper 1 Functional knock down of PlexinD1 in chicken results in misguidance of motor axons and alterations of dorsal and ventral roots organization 2 Dummyheading 35 Functional knock down of PlexinD1 in chicken results in misguidance of motor axons and alterations of dorsal and ventral roots organization 1 Joelle Gemayel, 2Rejina Sadhu, 1Regis Babey, 2Esther T. Stoeckli, and 1Matthias Gesemann 1 Brain Research Institute, University of Zurich, and department of biology, ETH Zurich, and 2 Institute of Zoology, University of Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland correspondence to: [email protected] phone: +41 44 635 3283 fax: +41 44 635 3303 key words: 36 chicken embryo, PlexinD1, motoneurons, vascular system, crest cells Abstract Plexins belong to a large family of molecules mainly implicated in directing axonal outgrowth, tissue vascularization, and cell migration. PlexinD1 (PD1), a unique member of the plexin super family, has been extensively studied in vascular system development. However, very little is known about its potential function in nervous system formation. In this paper, we show that PlexinD1 mRNA is expressed in chicken spinal motor neurons at a time corresponding to the period of motor axon sorting in the limb plexus. PlexinD1 knock down using in ovo RNAi lead to several defects in motor axons pathfinding of the crural nerve trunk. Interestingly, we also observed abnormalities in axon sorting at the dorsal root entry zones and the ventral motor root exit points, phenotypes that cannot simply be explained by the knock down of PlexinD1 in motor neurons. Our data, combined with earlier reports demonstrating the role of PlexinD1 in intersomitic vessels formation in mice, suggest a novel role for PlexinD1 in the migration of a subpopulation of neural crest cell derivatives and/or a tight relation between intersomitic vessels sprouting and spinal root formation. 2.1 Introduction The formation of precise and accurate sensory and motor connections, are a prerequisite for perceiving and integrating environmental information and for performing complex motor tasks. The fundamental basis for the formation of sensory and motor nerves are neural crest cells that delaminate from the neural crest shortly before or during neural tube closure, and motor neurons that are formed by progenitor cells located in the ventral half of the spinal cord (deLapeyriere and Henderson, 1997; Baker and Bronner-Fraser, 1997; Jessell, 2000). Motor neurons located along the antero-posterior axis of the embryo sort into different motor neuron pools, such as the lateral motor column (LMC), the medial motor column (MMC) or the column of Terni (CT) (Landmesser, 1978b). Within the different subgroups, neurons can be easily identified either based on their position within the developing spinal cord and/or their axonal trajectory in the limb bud. Despite various studies attempting to explain motoneuron pathfinding in the chicken embryo, our understanding of the molecular mechanisms implicated in axonal targeting is still fragmentary. Initial axon outgrowth occurs away from the floor plate towards a region called the motor exit point (Guthrie and Pini, 1995). After exiting the spinal cord, motor axons converge to reach the limb base where they pause for 24 hours in the limb plexus, undergoing intensive sorting prior to dividing into a dorsal and ventral nerve trunks (Varela-Echavarria et al., 1997). In the hindlimb, a lateral 37 38 Part II: Paper 1 population of LMC motoneurons, called LMCL send axons through the dorsal nerve trunk to innervate dorsal muscles, whereas axons of motor neurons located in the medial part of the LMC (LMCM) navigate along a ventral path to innervate the ventrally developing limb muscles (Tosney and Landmesser, 1985b). The crural nerve trunk, which innervates anterior muscles, further subdivides into two major branches innervating the anterior dorsal thigh and the anterior ventral thigh respectively. The final targeting towards the different peripheral muscles is accomplished by additional branching events increasing the complexity of motor axon guidance. For some parts of their path, motor axons intermingle with axons of sensory neurons that have their cell bodies in the dorsal root ganglia (DRG). Sensory neurons are generated from neural crest cells, which can give rise to a wide variety of cells ranging from neurons, to smooth muscle cells, cardiac cells, melanocytes and a specific type of glia located at the dorsal root entry zones (DREZs) and the motor exit points called the boundary cap cells (Knecht and Bronner-Fraser, 2002; Golding and Cohen, 1997). Precursors of sensory and sympathetic neurons populating the DRG and the sympathetic ganglia the migrate between stage 14 (E2) and stage 18 (E3) along a ventrolateral pathway, passing through the rostral half avoiding the caudal half. In contrast to this melanocyte precursors migrate after stage 18 (E3) along a dorsal pathway (Krull, 2001). After completing migration, dorsal root sensory neurons extend axons bi-directionally forming the dorsal roots and together with motor axons the spinal nerve. Dorsal root afferent axons reache the primordium of the dorsal funiculus (DF) at stage 23 (E4) entering the spinal cord via the dorsal root entry zones (DREZs) (Davis et al., 1989). The dorsal roots acquire a highly organized pattern during development in which adjacent dorsal roots are equally spaced exhibiting a stereotypic segmental organization. So far, little is known about molecules and mechanisms that are responsible for the formation of highly stereotypic sensory and motor networks. While some axon guidance molecules (ephrinA4) (Helmbacher et al., 2000) and transcription factors (lhx3 and lim-1) (Sharma et al., 1998; Sharma et al., 2000) have been shown to direct the ventral or dorsal path of motor axons, very little is known about molecules involved in axon sorting at the limb bud plexus. The plant lectin peanut agglutinin (PNA)-binding glycoproteins, ephrins, netrins, Hepatocyte growth factor (HGF), extracellular matrix molecules including F-spondin and tenascin (Davies et al., 1990; Debby-Brafman et al., 1999; Faissner and Kruse, 1990; Flanagan and Vanderhaeghen, 1998) have been shown to exert certain repulsive and attractive influences on motor axons. Interestingly, Semaphorins have been shown to influence pathfinding and cell migration of a broad range of neurons in both the central as well as peripheral nervous system 2.1 Introduction 39 (Shepherd et al., 1996; Puschel et al., 1995; Eickholt et al., 1999). Two receptor families have been shown to bind semaphorins directly or indirectly: plexins and neuropilins (Puschel, 2002). While plexins are transmembrane proteins with an extracellular semaphorin (Sema) domain, a cysteines-rich motif and a conserved intracellular plexin-specific sex-plexin (SP) domain (Tamagnone and Comoglio, 2000), neuropilins have a highly conserved extracellular MAM domain but only a very short cytoplasmic tail (Takahashi et al., 1999). Due to the absence of an intracellular neuropilin domain implicated in signal transduction, it is believed that semaphorin signaling is mainly transmitted through the activation of plexins (Raper, 2000). In vertebrates, nine different plexins have been identified so far, that based on their structural similarities have been grouped into four subfamilies (A-D) (Tamagnone et al., 1999). Plexins have been reported to mediate multiple biological functions including axon guidance and cell migration (Rohm et al., 2000; Winberg et al., 1998b; Cheng et al., 2001; Brown et al., 2001; Gitler et al., 2004). PlexinA3 for example, mediates a semaphorin dependent fasciculating activity in the hippocampal commissural pathway (Cheng et al., 2001), whereas PlexinA2 is implicated in semaphorin dependent cardiac crest cell migration (Brown et al., 2001). Interestingly, plexins and semaphorins are also expressed in many tissues outside the nervous system, especially in endothelial cells and the heart, a finding that goes hand in hand with recent studies demonstrating defects in vascular system development in plexin and semaphorin knockouts. Disruption of the PlexinD1 (PD1) gene in mouse leads to severe cardiovascular defects such as outflow tract septation, pharyngeal arch arteries malformation and intersomitic vessels disorganization (Gitler et al., 2004). Similarly, PD1 knock down in fish leads to angiogenesis defects of intersomitic vessels although heart and main vessels formation are otherwise normal (Torres-Vazquez et al., 2004). While these two initial studies suggested a role for a PD1-neuropilin-sema3A complex in intersomitic vessel formation, a more recent study demonstrates that a neuropilin independent Sema3E-PD1 interaction is required for accurate vascularization of the intersomitic vessels (Gu et al., 2004). Surprisingly, although PlexinD1 message has been shown to be expressed not only in the vascular system but also in several regions of the developing nervous system (van der Zwaag et al., 2002 161), no potential role for this protein in nervous system development has been documented. In this paper, we now show that in the chicken embryo, PlexinD1 is expressed in spinal motor neurons around the time motor axons reach and initiate sorting in the limb plexus. PlexinD1 knock down using in ovo RNAi leads to motor axons misguidance and disorganization of the 40 Part II: Paper 1 dorsal and ventral spinal roots. This suggests that PlexinD1 besides its important role in vascularization is implicated in multiple new functions during nervous system development. 2.2 Material and methods 2.2.1 in ovo RNAi 2.2.1.1 Production of long double stranded RNA (ds RNA) Double stranded RNA for targeting PlexinD1 was produced from the following chicken EST clones ChEST697F11 and ChEST867E24. ESTs were linearized using NotI or EcoRI and. Two µg of linearized plasmid were mixed with a final concentration of 4mM dNTPs (Roche), two µl T3 or T7 RNA polymerase (15U/ µl; Roche), 4 µl of 5X transcription buffer and 0.5 µl RNasin (30 U; Promega) in a total volume of 20 µl. After completion of transcription (37°C for 4 hours), DNAse I (Roche) was added and the RNA was extracted using once acidic phenol-chloroform (25:24:1 vol/vol/vol phenol/chloroform/isoamyl alcohol) and once chloroform/isoamyl alcohol (24:1 vol/vol). Following precipitation with ethanol, the RNA was dissolved in 20 µl PBS. Subsequently, equal amounts of sense and anti-sense RNA were mixed, heated to 95°C for 5 minutes, and double stranded RNAs were allowed to form by slow cooling of the reaction for several hours. 2.2.1.2 Electroporation of long ds RNA Electroporation of double stranded RNA was performed as earlier described by Perrin and Stoeckli (Perrin and Stoeckli, 2000). In summary, 0.1-0.5µl phosphate-buffered saline (PBS), containing either a mixture of, PlexinD1 ds RNA (200-500 ng/µl) and YFP plasmid (under the control of β-actin promoter), or plasmids encoding YFP alone were injected into the central canal of the chicken spinal cord. Before and after electroporation a few drops of sterile PBS were added to cool the embryo. Platinum electrodes (BTX, Genotronics) of 4 mm length with a distance of 4 mm between anode and cathode were used. The number of pulses and the voltage were chosen dependent on the embryo age. After two to three days, embryos were sacrificed and processed for immunofluorescence analysis. 2.2.1.3 Whole mount immunostaining Embryos for whole mount immunostaining were taken between stages 24-26. Internal organs and non-required body parts were removed and the clean embryo was fixed in 4% PFA-PBS for 1 hour. After several washes with PBS, the embryos were permeabilized at ambient temperature for 1 hour with constant but gentle agitation, using PBS containing 1 % Triton. 2.2 Material and methods 41 Following two washes with PBS, 20 mM lysine in 0.1 M sodium phosphate (pH 7.3 – 7.4) was added for 1 hour and was subsequently replaced by PBS containing 10% goat serum (blocking buffer). After two hours, the first antibody mixture containing a mouse monoclonal anti-neurofilament antibody at a dilution of 1:1500 (RMO-270; Zymed laboratories) and a rabbit anti-YFP polyclonal antibody (ab 290; Abcam) in a 1:500 dilution was added and incubated at 4°C for at least 48 hours. Several PBS washes were performed before incubating the embryos in blocking buffer at 4°C overnight. Embryos were than incubated with the secondary antibodies, a Cy3-conjugated goat anti-mouse IgG (Jackson ImmunoResearch) and an Alexa-conjugated goat anti-rabbit IgG. Following an overnight incubation at 4°C, embryos were washed several times with PBS (5 times for at least 1 hour), and subsequently dehydrated in methanol and immersed in BBDA. Specimen, were analyzed using indirect Immunofluorescence on a Zeiss axiovert. 2.2.1.4 Immunofluorescence staining Embryos were dissected at the designed stages in PBS and fixed for 1 hour in 4% PFA-PBS. After 30 minutes washing in PBS, tissues were embedded in OCT and quickly frozen in isopentane on dry ice. Cryostat sections, cut at 25-µm thickness were washed for 3 minutes in PBS at 37°C and subsequently transferred for 30 minutes to 20 mM Lysine. Sections were than washed with PBS and incubated for 1 hour at room temperature in PBS containing 10% goat serum (blocking buffer). The primary antibody mixture (a rabbit anti-YFP polyclonal antibody (ab 290; Abcam) 1:500 mixed with either a mouse monoclonal anti-neurofilament antibody 1:1500 (RMO-270; Zymed laboratories), or a mouse monoclonal anti-HNK-1 antibody (sigma) 1:500 or mouse monoclonal anti-1E8 were added overnight at 4°C. Slides were than washed in three changes of PBS and incubated for 1 hour in blocking buffer. 1:250 diluted fluorochrome-conjugates secondary antibodies (Cy3-conjugated goat anti-mouse IgG (Jackson ImmunoResearch laboratories) and Alexa 488-conjugated goat anti-rabbit IgG (molecular probes) were added subsequently for 2 hour at room temperature. Finally, the sections were washed 5 times in PBS and mounted in Fluoromount before being examined by indirect Immunofluorescence on a Zeiss axiovert. 2.2.2 In situ hybridization 2.2.2.1 cRNA probe labeling The Chicken DNA plasmid derived from the two EST clone (ChEST697F11 and ChEST867E24), found by data base search, were linearized using restriction endonucleases 42 Part II: Paper 1 (NotI or EcoRI; Roche). The linearized PD1 plasmid was DIG labeled by incubating 2 µg of DNA with 2 µl digoxigenin (DIG) labeling mix (Roche), 2 µl of T3 or T7 RNA polymerase (Roche), 2 µl of 10 X transcription buffer (Roche), and H2O added to a final volume of 20 µl, at 37°C for 2 hours. After incubation, 2 units of Rnase free DNaseI (Roche, 10U/µl) was added to the mix, and incubated at 37°C for 30 min, after which 2 µl of 0.2 M EDTA, pH 8.0, was added to stop the nuclease treatment. The cRNA probe was ethanol-precipitated and dissolved in 50 µl of Rnase-free H2O. 2.2.2.2 RNA in Situ Hybridization Chick embryos at the designated stages were dissected in PBS and fixed in 4% PFA-PBS for 1 hour. After 30 minutes washing in PBS, tissues were embedded in OCT and quickly frozen in isopentane on dry ice. Sections of 25-µm thick were cut, collected on Super frost Plus (Fisher Scientific) microscope slides, dried at room temperature and stored at -20°C until use. Alternatively, embryonic tissue from different stages were collected after dissection in PBS and immediately embedded in OCT prior to quick freezing in isopentane on dry ice. Tissue sections were post-fixed half an hour in 4% PFA-DEPC PBS before a single 5 minutes wash in PBS followed by a 5 minutes wash in DEPC water were carried out. Sections were subsequently acetylated for 10 minutes, washed for 5 minutes once in PBS and once in 2X SSC-DEPC and subsequently incubated with the prehybridization buffer containing 40% formamid , 5X SSC-DEPC, 5X denhardts’ solution, 0.5 mg/ml yeast tRNA, 0.5 mg/ml salmon sperm DNA at 54°C for 3 hours. cRNA probes, diluted in prehybridization buffer at final concentration of 3 ng/µl, were added to the slides and incubated over night at 54°C. The next morning, slides were washed as following: 5 minutes in 5X SSC, 5 minutes in 2X SSC, 5 minutes in 0.2X SSC, 20 minutes in 0.2X SSC containing 40% formamid at 54°C followed by one wash for 5 minutes in 2X SSC at room temperature. All the following steps were carried out at ambient temperature. Slides were afterwards washed twice for 10 minutes in detection buffer (0.1 M Tris-base, 15 mM NaCl, pH 7.5) before incubation in blocking buffer (3% milk in detection buffer) to block non specific binding. The anti-DIG phosphatase-conjugated antibody diluted in blocking buffer at 1:2000, was added to slides and left for 1 hour at room temperature prior to washing twice in detection buffer and one wash in alkaline phosphatase buffer (0.1 M Tris-Base, pH 9.5, 0.1 M NaCl and 50 mM MgCl2) for 5 minutes each wash. The bound probe was detected by adding NBT/BCIP substrate (Roche). For each ml of alkaline phosphatase buffer, 4.5 µl NBT and 2.3 Results 43 3.5µl BCIP were added and the mixture was added to tissue sections and developed in the dark over night at 4°C. Images were recorded on a Zeiss Axioskop. 2.3 Results 2.3.1 Expression of PlexinD1 in chicken spinal cord during motor axon sorting in the limb plexus In embryonic mice and zebra fish, PlexinD1 is expressed in endothelial cells where it is implicated in intersomitic vessels formation and sprouting (Torres-Vazquez et al., 2004), (Gitler et al., 2004). However, earlier reports have shown that PlexinD1 is also expressed in the developing nervous system (van der Zwaag et al., 2002), suggesting that it might have additional functions than directing vascularization. Therefore, we analyzed PlexinD1 expression in the chicken embryo between stage 17 (E3) and stage 37 (E12). Similarly, to mice and fish, PlexinD1 mRNA is present in endothelial cells. First, transcript can be detected around stage 17 (Fig. 1A, arrow) in cells of the developing perineural vascular plexus that sprouts and extends around the ventral half of the neural tube. By stage 20 PlexinD1 expression, has spread to the perineural vascular plexus surrounding the spinal cord (Fig. 1B) and DRGs (Fig. 1C, arrows). PD1 expression in this region persists until stage 36 (data not shown). Around stage 24, PD1 transcripts are detected additionally in endothelial cells of many developing and sprouting vessels throughout the entire embryo (Fig. 1E and 1F, arrowheads) and specifically in endothelial cells sprouting from the perineural vascular plexus to vascularize the ventral part of the spinal cord (Fig. 1E and 1F, arrows). Interestingly, between stage 23 and 26, PlexinD1 expression can also be detected in the spinal cord, displaying a very specific and restricted spatio-temporal pattern (Fig. 1C, 1D, and 1E). Expression of PlexinD1 in the spinal cord is limited to neurons of the lateral motor column (LMC) that send axons towards different areas of the developing limb bud. The onset and offset of PD1 expression in motor neurons correlates very well with the period of motor axon sorting in the limb plexus, suggesting that PlexinD1 might be important for nerve segregation. Interestingly PlexinD1 is not expressed equally throughout the entire rostrocaudal axis of the spinal cord, suggesting that PlexinD1 might be involved only in motor axons sorting at the hindlimb plexus. 44 Part II: Paper 1 Fig. 1: Chicken PlexinD1 mRNA is expressed in spinal motor neurons and endothelial cells. In situ hybridizations of transverse section through the developing chicken neural tube (nt) between stage 17 and stage 25 using a Dig labeled PlexinD1 anti-sense probe are shown. At stage 17 (A; arrows) PlexinD1 mRNA is expressed in endothelial cells of the perineural vascular plexus, surrounding the spinal cord and the developing DRGs (B and C; arrows). At stage 24 and stage 25 (E and F), PlexinD1 transcripts are also seen in endothelial cells inside the spinal cord and in sprouting vessels in throughout the entire embryo. Between stage 23 and stage 25 (D, E and F) the time motor axon sort in the limb plexus, PlexinD1 is transiently expressed in motor neurons and this expression persist until stage 26 (data not shown). Abbreviations: nt, neural tube; mc, motor column; drg, dorsal root ganglia. Scale bars are equivalent to 200 µm. 2.3.2 PD1 knock down specifically in chicken embryo spinal cord leads to motor axon pathfinding errors at the hindlimb level To test the functional role of PlexinD1 in motor axon pathfinding, we performed RNA knock down experiments in the chicken embryonic spinal cord using in ovo RNAi. Injection and subsequent electroporation of PD1 long double stranded RNA into the central canal of stage 14 to 18 chicken embryos resulted in specific targeting of cells in one half of the spinal cord, allowing the comparison of lumbosacral nerve patterns at the right side and the left side of the same embryo (Fig. 2). While in control-injected embryos the crural nerve displays a highly reproducible branching pattern, PlexinD1 knock down embryos reveal several axon pathfinding defects in several subdivisions of the crural nerve. At stage 25 (Fig. 2C, arrow) axons of the dorsal crural trunk begin to arrange and organize in a specific and highly structured manner. The nerve starts to defasciculate and forms small branches and each newly formed branch elongates and defasciculates later to give rise to the highly organized structures we observe at stage 26 (Fig. 2D, arrow). At stage 25 this highly structured arrangement of fibers appears strikingly disturbed in one side of PD1 injected embryos (Fig. 2C, arrowhead) 2.3 Results 45 where bundles of axons emerge highly defasciculated and disordered compared the control side of the embryo. At stage 26, we can clearly observe that different bundles of the main nerve fail to correctly defasciculate and branch, in the PD1 dsRNA electroporated side (Fig. 2D, arrowhead). Additionally, in embryos lacking PlexinD1 in the spinal cord, the ventral branch that emerges from the crural nerve and forms the branch innervating the anterior ventral thigh, fails to enter the limb bud in many embryos. Motor axons of the ventral crural nerve trunk at the injected side seem to stall completely in the plexus region or only partially leave the plexus to enter the limb. These defects are also observed when we compare the right limb innervations in PD1 ds RNA injected embryos to control injected embryos (Fig. 2 control/dorsal crural trunk and control/ventral crural trunk, arrows; compared to arrowheads in PlexinD1 knock down embryos). Fig. 2: PD1 loss of function in chicken spinal motor neurons, results in aberrant axonal outgrowth of crural nerve branches. Whole mount anti-neurofilament immunostaining of chicken embryos co-injected with long ds RNA and YFP at lumbosacral levels, are shown. Within the same embryo, PlexinD1 is knocked down at the right half of the spinal cord (arrowheads) whereas the left side is unaffected serving as a control (arrows). The injection side is confirmed by immunostaining using an antibody against YFP (not shown). At stage 25 (A) and stage 26 (B) the ventral branch of the crural trunk at the right half of the embryo (arrowhead) fails to grow into the right limb in contrary to the left side (arrow) where the same nerve navigate through the limb and form branches (A, B). In addition, these embryos exhibit defects at one branch of the dorsal crural nerve trunk at stage 25 (C, arrowhead) and stage 26 (D, arrowhead); the nerve fail to segregate, elongate and branch correctly (arrowheads in C and D) compared to the control sides (arrows in C and D). These defects are also visible when we compare the ventral crural nerve trunk at the left side of PlexinD1 knock down embryo (arrowheads in A and B) to control injected embryos (arrow in control/ventral crural trunk). The same abnormalities are observed comparing the dorsal crural nerve trunk of PlexinD1 knock down embryos (arrowheads in C and D) and control injected embryos (arrow in control/dorsal crural trunk). Scale bars are equivalent to 200 µm. 46 Part II: Paper 1 These data suggest that the lack of PlexinD1 expression in motor neurons is responsible for the complete or partial retention of ventral crural trunk axons at the plexus level and for the erroneous segregation of dorsal crural trunk axons in the limb bud. 2.3.3 PD1 knock down animals show fusion of the dorsal root entry zones The electroporation of PlexinD1 long double stranded RNA co-injected with an YFP expression vector into the central canal of embryos around stage 14 (E2) results in addition to single side targeting of cells within the spinal cord also in targeting of neural crest cells destined to populate both sides of the developing neural tube. However, YFP expression in migrating crest cells appears to be much stronger on the right side of the injected embryo, suggesting that most crest cells are staying on the ipsilateral side. Fig 3: DREZs in PlexinD1 knock down animals loose their stereotypic segmental pattern. Whole mount antineurofilament immunostainings of dorsal roots at the lumbosacral levels of chicken embryos co-injected with long ds RNA and YFP plasmid are shown. At stage 24 (B, arrowheads), stage 25 (D, arrowheads) and stage 26 (F, arrowheads) PlexinD1 knock down animals exhibit imperfections in the organization of the dorsal roots. The normal evenly spacing between adjacent dorsal root entry zones in control animals at stage 24 (A), stage 25 (C) and stage 26 (E) is disturbed in treated animals. Scale bars are equivalent to 200 µm. Unexpectedly, PlexinD1 knock down embryos exhibit defects at the dorsal root entry zones. The segmental patterning of the dorsal roots and the equal spacing between neighboring dorsal root entry zones, usually observed in control animals between stages 24 and 26 (Fig. 3A, 3C and 3E), is no longer visible in injected embryos. Several dorsal roots appear completely or partially fused (Fig. 3B, arrowhead), (Fig. 3D, arrowhead) and (Fig. 3E, 2.3 Results 47 arrowhead). Interestingly, the aberrant DREZ formation in PD1 knock down embryos was often detected on both side of the spinal cord, suggesting the involvement of migratory cells, traveling bilaterally through the embryo. This phenotype strongly suggests a functional role for PlexinD1 not only in motor axon guidance but also in a subpopulation of crest cells derivatives, possibly Schwann cells, and/or boundary cap cells. 2.3.4 PD1 knock down animals exhibit defects at the motor exit points of the ventral roots Encouraged by the circumstantial evidence that PlexinD1 might be expressed by a subpopulation of migratory crest cells, we started to analyze the integrity of the ventral roots at the motor neuron exit points. Immunostaining on cross sections of chicken spinal cord at hindlimb levels, using different markers such as anti-HNK-1, anti-neurofilament and anti-1E8 an antibody that stain Schwann cells and boundary cap cells (Fig. 4), demonstrates clear abnormalities at the ventral root level in injected embryos lacking PD1. When compared to control embryos, the ventral roots in PlexinD1 knock down animals appear wider and less compact as seen by immunostaining using the HNK-1 antibody (Fig. 4A arrow, 4B arrowhead). In knock down animals at the motor exit point we can observe a split appearance of the ventral roots that elongate away from the motor exit points in apparently two separate bundles (Fig. 4B, arrowhead). This is in clear contrast to control embryos where the ventral roots grow and extend as a single branch (Fig. 4A, arrow). Additionally, some fiber bundles seem to navigate apart from the main motor ventral root or exit away from the ventral exit point when visualized by anti-neurofilament immunostaining (Fig. 4C arrow, 4D arrowheads). In contrast to the straight route adopted by motor fibers in control animals, in injected embryos several axon bundles clearly follow an aberrant path after exiting the spinal cord (Fig. 4D, arrowheads). Interestingly, the aberrant axonal trajectory is apparent on both sides of the spinal cord at the ventral root level prior to the fusion between the motor ventral root and the sensory root. Additionally, some motor fibers in PlexinD1 knock down animals exit the spinal cord at erroneous locations away from the usual normal motor exit point. Furthermore, in injected embryos aggregated boundary cap cells stained by the anti-1E8 antibody seem to have an altered morphology where the ventral root appears split and often defasciculated (Fig. 4F, arrowheads). Interestingly these defects are also obvious on both sides (right and left) of the spinal cord again suggesting that an aberrant positioning of neural crest derivatives is responsible of the observed sorting errors at the ventral roots, rather than a knock down of PlexinD1 in motor neurons. 48 Part II: Paper 1 Fig. 4: PlexinD1 knock down in chicken spinal cord resulting in abnormalities at the motor exit points. A comparison of ventral motor root axons between PlexinD1 knock down embryos and control injected embryos is shown by immunostaining performed on cross sections of chicken spinal cord at stage 26 using anti-HNK-1 (A, B), anti-neurofilament (C, D) and anti-1E8 antibodies (E, F). Crest cells expressing HNK-1 at stage 26, appear less compact at the ventral motor bundle in injected animals (arrowhead in B) when compared to control animals (arrow in A). Similarly, motor axons of PlexinD1 knock down animals exhibit less compact bundles after exiting the spinal cord (D, arrowheads) in comparison to controls (C, arrow). Additionally, boundary cap cells stained by the 1E8 antibody appear disorganized in animals lacking PlexinD1 and axon bundles seem defasciculated (F, arrowheads) on both sides of the spinal cord. Scale bars are equivalent to 200 µm. 2.4 Discussion 49 2.4 Discussion 2.4.1 Chicken spinal motor neurons express PD1 mRNA Several studies demonstrate contradictory results concerning the expression of PD1. While studies investigating the role of PD1 in the vascular system, report the exclusive expression of PD1 in endothelial cells (Torres-Vazquez et al., 2004; Gu et al., 2004; Gitler et al., 2004), another laboratory describes PD1 expression, in addition to endothelial cells, in several structures of the mouse central nervous system (van der Zwaag et al., 2002). Neither group reported expression of PD1 in developing spinal motor neurons of embryonic mice. However, our data clearly demonstrate the expression of PlexinD1 mRNA in chicken spinal motor neurons during a time window that corresponds to hindlimb innervation. This raises the question whether the previously reported expression patterns in mice are incomplete or whether motor neuron specific expression is avian specific. The fact that PD1 can indeed be expressed by neurons in the developing mouse or rat brain(Gesemann et al., 2001; van der Zwaag et al., 2002), suggest that a functional role for PD1 in nervous system development is not restricted to the chicken embryo. However, only a careful analysis of PlexinD1 expression in the spinal cord of other species will answer the question whether motor axon guidance in other species is at least in part PlexinD1 dependent. 2.4.2 PD1 involvement in the navigation of the crural nerve toward the limb Motor axons grow in a direct manner and commit few mistakes during their pathfinding journey. Once they exit the spinal cord, motor axons reach the base of the limb where they sort in the limb plexus. Axon arrangement is highly specific and determines subsequently the diverse trajectory choices that nerves innervating distinct ventral or dorsal muscles will take. Already at the plexus region, various axon bundles innervating the same muscle assemble and occupy particular spatial locations (ventral, dorsal medial or lateral) within the proximal nerve trunks (Lance-Jones and Landmesser, 1981; Tosney and Landmesser, 1985c). These reorganization events are decisive not only for ensuing dorsal or ventral trajectory choices distinct nerves make but also for recognition of appropriate choice points where a muscle nerve branches away from the main nerve trunk before continuing the journey to its final target location. Selective adhesion between axons belonging to a single motor neurons pool may contribute to this sorting process. However, axons also group into specific spatial locations within the nerve trunks, therefore probably responding to extrinsic guidance cues located at the base of the limb (Tosney and Landmesser, 1984). 50 Part II: Paper 1 Our in situ data demonstrate that PD1 mRNA is expressed by all motor neurons of the LMC. Although we require whole mount in situ hybridization experiments to confirm the precise spinal levels of PD1 expression, we noticed on cross sections that PlexinD1 mRNA is restricted to the most rostral lumbar and the most caudal thoracic levels while being completely absent from all cervical and sacral motor neurons. Interestingly, PlexinD1 knock down in chicken embryos specifically affect the normal patterning of those branches of the crural nerve which emerge from motor neurons of the LMC at lumbar spinal levels (L1-L3) (Landmesser, 1978a). In PlexinD1 knock down embryos, a part of the axons in the dorsal crural nerve trunk motor axons cross the limb base to enter the limb mass, but assemble in a disordered manner and without respecting the highly organized normal structure, whereas axons in the ventral nerve trunk fail completely or partially to navigate correctly within the limb bud. Thus, the differential expression of PD1 mRNA at different spinal levels and its restricted temporal expression between stage 23 and stage 26 in combination with the motor axon defects noticed in PlexinD1 knock down embryos at specifically two subdivisions of the crural nerve strongly suggest the involvement of PlexinD1 in motor axon guidance of the dorsal and ventral crural trunk. Our in situ data only give information about the presence of PlexinD1 mRNA, and we can only speculate about the subcellular localization of PlexinD1 protein. As observed for other axon guidance receptors, it seems likely that PD1 is present on advancing growth cones and /or elongating axons. The absence of this protein in the dorsal and ventral crural nerve trunk of knock down embryos may reduce the capability of normally PlexinD1 positive-growth cones to react to specific cues in the surrounding tissues and subsequently results in inaccuracy of the axonal outgrowth. Alternatively, PD1 protein could be responsible for axon sorting in the limb plexus. At stage 23, motor axons of the LMCL and LMCM defasciculate at the plexus level and rearrange spatially into the proximal nerve tracts before extending further dorsally or ventrally in the limb muscle mass. The lack of PlexinD1 protein could prevent correct defasciculation between lateral and medial lateral motor column axons, leading to aberrant arrangements of nerve bundles at the plexus level that can subsequently position axons at irregular locations within the limb bud. Thus, the defects observed in the course of the dorsal and the ventral crural trunk of PlexinD1 knock down animals might be a consequence of abnormal axon sorting at the plexus. The growth cones of abnormally positioned axons might not be able to interpret or react to environmental cues present in their unusual location and commit various pathfinding errors. Our in situ data combined with the 2.4 Discussion 51 aberrant projections observed in the dorsal and ventral proximal crural trunks of PlexinD1 deficient embryos support a potential role for PlexinD1 in axon guidance or alternatively in axon sorting (involving fasciculation and defasciculation events) of the crural trunk nerves at the limb plexus and /or proximal nerve trunks. 2.4.3 PD1 knock down affects the dorsal roots entry zones and the motor exit points In addition to the aberrant outgrowth of the crural trunk, PD1 knock down in chicken spinal cord shows unexpected defects at the dorsal root entry zones (DREZs) and the motor exit points. Surprisingly, the abolished segmental pattern noticed at the dorsal root entry zones and the abnormalities observed at the motor ventral roots are perceived on both sides of the spinal cord. PlexinD1 mRNA expression was never detected in dorsal root sensory neurons in chicken embryos, eliminating a possible direct involvement of PlexinD1 in axon guidance of sensory neurons to the dorsal root entry zones. Moreover, defects observed at the motor exit points are unlikely due to PlexinD1 knock down in motor neurons, since changes in axon guidance are observed on both sides of the injected embryo, while electroporation targets only motor neurons located at one side of the spinal cord. Conversely, a Schwann cell marker (1E8) and a neural crest marker (HNK-1) demonstrate a disorganization of boundary cap cells in the region of the motor exit points and a defasciculation of ventral roots. Based on these intriguing data, we can propose two models to explain the defects observed in PlexinD1 knock down animals. The first hypothesis, suggests that the observed alterations in spinal root formation in PlexinD1 knock down animals are based on migration defects of a subpopulation of neural crest cell derivatives, notably Schwann cell precursors and/or boundary cap cell precursors. These crest derivatives might migrate erroneously and would fail to reach their appropriate final locations. The abnormal positioning of Schwann cells and/or the boundary cap cells in PlexinD1 deficient embryos might be responsible for the fusions observed at the dorsal root entry zones as well as the defects seen at the motor exit points. Certainly, additional experiments are required to show an expression of PlexinD1 in crest cell derivatives such as double in situ hybridization for PlexinD1 and a neural crest cell marker (sox 10, slug, HNK, etc). Nevertheless, to date all studies (in mice and zebra fish) related to PD1 expression reported its presence exclusively in endothelial cells and excluded a role for PlexinD1 in neural crest cell migration (Gitler et al., 2004; Torres-Vazquez et al., 2004). While, Gilter and colleges claimed to have eliminate a possible involvement of cardiac neural crest cells for the defects 52 Part II: Paper 1 observed in PD1 mutant animals on E10.5 frontal sections. Our data clearly suggest a potential involvement of crest cells in axon guidance. This discrepancy between our data and earlier published reports can be due to species differences as well as to divergence in the analyzed spinal levels. Our study mainly focuses on lumbosacral level of the spinal cord, whereas the other study describes a more rostral region at the pharyngeal arch level in the mouse embryo. In addition, while we analyzed axonal outgrowth at the DREZ using whole mount immunostaining, the study by Gilter and al exclusively relies on the use of tissue sections, which make the defects at the DREZ quite difficult to assess. Alternatively, another hypothesis to explain our present findings derives from the potential interaction between growing axons and blood vessels. PlexinD1 has been shown to play a crucial role in intersomitic vessels sprouting and formation in vivo. Our neural defects can therefore be seen as secondary effects resulting from PlexinD1 knock down in endothelial cells that might support or guide growing axons. This would be consistent with a recent and very elegant study in xenopus where Levine and colleges demonstrated that intersomitic arteries form in tight relation to spinal nerves during development (Levine et al., 2003). In our case, the electroporation of PlexinD1 long double stranded RNA in chicken spinal cord at stage 14 might as well knock down PlexinD1 in the intersomitic vessels precursors leading to defects in the patterning of these vessels as earlier reported in the zebra fish and mice (Gitler et al., 2004; Torres-Vazquez et al., 2004). Subsequently, outgrowth of axons in the area of the dorsal root is perturbed by abnormally formed intersomitic vessels leading to aberrant arrangement of the DREZs. According to this model, defects in vascular patterning would lead to axon misguidance. Although this model appears interesting and eliminates contradiction between our current study and earlier published work, it certainly requires many additional experiments to determine its validity. To date we do not have evidence about the possibility of targeting nonneural cells with our electroporation technique. In addition, we did not examine the eventual disturbances of the vascular system especially in intersomitic vessels in PD1 knock down animals. Importantly, we also could not observe any YFP positive cells in the peripheryoutside the spinal cord and the DRG- in double injected (ds RNA and YFP vector) or control animals (injected only with YFP vectors). This argues for the specificity in targeting only neural cells. However, it is essential to mention that DNA vector and long double stranded RNA do neither have the same size and nor the same charges. Thus, the possibility of a wider diffusion of long double stranded RNA comparing to the YFP-vector cannot be absolutely excluded. 2.4 Discussion 53 Despite the enormous advances achieved in the field of axon guidance, our knowledge about pathfinding choices taken by different neurons in vivo remains very fragmentary. 2.4.4 PD1 potential binding partner(s) in chicken embryos during development PD1 knock down in spinal cord leads to aberrant motor axon pathfinding as well as defects in the dorsal root entry zones and ventral motor roots formation. Our current study does not give indications about the potential binding partner(s) of PlexinD1 that might be responsible for mediating its biological functions in vivo. Earlier reports, demonstrating a biological function for PD1 in vivo (Gitler et al., 2004; Gu et al., 2004), suggest the involvement of different signaling pathways using different ligands or ligand complexes. In the first report, PlexinD1 was found to co-immunoprecipitates with NP-1 and NP-2 proteins that bind with high affinity to Sema3A and Sema3C respectively (Gitler et al., 2004). In contrast, Gu and colleagues confirm that PlexinD1 and Sema3E display complementary expression patterns in mice as well as similar defects in PlexinD1 and Sema3E knock out animals. Interestingly, Sema3E-PD1 signaling did not require neuropilins (Gu et al., 2004), presumed obligate co-receptor for Semaphorins Class III members (Tamagnone et al., 1999; Raper, 2000). In a parallel work, we investigated the spatio-temporal expression pattern of chicken semaphorins and plexins during development (chapter 3 and 4). Our study shows, that between stage 22 and stage 26, the period chicken spinal motor neurons express PlexinD1 messages, neuropilin-1, neuropilin-2, Sema3A, Sema3C, and Sema3E messages are also detected in chicken motor neurons (chapter 3 and chapter 4). It is tempting to speculate that these semaphorin individually or combined in a complex, might act also in chicken as potential ligands to PlexinD1. In this case, secreted semaphorin in combination with neuropilins or independently of it, can bind PlexinD1 driving defasciculation events at the crural nerve trunk. Surprisingly, at stage 22 and stage 25 PlexinC1 (PC1) is detected in endothelial cells outside the spinal cord, however to a lesser extend in comparison to PlexinD1 expression (chapter 4), suggesting a possible binding between these plexins to mediate signaling in endothelial cells. However, our current data are not sufficient to conclude whether the same binding partners described earlier in mouse, activate also PlexinD1 signaling in chicken embryo. Supplementary experiments, using biochemical assays are essential to elucidate the potential 54 Part II: Paper 1 binding of chicken PlexinD1 to one or more of the different semaphorin class III and neuropilins members. Additionally, functional knock down of sema3A, 3C and 3E and neuropilins genes in the chicken spinal cord, can give insights about the possible interactions in vivo between PlexinD1 and these different semaphorins. Part III: Paper 2 Developmental regulation of semaphorin expression in the chicken spinal cord and peripheral nervous system suggests different functions and interactions 3 Dummyheading 55 Developmental regulation of semaphorins in spinal cord and peripheral nervous system suggests different semaphorin functions and interactions in the chicken embryo 1 Joelle Gemayel, 2Rejina Sadhu, 2Olivier Mauti, 2Esther T. Stoeckli, and 1Matthias Gesemann 1 Brain Research Institute, University of Zurich, and department of biology, ETH Zurich, and 2 Institute of Zoology, University of Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland correspondence to: [email protected] phone: +41 44 635 3283 fax: +41 44 635 3303 [email protected] phone: +41 44 635 4840 fax: +41 44 635 6879 key words: 56 chicken embryo, semaphorin, motoneurons, interneurons, DRG Abstract Semaphorins are axon guidance molecules with mainly repulsive activities that exist in a variety of different subclasses. While functions and expression patterns of members in the secreted subclass III have been described in detail, far less is known about expression and function of molecules in the subclasses IV, V, VI and VII and few data are available about the expression of these molecules in the developing chicken embryo. We have now performed a combined EST and genomic database search to identify chicken semaphorins and to analyze their expression patterns in the developing chicken spinal cord. In contrast to mouse, human and rat, the chicken genome contains a reduced number of semaphorins, as no homologues have been found for the class IV semaphorins Sema4A, Sema4C and Sema4F. In addition, no counterpart could be identified for Sema6C. Interestingly, the chicken genome contains a class IV semaphorin that cannot be assigned to class IV semaphorins found in other species. Our databank search also revealed a yet non-described novel semaphorin Sema3G that can be found in chicken as well as other vertebrate species. Expression analysis of the different chicken semaphorins in the spinal cord revealed that while semaphorins of the subclass III seemed to have conserved expression patterns across species, expression domains for class V and VI semaphorins seemed to be altered compared to mice or rats and that class IV semaphorins are not expressed in the developing chicken spinal cord. Taken together, our results demonstrate that the chicken genome shows a slightly altered composition of semaphorins compared to other species, resulting in a changed expression pattern for several members that might most likely reflect changes in their binding partners and functions. 3.1 Introduction Directed cell migration, accurate axonal navigation, and correct target recognition are crucial for the proper functioning of the nervous system. Migration and axon outgrowth occur along specific and highly stereotypic pathways that are determined by interactions of specific cell surface receptors with environmental cues that confer positional and guidance information to the migrating neuron or advancing axon (Tessier-Lavigne and Goodman, 1996). Therefore, defining mechanisms as well as identifying molecules by which migrating neurons and growing axons select their pathways, maintain a directed growth along them, and later recognize and innervate the appropriate target areas, are fundamental issues in neurobiology. Several different families of axon guidance molecules have been identified within the last decade(Dickson, 2002; Huot, 2004; Zou, 2004). One of the largest families of axon guidance 57 58 Part III: Paper 2 cues, being mainly involved in redirecting axonal outgrowth by selectively collapsing the axon growth cone, is the semaphorin family (Raper, 2000). Today, more than 20 different semaphorins are known which based on their structural similarities can be grouped into seven different subclasses. While Semaphorins falling into the subclasses I and II are exclusively found in insects, semaphorins belonging to the higher subclasses are found in many different vertebrate species ranging from fish to avian to men (Fiore and Puschel, 2003; Kolodkin et al., 1992; Matthes et al., 1995). Initial functional experiments suggested that Semaphorins are mainly repulsive axon guidance molecules. However, it has become increasingly clear that semaphorins serve a far wider range of functions such as chemoattraction, directing cell migration, being involved in immune responses and in the formation of the vascular system (Fujisawa, 2004; Pasterkamp et al., 2003; Tamagnone and Comoglio, 2004). Nevertheless, the most prominent function of semaphorin family members remains their role in conferring guidance information for growing axons. In the spinal cord, a great number of precise connections must be established to enable proper integration of sensory information from and accurate delivery of motor impulses to the periphery. While nociceptive and thermoceptive sensory neurons establish connections with the dorsal-most layer of the spinal cord, proprioceptive neurons form synapses with interneurons and motoneurons. Survival and axonal outgrowth of different functional subtypes of sensory neurons critically depend on the presence of different neurotrophins, with nociceptive and thermoceptive neurons requiring NGF/TrkA signaling (Patel et al., 2000), whereas proprioceptive sensory neurons clearly depend on NT-3/TrkC activation (Patel et al., 2003). The majority of sensory input is relayed to the brain via spinal interneurons. While many different subtypes of interneurons exist in the spinal cord, the best-studied neuronal subtype are commissural interneurons, which are located in the dorsal part of the spinal cord (Yaginuma et al., 1994). This particular subtype of interneurons is characterized by a highly complex but stereotypic axonal trajectory. Initial commissural axon outgrowth occurs towards the ventral midline of the spinal cord, resulting in axons crossing a particular structure called the floor plate. Once on the contralateral side commissural axons perform a sharp orthogonal turn projecting rostrally towards the brain finally making connections with higher brain regions (Kaprielian et al., 2001). However, sensory neurons not only form functional connections with spinal interneurons but in special cases also with motoneurons located in the ventral horn of the spinal cord (Davis et al., 1989). As for sensory neurons and spinal interneurons, motor neurons can acquire a variety of different identities that are reflected in the choice of specific axon pathways and synaptic targets (Schneider and Granato, 2003). 3.1 Introduction 59 While initially axons from neurons of different pools project along a common pathway, their arrival in the peripheral mesoderm is accompanied by multiple sorting events into different motor axon trajectories. The trajectory of all the above mentioned neuronal subtypes seems to be determined by a variety of different guidance cues and cell adhesion molecules, including slits (Wang et al., 1999), different members of the IgCam super family (Perrin et al., 2001) and semaphorins (Masuda et al., 2003). Sema3A was isolated more than a decade ago based on its collapse inducing activity for axons of DRG neurons (Kolodkin et al., 1993; Luo et al., 1993). Subsequently, other experiments demonstrated that Sema3A selectively repels NGF responsive axons that normally terminate in the dorsal half of the spinal cord (Messersmith et al., 1995). Additionally, a zebrafish homologue of Sema3A has been shown to inhibit growing motor axons (Roos et al., 1999). However, the analyses of Sema3A function in vivo using gene disruption in mice gave rather surprising results in respect to nervous system development. Mutant mice exhibit no or only modest defects in sensory, commissural or motor axon projections, suggesting that either the functional significance for Sema3A in axon guidance has been over estimated or that due to functional redundancy, other class III semaphorins can compensate for the loss of Sema3A (Taniguchi et al., 1997). So far, only few or no additional semaphorins have been shown to be involved in guiding sensory, commissural or motor axons (Kikuchi et al., 1999). Zebrafish Sema4E has been demonstrated to act as a repulsive boundary, guiding brachiomotor axons to their targets, and Sema6A acts as a repellent on sensory NT-3 and NGF sensitive DRG neurons, whereas Sema4D seems to stimulate axonal outgrowth of sensory neurons in a autocrine way, (Masuda et al., 2004; Xiao et al., 2003; Zhou et al., 1997). While the number of semaphorins with demonstrated functional roles in sensory, motor and interneuron guidance is limited, several other family members are also expressed in and around the spinal cord about which no functional data is available (Bagnard et al., 1998; Giger et al., 2000; Takahashi et al., 1998). Chicken Sema3D as well as mouse Sema3C are highly expressed in developing motor neurons whereas high levels of Sema3B message are found in sensory neurons, (Bagnard et al., 1998; Takahashi et al., 1998). Several of these transcripts are also expressed in adjacent mesodermal tissue, suggesting that they might also influence outgrowth of sensory and motor axons. Over the last couple of years, several Semaphorin-binding proteins have been identified. While neuropilins serve mainly as co-receptors transmitting guidance responses of Class III semaphorins in a PlexinA/neuropilin/semaphorin complex, different members of the plexin family have been shown to convey semaphorin signaling in a neuropilin independent manner 60 Part III: Paper 2 (Tamagnone et al., 1999). As for semaphorins, neuropilins and plexins are widely expressed throughout the entire nervous system, suggesting functional roles in a variety of guidance events. Despite various studies describing a wide range of interactions between semaphorins, neuropilins, and plexins, our understanding of how these interactions are implicated in axon targeting is still fragmentary. Moreover, for some semaphorins and plexins, binding partners have not yet been described, suggesting that a certain number of interactions still await its discovery. Finally, it seems clear that several plexins have to form interactions with multiple semaphorins, as already described for different class A plexins and PlexinB1, and that some semaphorins might have additional binding partners besides the described ones. As a first step towards a more complete understanding of semaphorin functions, we performed a detailed spatiotemporal expression analysis in developing spinal cord and surrounding tissue. Surprisingly, the chicken genome only encodes seventeen different semaphorins, whereas mammalian genomes code for twenty different semaphorin genes. Interestingly the number of class III semaphorin genes is identical in avian and mammals, however conservation within the subclass IV is very low, and the avian genome lacks a Sema6C homologue. The absence of several semaphorins in chicken is reflected by an altered expression pattern of chicken semaphorins in the different subclasses, suggesting that these semaphorins have slightly altered functions and binding properties in higher vertebrates. 3.2 Material and Methods 3.2.1 Assembly of chicken semaphorin cDNAs cDNA sequences for chicken semaphorins were assembled using the combined information from the chicken EST (http://www.chick.umist.ac.uk) and the chicken genomic database (http://www.ensembl.org/Multi/blastview?species=Gallus_gallus). 17 different genomic regions coding for semaphorins were identified using the tblastx alignment algorithms on available vertebrate semaphorins or existing chicken semaphorins. The corresponding genomic fragments were downloaded and analyzed using the genescan gene prediction program (http://genes.mit.edu/GENSCAN.html). The obtained putative cDNA and protein sequences were compared to the corresponding mammalian homologues and genescan prediction errors were corrected by manual inspection of the intron/exon boundaries in false predicted regions. Gaps in the assembled sequences due to inaccurate or incomplete genome sequencing were wherever possible filled by corresponding EST sequences. A total of 162 chicken ESTs encoding different semaphorins were identified. Among these 55 covered only 3.2 Material and Methods 61 parts of the 3’UTR sequence whereas the rest contained part of the coding sequence. Sequence alignment of genomic and EST sequences was done using the SeqMan software (Lasergene, DNASTAR, Madison WI). 3’UTR sequences were added to the coding sequence based on overlapping EST sequences that were supplemented with genomic sequences. UTR sequences were terminated at the first polyadenylation AATAAA/ATTAAA sequence that followed verified chicken UTR EST sequences. Using this combined approach 12 complete and 5 partial cDNA sequences for semaphorins could be assembled. 3.2.2 Phylogenetic tree analysis The domain structure of representative members of the chicken and mouse semaphorin super family were obtained using the smart program (http://smart.embl-heidelberg.de). Individual domains were extracted from the sequence using domain boundaries as predicted. Conserved domains from the different semaphorin subfamilies were aligned using the CLUSTAL W alignment algorithm (Higgens and Sharp, 1989; Thompson et al. 1994) provides by the MagAlign software (Lasergene, DNASTAR, Madison WI). Obvious mistakes in domain boundary prediction were manually adjusted. For better representation, alignment files were exported into TREEVIEW software, enabling the graphical representation of the unrooted tree (Page, 1996). 3.2.3 Cloning of semaphorin cDNA fragments cDNA fragments of semaphorin genes with no hits in the EST database were cloned using RT-PCR. A putative cDNA assembled based on genomic information was used as a template to design sense and antisense primers. Total RNA was prepared using spinal cord and DRG tissue isolated from stage 30 chicken embryos. Random and oligodT primed first strand cDNAs were generated using the SuperscriptII reverse transcription kit (Invirtogen, Carlsbad CA) according to the manufacturers’ instruction. A 669 bp long cDNA fragment for Sema3G was amplified by PCR using the following sense 5’CTGTCAAGCGCCAAAAGC and antisense 5’GGCACTGCTCCTCCACC primers. In analogy to this, a 656 bp long fragment for Sema6B was amplified using the following two primers 5’ ATCCAGCGCATCCTCAAG (sense) and 5’ CCCATGTCGTTCTTGCAC (antisense). Obtained PCR fragments were cloned into the Topo TA cloning vector (Invirtogen, Carlsbad CA) and subsequently sequenced to verify the identity of the insert. 62 Part III: Paper 2 3.2.4 In situ hybridization 3.2.4.1 cRNA probe labeling Chicken DNA plasmids derived from the EST clones ChEST399E25, ChEST771A21, ChEST578C18, ChEST375M12, ChEST7787C4, ChEST659J12, ChEST986M15, ChEST585F20, ChEST225N10 and ChEST1004D4 corresponding to Sema3A, Sema3B, Sema3C, Sema3D, Sema3E, Sema3F, Sema5A, Sema5B, Sema6D and Sema7A, found by database searches, were linearized using the restriction endonucleases NotI or EcoRI. Linearized plasmids were DIG labeled by incubating 2 µg of each DNA with 2 µl digoxigenin (DIG) labeling mix (Roche), 2 µl of T3 or T7 RNA polymerase (Roche), 2 µl of 10 X transcription buffer (Roche), and H2O added to a final volume of 20 µl for each reaction, at 37°C for 2 hours. After incubation, 2 units of Rnase free DNaseI (Roche, 10U/µl) was added to the mix, and incubated at 37°C for 30 min, after which 2 µl of 0.2 M EDTA, pH 8.0, was added to stop the nuclease treatment. The cRNA probe was ethanol-precipitated and dissolved in 50 µl of Rnase-free H2O. 3.2.4.2 RNA in Situ Hybridization Chick embryos at the designated stages were dissected in PBS and fixed in 4% PFA-PBS for 1 hour. After 30 minutes washing in PBS, tissues were embedded in OCT and quickly frozen in isopentane on dry ice. Sections of 25 µm thick were cut, collected on Super frost Plus (Fisher Scientific) microscope slides, dried at room temperature and stored at -20°C until use. Alternatively, embryonic tissue from different stages were collected after dissection in PBS and immediately embedded in OCT prior to quick freezing in isopentane on dry ice. Tissue sections were post-fixed half an hour in 4% PFA-DEPC PBS before a single 5 minutes wash in PBS followed by a 5 minutes wash in DEPC water were carried out. Sections were subsequently acetylated for 10 minutes, washed for 5 minutes once in PBS and once in 2X SSC-DEPC and subsequently incubated with the prehybridization buffer containing 40% formamid , 5X SSC- DEPC, 5X denhardts’ solution, 0.5 mg/ml yeast tRNA, 0.5 mg/ml salmon sperm DNA at 54°C for 3 hours. cRNA probes, diluted in prehybridization buffer at final concentration of 3 ng/µl, were added to the slides and incubated over night at 54°C. The next morning, slides were washed as following: 5 minutes in 5X SSC, 5 minutes in 2X SSC, 5 minutes in 0.2X SSC, 20 minutes in 0.2X SSC containing 40% formamid at 54°C followed by one wash for 5 minutes in 2X SSC at room temperature. All the following steps were carried out at ambient temperature. Slides were afterwards washed twice for 10 minutes in detection buffer (0.1 M Tris-base, 15 mM 3.3 Results 63 NaCl, pH 7.5) before incubation in blocking buffer (3% milk in detection buffer) to block non specific binding. The anti-DIG phosphatase-conjugated antibody diluted in blocking buffer at 1:2000, was added to slides and left for 1 hour at room temperature prior to washing twice in detection buffer and one wash in alkaline phosphatase buffer (0.1 M Tris-Base, pH 9.5, 0.1 M NaCl and 50 mM MgCl2) for 5 minutes each wash. The bound probe was detected by adding NBT/BCIP substrate (Roche). For each ml of alkaline phosphatase buffer, 4.5 µl NBT and 3.5µl BCIP were added and the mixture was added to tissue sections and developed in the dark over night at 4°C. Images were recorded on a Zeiss Axioskop. 3.3 Results 3.3.1 Identification of Sema3G, a novel member of the class III semaphorins Due to its easy accessibility and its high regeneration capability, the chicken embryo has been the model organism of choice to study a variety of developmental phenomena, ranging from neuronal differentiation to axon guidance and synaptogenesis. However, the lack of appropriate tools to knockout or knockdown gene function has strongly limited the use of chicken embryos in recent years. Lately, there has been a revival in the use of the chicken systems, mainly due to three different reasons. While the establishment of an extraordinarily large EST database and the near completion of the chicken genome sequencing have provided the molecular bases for obtaining a great variety of chicken gene sequences, the combination of in ovo electroporation with functional RNAi knockdown have overcome the limitations for chicken research (Bourikas and Stoeckli, 2003; Krull, 2004; Pekarik et al., 2003). Despite these advances, the use of chicken embryos as a model system for the study of molecular mechanisms of axon guidance is still limited compared to other experimental systems such as mouse and rat. This is represented also by the fact that even if the first vertebrate semaphorin to be identified has been chicken Sema3A (Luo et al., 1993), information about the existence and the expression of different chicken semaphorins remains fragmentary. Until now, only five different avian semaphorins have been characterized (Luo et al., 1995), preventing a careful analysis of chicken semaphorin expression and function. To narrow this gap we have now performed extensive databank searches using the combined information from the EST and the genomic database to predict the number of chicken semaphorin genes. Searches for different class III semaphorins resulted in the identification of all previously described mammalian semaphorins (Fig. 1). Both databases provided a very complete set of information as seen by the fact that neither the genomic nor the EST database failed to identify the 64 Part III: Paper 2 different class III semaphorins (Table 1). Indeed, the genomic database reveals the presence of an additional class III semaphorin member that we called Sema3G. Subsequent databank searches in mouse, rat, and human confirmed the existence of this novel semaphorin member. Phylogenetic tree analysis of the conserved semaphorin domain revealed that the novel Sema3G is most closely related to Sema3E, however the conservation of the sema domain of chicken Sema3G and mouse Sema3G is the lowest between all class III semaphorins (Fig. 1). Sema3A Sema3B Sema3C Sema3D Sema3E Sema3F Sema3G ESTs 8 2 10 14 1 6 0 Chicken Ch 1 Un 1 1 1 Un 12 5a2 9f2 5a2 5a2 5a2 9f1 14a3 Mouse Ch Sema4A Sema4B Sema4C Sema4D Sema4F Sema4G Sema4X ESTs n.d 0 n.d 12 n.d 5 7 Chicken Ch n.d 28 n.d Un n.d 6 10 Mouse Ch 3f1 7d1 1b 13a5 6c3 19c3 n.d Sema5A Sema5B Sema6A Sema6B Sema6C Sema6D Sema7A ESTs 22 15 14 0 n.d 29 10 Chicken Ch 2 7 z 28 n.d 10 10 15b2 16b1 18c 17c 3f2.1 2e5 9b Mouse Ch Table 1: Chromosomal localization and EST representation of different semaphorin gene products. The number of identified chicken ESTs for each gene is indicated and chromosomal location of the chicken and its corresponding mouse gene are given. Not detected genes are abbreviated by n.d, whereas identified chromosomal sequences on not yet localized chromosomes are indicated by the letters Un. 3.3.2 The chicken genome has a reduced number of semaphorin genes In contrast to the class III semaphorins, the number of class IV semaphorins in chicken was significantly reduced (Fig. 1; Table 1). While in rat, mouse and human we could identify six different class IV semaphorins, only four such family members could be identified in chicken. Moreover, these semaphorins were neither as highly conserved as the class III semaphorins, 3.3 Results 65 nor could we align the identified members to a mammalian counterpart. We were unable to obtain a corresponding sequence for chicken Sema4A, Sema4C, and Sema4F; however, we identified an additional class IV semaphorin that shows the highest homology to mouse Sema4D and Sema4A, without being a clear homologue. Fig. 1: Phylogenetic tree of the members of the semaphorin super family. Alignment is done on the bases of the semaphorin domain of each family member using the CLUSTAL W program. The scale bar represents the substitution rate of 10 amino acids per 100 amino acid residues. Please note that only mouse (m) and chicken (c) semaphorin sequences are depicted, but that alignment of rat or human sequences gave similar results (data not shown). 66 Part III: Paper 2 In contrast to semaphorin members found in the subclass IV, no changes in gene number have been observed for members of the subclasses V and VII. Nevertheless, as already observed for class IV semaphorins the number of class VI semaphorin genes in chicken is lower compared to mammals. While Sema6A, 6B, and 6D are present and highly conserved in the chicken genome, Sema6C gene is absent. 3.3.3 Class III semaphorins are highly expressed in developing motoneurons Seven secreted Class III semaphorins have been identified in the chicken genome. At stage 18, the earliest stage we have analyzed, no expression of sema3A, 3D, 3E, and 3F is detectable in the embryonic chicken spinal cord (Fig. 2). Sema3B and 3C are however expressed in motoneurons at this early developmental stage (Fig. 2, 3B and 3F white arrows). By stage 22, motoneurons also express sema3A (Fig. 3, 3A white arrow) and 3D (Fig. 3, 3D black arrow) and at stage 30 Sema3E transcripts are up regulated in these cells (Fig. 4, 3E white arrow). However, Sema3F transcripts are never expressed in motoneurons at all analyzed stages (Fig. 2, 3, 4, 5, and 6). In addition to motoneurons, Sema3A is expressed also in cells of the ventral ventricular zone (Fig. 3, 3A white arrowhead) and transiently in interneurons of the intermediate area of the spinal cord (Fig. 3, 3A star) and (Fig. 4, 3A). In addition, by stage 30 ventral interneurons also express Sema3D (Fig. 5, 3D white arrowhead) and Sema3F transcripts (Fig. 4, 3F white arrowhead) are expressed in a small population of cells adjacent to the roof plate at stage 26. Interestingly, at this stage another semaphorin Class III member, Sema3D is substantially expressed in sensory neurons of the dorsal root ganglia (Fig. 4, 3D white arrowhead) and this expression remains high until stage 35 (Fig. 6, 3D white arrowhead). 3.3.4 Floor plate cells expresses high levels of semaphorin V transcripts Until now, little is known about functions, and expression patterns of chicken Class V semaphorins. We have looked at the distribution of Sema5A and Sema5B during development of the spinal cord. At stage 18, shortly before commissural neurons located in the dorsolateral spinal cord start to extend their axons toward the ventral midline, both semaphorin Class V members are expressed in the floor plate (Fig. 2, 5A and 5B black arrows). Strong expression of both Sema5A and Sema5B in the floor plate persists through stage 24 (data not shown). While the level of Sema5A transcripts starts to decrease after stage 24, Sema5B RNA is still expressed very strongly in the floor plate at stages 26 (Fig. 4, 5B) and are still detectable at stage 30 (Fig. 5, 5B). In contrast to Sema5A, whose expression is restricted to the floor plate, 3.3 Results 67 Sema5B is also transiently expressed in neuroblasts of the dorsal spinal cord (Fig. 2, 5B white star) and commissural interneurons. By stage 26, when commissural axons have crossed the floor plate and are just about to exit and to turn into the longitudinal axis Sema5B mRNA is strongly down regulated in commissural neurons (Fig. 4, 5B). Motoneurons become post mitotic between stages 18 and 26 (Krull and Koblar, 2000). While Sema5A expression can be detected as early as stage 20 in motor neurons (data not shown) and remains later in specific subpopulations (Fig. 4, 5A), Sema5B mRNA is never detected in this neuronal population. Fig. 2: Expression patterns of chicken semaphorins in spinal cord and peripheral nervous system at stage 18. Cross sections of lumbosacral chicken spinal cord were incubated with Dig labeled RNA antisense probes. Semaphorin probes used are indicated. While Sema3B (3B, white arrow) and Sema3C (3C, white arrow) transcripts are detected in the motor neurons, the floor plate expresses Sema5A (5A, black arrow), Sema5B (5B, black arrow) and Sema7A (7A, black arrow). Additionally, Sema5B is expressed in dorsal interneurons (white star). Scale bar correspond to 200 µm. 68 Part III: Paper 2 Fig. 3: Expression patterns of chicken semaphorin members in spinal cord and peripheral nervous system at stage 22. Cross sections of lumbosacral chicken spinal cord were incubated with Dig labeled RNA antisense probes. Semaphorin probes used are indicated. By stage 22 Sema3A expression appears in motoneurons (3A, white arrow) and in cells of the ventral ventricular zones (3A, white arrowhead) as well as in interneurons of the intermediate ventricular zone (3A, star). While Sema3D (3D, black arrow) and Sema3C (3C, white arrow) expression persists in motoneurons, Sema5A mRNA is now up regulated in motoneurons (5A, white arrow), Sema5B is strongly expressed in ventricular zone (5B, white arrow) and Sema7A is very weakly expressed in post mitotic motoneurons (7A, white arrow). Scale bar correspond to 200 µm. Moreover, sensory neurons in the dorsal root ganglia appear to express Sema5A transcript during the time when they form collaterals to target their specific layers in the gray matter (Fig. 5, 5A black arrow). While this expression is subpopulation specific, Sema5B transcript was never detected in any sensory neurons (Fig. 2, 3, 4, 5, 6). 3.3.5 Class VI semaphorins are highly expressed in boundary cap cells In contrast to mammals, which contain four different class VI semaphorin genes, the chicken genome encodes only three of them (Fig. 1). While Sema6A has a striking expression pattern, 3.3 Results 69 being detectable only in boundary cap cells (Fig. 4, 6A black arrows and Fig. 5, 6A), Sema6B is diffusely expressed in motoneurons starting at stage 22 (data not shown). Additionally Sema6D transcripts are also detected in motoneurons as early as stage 22 (data not shown). Fig. 4: Expression patterns of chicken semaphorins in spinal cord and peripheral nervous system at stage 26. Cross sections of lumbosacral chicken spinal cord were incubated with Dig labeled RNA antisense probes. Semaphorin probes used are indicated. Expression levels of Sema3A (3A white arrow) and Sema3D (3D, white arrow) transcripts remain unaltered between stage 22 and stage 26 in motoneurons. In contrast Sema3C (3C, white arrow) expression has been strongly down regulated in motoneurons and Sema3D RNA levels are up regulated in DRGs (3D, white arrowhead). Moreover, Sema3F transcripts start to be expressed in a population of cells adjacent to the roof plate, most likely representing associational interneurons (3F, white arrowheads). Sema5A and Sema5B expression persist in the floor plate (5A and 5B arrowheads), in addition Sema5A is expressed in motoneurons (5A, white arrowhead) and Sema5B remains highly expressed in the ventricular zone (5B white arrowhead). In addition strong expression of Sema6A and Sema6D is now seen in boundary cap cells (6A, black arrowheads). High levels of Sema6D are also detected throughout the gray matter of the spinal cord (6D). Finally, Sema7A expression is strongly up regulated in motoneurons and in the dorsal lateral spinal cord (7A, white arrowhead). Scale bar correspond to 200 µm. 70 Part III: Paper 2 Sema6D expression broadens as transcripts can be detected during later stages of development (Fig. 4, 6D) in the dorsal spinal cord, the dorsal root ganglia, and interestingly also in the boundary cap cells (Fig. 4, 6D, black arrow) and (Fig. 5, 6D). Subsequently, Sema6D expression extends throughout the gray matter, with stronger signals in lamina І of the dorsal horn (Fig. 6, 6D white arrowhead). Fig. 5: Expression patterns of chicken semaphorin members in spinal cord and peripheral nervous system at stage 30. Cross sections of lumbosacral chicken spinal cord were incubated with Dig labeled RNA antisense probes. Semaphorin probes used are indicated. Little changes in expression patterns of different semaphorin transcripts are detected between stage 26 and stage 30. The most obvious change can be seen for Sema3D (3D, white arrowhead) and Sema5A (5A, white arrow) where transcripts are also detected now in ventrally located interneurons. Note that Sema3E transcripts are transiently located in a subpopulation of motoneurons (3E, white arrow) and Sema5A is expressed in a subpopulation of sensory neurons (5A, black arrow). Scale bar correspond to 200 µm. 3.3 Results 71 Fig. 6: Expression patterns of chicken semaphorins in spinal cord and peripheral nervous system at stage 35. Cross sections of lumbosacral chicken spinal cord were incubated with Dig labeled RNA antisense probes. Semaphorin probes used are indicated. By this stage of spinal cord development most semaphorins transcript are strongly down regulated, with the exception of Sema3D whose expression is maintained in motoneurons as well as sensory neurons (3D, white arrowhead) and Sema6D, which is now detected in lamina I of the dorsal horn (3D, white arrowhead). Interestingly, a subpopulation of motoneurons still express Sema3C and low levels of Sema3B are still detectable in the gray matter and in DRGs. Scale bar correspond to 200 µm. 3.3.6 Semaphorin7A is expressed in endothelial cells and motoneurons Only one member, Sema7A is found in semaphorin Class VII. Sema7A expression in the nervous system is highly regulated and starts at stage 18 (Fig. 2, 7A black arrow) with a very weak signal in the floor plate that is maintained through stage 26 (Fig. 4, 7A), but is no longer visible at stage 30. Post mitotic motoneurons in the ventral spinal cord express Sema7A weakly at stage 22 (Fig. 3, 7A white arrow), later the expression levels are up regulated very 72 Part III: Paper 2 strongly (Fig. 4, 7A) and persists until stage 30 (Fig. 5, 7A). Sensory neurons in DRGs express Sema7A only during the time when they form collaterals (data not shown). Interestingly, a transient expression of Sema7A is also seen in the dorsolateral spinal cord at stage 26 (Fig. 4, 7A white arrowhead) corresponding to the location of associational neurons. 3.4 Discussion 3.4.1 The chicken genome has fewer semaphorin genes that the mammalian genome While the mammalian genome encodes for at least 20 different semaphorin proteins, our databank searches only identified 17 chicken counterparts. One of the major questions in this respect is whether we were simply unable to identify corresponding chicken fragments or whether the chicken genome indeed contains a reduced number of semaphorin genes. The fact that we did not obtain EST sequences for several chicken semaphorins while identifying a genomic sequence for it, demonstrates that the EST database coverage is not yet complete, making it possible that the failure of identifying chicken counterparts for Sema4A, Sema4C, Sema4F and Sema6C could be due to the incompleteness of the available databases. While this is certainly a possibility, it seems rather unlikely, since we were able to identify a genomic sequence for each of the chicken semaphorins that showed a successful EST hit. Several other circumstances suggest that we indeed identified all chicken semaphorins. Despite the fact that the haploid chicken genome contains 38 autosomes plus the Z and W sex chromosomes, compared to only 19 autosomes and the X and Y sex chromosomes in mouse, the chicken genome is about 50% (1.2 x 109 bp) reduced in size when compared to the mouse genome (2.56 x 109 bp). This reduction in size is mainly based on the fact that 30 of the 38 chicken autosomes, called microchromosomes, are extremely small, ranging only between 5 and 20 Mb (Venkatesh et al., 2000; Vinogradov, 1999). While a reduction in size is not per se an indication for a reduced number of genes, as demonstrated for many teleost, two observations we made suggest that in chicken the reduced size goes hand in hand with a partial loss of genes. First, the overall gene length between corresponding chicken and mouse genes is not significantly reduced (Vinogradov, 1999), second when we analyzed the conservation of chicken genes belonging to either the cadherin or the IgCam super family, we observed a similar reduction in genes (data not shown). Attempts to clone chicken homologues that could not be identified by database screening using degenerate primers in RT-PCR reactions never did yield to a corresponding gene product. Finally, yet importantly, 3.4 Discussion 73 two of the four mouse genes absent in the chicken genome are located in the same chromosomal area, suggesting that a homologous part for this chromosome is not present in the chicken genome. Nevertheless only the completion of the chicken genome sequencing project as well as the extension of the number of sequenced chicken EST can finally answer the question whether the chicken genome indeed encodes fewer class IV semaphorins than the mammalian genome does. 3.4.2 Expression patterns of semaphorins are dynamically regulated during spinal cord development 3.4.2.1 Motoneurons express high transcript levels of different class III semaphorins Sema3A was isolated more than a decade ago based on its collapse inducing activity for sensory axons o (Kolodkin et al., 1993; Luo et al., 1993). Initial experiments demonstrated that Sema3A selectively repels NGF responsive axons that normally terminate in the dorsal half of the spinal cord (Messersmith et al., 1995). While many semaphorin family members have been isolated, most studies carried out to learn more about functional roles of semaphorins in vivo and in vitro still concentrate on secreted members belonging to the subclass III (for reviews see , (Fiore and Puschel, 2003). Class III semaphorins mainly signal through the formation of large signaling complexes involving neuropilins and plexinAs (Tamagnone et al., 1999). However there is also evidence that class III semaphorins can act in a neuropilin independent way (Gu et al., 2004). In chicken spinal cord and adjacent sensory systems, we demonstrated the expression of several different class III semaphorins. While almost all family members are expressed in developing motoneurons, Sema3s are never found in the floor plate, although commissural axons extending toward the floor plate express plexinAs. This is, however, consistent with the absence of neuropilins in commissural axons that would be expected as co-receptors to bind secreted semaphorins (Tamagnone et al., 1999). Interestingly, the subpopulation specific expression of Class III semaphorins in motoneurons supports a potential implication of these molecules in guiding motor axons. An interesting hypothesis would be that SemaIIIs are constantly secreted by different motor nerves and that the presence of different semaphorins will be used to drive defasciculation events of selective nerves at particular choice points, when attractive or adhesive forces within the fascicle are weakened. Such a mechanism would be in analogy to results seen in the Drosophila embryo, where the transmembrane Semaphorin Іa is required for the selective defasciculation of specific motoneurons (Yu et al., 1998). Even 74 Part III: Paper 2 if class III semaphorins are supposed to be secreted they have been shown to bind strongly to neuronal surfaces via their highly charged C-terminal end, making it possible that these molecules are indeed highly concentrated on axon fascicles driving selective defasciculation events at various choice points (Bagnard et al., 2000). 3.4.2.2 Expression of class V semaphorins in the floor plate, commissural interneurons and sensory neurons suggest conserved as well as novel functions for these molecules We have shown that class V semaphorins are highly expressed at the ventral midline, suggesting several potential functions for these molecules. While they can simply serve as repulsive or attractive guidance cues, providing a signal for contralaterally and ipsilaterally projecting neurons they may also be important as floor plate intrinsic molecules taking part in determining floor plate morphology. Such a function would mainly be achieved when class V semaphorins would act as cell adhesion molecules rather than as axon guidance cues. However, such a function has never been shown so far, and no semaphorin/semaphorin interactions have been documented. Nevertheless, the facts that binding between semaphorins and their plexin receptors occur most likely through functional interactions between their sema domains (Raper, 2000) and that plexins are capable of forming homophilic complexes, make a speculation of homophilic semaphorin interactions quite plausible. However, only functional assays using either RNA interference or knockout techniques will provide evidence about the functional roles class V semaphorins play in floor plate cells. Interestingly, during initial stages of commissural axon outgrowth Sema5B mRNA is also transiently expressed in this interneuron subpopulation. However, the exact timing of Sema5B transcript regulation in commissural neurons has not been carefully determined. It seems feasible that Sema5B is expressed in commissural interneurons while they extend axons toward the floor plate, but that this expression is down regulated upon contact with floor plate cells. Analysis of Sema5B expression in stage 24 embryos (data not shown) indeed points in this direction as significantly lower expression levels are observed in these embryos when compared to stage 22 animals. Given the fact that growing axons seem to express Sema5B during periods of axonal growth could suggest a novel function for this molecule potentially serving as a receptor mediating reverse signaling. Such a function would also explain the presence of Sema5A in a subpopulation of sensory neurons. While reverse signaling has been already shown for many Ephrin ligands (Murai and Pasquale, 2003), this concept is rather new for semaphorins. Nevertheless, recent reports have confirmed that semaphorins 3.4 Discussion 75 belonging to another subfamily of transmembrane semaphorins have indeed the capacity to serve as receptors (Toyofuku et al., 2004b). However, so far no interactions for Sema5B have been documented. Interestingly a recent report suggests Sema5A might indeed be bifunctional, serving as attractive as well as repulsive axon guidance cue (Kantor et al., 2004). This dual function strongly depends on interactions of heparan and chondroitin sulfate proteoglycans with the conserved thrombospondin repeats of Sema5A (Kantor et al., 2004), but it remains to be determined whether Sema5B is capable of similar interactions. However, while it seems clear that either proteoglycan loaded Sema5A can serve as an attractant or repellant in Trans, it would be interesting to see whether these complexes also function in a Cis interaction, conferring either positive or negative guidance for commissural neurons. 3.4.2.3 Class VI semaphorins as potential gatekeeper between neurons located in the central and peripheral nervous system Very striking expression patterns are especially found for two class VI semaphorin members, namely Sema6A and Sema6D. While Sema6D is expressed in a subpopulation of motoneurons as well as in sensory neurons, exceptionally high mRNA levels for Sema6D and Sema6A are detected in a special subpopulation of migrating neural crest cells called boundary cap cells. Boundary cap cells settle at four different locations along the spinal cord in the area of the dorsal root entry zones and the motor neuron exit points (Golding and Cohen, 1997). The fact that Sema6A and Sema6D transcript up regulation occurs only after neural crest cells have initiated their migration suggests that Sema6A and 6D are not required during early events of boundary cap cell development. However, transcript levels for both molecules are strongly increased by the time these cells aggregate in the area of the ventral and dorsal roots. This expression pattern suggests several possible functions for these class VI semaphorin family members. Thus, it seems reasonable to speculate about a possible role for Sema6A and Sema6D in initial events of boundary cap cells aggregation, by means of conferring either attractive or adhesive interactions. However, another appealing possibility remains to be considered about the potential role class VI semaphorin in boundary cap cells as “gate keepers” keeping neuronal cell bodies confined within specific locations by repulsive interactions. The latter hypothesis is strengthened by the fact that a potential Sema6D receptor, PlexinA1, is highly expressed in developing motor and sensory neurons (Toyofuku et al., 2004b), (see also chapter 3). Conversely, class VI semaphorins might also act as attractants for motor and sensory neurons, specifically guiding these axons towards the entry zones or exit points of the spinal cord. That semaphorins can indeed act as attractant has been 76 Part III: Paper 2 shown in several recent reports (Dent et al., 2004; Wolman et al., 2004). However, only functional assays will provide evidence about the role of class VI semaphorins in boundary cap cells. Part IV: Paper 3 Expression patterns of plexins and neuropilins suggest cooperative and separate functions in spinal cord development 4 Dummyheading 77 Expression patterns of plexins and neuropilins suggest cooperative and separate functions in spinal cord development 1 Olivier Mauti, 2Joelle Gemayel 1Rejina Sadhu, 2Matthias Gesemann, and 1Esther T. Stoeckli 1 Brain Research Institute, University of Zurich, and department of biology, ETH Zurich, and 2 Institute of Zoology, University of Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland correspondence to: [email protected] phone: +41 44 635 4840 fax: +41 44 635 6879 [email protected] phone: +41 44 635 3283 fax: +41 44 635 3303 key words: 78 chicken embryo, dorsal root ganglia, motoneurons, commissural neurons Abstract Semaphorins and their receptors, neuropilins and plexins, play important roles in providing positional and guidance information for growing axons. While class III semaphorins are mainly using a plexin/neuropilin complex as signaling platform, other described interactions suggest that most semaphorins act as guidance cues in a neuropilin independent way. However, only few interactions between plexins and semaphorins have been described so far, suggesting that additional cross talk exists between these molecular groups. In order to analyze potential functions and interactions of plexins and neuropilins, we now performed a combined EST and genomic database search to identify chicken plexins and analyze their expression patterns in the developing chicken spinal cord. As already observed for semaphorin genes, the chicken genome contains a reduced number of plexins when compared to mouse, human and rat. Only seven plexins were found in chicken, whereas nine plexins were identified in mammals. While both plexins in the single protein subfamilies C and D are conserved between species, the A subfamily and B subfamily each lack one family member as neither PlexinA3 nor PlexinB3 could be identified in the chicken genome. Expression analysis of chicken plexins and neuropilins during spinal cord development revealed a strikingly different expression for plexins and neuropilins. While PlexinAs are widely expressed throughout the entire spinal cord, PlexinBs seemed to be most abundant in glia and PlexinC1 seems to be expressed only during late stages of neuronal development. Interestingly PlexinD1, which has previously been reported to be exclusively expressed in endothelial cells of the vascular system, seems also be expressed by developing motoneurons. In contrast to this, neuropilin expression domains are far more restricted, suggesting in addition to cooperative also separate function for both molecular groups. 4.1 Introduction Semaphorins constitute a large family of secreted and membrane bound molecules that can function as repellents or attractants regulating fasciculation, axonal growth and branching as well as terminal arborisation (Fujisawa, 2004). Two different classes of receptors, called neuropilins and plexins, mediate these described functions. While plexins constitute a large family of molecules that can be grouped into four major subfamilies, PlexinA, B, C and D, neuropilins form only a small family with two known members (Tamagnone et al., 1999), (Puschel, 2002). Neuropilins have been shown to be involved in mediating various sema III 79 80 Part IV: Paper 3 signals by forming a complex with different plexins. However, class IV and VII semaphorins have been shown to act in a neuropilin independent manner (Tamagnone et al., 1999). Neuropilins are transmembrane proteins acting as co-receptors conferring ligand specificity in a larger signaling complex. Due to the lack of any obvious signaling component in the neuropilin cytoplasmic part, activity is mediated by complex formation with either plexins, L1, or off track tyrosine kinase (Castellani, 2002; Castellani and Rougon, 2002). While neuropilin-1 (NP-1) is expressed in peripheral sensory neurons, autonomic neurons of the sympathetic ganglia and a subpopulation of ventral motoneurons, neuropilin-2 is expressed throughout the entire ventral half of the spinal cord including motoneurons and several subpopulations of interneurons (He and Tessier-Lavigne, 1997; Giger et al., 2000). In sensory neurons, neuropilin-2 (NP-2) seems only to be transiently expressed during initial stages of sensory axon outgrowth, as no expression has been documented in mice older than E13 (Chen et al., 1997; He and Tessier-Lavigne, 1997; Kolodkin et al., 1997). In accordance with the expression of the two neuropilin RNAs, neuropilin-2 mutant mice display a strikingly reduced size of the dorsal funiculus, whereas Neuropilin-1 deficient mice are embryonic lethal showing abnormal defasciculation of cranial nerves and peripheral nerves as well as a looser DRG packaging (Giger et al., 2000; Kitsukawa et al., 1997). Among the different plexin subfamilies, members of the A group have been best characterized in terms of expression and function. All four mammalian PlexinAs are widely but somehow complementarily expressed in the developing spinal cord and adjacent sensory ganglia (Murakami et al., 2001). While PlexinA3 seems strongly expressed throughout the entire spinal cord, PlexinA2 is selectively expressed in the dorsal spinal cord. Moreover, PlexinA3 and PlexinA4 are most abundant in DRGs, whereas PlexinA1 and PlexinA2 are also expressed in sensory neurons, but the expression is limited to a small subpopulation and expression is absent from sympathetic ganglia (Murakami et al., 2001). Finally, PlexinA3 appears to be expressed in all peripheral ganglia, including trigeminal, vagal and otic in addition to dorsal root and sympathetic ganglia (Cheng et al., 2001). Interestingly, PlexinA2expressing cardiac neural crest cells are patterned abnormally in Sema3C mutant mice whereas PlexinA3 null mice display no gross defects in pathfinding of dorsal root sensory neurons but show fasciculation defects in the ophthalmic branch of the trigeminal nerve and aberrant development of hippocampal projections(Brown et al., 2001; Cheng et al., 2001). In contrast to PlexinAs, PlexinBs transcripts are only expressed at low levels in different regions of the spinal cord and peripheral nervous system. While PlexinB2 and B3 transcripts are absent in sensory neurons, PlexinB1 mRNA can be found in both, DRG as well as 4.2 Material and methods 81 trigeminal ganglia (Worzfeld et al., 2004). In the spinal cord, PlexinB1 and B2 messages are strongly expressed in the ependymal neuroepithelium whereas no PlexinB3 transcript is detectable in the spinal cord before birth. Until now, also no expression in the developing spinal cord has been reported for the unique PlexinC and PlexinD subfamily members. Despite the fact that neuropilins and plexins are expressed in various populations of spinal inter- and motoneurons as well as in sensory neurons of the peripheral nervous system, only few functional interactions between these molecules have been shown to be of biological significance (Chedotal et al., 1998; Kawasaki et al., 2002). In addition, no binding partners for several plexins and a large number of semaphorins have been found, suggesting that a large number of plexin/semaphorin ligand complexes are not yet identified. In order to learn more about potential functions and interactions of neuropilins, plexins and semaphorins participating in spinal cord and peripheral nervous system development, we have performed a detailed spatiotemporal expression study of all semaphorins (see chapter 2), neuropilins, and plexins present in the chicken genome. Interestingly, when compared to the mammalian genome the chicken genome lacks two plexin family members, namely PlexinA3 and PlexinB3 and the conservation within the PlexinB subclass is rather low. Expression of chicken plexins and neuropilins during spinal cord development revealed that neuropilin mRNAs are expressed only in restricted regions, only partially overlapping with the expression of PlexinA, B, C, and D, suggesting in addition to cooperative also separate functions for both molecular groups. 4.2 Material and methods 4.2.1 Assembly of chicken plexin cDNAs cDNA sequences for chicken plexins were assembled using the combined information from the chicken EST (http://www.chick.umist.ac.uk) and the chicken genomic database (http://www.ensembl.org/Multi/blastview?species=Gallus_gallus). 7 different genomic regions coding for plexins were identified using the tblastx alignment algorithms on available vertebrate plexins. The corresponding genomic fragments were downloaded and analyzed using the genescan gene prediction program (http://genes.mit.edu/GENSCAN.html). Putative cDNA and protein sequences were compared to the corresponding mammalian homologues and genescan prediction errors were corrected by manual inspection of the intron/exon boundaries in false predicted regions. Gaps in the assembled sequences due to inaccurate or incomplete genome sequencing were wherever possible filled by corresponding EST 82 Part IV: Paper 3 sequences. A total number of 85 chicken ESTs covering parts of 7 different plexins were identified. Among these 65 contained part of the coding sequence whereas the rest covered only parts of the 3’UTR sequence. Sequence alignment of genomic and EST sequences was done using the SeqMan software (Lasergene, DNASTAR, Madison WI). 3’UTR sequences were added to the coding sequence based on overlapping EST sequences that were supplemented with genomic sequences. UTR sequences were terminated at the first polyadenylation AATAAA/ATTAAA sequence that followed verified chicken UTR EST sequences. Using this combined approach a total of 5 complete and 2 partial cDNA sequences for plexins could be assembled. 4.2.2 Phylogenetic tree and domain identity analysis The domain structure of representative members of the chicken and mouse plexin super family was obtained using the smart program (http://smart.embl-heidelberg.de). Individual domains were extracted from the sequence using domain boundaries as predicted. Conserved domains from the different plexin subfamilies were aligned using the CLUSTAL W alignment algorithm (Higgens and Sharp, 1989; Thompson et al. 1994) provides by the MagAlign software (Lasergene, DNASTAR, Madison WI). Obvious mistakes in domain boundary prediction were manually adjusted. For better representation, alignment files were exported into TREEVIEW software, enabling the graphical representation of the unrooted tree (Page, 1996). Identical and conserved amino acids within individual domains were determined by pairwise alignment using the bl2seq blast program (http://www.ncbi.nlm.nih.gov/-blast/bl2seq/wblast2.cgi). 4.2.3 In situ hybridization 4.2.3.1 cRNA probe labeling The Chicken DNA plasmids derived from the EST clones (ChEST53D13, ChEST128L21, ChEST1014M19, ChEST890P9, ChEST799I19, ChEST860K1, ChEST110K21 and ChEST675H12) corresponding respectively to PlexinA1, PlexinA2, PlexinA4, PlexinB1, PlexinB2, PlexinC1, Neuropilin-1 and Neuropilin-2, found by data base search, were linearized using restriction endonucleases (NotI or EcoRI; Roche). The linearized plasmids were DIG labeled by incubating 2 µg of each DNA with 2 µl digoxigenin (DIG) labeling mix (Roche), 2 µl of T3 or T7 RNA polymerase (Roche), 2 µl of 10 X transcription buffer (Roche), and H2O added to a final volume of 20 µl for each reaction, at 37°C for 2 hours. After incubation, 2 units of Rnasefree DNaseI (Roche, 10U/µl) was added to the mix, and 4.2 Material and methods 83 incubated at 37°C for 30 min, after which 2 µl of 0.2 M EDTA, pH 8.0, was added to stop the nuclease treatment. The cRNA probe was ethanol-precipitated and dissolved in 50 µl of Rnase-free H2O. 4.2.3.2 RNA in Situ Hybridization Chick embryos at the designated stages were dissected in PBS and fixed in 4% PFA-PBS for 1 hour. After 30 minutes washing in PBS, tissues were embedded in OCT and quickly frozen in isopentane on dry ice. Sections of 25 µm thick were cut, collected on Super frost Plus (Fisher Scientific) microscope slides, dried at room temperature and stored at -20°C until use. Alternatively, embryonic tissue from different stages were collected after dissection in PBS and immediately embedded in OCT prior to quick freezing in isopentane on dry ice. Tissue sections were post-fixed half an hour in 4% PFA-DEPC PBS before a single 5 minutes wash in PBS followed by a 5 minutes wash in DEPC water were carried out. Sections were subsequently acetylated for 10 minutes, washed for 5 minutes once in PBS and once in 2X SSC-DEPC and subsequently incubated with the prehybridization buffer containing 40% formamid , 5X SSC- DEPC, 5X denhardts’ solution, 0.5 mg/ml yeast tRNA, 0.5 mg/ml salmon sperm DNA at 54°C for 3 hours. cRNA probes, diluted in prehybridization buffer at final concentration of 3 ng/µl were added to the slides and incubated over night at 54°C. The next morning, slides were washed as following: 5 minutes in 5X SSC, 5 minutes in 2X SSC, 5 minutes in 0.2X SSC, 20 minutes in 0.2X SSC containing 40% formamid at 54°C followed by one wash for 5 minutes in 2X SSC at room temperature. All the following steps were carried out at ambient temperature. Slides were afterwards washed twice for 10 minutes in detection buffer (0.1 M Tris-base, 15 mM NaCl, pH 7.5) before incubation in blocking buffer (3% milk in detection buffer) to block non specific binding. The anti-DIG phosphatase-conjugated antibody diluted in blocking buffer at 1:2000, was added to slides and left for 1 hour at room temperature prior to washing twice in detection buffer and one wash in alkaline phosphatase buffer (0.1 M Tris-Base, pH 9.5, 0.1 M NaCl and 50 mM MgCl2) for 5 minutes each wash. The bound probe was detected by adding NBT/BCIP substrate (Roche). For each ml of alkaline phosphatase buffer, 4.5 µl NBT and 3.5µl BCIP were added and the mixture was added to tissue sections and developed in the dark over night at 4°C. Images were recorded on a Zeiss Axioskop. 84 Part IV: Paper 3 4.3 Results 4.3.1 Plexin and Neuropilin genes in chicken 4.3.1.1 The avian genome lacks homologues for two mammalian plexin counterparts In order to identify chicken plexins and neuropilins we have performed extensive databank searches using the combined information from the EST and the genomic database. As previously observed for chicken semaphorin genes, database searches provide evidence that the chicken genome encodes a reduced number of plexins compared to its mammalian counterpart. While homologous chicken genes for the two unique family members PlexinC1 and PlexinD1 could be readily identified, neither an identical number of chicken PlexinAs nor PlexinBs is present in the chicken genome when compared to mammalians (Table 2). While the PlexinA subfamily contains no matching chicken sequence for PlexinA3, no counterpart for PlexinB3 could be extracted from chicken databases (Fig. 1 and Table 1). Interestingly, while conservation between different chicken PlexinAs was in the range of 70 to 90%, depending on the plexin parts used for alignment, conservation between chicken PlexinBs was only around 50% (Table 2). This value is not much higher than the values obtained when PlexinBs were compared to plexins in other subclasses, suggesting that even when put into the same subclass PlexinB1 and PlexinB2 might actually be members of different subclasses. PA1 PA2 PA3 PA4 PB1 PB2 PB3 PC1 PD1 ESTs 16 23 n.d 4 4 23 n.d 6 9 Chicken Ch 12 26 n.d 1 12 Un n.d 1 12 Mouse Ch 6d1 1h6 Xa7.1 6a3.3 9f2 15e3 Xa7.1 10c2 6e3 Table 1: Chromosomal localization and EST representation of different plexin gene products. Numbers of identified chicken ESTs for each gene are indicated and chromosomal location of the chicken and its corresponding mouse gene are given. Not detected genes are abbreviated by n.d, whereas identified chromosomal sequences on not yet localized chromosomes are indicated by the letters Un. 4.3 Results 85 Sema Domain cPlexinA1 cPlexinA2 57 / 71 cPlexinA4 56 / 70 59 / 76 cPlexinB1 28 / 47 28 / 48 27 / 47 cPlexinB2 28 / 47 30 / 49 30 / 48 36 / 55 cPlexinC1 13 / 23 12 / 22 13 / 22 12 / 20 12 / 21 cPlexinD1 24 / 38 23 / 40 23 / 40 27 / 43 24 / 40 15 / 28 cPlexinA1 cPlexinA2 cPlexinA4 cPlexinB1 cPlexinB2 cPlexinC1 cPlexinD1 PSI Domain cPlexinA1 cPlexinA2 64 / 73 cPlexinA4 74 / 87 62 / 77 cPlexinB1 50 / 61 44 / 55 50 / 58 cPlexinB2 44 / 65 39 / 56 39 / 60 48 / 66 cPlexinC1 42 / 70 42 / 76 46 / 65 48 / 65 38 / 57 cPlexinD1 37 / 53 41 / 61 37 / 55 40 / 51 39 / 53 39 / 56 cPlexinA1 cPlexinA2 cPlexinA4 cPlexinB1 cPlexinB2 cPlexinC1 cPlexinD1 Table 2: Conservation of different regions between plexin super family members. Pairwise alignment of different individual domains within the plexin proteins indicate that the intracellular SCOP domain is highly conserved between the different plexin family members, whereas the two analyzed extracellular domains, sema and PSI, show far less conservation. Interestingly the conservation between individual domains of chicken PlexinB1 and PlexinB2 is only slightly higher than the homology between class A plexins and class B plexins, suggesting that chicken PlexinB genes are evolutionary quite distinct. The left numbers indicate the identical amino acids, whereas the right numbers indicate the conserved amino acids. 86 Part IV: Paper 3 SCOP Domain cPlexinA1 cPlexinA2 88 / 95 cPlexinA4 82 / 94 86 / 95 cPlexinB1 56 / 72 56 / 73 53 / 73 cPlexinB2 56 / 74 56 / 75 56 / 76 66 / 79 cPlexinC1 49 / 69 48 / 68 49 / 70 45 / 65 48 / 66 cPlexinD1 56 / 75 56 / 74 58 / 75 55 / 74 55 / 74 57 / 77 cPlexinA1 cPlexinA2 cPlexinA4 cPlexinB1 cPlexinB2 cPlexinC1 cPlexinD1 Table 2 (continued): Conservation of different regions between plexin super family members. Pairwise alignment of different individual domains within the plexin proteins indicate that the intracellular SCOP domain is highly conserved between the different plexin family members, whereas the two analyzed extracellular domains, sema and PSI, show far less conservation. Interestingly the conservation between individual domains of chicken PlexinB1 and PlexinB2 is only slightly higher than the homology between class A plexins and class B plexins, suggesting that chicken PlexinB genes are evolutionary quite distinct. The left numbers indicate the identical amino acids, whereas the right numbers indicate the conserved amino acids. 4.3.1.2 Plexins and Neuropilins expression at early developmental stages of the chicken spinal cord During early developmental stages, (stage 18 to stage 22), axons of motoneurons start to leave the spinal cord to form the ventral roots and neural crest cells migrate ventrolaterally aggregating into bilaterally paired DRGs (Krull, 2001; Krull and Koblar, 2000). At stage 18, only PlexinA1 (Fig. 2, A1 black arrow) and PlexinA2 (Fig. 2, A2 black arrow) mRNAs are detected in motoneurons. While the expression patterns of these two plexins are originally similar in motoneurons by stage 22, PlexinA1 and PlexinA2 transcript levels are differently regulated. PlexinA1 transcript levels remain high in motoneurons whereas PlexinA2 mRNA expression decreases and subsists in only a subgroup of motoneurons (Fig. 3, A1, and A2). Interestingly, PlexinA4 mRNA expression in motoneurons is also detected at stage 22 (Fig. 3, A4 white arrow) whereas no other plexins subfamily members are expressed during early motor axonal outgrowth (Fig. 2 and Fig. 3). Neuropilin-1 (NP-1) and Neuropilin2 (NP-2) are also expressed during early motor neuron development (Fig. 3, NP-1 and NP-2 white arrows), however only neuropilin-1 can be detected at stage 18 (Fig. 2, NP-1 black arrow). 4.3 Results 87 Fig. 1: Phylogenetic tree of the members of the plexin super family. Alignments are done on the bases of the plexin sema domain (A) or the first PSI domain of each family member using the CLUSTAL W program. The scale bar represents the substitution rate of 10 amino acids per 100 amino acid residues. Please note that only mouse (m) and chicken (c) semaphorin sequences are depicted, but that alignment of rat or human sequences gave similar results (data not shown). However, PlexinA1 and PlexinA2 expression is not restricted to the ventral spinal cord but extends to the dorsal spinal cord, where PlexinA1 and PlexinA2 mRNA are also detected in commissural interneuron precursors (Fig. 3, A1, and A2, white arrowheads). Commissural neurons are generated at stage 19 (Stoeckli and Landmesser, 1995) and express all PlexinA members by stage 22 (Fig. 3, A1, A2 and A3, white arrowheads), the time when the majority of commissural axons reach the floor plate. Surprisingly, around stage 22 all PlexinAs, 88 Part IV: Paper 3 PlexinB1 (PB1) and PlexinC1 (PC1) are also expressed in the floor plate (Fig. 3, A1, A2, A4, B1 and C1, black arrows), a structure that represents an intermediate target for commissural interneurons. Interestingly, PlexinA1 (PA1) is only expressed in lateral but not medial floor plate cells, and PlexinB1 is expressed in the entire ventricular zone (Fig. 3, B1 and Fig. 6, B1), PlexinB2 is strongly expressed in cells surrounding the spinal cord and this expression is rapidly down regulated later. In addition, PlexinA1, PlexinA4 (PA4), and both neuropilins transcripts are detected in dorsal root ganglia at stage 22. Fig. 2: Plexin and Neuropilin expression patterns in chicken spinal cord and peripheral nervous system at stage 18. Cross sections of chicken lumbosacral spinal cord were incubated with Dig labeled antisense probes. Different probes used are indicated. At this early stage of development, only PA1, PA2, and NP-1 transcripts are detected in the ventral spinal cord (A1, A2 and NP-1 black arrow). Scale bars correspond to 200 µm. 4.3.1.3 Plexins and Neuropilins expression during hindlimb innervation After stage 23, motor axons reach the plexus region where they sort out extensively according to their muscle targets. While the first decision is primarily a choice to grow either dorsally or ventrally, pathways that are more refined are chosen by stages 25/26, when individual muscle nerves begin to form (Landmesser, 2001). At that time, motoneurons can based on their rostrocaudal position within the spinal cord axis and their transcriptional profile be separated into different subpopulations (Jessell, 2000). 4.3 Results 89 Fig. 3: Plexin and Neuropilin expression patterns in chicken spinal cord and peripheral nervous system at stage 22. Cross sections of chicken lumbosacral spinal cord were incubated with Dig labeled antisense probes. Different probes used are indicated. At stage 22 all PlexinAs PB1 and PC1 are expressed in the floor plate (A1, A2, A4, B1 and C1 black arrows). While, PlexinAs are expressed in dorsal interneurons (A1, A2 and A4 white arrowheads), PA4 and NP-2 messages are also detected in motoneurons (A4, NP-2 white arrows). PA1, PA2, and NP-1 (NP-1 white arrow) messages persist in motoneurons at this stage and all PlexinAs and both neuropilins are expressed in DRGs. PlexinBs transcripts are also up regulated at this time. PB1 (B1) is highly expressed in the ventricular zone and PB2 messages (B2) are detected transiently in cells surrounding the spinal cord. Scale bars correspond to 200 µm. At the lumbosacral level, the region we studied, the motor column can mainly be divided into the lateral motor column (LMC) and medial motor column (MMC), with the later motor column being further subdivided into lateral LMCL and medial LMCM parts depending on the motoneurons position within the LMC (Jessell, 2000; Landmesser, 1978b). At stage 25, we can clearly detect neuropilin-2 and PlexinA1 messages in a subpopulation of motoneurons belonging to the LMCL (Fig. 4, NP-2, and A1, black arrows), whereas neuropilin-1, PlexinA4 and PlexinD1 (PD1) display a broader expression corresponding to all LMCs (Fig. 4, NP-1, A4 and D1, white arrows). Additionally, while PlexinB1 transcripts are not detected in motoneurons at this stage, PlexinB2 and PlexinC1 display weak expression in some motoneurons that do not appear to be clearly segregated into pools (Fig. 4, B2, and C1, black arrowheads). 90 Part IV: Paper 3 Plexin and neuropilin expression in different motoneuron subpopulations is also observed at stage 30, although the distribution of individual plexins has changed (Fig. 5). PlexinA4 mRNA is no longer expressed homogenously in motoneurons but rather in a gradient with increasing transcription levels from the medial to the lateral motoneuron pool (Fig. 5, A4 white arrow). The expression of PlexinD1 is lost in motoneurons and remains only in interneurons along the ventricular zone that overlaps with NP-1 expression (Fig. 5, D1, and NP-1, black arrows). Strikingly NP-1 and NP-2 expression shifts to distinct subsets of neurons with almost complementary patterns (Fig. 5, NP-1, and NP-2, white arrowheads). While the expression of PlexinB1 changes little compared to earlier stages, PlexinB2 expression is down regulated in cells surrounding the spinal cord but is weakly up regulated in motoneurons. Such an up regulation in the lateral part of the motor column is also seen for PlexinC1 (Fig. 5, C1). Fig. 4: Plexin and Neuropilin expression patterns in chicken spinal cord and peripheral nervous system at stage 26. Cross sections of chicken lumbosacral spinal cord were incubated with Dig labeled antisense probes. Different probes used are indicated. At stage 26, PA1 and NP-2 expression shift to a subpopulation of the lateral motor column (A1, black arrow), whereas PA2 transcripts are highly detected in interneurons. PA4, NP-1, and PD1 display a broader expression in the lateral motor column (A4, NP-1 and D1 white arrows). PB2 and PC1 messages (B2 and C1 black arrowheads) are also detectable in motoneurons although at a lower level. PlexinA and neuropilin transcript are also detected in DRGs (A1, A2, A4, NP-1 and NP-2) and PB1 (B1) messages is still detectable in the ventricular zone. In addition, a weak expression of PB1 is detected in a subpopulation of sensory neurons in DRGs. Scale bars correspond to 200 µm. 4.3 Results 91 PlexinA2 and PlexinC1 expression in the floor plate persist through stage 26, when commissural axons have crossed the midline and turned into the longitudinal axis (Fig. 4, A2 and C1). At the same time, PlexinC1 expression is also seen in the dorsal spinal cord in a position that overlaps with the position of the dorsolateral commissural neurons. During this stage PlexinA1, PlexinA4, neuropilin-1 and neuropilin-2 expression is unchanged in DRGs compared to stage 22. Fig. 5: Plexin and Neuropilin expression patterns in chicken spinal cord and peripheral nervous system at stage 30. Cross sections of chicken lumbosacral spinal cord were incubated with Dig labeled antisense probes. Different probes used are indicated. At stage 30, PlexinA, PlexinB and PC1 expression patterns remain largely unchanged however PA4 transcript are now detectable only in a lateral subpopulation of motoneurons (A4 white arrowhead). Strikingly, NP-1 and NP-2 display complementary expression patterns in motoneurons (NP-1 and NP-2 white arrowheads), whereas PD1 messages are now only detected in interneurons close the ventricular zone (D1 black arrow). Only PA1, PA4, and NP-1 continue to be expressed in DRGs. Scale bars correspond to 200 µm. 4.3.1.4 Plexins and Neuropilins at late stages of spinal cord development During late stages of spinal cord development, between stages 35 (Fig. 6) and 40 (data not shown), when motoneurons have reached their muscle targets and sensory afferents terminate in their specific target layers of the gray matter, expression patterns of plexins and neuropilins are still changing. 92 Part IV: Paper 3 Fig. 6: Plexin and Neuropilin expression patterns in chicken spinal cord and peripheral nervous system at stage 35. Cross sections of chicken lumbosacral spinal cord were incubated with Dig labeled antisense probes. Different probes used are indicated. During this late stage of spinal cord development only PA1 transcripts are highly expressed in motoneurons (A1), whereas PA2, PA4 and NP-1 are expressed in the dorsal horn of the spinal cord (A2 white arrowhead, A4 white arrow and NP-1 black arrow). PA2 is strongly detected in layer І to layer III (A2 white arrowhead) whereas NP-1 and PA4 expression are less intense. NP-2 messages become restricted to few cells in the ventral horns of the spinal cord (NP-2 black arrow). While PB2 and PD1 are not detected any more at this stage, PB1 expression remain unchanged and PC1 messages are up regulated in DRGs. Scale bars correspond to 200 µm. The expression of neuropilins gets more and more restricted to only very distinct populations of cells. Neuropilin-1 is expressed in the dorsal horn, whereas neuropilin-2 becomes restricted to cells in the ventral horns of the spinal cord (Fig. 6, NP-1 and NP-2, black arrows). Similarly, the expression patterns of PlexinA get more restricted, where PlexinA4 is detected only in the dorsal horn at stage 35 (Fig. 6, A4 white arrow) but disappears between stages 35 and 40 (data not shown). In contrast, PlexinA1 remains more or less diffusely expressed in the gray matter. However, a slight expression is seen also in a small area of the LMC. PlexinA2 is expressed predominantly in layers I to III (Fig. 6, A2 white arrowhead), whereas PlexinC1, at stage 26 widely expressed in the gray matter, retracts more and more dorsally to become totally absent from the intermediate and ventral spinal cord at stage 40 (data not shown). In contrast to PlexinAs, which are expressed during the time when neurons extend their axons, PlexinC1 is expressed only during late stages of neuronal development. Neither 4.4 Discussion 93 commissural neurons nor motor neurons express PlexinC1 during the time when they approach their first intermediate target, the floor plate, and the plexus region respectively. After stage 30, PlexinC1 is expressed transiently throughout the gray matter (Fig. 6, C1) before it becomes restricted to the dorsal horns at stage 40 (data not shown), where it overlaps with PlexinA1 and PlexinA2 (data not shown). Interestingly, a massive increase in PlexinC1 expression in dorsal root ganglia at stage 35 (Fig. 6, C1) is observed, and this high expression persists at stage 40 (data not shown). At these late stages, PlexinD1 expression in endothelial cells in no longer detected (Fig. 6, D1 and data now shown) and neuropilin-2 expression is down regulated in DRGs (Fig. 6, NP-2). In contrast, PlexinA1, PlexinA4 and neuropilin-1 expression remain unchanged in DRGs until stage 35 (Fig. 6, A1, A4 and NP-1). 4.4 Discussion 4.4.1 Mouse plexin genes located on the X sex chromosome are absent in chicken Based on their sequence homology and domain organization plexins can be grouped into 4 different subclasses namely A to D. While the C and D subfamilies contain only a single representative, four different PlexinAs and three different PlexinBs are present in the mammalian genome. However, extensive databank searches only identified 7 chicken counterparts compared to the nine mammalian plexins. One of the major questions in this respect is whether we were simply unable to identify corresponding chicken fragments or whether the chicken genome indeed contains a reduced number of plexin genes. The fact that we did identify all the seven chicken plexins using either the EST or the chicken genomic database demonstrates that the coverage of these databases is, even while not complete, quite high. Interestingly the mouse variants of the two missing plexin genes are located on the X chromosome, implying that a homologous region of this chromosome is absent in chicken. This hypothesis is strengthened by the fact that avian use a Z/W sex determination system instead of the normal X/Y system found in mammals (Gilgenkrantz, 2004). Moreover, in contrast to mammals, in chicken the female is the heterogametic sex (ZW), whereas the male is homogametic (ZZ), further suggesting that intense chromosomal rearrangements did happen during evolutionary development (Smith and Sinclair, 2004). These changes are also reflected by the fact that the haploid chicken genome contains 38 autosomes compared to only 19 autosomes in mouse (Vinogradov, 1999). Remarkably, the chicken genome is still only about half the size (1.2 x 109 bp) of the mouse genome (2.56 x 109 bp), a difference that can 94 Part IV: Paper 3 be largely explained by the finding that 30 of the 38 chicken autosomes, called microchromosomes are extremely small in size ranging only between 5 and 20 Mb (Venkatesh et al., 2000; Vinogradov, 1999). While a reduction in size is not per se an indication for a reduced number of genes as demonstrated for many teleost, two indications point towards a reduced number of active chicken genes. In contrast to the pufferfish Fugu rubripes which has been shown to have extremely small gene sizes in which intron sequence length has been reduced to a minimum (Taylor and Semple, 2002), the overall gene length in chicken and mouse are not significantly different . Moreover, all attempts to clone chicken homologues that could not be identified by database screening using degenerate primers in RT-PCR reactions failed, again implying that the missing genes are indeed not represented in the chicken genome. Nevertheless only the completion of the chicken genome sequencing project as well as the extension of the number of sequenced chicken ESTs can finally answer the question whether the chicken genome indeed encodes fewer plexins than the mammalian genome does. 4.4.2 Expression patterns of plexins and neuropilins are dynamically regulated during spinal cord development 4.4.2.1 PlexinAs and neuropilins expression patterns in chicken suggest possible functions for PlexinAs independently from neuropilins A variety of studies report about the role of plexins during development (Fujisawa, 2004). While the first plexin was described quite some time ago (Ohta et al., 1995) their expression patterns are not known in details. In particular, little is know about temporal changes in the expression of different plexins. We have now performed a detailed spatiotemporal expression analysis for all different chicken plexins. Most plexins and neuropilins are highly expressed during different stages of motoneuron development, suggesting multiple roles for these molecules in motor nerve segregation and/or defasciculation. Interestingly, motoneurons also express different class III semaphorins and high levels of Sema5A and Sema7A (see chapter 2), implying that in analogy to results seen in the Drosophila embryo, where the transmembrane semaphorin Іa is required for the selective defasciculation of specific motoneurons (Yu et al., 1998), plexin- semaphorin interactions are required for similar events in higher vertebrates. While the function of PlexinAs has been studied predominantly in the context of their role as a co-receptor together with neuropilins (Tamagnone and Comoglio, 2000), it is clear that in chicken PlexinAs must have functions that are independent of neuropilins, because they are 4.4 Discussion 95 distributed much more widely in the developing nervous system than neuropilin-1 and -2. In the dorsal spinal cord for example, commissural neurons express all three members of the plexin-A class but neither neuropilin-1 nor neuropilin-2. This expression pattern is divergent from the widespread expression of neuropilin-2 in the spinal cord of the E10.5-12.5 mouse (Brown et al., 2001; Chen et al., 2000), where neuropilin-2 is expressed very strongly in dorsal commissural neurons and in all ventral populations of interneurons. Interestingly, all PlexinAs are expressed also in the floor plate. PlexinA1 is expressed only in lateral but not medial floor plate cells. Expression of plexins in the floor plate is rather surprising, as the floor plate is the intermediate target of commissural axons, and therefore, the site where ligands for axonal receptors are expected. A receptor function of plexins in floor plate cells at stage 22 is less obvious, as these cells do not migrate or undergo structural remodeling at this time. Floor plate development seems to be terminated much earlier (Briscoe and Ericson, 2001; Briscoe et al., 2000; Jessell, 2000; Wijgerde et al., 2002). Interestingly, the floor plate is also the earliest site of PlexinC1 expression in the spinal cord where the expression is observed by stage 22 and last until stage 30, although much weaker. The expression of several plexins in the floor plate might suggest a possible role for these plexins as a ligand rather than a receptor as always presumed. 4.4.3 PlexinBs and PD1 are only transiently expressed in neurons The expression of class-B plexins is difficult to link to any specific function. Both plexin B members are expressed in neurons. The most prominent and longest lasting expression for a class-B plexin is seen in the ventricular zone, where PlexinB1 is expressed from stage 22 to stage 35. This pattern is consistent with PlexinB1 expression seen in mouse, where PB1 transcript was found in the ventricular zone of the spinal cord at E13.5 (Worzfeld et al., 2004). The expression of chicken PlexinB2 is generally very weak with the exception of cells surrounding the spinal cord at stage 22. These cells can be crest cell derivatives, however no expression in sensory ganglia neurons is observed at any stage, or endothelial cells of the perineural vascular vessels sprouting from the primitive arterial tract. In contrast to the observations in mouse, PlexinBs are not expressed during the time when sensory afferents target their specific layers in gray matter, as collaterals of primary sensory axons in chicken do not form before stage 29 (Perrin et al., 2001), an age where PlexinB1 and PlexinB2 are absent in DRG. Thus, although a contribution of both PlexinBs to the formation of the central sensory connections is still possible it seems rather unlikely. Taken together, chicken PlexinBs seem to have slightly altered expression patterns and functions when compared to 96 Part IV: Paper 3 mouse PlexinBs a result that is hardly surprising since the chicken genome lacks one PlexinB family member. PlexinD1 has been reported to be expressed in endothelial cells in mice and zebrafish and play an important role in vivo in vascular system development (Gitler et al., 2004; TorresVazquez et al., 2004). In chicken, PlexinD1 is expressed as early as stage 18 in endothelial cells of the perineural vascular plexus and this expression persists until stage 30. In addition PlexinD1 is expressed in developing vessels throughout the entire embryo suggesting an important role for this protein during early stages of angiogenesis as earlier observed in mice and zebrafish (Gitler et al., 2004; Gu et al., 2004; Torres-Vazquez et al., 2004). However, PlexinD1 in chicken is also expressed in the central nervous system in spinal motor neurons at the time motor axons sort in the limb plexus before entering the limb muscle mass. This is in concordance with earlier published report demonstrating the expression of PlexinD1 in mouse brain during embryonic development (van der Zwaag et al., 2002). 4.4.4 PlexinC1 is not expressed in early stages of neuronal development In contrast to PlexinAs, which are expressed during the time when neurons extend their axons, PlexinC1 is expressed only during late stages of neuronal development. Neither commissural neurons nor motor neurons express PlexinC1 during the time when they approach their first intermediate target the floor plate and the plexus region respectively. Interestingly, strong expression of PlexinC1 is seen in the floor plate at stage 22, i.e. when the majority of the axons from dorsolateral commissural neurons are in the floor plate. The expression in the floor plate persists through stage 25, when commissural axons have crossed the midline and turned into the longitudinal axis (Bourikas, 2005). At that time expression is also seen in the dorsal spinal cord in a position that overlaps with the position of the dorsolateral commissural neurons (Bourikas, 2005; Stoeckli and Landmesser, 1995). While PlexinC1 expression in commissural interneurons is consistent with the expression of Sema7A in the floor plate, the semaphorin known to bind to PlexinC1 (Tamagnone et al., 1999), PC1 messages in the floor plate cells suggest homo or heterophilic plexin interactions rather than binding with Sema7A. Interestingly, there is a massive increase in PlexinC1 expression in dorsal root ganglia at stage 35, a time when all other plexins are down regulated and expression persists until stage 40. Thus, PC1 starts to be expressed when axons have already completed the navigation to their targets suggesting that PC1 might be involved in target recognition or synaptogenesis rather than pathfinding. Part V: Conclusion and Outlook 97 Conclusion and Outlook The adult nervous system is organized in remarkably complex networks capable of achieving a broad range of activities including very complicated functions such as learning. The proper functioning of this elaborate network results from the establishment of functional connections between the numerous neuronal cells that are generated during development. Since the early 1990s, a constantly increasing number of studies describe axon guidance molecules and their implication in vitro as well as in vivo in steering different types of neural axons or in affecting the migratory paths of neural cells (Johansen and Johansen, 1997; Young et al., 2004). However, the precise details of how neuronal wiring takes place remain to a large part unknown. This is due in part to the fact that the path of a growing axon appears to be very complex in vivo and implicates several families of guidance molecules. Moreover, several members of the same family of axon guidance may act in concert to direct the growth cone, through a series of attractive and/or repulsive events. Additionally, the expression of guidance molecules outside the nervous system and their implication in the growth of other tissues makes the understanding of how axon guidance works even harder. Plexins and their ligands, the Semaphorins, have been intensively characterized during the last years. They have been demonstrated to play crucial roles in directing axonal outgrowth as well as neural crest cells migration during brain and/or spinal cord development (Fiore and Puschel, 2003; Fujisawa, 2004). However, their expression patterns extend beyond the nervous system, suggesting their involvement in heart, bone, lung, intestine, immune system and vascular system development (Park et al., 2004; Serini and Bussolino, 2004). Such wide spread expression in several vital tissues made the generation of mutant mice very complicated, since conventional knock out techniques often resulted in early embryonic lethality related to vital function disruptions (Behar et al., 1996; Gitler et al., 2004). Recent progresses, in alternative procedures to knock down a particular gene function, have provided new possibilities to study plexin and semaphorin activities. We employed the chicken embryo as a model system to knock down PlexinD1 transcripts using in ovo RNAi, a technique based on the electroporation of long double stranded RNA injected into the central canal of the chicken embryo (Pekarik et al., 2003). This system provides a powerful tool to explore exclusively the role of axon guidance molecules within the spinal cord, circumventing the problem of embryonic lethality due to severe defects in vital tissues. This seems certainly necessary since our spatio-temporal expression analysis demonstrates that semaphorin and 98 Conclusion and Outlook 99 plexin expression in the chicken spinal cord is very complex, suggesting multiple functional interactions within the same system (chapter 2 and chapter 3). PlexinD1 activity in vivo has been extensively investigated during vasculogenesis. Nevertheless, PlexinD1 is also expressed in the central nervous system, however no functional role in nervous system development has been reported (Gesemann et al., 2001; van der Zwaag et al., 2002). Our study demonstrates that PlexinD1 is expressed in spinal motor neurons of chicken embryos between stages 23 and 26, the period motor axons sort within the limb plexus before initiating the limb innervation. PD1 knock down results in severe defects in the formation of the crural nerve, suggesting a functional role for PlexinD1 in axon sorting at the plexus level (chapter 1). Immunostaining of chicken spinal motoneurons on cross sections using MNR2 antibody that marks all motoneurons subtypes show no differences in the positioning of motoneurons within the spinal cord in PlexinD1 knock down embryos. However, MNR2 marker is a general marker for all motoneurons subtypes and does not give any indication about the arrangement of different motoneuron pools. To exclude that PD1 knock down in chicken spinal indeed causes incorrect positioning of motoneurons in different motor pools which could lead to abnormal motor axon outgrowth, supplementary experiments using specific markers for each motor pool such as different cadherins or ETS transcription factors are required. However, a function for PlexinD1 in motoneurons sorting seems unlikely because PD1 mRNA is expressed in motoneurons only after stage 23, at this time all motoneurons are post mitotic and have already migrated to their final position within different pools. Surprisingly, PlexinD1 loss of function also leads to abnormalities in the outgrowth of sensory afferents although PlexinD1 was never detected in the dorsal sensory neurons. Additionally, PD1 knock down embryos exhibit defasciculation defects at the ventral motor roots and disorganization of the boundary cap cells in the region of ventral motor exit points. Two different hypotheses could be raised to explain the observed phenotypes. In the first hypothesis, PlexinD1 positive cells outside the spinal cord are not exclusively endothelial cells but include a migratory subpopulation of crest cells, notably Schwann cells and/or boundary cap cells. PlexinD1 knock down in crest cell derivatives might affect the normal migration of these cells and subsequently lead to their incorrect positioning which can explain the dorsal root abnormal arrangement as well as the incorrect ventral motor roots formation. A potential role for PlexinD1 in cell migration goes along with earlier observations demonstrating that PD1 in postnatal rats is exclusively expressed in three brain nuclei, the 100 Conclusion and Outlook pons, the inferior olive and the dorsal column nuclei; structures that form by tangentialy migrating cells originating from the precerebellar neuroepithelium (Altman and Bayer, 1987). To verify this hypothesis, several additional experiments are required. In a first step and in the absence of available antibody against PlexinD1 protein, double in situ hybridization experiments to co-localize PlexinD1 expression outside the spinal cord with a crest cell marker such as Sox-10 or HNK-1 are required. Additionally, these same markers can be used also to investigate the migratory pattern of crest cells derivatives in PD1 knock down embryos in comparison to controls. In addition, several specific boundary cap cells markers can be used such as Krox20 or Sema6A (that is exclusively expressed in boundary cap cells in chicken) to check the integrity of boundary cap cells formation in PlexinD1 knock down animals. Alternatively, the aberrant dorsal roots formation as well as the ventral motor roots defects could result from the abnormal formation of certain vessels running in close contact and/or interacting with growing axons. This hypothesis is based on observations reported in many recent publications demonstrating an implication of PlexinD1 in vascular system formation (Gitler et al., 2004; Gu et al., 2004; Torres-Vazquez et al., 2004). PlexinD1 knock out mice, exhibit a severe disorganization of the normally highly structured and ordered intersomitic vessels. In the case that PD1 is exclusively expressed in endothelial cells, PlexinD1 knock down in chicken embryos might lead to the disorganization of intersomitic vessels, and consequently these alterations may cause abnormal sensory axonal growth. This idea is further supported by evidence that a tight cellular and/or molecular interaction between vasculogenesis and axonal outgrowth especially between intersomitic arteries and outgrowing sensory axons in xenopus embryos exist (Levine et al., 2003). To investigate this hypothesis, a very important issue is to check whether in our system PlexinD1 loss of function is achieved also in endothelial cells surrounding the neural tube. To date we do not have any indication about the possible targeting of intersomitic vessels cells using in ovo RNAi. To the contrary, our data seems to demonstrate a restricted targeting of cells, exclusively within the spinal cord and dorsal root ganglia, as seen by the electroporation of YFP plasmids. However, we cannot exclude the possibility that long double stranded RNA targets more cells than cDNA constructs, as it is smaller and might have a different charge compared to DNA plasmids. In order to resolve this issue, we need to use a different approach replacing the use of long ds RNA by a recombinant construct, containing an YFP reporter gene followed by a PlexinD1 Conclusion and Outlook 101 small interfering RNA. However, to date we did not assess the efficiency of small interfering RNA to knock down PlexinD1 expression. Another important matter required for verifying the plausibility of this hypothesis, is to investigate the integrity of intersomitic vessels patterning in PD1 knock down animals. Unfortunately, no antibodies recognizing the chicken vascular system are yet available making the use of alternative staining methods, such as the perfusion of certain lectins that can bind to endothelial cells and allow the staining of the chicken vascular system, necessary (Hagedorn et al., 2004). Furthermore, it is of high interest to investigate whether the identified PlexinD1 binding partners (Sema3A, Sema3C, and Sema3E) are used also in chicken embryo and are implicated in PlexinD1 signaling in spinal motoneurons. In this regard, the chicken embryo represents an excellent model system, as the RNAi technique offers a major advantage to gene disruption in mice, which is the possibility of knocking down several genes simultaneously. This is important in order to avoid redundancy and to tackle the combinatorial activity of multiple axon guidance molecules acting in chorus. In this respect, it will be interesting to investigate axon outgrowth in single and multiple semaphorins knock down embryos. However, in order to assess the binding between PlexinD1 and Sema3A, 3C and 3E in chicken, the cloning of these genes is required. The cloning of chicken PlexinD1 recombinant construct offers also the possibility to over express PlexinD1 in chicken spinal cord in vivo in order to unravel the potential function of this protein during development. In conclusion, our current results demonstrate clearly the expression of PD1 mRNA in chicken spinal motoneurons during the time motor axons sort in the limb plexus, suggesting a potential role for this protein in motor axon guidance. Indeed PlexinD1 knock down in chicken embryos using in ovo RNAi lead to motor axon pathfinding defects in the crural nerve trunk. Further experiments will allow us: to explain the aberrant formation of the dorsal and ventral roots in PlexinD1 knock down animals and to unravel whether PlexinD1 plays a novel role in the correct migration of a subpopulation of neural crest cell derivatives and/or a tight relation between intersomitic vessels sprouting and dorsal and motor root formation. 102 References Adams, D. H., and Scott, S. A. (1998). Response of "naive" cutaneous and muscle afferents to potential targets in vitro. Dev Biol 203, 210-220. Adams, R. H., Betz, H., and Puschel, A. W. (1996). A novel class of murine semaphorins with homology to thrombospondin is differentially expressed during early embryogenesis. Mech Dev 57, 33-45. Altman, J., and Bayer, S. A. (1987). Development of the precerebellar nuclei in the rat: IV. The anterior precerebellar extramural migratory stream and the nucleus reticularis tegmenti pontis and the basal pontine gray. J Comp Neurol 257, 529-552. Bagnard, D., Lohrum, M., Uziel, D., Puschel, A. W., and Bolz, J. (1998). Semaphorins act as attractive and repulsive guidance signals during the development of cortical projections. Development 125, 5043-5053. Bagnard, D., Thomasset, N., Lohrum, M., Puschel, A. W., and Bolz, J. (2000). Spatial distributions of guidance molecules regulate chemorepulsion and chemoattraction of growth cones. J Neurosci 20, 1030-1035. Baker, C. V., and Bronner-Fraser, M. (1997). The origins of the neural crest. Part I: embryonic induction. Mech Dev 69, 3-11. Basler, K., Edlund, T., Jessell, T. M., and Yamada, T. (1993). Control of cell pattern in the neural tube: regulation of cell differentiation by dorsalin-1, a novel TGF beta family member. Cell 73, 687-702. Bastiani, M. J., Raper, J. A., and Goodman, C. S. (1984). Pathfinding by neuronal growth cones in grasshopper embryos. III. Selective affinity of the G growth cone for the P cells within the A/P fascicle. J Neurosci 4, 2311-2328. Battye, R., Stevens, A., and Jacobs, J. R. (1999). Axon repulsion from the midline of the Drosophila CNS requires slit function. Development 126, 2475-2481. Behar, O., Golden, J. A., Mashimo, H., Schoen, F. J., and Fishman, M. C. (1996). Semaphorin III is needed for normal patterning and growth of nerves, bones and heart. Nature 383, 525528. Bermingham, N. A., Hassan, B. A., Wang, V. Y., Fernandez, M., Banfi, S., Bellen, H. J., Fritzsch, B., and Zoghbi, H. Y. (2001). Proprioceptor pathway development is dependent on Math1. Neuron 30, 411-422. Bourikas, D. (2005). Sonic hedgehog guides commissural axons along the longitudinal axis of the spinal cord. Nature Neuroscience In press. Bourikas, D., and Stoeckli, E. T. (2003). New tools for gene manipulation in chicken embryos. Oligonucleotides 13, 411-419. Bovolenta, P., and Dodd, J. (1990). Guidance of commissural growth cones at the floor plate in embryonic rat spinal cord. Development 109, 435-447. Briscoe, J., and Ericson, J. (2001). Specification of neuronal fates in the ventral neural tube. Curr Opin Neurobiol 11, 43-49. Briscoe, J., Pierani, A., Jessell, T. M., and Ericson, J. (2000). A homeodomain protein code specifies progenitor cell identity and neuronal fate in the ventral neural tube. Cell 101, 435445. 103 104 References Bronner-Fraser, M. (1993). Mechanisms of neural crest cell migration. Bioessays 15, 221230. Brown, C. B., Feiner, L., Lu, M. M., Li, J., Ma, X., Webber, A. L., Jia, L., Raper, J. A., and Epstein, J. A. (2001). PlexinA2 and semaphorin signaling during cardiac neural crest development. Development 128, 3071-3080. Buchman, V. L., and Davies, A. M. (1993). Different neurotrophins are expressed and act in a developmental sequence to promote the survival of embryonic sensory neurons. Development 118, 989-1001. Cano, A., Perez-Moreno, M. A., Rodrigo, I., Locascio, A., Blanco, M. J., del Barrio, M. G., Portillo, F., and Nieto, M. A. (2000). The transcription factor snail controls epithelialmesenchymal transitions by repressing E-cadherin expression. Nat Cell Biol 2, 76-83. Castellani, V. (2002). The function of neuropilin/L1 complex. Adv Exp Med Biol 515, 91102. Castellani, V., and Rougon, G. (2002). Control of semaphorin signaling. Curr Opin Neurobiol 12, 532-541. Chedotal, A., Del Rio, J. A., Ruiz, M., He, Z., Borrell, V., de Castro, F., Ezan, F., Goodman, C. S., Tessier-Lavigne, M., Sotelo, C., and Soriano, E. (1998). Semaphorins III and IV repel hippocampal axons via two distinct receptors. Development 125, 4313-4323. Chen, H., Bagri, A., Zupicich, J. A., Zou, Y., Stoeckli, E., Pleasure, S. J., Lowenstein, D. H., Skarnes, W. C., Chedotal, A., and Tessier-Lavigne, M. (2000). Neuropilin-2 regulates the development of selective cranial and sensory nerves and hippocampal mossy fiber projections. Neuron 25, 43-56. Chen, H., Chedotal, A., He, Z., Goodman, C. S., and Tessier-Lavigne, M. (1997). Neuropilin2, a novel member of the neuropilin family, is a high affinity receptor for the semaphorins Sema E and Sema IV but not Sema III. Neuron 19, 547-559. Cheng, H. J., Bagri, A., Yaron, A., Stein, E., Pleasure, S. J., and Tessier-Lavigne, M. (2001). Plexin-A3 mediates semaphorin signaling and regulates the development of hippocampal axonal projections. Neuron 32, 249-263. Christensen, J. G., Gonzales, A. J., Cattley, R. C., and Goldsworthy, T. L. (1998). Regulation of apoptosis in mouse hepatocytes and alteration of apoptosis by nongenotoxic carcinogens. Cell Growth Differ 9, 815-825. Cloutier, J. F., Sahay, A., Chang, E. C., Tessier-Lavigne, M., Dulac, C., Kolodkin, A. L., and Ginty, D. D. (2004). Differential requirements for semaphorin 3F and Slit-1 in axonal targeting, fasciculation, and segregation of olfactory sensory neuron projections. J Neurosci 24, 9087-9096. Colamarino, S. A., and Tessier-Lavigne, M. (1995a). The axonal chemoattractant netrin-1 is also a chemorepellent for trochlear motor axons. Cell 81, 621-629. Colamarino, S. A., and Tessier-Lavigne, M. (1995b). The role of the floor plate in axon guidance. Annu Rev Neurosci 18, 497-529. Comeau, M. R., Johnson, R., DuBose, R. F., Petersen, M., Gearing, P., VandenBos, T., Park, L., Farrah, T., Buller, R. M., Cohen, J. I., et al. (1998). A poxvirus-encoded semaphorin induces cytokine production from monocytes and binds to a novel cellular semaphorin receptor, VESPR. Immunity 8, 473-482. Culotti, J. G., and Merz, D. C. (1998). DCC and netrins. Curr Opin Cell Biol 10, 609-613. References 105 Davies, J. A., Cook, G. M., Stern, C. D., and Keynes, R. J. (1990). Isolation from chick somites of a glycoprotein fraction that causes collapse of dorsal root ganglion growth cones. Neuron 4, 11-20. Davis, B. M., Frank, E., Johnson, F. A., and Scott, S. A. (1989). Development of central projections of lumbosacral sensory neurons in the chick. J Comp Neurol 279, 556-566. Debby-Brafman, A., Burstyn-Cohen, T., Klar, A., and Kalcheim, C. (1999). F-Spondin, expressed in somite regions avoided by neural crest cells, mediates inhibition of distinct somite domains to neural crest migration. Neuron 22, 475-488. Delaire, S., Elhabazi, A., Bensussan, A., and Boumsell, L. (1998). CD100 is a leukocyte semaphorin. Cell Mol Life Sci 54, 1265-1276. deLapeyriere, O., and Henderson, C. E. (1997). Motoneuron differentiation, survival and synaptogenesis. Curr Opin Genet Dev 7, 642-650. Dent, E. W., Barnes, A. M., Tang, F., and Kalil, K. (2004). Netrin-1 and semaphorin 3A promote or inhibit cortical axon branching, respectively, by reorganization of the cytoskeleton. J Neurosci 24, 3002-3012. Dickson, B. J. (2002). Molecular mechanisms of axon guidance. Science 298, 1959-1964. Dorsky, R. I., Moon, R. T., and Raible, D. W. (1998). Control of neural crest cell fate by the Wnt signalling pathway. Nature 396, 370-373. Durston, A. J., Timmermans, J. P., Hage, W. J., Hendriks, H. F., de Vries, N. J., Heideveld, M., and Nieuwkoop, P. D. (1989). Retinoic acid causes an anteroposterior transformation in the developing central nervous system. Nature 340, 140-144. Eckhardt, F., Behar, O., Calautti, E., Yonezawa, K., Nishimoto, I., and Fishman, M. C. (1997). A novel transmembrane semaphorin can bind c-src. Mol Cell Neurosci 9, 409-419. Eickholt, B. J., Mackenzie, S. L., Graham, A., Walsh, F. S., and Doherty, P. (1999). Evidence for collapsin-1 functioning in the control of neural crest migration in both trunk and hindbrain regions. Development 126, 2181-2189. Ericson, J., Briscoe, J., Rashbass, P., van Heyningen, V., and Jessell, T. M. (1997). Graded sonic hedgehog signaling and the specification of cell fate in the ventral neural tube. Cold Spring Harb Symp Quant Biol 62, 451-466. Erskine, L., Williams, S. E., Brose, K., Kidd, T., Rachel, R. A., Goodman, C. S., TessierLavigne, M., and Mason, C. A. (2000). Retinal ganglion cell axon guidance in the mouse optic chiasm: expression and function of robos and slits. J Neurosci 20, 4975-4982. Faissner, A., and Kruse, J. (1990). J1/tenascin is a repulsive substrate for central nervous system neurons. Neuron 5, 627-637. Farinas, I., Cano-Jaimez, M., Bellmunt, E., and Soriano, M. (2002). Regulation of neurogenesis by neurotrophins in developing spinal sensory ganglia. Brain Res Bull 57, 809816. Farinas, I., Yoshida, C. K., Backus, C., and Reichardt, L. F. (1996). Lack of neurotrophin-3 results in death of spinal sensory neurons and premature differentiation of their precursors. Neuron 17, 1065-1078. Feiner, L., Webber, A. L., Brown, C. B., Lu, M. M., Jia, L., Feinstein, P., Mombaerts, P., Epstein, J. A., and Raper, J. A. (2001). Targeted disruption of semaphorin 3C leads to persistent truncus arteriosus and aortic arch interruption. Development 128, 3061-3070. 106 References Fiore, R., and Puschel, A. W. (2003). The function of semaphorins during nervous system development. Front Biosci 8, s484-499. Fitzgerald, M., Kwiat, G. C., Middleton, J., and Pini, A. (1993). Ventral spinal cord inhibition of neurite outgrowth from embryonic rat dorsal root ganglia. Development 117, 1377-1384. Flanagan, J. G., and Vanderhaeghen, P. (1998). The ephrins and Eph receptors in neural development. Annu Rev Neurosci 21, 309-345. Fujisawa, H. (2004). Discovery of semaphorin receptors, neuropilin and plexin, and their functions in neural development. J Neurobiol 59, 24-33. Gagliardini, V., and Fankhauser, C. (1999). Semaphorin III can induce death in sensory neurons. Mol Cell Neurosci 14, 301-316. Garcia-Castro, M. I., Marcelle, C., and Bronner-Fraser, M. (2002). Ectodermal Wnt function as a neural crest inducer. Science 297, 848-851. Gesemann, M., Litwack, E. D., Yee, K. T., Christen, U., and O'Leary, D. D. (2001). Identification of candidate genes for controlling development of the basilar pons by differential display PCR. Mol Cell Neurosci 18, 1-12. Giger, R. J., Cloutier, J. F., Sahay, A., Prinjha, R. K., Levengood, D. V., Moore, S. E., Pickering, S., Simmons, D., Rastan, S., Walsh, F. S., et al. (2000). Neuropilin-2 is required in vivo for selective axon guidance responses to secreted semaphorins. Neuron 25, 29-41. Giger, R. J., Wolfer, D. P., De Wit, G. M., and Verhaagen, J. (1996). Anatomy of rat semaphorin III/collapsin-1 mRNA expression and relationship to developing nerve tracts during neuroembryogenesis. J Comp Neurol 375, 378-392. Gilgenkrantz, S. (2004). [Bird sex determination]. Med Sci (Paris) 20, 1004-1008. Gitler, A. D., Lu, M. M., and Epstein, J. A. (2004). PlexinD1 and semaphorin signaling are required in endothelial cells for cardiovascular development. Dev Cell 7, 107-116. Godsave, S. F., and Slack, J. M. (1989). Clonal analysis of mesoderm induction in Xenopus laevis. Dev Biol 134, 486-490. Goldberg, J. L., Vargas, M. E., Wang, J. T., Mandemakers, W., Oster, S. F., Sretavan, D. W., and Barres, B. A. (2004). An oligodendrocyte lineage-specific semaphorin, Sema5A, inhibits axon growth by retinal ganglion cells. J Neurosci 24, 4989-4999. Golding, J. P., and Cohen, J. (1997). Border Controls at the Mammalian Spinal Cord: LateSurviving Neural Crest Boundary Cap Cells at Dorsal Root Entry Sites May Regulate Sensory Afferent Ingrowth and Entry Zone Morphogenesis. Mol Cell Neurosci 9, 381-396. Gross, M. K., Dottori, M., and Goulding, M. (2002). Lbx1 specifies somatosensory association interneurons in the dorsal spinal cord. Neuron 34, 535-549. Gu, C., Yoshida, Y., Livet, J., Reimert, D. V., Mann, F., Merte, J., Henderson, C. E., Jessell, T. M., Kolodkin, A. L., and Ginty, D. D. (2004). Semaphorin 3E and Plexin-D1 Control Vascular Pattern Independently of Neuropilins. Science. Guthrie, S., and Pini, A. (1995). Chemorepulsion of developing motor axons by the floor plate. Neuron 14, 1117-1130. Hagedorn, M., Balke, M., Schmidt, A., Bloch, W., Kurz, H., Javerzat, S., Rousseau, B., Wilting, J., and Bikfalvi, A. (2004). VEGF coordinates interaction of pericytes and endothelial cells during vasculogenesis and experimental angiogenesis. Dev Dyn 230, 23-33. References 107 Hall, K. T., Boumsell, L., Schultze, J. L., Boussiotis, V. A., Dorfman, D. M., Cardoso, A. A., Bensussan, A., Nadler, L. M., and Freeman, G. J. (1996). Human CD100, a novel leukocyte semaphorin that promotes B-cell aggregation and differentiation. Proc Natl Acad Sci U S A 93, 11780-11785. Hatten, M. E. (1999). Central nervous system neuronal migration. Annu Rev Neurosci 22, 511-539. He, Z., and Tessier-Lavigne, M. (1997). Neuropilin is a receptor for the axonal chemorepellent Semaphorin III. Cell 90, 739-751. Hedgecock, E. M., Culotti, J. G., and Hall, D. H. (1990). The unc-5, unc-6, and unc-40 genes guide circumferential migrations of pioneer axons and mesodermal cells on the epidermis in C. elegans. Neuron 4, 61-85. Helmbacher, F., Schneider-Maunoury, S., Topilko, P., Tiret, L., and Charnay, P. (2000). Targeting of the EphA4 tyrosine kinase receptor affects dorsal/ventral pathfinding of limb motor axons. Development 127, 3313-3324. Helms, A. W., and Johnson, J. E. (2003). Specification of dorsal spinal cord interneurons. Curr Opin Neurobiol 13, 42-49. Hemmati-Brivanlou, A., and Melton, D. A. (1994). Inhibition of activin receptor signaling promotes neuralization in Xenopus. Cell 77, 273-281. Hill, J., Clarke, J. D., Vargesson, N., Jowett, T., and Holder, N. (1995). Exogenous retinoic acid causes specific alterations in the development of the midbrain and hindbrain of the zebrafish embryo including positional respecification of the Mauthner neuron. Mech Dev 50, 3-16. Hollyday, M. (2001). Neurogenesis in the vertebrate neural tube. Int J Dev Neurosci 19, 161173. Honig, M. G. (1982). The development of sensory projection patterns in embryonic chick hind limb. J Physiol 330, 175-202. Huot, J. (2004). Ephrin signaling in axon guidance. Prog Neuropsychopharmacol Biol Psychiatry 28, 813-818. Jessell, T. M. (2000). Neuronal specification in the spinal cord: inductive signals and transcriptional codes. Nat Rev Genet 1, 20-29. Johansen, J., and Johansen, K. M. (1997). Molecular mechanisms mediating axon pathway formation. Crit Rev Eukaryot Gene Expr 7, 95-116. Kameyama, T., Murakami, Y., Suto, F., Kawakami, A., Takagi, S., Hirata, T., and Fujisawa, H. (1996a). Identification of a neuronal cell surface molecule, plexin, in mice. Biochem Biophys Res Commun 226, 524-529. Kameyama, T., Murakami, Y., Suto, F., Kawakami, A., Takagi, S., Hirata, T., and Fujisawa, H. (1996b). Identification of plexin family molecules in mice. Biochem Biophys Res Commun 226, 396-402. Kantor, D. B., Chivatakarn, O., Peer, K. L., Oster, S. F., Inatani, M., Hansen, M. J., Flanagan, J. G., Yamaguchi, Y., Sretavan, D. W., Giger, R. J., and Kolodkin, A. L. (2004). Semaphorin 5A is a bifunctional axon guidance cue regulated by heparan and chondroitin sulfate proteoglycans. Neuron 44, 961-975. 108 References Kaprielian, Z., Runko, E., and Imondi, R. (2001). Axon guidance at the midline choice point. Dev Dyn 221, 154-181. Kawakami, A., Kitsukawa, T., Takagi, S., and Fujisawa, H. (1996). Developmentally regulated expression of a cell surface protein, neuropilin, in the mouse nervous system. J Neurobiol 29, 1-17. Kawasaki, T., Bekku, Y., Suto, F., Kitsukawa, T., Taniguchi, M., Nagatsu, I., Nagatsu, T., Itoh, K., Yagi, T., and Fujisawa, H. (2002). Requirement of neuropilin 1-mediated Sema3A signals in patterning of the sympathetic nervous system. Development 129, 671-680. Kawasaki, T., Kitsukawa, T., Bekku, Y., Matsuda, Y., Sanbo, M., Yagi, T., and Fujisawa, H. (1999). A requirement for neuropilin-1 in embryonic vessel formation. Development 126, 4895-4902. Keleman, K., and Dickson, B. J. (2001). Short- and long-range repulsion by the Drosophila Unc5 netrin receptor. Neuron 32, 605-617. Keynes, R. J., and Stern, C. D. (1984). Segmentation in the vertebrate nervous system. Nature 310, 786-789. Kidd, T., Brose, K., Mitchell, K. J., Fetter, R. D., Tessier-Lavigne, M., Goodman, C. S., and Tear, G. (1998). Roundabout controls axon crossing of the CNS midline and defines a novel subfamily of evolutionarily conserved guidance receptors. Cell 92, 205-215. Kikuchi, K., Chedotal, A., Hanafusa, H., Ujimasa, Y., de Castro, F., Goodman, C. S., and Kimura, T. (1999). Cloning and characterization of a novel class VI semaphorin, semaphorin Y. Mol Cell Neurosci 13, 9-23. Kikuchi, K., Ishida, H., and Kimura, T. (1997). Molecular cloning of a novel member of semaphorin family genes, semaphorin Z. Brain Res Mol Brain Res 51, 229-237. Kitsukawa, T., Shimizu, M., Sanbo, M., Hirata, T., Taniguchi, M., Bekku, Y., Yagi, T., and Fujisawa, H. (1997). Neuropilin-semaphorin III/D-mediated chemorepulsive signals play a crucial role in peripheral nerve projection in mice. Neuron 19, 995-1005. Klostermann, A., Lohrum, M., Adams, R. H., and Puschel, A. W. (1998). The chemorepulsive activity of the axonal guidance signal semaphorin D requires dimerization. J Biol Chem 273, 7326-7331. Klostermann, A., Lutz, B., Gertler, F., and Behl, C. (2000). The orthologous human and murine semaphorin 6A-1 proteins (SEMA6A-1/Sema6A-1) bind to the enabled/vasodilatorstimulated phosphoprotein-like protein (EVL) via a novel carboxyl-terminal zyxin-like domain. J Biol Chem 275, 39647-39653. Knecht, A. K., and Bronner-Fraser, M. (2002). Induction of the neural crest: a multigene process. Nat Rev Genet 3, 453-461. Kobayashi, H., Koppel, A. M., Luo, Y., and Raper, J. A. (1997). A role for collapsin-1 in olfactory and cranial sensory axon guidance. J Neurosci 17, 8339-8352. Kolodkin, A. L., Levengood, D. V., Rowe, E. G., Tai, Y. T., Giger, R. J., and Ginty, D. D. (1997). Neuropilin is a semaphorin III receptor. Cell 90, 753-762. Kolodkin, A. L., Matthes, D. J., and Goodman, C. S. (1993). The semaphorin genes encode a family of transmembrane and secreted growth cone guidance molecules. Cell 75, 1389-1399. References 109 Kolodkin, A. L., Matthes, D. J., O'Connor, T. P., Patel, N. H., Admon, A., Bentley, D., and Goodman, C. S. (1992). Fasciclin IV: sequence, expression, and function during growth cone guidance in the grasshopper embryo. Neuron 9, 831-845. Koppel, A. M., Feiner, L., Kobayashi, H., and Raper, J. A. (1997). A 70 amino acid region within the semaphorin domain activates specific cellular response of semaphorin family members. Neuron 19, 531-537. Koppel, A. M., and Raper, J. A. (1998). Collapsin-1 covalently dimerizes, and dimerization is necessary for collapsing activity. J Biol Chem 273, 15708-15713. Krull, C. E. (2001). Segmental organization of neural crest migration. Mech Dev 105, 37-45. Krull, C. E. (2004). A primer on using in ovo electroporation to analyze gene function. Dev Dyn 229, 433-439. Krull, C. E., and Koblar, S. A. (2000). Motor axon pathfinding in the peripheral nervous system. Brain Res Bull 53, 479-487. Lamb, T. M., and Harland, R. M. (1995). Fibroblast growth factor is a direct neural inducer, which combined with noggin generates anterior-posterior neural pattern. Development 121, 3627-3636. Lamb, T. M., Knecht, A. K., Smith, W. C., Stachel, S. E., Economides, A. N., Stahl, N., Yancopolous, G. D., and Harland, R. M. (1993). Neural induction by the secreted polypeptide noggin. Science 262, 713-718. Lance-Jones, C., and Landmesser, L. (1981). Pathway selection by chick lumbosacral motoneurons during normal development. Proc R Soc Lond B Biol Sci 214, 1-18. Landmesser, L. (1978a). The development of motor projection patterns in the chick hind limb. J Physiol 284, 391-414. Landmesser, L. (1978b). The distribution of motoneurones supplying chick hind limb muscles. J Physiol 284, 371-389. Landmesser, L. T. (2001). The acquisition of motoneuron subtype identity and motor circuit formation. Int J Dev Neurosci 19, 175-182. Laskowski, M. B., and Sanes, J. R. (1987). Topographic mapping of motor pools onto skeletal muscles. J Neurosci 7, 252-260. Le Douarin, N. M., and Dupin, E. (2003). Multipotentiality of the neural crest. Curr Opin Genet Dev 13, 529-536. Lee, K. J., Dietrich, P., and Jessell, T. M. (2000). Genetic ablation reveals that the roof plate is essential for dorsal interneuron specification. Nature 403, 734-740. Leighton, P. A., Mitchell, K. J., Goodrich, L. V., Lu, X., Pinson, K., Scherz, P., Skarnes, W. C., and Tessier-Lavigne, M. (2001). Defining brain wiring patterns and mechanisms through gene trapping in mice. Nature 410, 174-179. Levine, A. J., Munoz-Sanjuan, I., Bell, E., North, A. J., and Brivanlou, A. H. (2003). Fluorescent labeling of endothelial cells allows in vivo, continuous characterization of the vascular development of Xenopus laevis. Dev Biol 254, 50-67. Liem, K. F., Jr., Tremml, G., Roelink, H., and Jessell, T. M. (1995). Dorsal differentiation of neural plate cells induced by BMP-mediated signals from epidermal ectoderm. Cell 82, 969979. 110 References Lin, J. H., Saito, T., Anderson, D. J., Lance-Jones, C., Jessell, T. M., and Arber, S. (1998). Functionally related motor neuron pool and muscle sensory afferent subtypes defined by coordinate ETS gene expression. Cell 95, 393-407. Lumsden, A., and Krumlauf, R. (1996). Patterning the vertebrate neuraxis. Science 274, 11091115. Lumsden, A. G., and Davies, A. M. (1983). Earliest sensory nerve fibres are guided to peripheral targets by attractants other than nerve growth factor. Nature 306, 786-788. Luo, Y., Raible, D., and Raper, J. A. (1993). Collapsin: a protein in brain that induces the collapse and paralysis of neuronal growth cones. Cell 75, 217-227. Luo, Y., Shepherd, I., Li, J., Renzi, M. J., Chang, S., and Raper, J. A. (1995). A family of molecules related to collapsin in the embryonic chick nervous system. Neuron 14, 1131-1140. Ma, Q., Fode, C., Guillemot, F., and Anderson, D. J. (1999). Neurogenin1 and neurogenin2 control two distinct waves of neurogenesis in developing dorsal root ganglia. Genes Dev 13, 1717-1728. Marin, O., and Rubenstein, J. L. (2003). Cell migration in the forebrain. Annu Rev Neurosci 26, 441-483. Masuda, K., Furuyama, T., Takahara, M., Fujioka, S., Kurinami, H., and Inagaki, S. (2004). Sema4D stimulates axonal outgrowth of embryonic DRG sensory neurones. Genes Cells 9, 821-829. Masuda, T., Tsuji, H., Taniguchi, M., Yagi, T., Tessier-Lavigne, M., Fujisawa, H., Okado, N., and Shiga, T. (2003). Differential non-target-derived repulsive signals play a critical role in shaping initial axonal growth of dorsal root ganglion neurons. Dev Biol 254, 289-302. Matthes, D. J., Sink, H., Kolodkin, A. L., and Goodman, C. S. (1995). Semaphorin II can function as a selective inhibitor of specific synaptic arborizations. Cell 81, 631-639. Matzuk, M. M., Lu, N., Vogel, H., Sellheyer, K., Roop, D. R., and Bradley, A. (1995). Multiple defects and perinatal death in mice deficient in follistatin. Nature 374, 360-363. Mendelson, B., Koerber, H. R., and Frank, E. (1992). Development of cutaneous and proprioceptive afferent projections in the chick spinal cord. Neurosci Lett 138, 72-76. Messersmith, E. K., Leonardo, E. D., Shatz, C. J., Tessier-Lavigne, M., Goodman, C. S., and Kolodkin, A. L. (1995). Semaphorin III can function as a selective chemorepellent to pattern sensory projections in the spinal cord. Neuron 14, 949-959. Miller, R. H. (2002). Regulation of oligodendrocyte development in the vertebrate CNS. Prog Neurobiol 67, 451-467. Ming, G. L., Song, H. J., Berninger, B., Holt, C. E., Tessier-Lavigne, M., and Poo, M. M. (1997). cAMP-dependent growth cone guidance by netrin-1. Neuron 19, 1225-1235. Morrison, S. J., Perez, S. E., Qiao, Z., Verdi, J. M., Hicks, C., Weinmaster, G., and Anderson, D. J. (2000). Transient Notch activation initiates an irreversible switch from neurogenesis to gliogenesis by neural crest stem cells. Cell 101, 499-510. Muller, T., Brohmann, H., Pierani, A., Heppenstall, P. A., Lewin, G. R., Jessell, T. M., and Birchmeier, C. (2002). The homeodomain factor lbx1 distinguishes two major programs of neuronal differentiation in the dorsal spinal cord. Neuron 34, 551-562. Murai, K. K., and Pasquale, E. B. (2003). 'Eph'ective signaling: forward, reverse and crosstalk. J Cell Sci 116, 2823-2832. References 111 Murakami, Y., Suto, F., Shimizu, M., Shinoda, T., Kameyama, T., and Fujisawa, H. (2001). Differential expression of plexin-A subfamily members in the mouse nervous system. Dev Dyn 220, 246-258. Nakamura, F., Tanaka, M., Takahashi, T., Kalb, R. G., and Strittmatter, S. M. (1998). Neuropilin-1 extracellular domains mediate semaphorin D/III-induced growth cone collapse. Neuron 21, 1093-1100. Nguyen, V. H., Trout, J., Connors, S. A., Andermann, P., Weinberg, E., and Mullins, M. C. (2000). Dorsal and intermediate neuronal cell types of the spinal cord are established by a BMP signaling pathway. Development 127, 1209-1220. Niederlander, C., and Lumsden, A. (1996). Late emigrating neural crest cells migrate specifically to the exit points of cranial branchiomotor nerves. Development 122, 2367-2374. Nieto, M. A., Sargent, M. G., Wilkinson, D. G., and Cooke, J. (1994). Control of cell behavior during vertebrate development by Slug, a zinc finger gene. Science 264, 835-839. Oakley, R. A., and Tosney, K. W. (1991). Peanut agglutinin and chondroitin-6-sulfate are molecular markers for tissues that act as barriers to axon advance in the avian embryo. Dev Biol 147, 187-206. O'Connor, D. B. a. T. P. (1992). The nerve growth cone, Eds (Raven, New York)). Ohta, K., Mizutani, A., Kawakami, A., Murakami, Y., Kasuya, Y., Takagi, S., Tanaka, H., and Fujisawa, H. (1995). Plexin: a novel neuronal cell surface molecule that mediates cell adhesion via a homophilic binding mechanism in the presence of calcium ions. Neuron 14, 1189-1199. Oster, S. F., Bodeker, M. O., He, F., and Sretavan, D. W. (2003). Invariant Sema5A inhibition serves an ensheathing function during optic nerve development. Development 130, 775-784. Ozaki, S., and Snider, W. D. (1997). Initial trajectories of sensory axons toward laminar targets in the developing mouse spinal cord. J Comp Neurol 380, 215-229. Panchision, D. M., Pickel, J. M., Studer, L., Lee, S. H., Turner, P. A., Hazel, T. G., and McKay, R. D. (2001). Sequential actions of BMP receptors control neural precursor cell production and fate. Genes Dev 15, 2094-2110. Park, H. T., Wu, J., and Rao, Y. (2002). Molecular control of neuronal migration. Bioessays 24, 821-827. Park, K. W., Crouse, D., Lee, M., Karnik, S. K., Sorensen, L. K., Murphy, K. J., Kuo, C. J., and Li, D. Y. (2004). The axonal attractant Netrin-1 is an angiogenic factor. Proc Natl Acad Sci U S A 101, 16210-16215. Pasterkamp, R. J., Peschon, J. J., Spriggs, M. K., and Kolodkin, A. L. (2003). Semaphorin 7A promotes axon outgrowth through integrins and MAPKs. Nature 424, 398-405. Patel, T. D., Jackman, A., Rice, F. L., Kucera, J., and Snider, W. D. (2000). Development of sensory neurons in the absence of NGF/TrkA signaling in vivo. Neuron 25, 345-357. Patel, T. D., Kramer, I., Kucera, J., Niederkofler, V., Jessell, T. M., Arber, S., and Snider, W. D. (2003). Peripheral NT3 signaling is required for ETS protein expression and central patterning of proprioceptive sensory afferents. Neuron 38, 403-416. Patten, I., Kulesa, P., Shen, M. M., Fraser, S., and Placzek, M. (2003). Distinct modes of floor plate induction in the chick embryo. Development 130, 4809-4821. 112 References Pekarik, V., Bourikas, D., Miglino, N., Joset, P., Preiswerk, S., and Stoeckli, E. T. (2003). Screening for gene function in chicken embryo using RNAi and electroporation. Nat Biotechnol 21, 93-96. Perrin, F. E., Rathjen, F. G., and Stoeckli, E. T. (2001). Distinct subpopulations of sensory afferents require F11 or axonin-1 for growth to their target layers within the spinal cord of the chick. Neuron 30, 707-723. Perrin, F. E., and Stoeckli, E. T. (2000). Use of lipophilic dyes in studies of axonal pathfinding in vivo. Microsc Res Tech 48, 25-31. Perris, R., and Perissinotto, D. (2000). Role of the extracellular matrix during neural crest cell migration. Mech Dev 95, 3-21. Pierani, A., Brenner-Morton, S., Chiang, C., and Jessell, T. M. (1999). A sonic hedgehogindependent, retinoid-activated pathway of neurogenesis in the ventral spinal cord. Cell 97, 903-915. Pituello, F. (1997). Neuronal specification: generating diversity in the spinal cord. Curr Biol 7, R701-704. Plump, A. S., Erskine, L., Sabatier, C., Brose, K., Epstein, C. J., Goodman, C. S., Mason, C. A., and Tessier-Lavigne, M. (2002). Slit1 and Slit2 cooperate to prevent premature midline crossing of retinal axons in the mouse visual system. Neuron 33, 219-232. Poh, A., Karunaratne, A., Kolle, G., Huang, N., Smith, E., Starkey, J., Wen, D., Wilson, I., Yamada, T., and Hargrave, M. (2002). Patterning of the vertebrate ventral spinal cord. Int J Dev Biol 46, 597-608. Puschel, A. W. (2002). The function of neuropilin/plexin complexes. Adv Exp Med Biol 515, 71-80. Puschel, A. W., Adams, R. H., and Betz, H. (1995). Murine semaphorin D/collapsin is a member of a diverse gene family and creates domains inhibitory for axonal extension. Neuron 14, 941-948. Qu, X., Wei, H., Zhai, Y., Que, H., Chen, Q., Tang, F., Wu, Y., Xing, G., Zhu, Y., Liu, S., et al. (2002). Identification, characterization, and functional study of the two novel human members of the semaphorin gene family. J Biol Chem 277, 35574-35585. Rabacchi, S. A., Solowska, J. M., Kruk, B., Luo, Y., Raper, J. A., and Baird, D. H. (1999). Collapsin-1/semaphorin-III/D is regulated developmentally in Purkinje cells and collapses pontocerebellar mossy fiber neuronal growth cones. J Neurosci 19, 4437-4448. Raper, J. A. (2000). Semaphorins and their receptors in vertebrates and invertebrates. Curr Opin Neurobiol 10, 88-94. Raper, J. A., Bastiani, M. J., and Goodman, C. S. (1984). Pathfinding by neuronal growth cones in grasshopper embryos. IV. The effects of ablating the A and P axons upon the behavior of the G growth cone. J Neurosci 4, 2329-2345. Reissmann, E., Ernsberger, U., Francis-West, P. H., Rueger, D., Brickell, P. M., and Rohrer, H. (1996). Involvement of bone morphogenetic protein-4 and bone morphogenetic protein-7 in the differentiation of the adrenergic phenotype in developing sympathetic neurons. Development 122, 2079-2088. Renzi, M. J., Feiner, L., Koppel, A. M., and Raper, J. A. (1999). A dominant negative receptor for specific secreted semaphorins is generated by deleting an extracellular domain from neuropilin-1. J Neurosci 19, 7870-7880. References 113 Rice, D. S., and Curran, T. (1999). Mutant mice with scrambled brains: understanding the signaling pathways that control cell positioning in the CNS. Genes Dev 13, 2758-2773. Richardson, W. D., Pringle, N. P., Yu, W. P., and Hall, A. C. (1997). Origins of spinal cord oligodendrocytes: possible developmental and evolutionary relationships with motor neurons. Dev Neurosci 19, 58-68. Robinson, V., Smith, A., Flenniken, A. M., and Wilkinson, D. G. (1997). Roles of Eph receptors and ephrins in neural crest pathfinding. Cell Tissue Res 290, 265-274. Rohm, B., Ottemeyer, A., Lohrum, M., and Puschel, A. W. (2000). Plexin/neuropilin complexes mediate repulsion by the axonal guidance signal semaphorin 3A. Mech Dev 93, 95-104. Roos, M., Schachner, M., and Bernhardt, R. R. (1999). Zebrafish semaphorin Z1b inhibits growing motor axons in vivo. Mech Dev 87, 103-117. Sahay, A., Molliver, M. E., Ginty, D. D., and Kolodkin, A. L. (2003). Semaphorin 3F is critical for development of limbic system circuitry and is required in neurons for selective CNS axon guidance events. J Neurosci 23, 6671-6680. Schneider, V. A., and Granato, M. (2003). Motor axon migration: a long way to go. Dev Biol 263, 1-11. Sefton, M., Sanchez, S., and Nieto, M. A. (1998). Conserved and divergent roles for members of the Snail family of transcription factors in the chick and mouse embryo. Development 125, 3111-3121. Sela-Donenfeld, D., and Kalcheim, C. (1999). Regulation of the onset of neural crest migration by coordinated activity of BMP4 and Noggin in the dorsal neural tube. Development 126, 4749-4762. Serini, G., and Bussolino, F. (2004). Common cues in vascular and axon guidance. Physiology (Bethesda) 19, 348-354. Sharma, K., Leonard, A. E., Lettieri, K., and Pfaff, S. L. (2000). Genetic and epigenetic mechanisms contribute to motor neuron pathfinding. Nature 406, 515-519. Sharma, K., Sheng, H. Z., Lettieri, K., Li, H., Karavanov, A., Potter, S., Westphal, H., and Pfaff, S. L. (1998). LIM homeodomain factors Lhx3 and Lhx4 assign subtype identities for motor neurons. Cell 95, 817-828. Shepherd, I., Luo, Y., Raper, J. A., and Chang, S. (1996). The distribution of collapsin-1 mRNA in the developing chick nervous system. Dev Biol 173, 185-199. Shirasaki, R., and Pfaff, S. L. (2002). Transcriptional codes and the control of neuronal identity. Annu Rev Neurosci 25, 251-281. Shirvan, A., Ziv, I., Fleminger, G., Shina, R., He, Z., Brudo, I., Melamed, E., and Barzilai, A. (1999). Semaphorins as mediators of neuronal apoptosis. J Neurochem 73, 961-971. Smith, C. A., and Sinclair, A. H. (2004). Sex determination: insights from the chicken. Bioessays 26, 120-132. Song, H. J., Ming, G. L., and Poo, M. M. (1997). cAMP-induced switching in turning direction of nerve growth cones. Nature 388, 275-279. Steup, A., Ninnemann, O., Savaskan, N. E., Nitsch, R., Puschel, A. W., and Skutella, T. (1999). Semaphorin D acts as a repulsive factor for entorhinal and hippocampal neurons. Eur J Neurosci 11, 729-734. 114 References Stoeckli, E. T., and Landmesser, L. T. (1995). Axonin-1, Nr-CAM, and Ng-CAM play different roles in the in vivo guidance of chick commissural neurons. Neuron 14, 1165-1179. Suto, F., Murakami, Y., Nakamura, F., Goshima, Y., and Fujisawa, H. (2003). Identification and characterization of a novel mouse plexin, plexin-A4. Mech Dev 120, 385-396. Takahashi, T., Fournier, A., Nakamura, F., Wang, L. H., Murakami, Y., Kalb, R. G., Fujisawa, H., and Strittmatter, S. M. (1999). Plexin-neuropilin-1 complexes form functional semaphorin-3A receptors. Cell 99, 59-69. Takahashi, T., Nakamura, F., Jin, Z., Kalb, R. G., and Strittmatter, S. M. (1998). Semaphorins A and E act as antagonists of neuropilin-1 and agonists of neuropilin-2 receptors. Nat Neurosci 1, 487-493. Tamagnone, L., Artigiani, S., Chen, H., He, Z., Ming, G. I., Song, H., Chedotal, A., Winberg, M. L., Goodman, C. S., Poo, M., et al. (1999). Plexins are a large family of receptors for transmembrane, secreted, and GPI-anchored semaphorins in vertebrates. Cell 99, 71-80. Tamagnone, L., and Comoglio, P. M. (2000). Signalling by semaphorin receptors: cell guidance and beyond. Trends Cell Biol 10, 377-383. Tamagnone, L., and Comoglio, P. M. (2004). To move or not to move? Semaphorin signalling in cell migration. EMBO Rep 5, 356-361. Tanabe, Y., and Jessell, T. M. (1996). Diversity and pattern in the developing spinal cord. Science 274, 1115-1123. Taniguchi, M., Yuasa, S., Fujisawa, H., Naruse, I., Saga, S., Mishina, M., and Yagi, T. (1997). Disruption of semaphorin III/D gene causes severe abnormality in peripheral nerve projection. Neuron 19, 519-530. Taylor, M. S., and Semple, C. A. (2002). Sushi gets serious: the draft genome sequence of the pufferfish Fugu rubripes. Genome Biol 3, reviews1025. Tessier-Lavigne, M., and Goodman, C. S. (1996). The molecular biology of axon guidance. Science 274, 1123-1133. Torres-Vazquez, J., Gitler, A. D., Fraser, S. D., Berk, J. D., Van, N. P., Fishman, M. C., Childs, S., Epstein, J. A., and Weinstein, B. M. (2004). Semaphorin-plexin signaling guides patterning of the developing vasculature. Dev Cell 7, 117-123. Tosney, K. W., and Landmesser, L. T. (1984). Pattern and specificity of axonal outgrowth following varying degrees of chick limb bud ablation. J Neurosci 4, 2518-2527. Tosney, K. W., and Landmesser, L. T. (1985a). Development of the major pathways for neurite outgrowth in the chick hindlimb. Dev Biol 109, 193-214. Tosney, K. W., and Landmesser, L. T. (1985b). Growth cone morphology and trajectory in the lumbosacral region of the chick embryo. J Neurosci 5, 2345-2358. Tosney, K. W., and Landmesser, L. T. (1985c). Specificity of early motoneuron growth cone outgrowth in the chick embryo. J Neurosci 5, 2336-2344. Tosney, K. W., and Oakley, R. A. (1990). The perinotochordal mesenchyme acts as a barrier to axon advance in the chick embryo: implications for a general mechanism of axonal guidance. Exp Neurol 109, 75-89. Toyofuku, T., Zhang, H., Kumanogoh, A., Takegahara, N., Suto, F., Kamei, J., Aoki, K., Yabuki, M., Hori, M., Fujisawa, H., and Kikutani, H. (2004a). Dual roles of Sema6D in References 115 cardiac morphogenesis through region-specific association of its receptor, Plexin-A1, with off-track and vascular endothelial growth factor receptor type 2. Genes Dev 18, 435-447. Toyofuku, T., Zhang, H., Kumanogoh, A., Takegahara, N., Yabuki, M., Harada, K., Hori, M., and Kikutani, H. (2004b). Guidance of myocardial patterning in cardiac development by Sema6D reverse signalling. Nat Cell Biol 6, 1204-1211. Tsuchida, T., Ensini, M., Morton, S. B., Baldassare, M., Edlund, T., Jessell, T. M., and Pfaff, S. L. (1994). Topographic organization of embryonic motor neurons defined by expression of LIM homeobox genes. Cell 79, 957-970. van der Zwaag, B., Hellemons, A. J., Leenders, W. P., Burbach, J. P., Brunner, H. G., Padberg, G. W., and Van Bokhoven, H. (2002). PLEXIN-D1, a novel plexin family member, is expressed in vascular endothelium and the central nervous system during mouse embryogenesis. Dev Dyn 225, 336-343. Varela-Echavarria, A., Tucker, A., Puschel, A. W., and Guthrie, S. (1997). Motor axon subpopulations respond differentially to the chemorepellents netrin-1 and semaphorin D. Neuron 18, 193-207. Venkatesh, B., Gilligan, P., and Brenner, S. (2000). Fugu: a compact vertebrate reference genome. FEBS Lett 476, 3-7. Vermeren, M., Maro, G. S., Bron, R., McGonnell, I. M., Charnay, P., Topilko, P., and Cohen, J. (2003). Integrity of developing spinal motor columns is regulated by neural crest derivatives at motor exit points. Neuron 37, 403-415. Vinogradov, A. E. (1999). Intron-genome size relationship on a large evolutionary scale. J Mol Evol 49, 376-384. Wang, K. H., Brose, K., Arnott, D., Kidd, T., Goodman, C. S., Henzel, W., and TessierLavigne, M. (1999). Biochemical purification of a mammalian slit protein as a positive regulator of sensory axon elongation and branching. Cell 96, 771-784. Wijgerde, M., McMahon, J. A., Rule, M., and McMahon, A. P. (2002). A direct requirement for Hedgehog signaling for normal specification of all ventral progenitor domains in the presumptive mammalian spinal cord. Genes Dev 16, 2849-2864. Wilkinson, D. G. (2001). Multiple roles of EPH receptors and ephrins in neural development. Nat Rev Neurosci 2, 155-164. Winberg, M. L., Mitchell, K. J., and Goodman, C. S. (1998a). Genetic analysis of the mechanisms controlling target selection: complementary and combinatorial functions of netrins, semaphorins, and IgCAMs. Cell 93, 581-591. Winberg, M. L., Noordermeer, J. N., Tamagnone, L., Comoglio, P. M., Spriggs, M. K., Tessier-Lavigne, M., and Goodman, C. S. (1998b). Plexin A is a neuronal semaphorin receptor that controls axon guidance. Cell 95, 903-916. Winnier, G., Blessing, M., Labosky, P. A., and Hogan, B. L. (1995). Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev 9, 2105-2116. Wodarz, A., and Nusse, R. (1998). Mechanisms of Wnt signaling in development. Annu Rev Cell Dev Biol 14, 59-88. Wolman, M. A., Liu, Y., Tawarayama, H., Shoji, W., and Halloran, M. C. (2004). Repulsion and attraction of axons by Semaphorin3D are mediated by different neuropilins in vivo. J Neurosci 24, 8428-8435. 116 References Worzfeld, T., Puschel, A. W., Offermanns, S., and Kuner, R. (2004). Plexin-B family members demonstrate non-redundant expression patterns in the developing mouse nervous system: an anatomical basis for morphogenetic effects of Sema4D during development. Eur J Neurosci 19, 2622-2632. Wright, D. E., White, F. A., Gerfen, R. W., Silos-Santiago, I., and Snider, W. D. (1995). The guidance molecule semaphorin III is expressed in regions of spinal cord and periphery avoided by growing sensory axons. J Comp Neurol 361, 321-333. Xiao, T., Shoji, W., Zhou, W., Su, F., and Kuwada, J. Y. (2003). Transmembrane sema4E guides branchiomotor axons to their targets in zebrafish. J Neurosci 23, 4190-4198. Yaginuma, H., Shiga, T., Homma, S., Ishihara, R., and Oppenheim, R. W. (1990). Identification of early developing axon projections from spinal interneurons in the chick embryo with a neuron specific beta-tubulin antibody: evidence for a new 'pioneer' pathway in the spinal cord. Development 108, 705-716. Yaginuma, H., Shiga, T., and Oppenheim, R. W. (1994). Early developmental patterns and mechanisms of axonal guidance of spinal interneurons in the chick embryo spinal cord. Prog Neurobiol 44, 249-278. Yamada, T., Endo, R., Gotoh, M., and Hirohashi, S. (1997). Identification of semaphorin E as a non-MDR drug resistance gene of human cancers. Proc Natl Acad Sci U S A 94, 1471314718. Yee, K. T., Simon, H. H., Tessier-Lavigne, M., and 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. Young, H. M., Anderson, R. B., and Anderson, C. R. (2004). Guidance cues involved in the development of the peripheral autonomic nervous system. Auton Neurosci 112, 1-14. Yu, H. H., Araj, H. H., Ralls, S. A., and Kolodkin, A. L. (1998). The transmembrane Semaphorin Sema I is required in Drosophila for embryonic motor and CNS axon guidance. Neuron 20, 207-220. Zhou, L., White, F. A., Lentz, S. I., Wright, D. E., Fisher, D. A., and Snider, W. D. (1997). Cloning and expression of a novel murine semaphorin with structural similarity to insect semaphorin I. Mol Cell Neurosci 9, 26-41. Zou, Y. (2004). Wnt signaling in axon guidance. Trends Neurosci 27, 528-532. List of Publications P1 J. Gemayel, A. Geloen, F. Mion, “Propofol-induced cytochrome P450 inhibition: an in vitro and in vivo study in rats“, Life Science, 68, 2957–65, 2001. P2 O. Matui, J. Gemayel, R. Sadhu, M. Gesemann, E. Stoeckli, „Expression patterns of plexins and neuropilins suggest cooperative separate functions in spinal cord development”. Submitted in June 2005 to Developmental Dynamics. P3 J. Gemayel, R. Sadhu, O. Matui, E. Stoeckli, M. Gesemann, „Developmental regulation of semaphorins in spinal cord and peripheral nervous system suggests additional semaphorin functions and interactions in chicken embryo. In preparation (will be submitted in July 2005 to Developmental Dynamics). P4 J. Gemayel, R. Sadhu, R. Babey, E. Stoeckli, M. Gesemann, „Functional knock down of Plexin D1 in chicken results in misguided motor axons and alterations in dorsal and ventral roots organization“. In preparation. 117 118 Curriculum Vitae Personal data Name: Joelle Gemayel Nationality: Lebanese Date of birth: May 27, 1974 E-mail: [email protected] Education 11/00 − 02/05: ETH Zürich, Zurich, Switzerland Ph. D. student in Neurobiology at Brain Research Institute Specialization: “Role of PlexinD1 in nervous system development” 09/98 − 10/00: Université Claude Bernard, Lyon, France DEA “Metabolism, Endocrinologie et Nutrition” 09/97 − 08/98: Université de Brest, Brest, France Maitrise in Physiology and Cell Biology 10/91 − 09/95: Lebanese University Fanar, Beirut, Lebanon Maitrise in Animal Physiology Professional experience 09/98 − 10/00: Université Claude Bernard, Lyon, France Teaching assistant in practical physiology courses for sports’ athletes 01/96 − 06/97: Tamer Group, Beirut, Lebanon Sales representative for Tamer Group dental division Responsible for marketing and sales of dental products to dentists 119