ZEBRAFISH DEADLY SEVEN: NEUROGENESIS, SOMITOGENESIS, AND NEURAL CIRCUIT FORMATION DISSERTATION

Transcription

1 ZEBRAFISH DEADLY SEVEN: NEUROGENESIS, SOMITOGENESIS, AND NEURAL CIRCUIT FORMATION DISSERTATION Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University By Michelle Gray, B. S. ***** The Ohio State University 2003 Dissertation Committee: Dr. Christine E. Beattie, Adviser Dr. Paul Henion Dr. Gail Herman Dr. Mark Seeger Approved by Adviser Molecular, Cellular and Developmental Biology Program

2 ABSTRACT Proper development of the vertebrate nervous system is essential for the overall function of the organism. The vertebrate nervous system is highly complex and contains an enormous number of distinct cell types. In order for the organism to carry out its normal behavior, it requires that all of the components of the nervous system be produced in appropriate numbers, at correct times, in the right locations and that they make the proper connections. The relative simplicity of the early zebrafish nervous system makes it an attractive model for studying vertebrate nervous system development. Mutagenesis screens in zebrafish have been undertaken to identify new genes involved in different aspects of nervous system development. We have characterized an ENU induced mutation in zebrafish deadly seven/notch 1a (des), which perturbs neurogenesis, somitogenesis, and motor axon outgrowth. The neurogenic defect is manifested as an increase in hindbrain interneurons and spinal motoneurons. In addition, we find a decrease in the number of spinal sensory neurons, and an increase in sensory neurons derived from neural crest cells. This data demonstrates that Notch signaling is important for determining the number of specific neuronal cell types during early nervous system development. ii

3 The somite defect in des mutants is revealed by abnormalities in somite/myotome boundary formation and somite/myotome gene expression in the mid- and posterior trunk and tail. Each somite/myotome in wild-type embryos contains an anterior and posterior domain. This anterior-posterior somite patterning is disrupted in des mutant embryos. Our studies reveal that this patterning defect causes aberrant motor axon outgrowth. Motor axons in wild-types obey domain restrictions, never entering the posterior domain. However, in des mutant embryos, motor axons are seen in both domains. Thus, proper patterning of the somite is essential for stereotyped motor axon pathfinding. The des mutation results in a dramatic increase in the hindbrain interneuron, Mauthner (Mth). This neuron is an integral part of a relatively simple neural circuit driving the escape response in zebrafish and thus presents an excellent opportunity to study properties of neural circuit formation. Due to the presence of supernumerary Mth cells in des mutants; we analyzed the affect of having one cellular component of this circuit dramatically increased on circuit formation and behavior. Our results indicate that all of the supernumerary Mth cells are integrated into the circuit and the circuit is functional. The escape behavior of des mutants is very similar to wild-type embryos. We found, however, that individual Mth cells in des mutants contacted fewer target cells in the spinal cord than Mth cell in wild-type larvae. These data indicate that when there are more Mth cells present, they divide up the territory thus incorporating iii

4 all cells into the circuit yet maintaining a normal escape response behavior. This study demonstrates that there is plasticity in the formation of the escape response circuit in zebrafish. iv

5 Dedicated to my mother and the Gray and McCalpine families v

6 ACKNOWLEDGMENTS I wish to thank my adviser, Christine Beattie, for her help in molding me scientifically, encouragement in the face of experimental difficulties, excitement about results and enthusiasm for science in general. I thank my committee members, Paul Henion, Gail Herman, and Mark Seeger for setting aside time for me in their busy schedules. I owe a debt of gratitude to my fellow lab members Louise Rodino-Klapac, Michelle McWhorter, Chunping Wang, and Emily Tansey and the members of the Henion lab for making the lab experience a fun and enjoyable one. I thank my fellow lab member Anil Challa and Vicki McGovern for stimulating discussions, and for being my sounding boards, and my friends throughout these years. I am grateful to the people who took care of the zebrafish, Dr. Cindy Herpolsheimer, Timothy Hammontree, and Matthew Frieda. I am also grateful to Joel Winston for supporting me as I pursued my goal. vi

7 VITA November 21, 1974 Born, Sawyerville, AL B.S. Biology, Alabama State University, Montgomery, AL Graduate Teaching Assistant, The Ohio State University Graduate Research Associate, The Ohio State University National Institutes of Health, Pre-doctoral Fellow, The Ohio State University Research Publication PUBLICATION 1. Michelle Gray, Cecilia B.Moens, Sharon L. Amacher, Judith S. Eisen and Christine E. Beattie. Zebrafish deadly seven functions in neurogenesis. Developmental Biology, 2001 Sep 15; 237(2): FIELD OF STUDY Major Field: Molecular, Cellular, and Developmental Biology Program vii

8 TABLE OF CONTENTS Page Abstract..ii Dedication..v Acknowledgments...vi Vita vii List of Tables..x List of Figures...xi Chapters: Chapter 1: Introduction 1 Neural specification 2 Patterning the nervous system.. 6 Patterning the spinal cord.. 9 Patterning along the DV axis of the spinal cord. 9 Proneural and neurogenic gene activity.. 13 Other roles for neurogenic genes during development. 15 Production of functional neural circuits.. 17 Connecting components of neural circuits.. 19 Chapter 2: Zebrafish deadly seven functions in neurogenesis Introduction.. 27 viii

9 Materials and Methods Fish In situ hybridization BrdU incorporation. 31 Single cell labels and transplants.. 32 RNA injections and analysis of injected embryos Results Somitogenesis is disrupted in des b420 mutants Spinal motor axons are disorganized in des b420 mutants. 36 Trunk neural crest migration is selectively disrupted in des b420 mutants.. 37 des 420 mutants have a neurogenic phenotype.. 39 Reciprocal effects on primary and secondary sensory neurons in des mutants Supernumerary Mauthner cells are born at the same stage in des b420 mutants and wild-type Activated Notch rescues the Mauthner cell phenotype in des b420 mutants Discussion. 45 Motor axon outgrowth and neural crest migration in des b420 mutants.. 47 des regulates neuronal cell number.. 49 des regulates the number of trunk sensory neurons The number of Mauthner cells is controlled by des-mediate lateral inhibition. 51 Multiple Notch-Delta signaling pathways.. 52 ix

10 Chapter 3: Mutations in deadly seven reveal developmental plasticity in the escape response circuit Introduction.. 73 Materials and Methods Fish Care and Identification Whole-mount antibody labeling Retrograde labeling of calcium green dextran Calcium imaging High-speed recording of behavior Retrograde labeling of calcium green dextran Single Mauthner axon labels.. 79 Results.. 81 The escape response circuit in des b420 mutants Supernumerary Mauthner cells are active during the escape response 82 Escape performance is similar in wild-type and des b420 mutants. 84 Mauthner axons in des mutants have fewer axon collaterals.. 86 Discussion. 88 Mauthner cell and the escape response Mauthner axon collaterals.. 90 Evolution of the escape response.. 92 Chapter 4: Motor axon behavior in abnormally patterned somites Introduction x

11 Materials and Methods Fish In situ hybridization and immunohistochemistry Blastula transplants Construction and overexpression of muscle actin promoter EphA4GFP construct Results. 111 Motor axons show defective pathfinding in des b420 mutant embryos Wild-type muscle cells can rescue the motor axon pathfinding defect in des b fss mutant embryos have defective somites and disorganized motor axons Motor axons pathfind abnormally along anteriorized somites/myotomes Discussion Motor axon pathfinding in somite patterning mutants A properly patterned somite is required for guidance molecules to be expressed in the correct domains Degree of motor axon rescue may be dependent on the position of cells within the somite 120 Juxtaposed anterior and posterior cells Chapter 5: Discussion List of References xi

12 LIST OF TABLES Table Page 2.1 Somite boundary formation is aberrant in des b Cell types unaffected in des b Two other des alleles have supernumary reticulospinal neurons Affect of Notch-ICD on Mauthner cell number Escape response performance in wild-type and des b420 mutant Larvae Axon collaterals in wild-type and des mutant larvae Mauthner axon collaterals from distinct axons rarely form in the same location 103 xii

13 LIST OF FIGURES Figure Page 1.1 Schematic representation of a cross-section of zebrafish during different stages of neurulation Schematic representation of primary motoneurons Schematic representation of subsets of hindbrain reticulospinal interneurons Somite boundary formation is aberrant in des b420 mutants Somite and myotome gene expression is perturbed in des b420 mutants Motor nerves are disorganized in des mutants CaP and MiP motor axons display abnormalities in des b420 mutants des b420 is non-cell autonomous for CaP motoneurons Migrating neural crest and neural crest derivatives are differentially affected in des b420 mutants des b420 mutants have excess primary motoneurons Number of primary motoneurons in des b420 mutant embryos. 63 xiii

14 2.9 Supernumerary reticulospinal neurons are present in des b Number of supernumerary reticulospinal neurons present in des b420 mutants Analysis of trunk sensory neurons in des b420 mutants Number of sensory neurons in des b420 mutants Supernumerary Mauthner cells in des b420 mutants are formed at the same time as Mauthner cells in wild-types Notch rescues the de sb420 Mauthner cell defect des mutants have supernumerary Mauthner cells Mauthner cells in des mutants are active during an escape Quantification of the increase in fluorescence intensities Mauthner cell lateral dendrites in des mutants are in close apposition to sensory nerves des b420 mutant larvae exhibited largely normal escape responses The number of collaterals per Mauthner axon is dramatically decreased in des mutants Collaterals from distinct Mauthner axons do not form in the same location.102 xiv

15 4.1 Schematic representation of wild-type CaP axon projections des and fss mutants have abnormal somites Gene expression is perturbed in somites of des and fss mutants CaP motor axons are abnormal in des and fss mutants Number of branches on CaP motor axons Position of growth cones of CaP motor axons at 26 h Wild-type muscle cells rescues CaP motor axons in des mutants sema3a2 is expressed throughout the somitic domain of fss mutants Overexpression of an α-actinepha4egfp results in abnormal CaP motor axons xv

16 CHAPTER 1 INTRODUCTION The vertebrate nervous system is highly complex, in both organization and function. The complexity of this system is achieved as development progresses through the specification of the neural ectoderm to the production of distinct cell types, in specific locations, at the correct times and in the right proportions. Proper function of the nervous system can only be accomplished if these cells make precise connections with the appropriate targets and thus produce the correct pattern of connections within the nervous system and in the periphery. Much of the early data concerning nervous system development comes from classical studies performed mainly in frog and chick embryos using overexpression experiments, tissue culture and grafting experiments. The production of knockout mice has provided additional knowledge about vertebrate nervous system development (Tanabe and Jessell, 1996; Sasai and De Robertis, 1997; Weinstein and Hemmati-Brivanlou, 1999). In more recent times, the zebrafish, which provides the advantage of genetics, molecular biology and cellular manipulation, has emerged as an excellent model system for understanding vertebrate neural development. The embryonic zebrafish has an accessible and relatively simple nervous system, with many of its components formed and functioning as early as 24 1

17 hours post fertilization (h). Early work performed on some aspects of nervous system development in zebrafish has laid the foundation for more exhaustive studies of nervous system development using this model. To help facilitate these studies, two large-scale mutagenesis screens and many smaller ones, designed to identify mutations that affect many aspects of development, have produced mutations that affect nervous system development. The characterization of these mutants reveals that many aspects of vertebrate nervous system development are well conserved in zebrafish and that the use of this model may be able to lend insights into the developmental process of vertebrates in general. This dissertation work has focused on the use of a mutation identified in a mutagenesis screen that affects nervous system development, in addition to other developmental processes. Neural specification The production of a functional nervous system begins with the specification of neural tissue. The specification of the neural ectoderm or neural plate formation is initiated during gastrulation in vertebrates. It is induced from the ectoderm that gives rise to the epidermis (non-neural ectoderm), central nervous system (CNS), peripheral nervous system (PNS) and the placodes. The production and patterning of these tissues from ectodermal precursors is due to inductive signals that originate from neighboring mesoderm and endoderm. The Spemann organizer, as it is termed in amphibians, can induce the formation of 2

18 the neural ectoderm (Spemann and Mangold, 1924). In addition, the organizer also influences anterioposterior (AP) and dorsal-ventral (DV) neural patterning. In chick and mice, the primitive node is considered the organizer. This structure can induce neural structures when grafted ectopically in an embryo (Waddington, 1933; Storey et al., 1992; Beddington, 1995; Kintner and Dodd, 1991; Blum et al., 1992). The organizer induces formation of neural tissue from dorsal ectoderm by expressing the molecules Chordin and Noggin, which antagonize BMP signaling. The newly induced tissue is thought to have anterior character, as it expresses molecules such as Otx-2 (forebrain) and does not express markers of more posterior neural identity. The production of the hindbrain and spinal cord is a result of the posteriorization of the neuroectoderm. Signals from the organizer in many vertebrates, which include Wnts, Fgfs, and retinoic acid, transform anterior neural tissue into posterior neural tissue (Ruiz i Altaba and Jessell,1991; Chen et al., 1994; Lamb and Harland, 1995; Cox and Hemmati-Brivanlou, 1995; Sasai et al., 1996; Takeda et al., 1994; reviewed by Doniach 1995). Like other vertebrates, the zebrafish nervous system begins as the neural plate. It is first manifested as a dorsomedial thickening of cells towards the end of gastrulation. The embryonic shield in zebrafish is sufficient for neural induction and is thought to be the equivalent of the organizer, as its transplantation to ventral regions induced axis formation, including neural tissue (Oppenheimer, 1936; Shih and Fraser, 1996, Driever et al., 1997; Koshida et al., 1998). Two classes of mutations have been described that exhibit expanded or 3

19 reduced neural domains, dorsalized and ventralized respectively (Hammershmidt et al., 1996a; Mullins et al., 1996). In the dorsalized class, two of the mutations encode components of Bmp signaling, swirl (swr) and snailhouse (snh) while a third encodes an intracellular component of the Bmp signal transduction pathway (Hild et al., 1999). Embryos mutant for these genes have expanded neural tissue and concomitant reduction of non-neural ectoderm, which is consistent with the inhibition of neural tissue formation by Bmp signaling. The gene dino (din), which encodes chordin, falls into the ventral class. Chordin normally antagonizes the ventralizing activity of Bmps and thus embryos mutant for din have reduced neuroectoderm and expanded nonneuroectoderm (Hammerschmidt et al., 1996a, Hammerschmidt et al., 1996b). In zebrafish the transformation of anterior neural tissue into tissue of more posterior character does not appear to be a result of the embryonic shield, which is considered the equivalent of the organizer. This posteriorizing activity seems to be from the lateral and ventral embryonic margin (Woo and Fraser, 1997; Koshida et al., 1998). Transplantation experiments at the beginning of gastrulation, in which cells were transplanted from different locations on the embryonic margin, revealed that this activity likely came from marginal cells or germ ring (lateral mesendoderm). Tissue grafting of the germ ring is able to transform tissue fated to become forebrain into tissue with morphological, cellular and molecular characteristics of the hindbrain (Woo and Fraser, 1997). Furthermore, Koshida et al., (1998) found that transplanting ventral marginal cells at early gastrulation stage to prospective anterior neural tissue inhibited 4

20 expression of anterior neural fates and induced expression of posterior neural fates. Together, these experiments indicate that ventral and lateral marginal cells are the source of a signal(s) that posteriorizes neural tissue in zebrafish. When gastrulation is completed in embryonic zebrafish, the cells of the future zebrafish nervous system form a plate of monostratified epithelium (Papan and Campos-Ortega, 1994). The cells converge and engage in extension movements during the process of neurulation, in which the neural plate becomes longer and narrower, transforming first into a neural keel (13 h) and then a solid neural rod (16 h). The neural rod then cavitates to form the neural tube (17 h) (Fig.1.1; Raible et al., 1992; Schmitz et al., 1993). This process has been compared to secondary neurulation, characteristic of caudal regions of amniote embryos. There are two waves of neurogenesis in zebrafish. The first wave produces primary neurons (Kimmel and Westerfield, 1990). Post-mitotic cells appear that will develop as primary neurons as gastrulation comes to an end. These neurons are large, few in number, and born early in development, starting about 9-10 h (Myers et al., 1986; Kimmel and Westerfield, 1990), and undergo axogenesis between h. These cells include sensory neurons, interneurons and motoneurons. Based on the above criteria, some of these cell types are individually identifiable. Secondary neurons include those neurons that are produced after this first wave. The characteristics of these neurons include their small sizes and large numbers, and later birthdate at about h (Myers et al., 1986; Kimmel and Westerfield, 1990; Appel et al., 2000). 5

21 Primary neurons are born within the neural plate at about 9-10 h, as revealed by the expression of a neuron-specific gene, HuC, a zebrafish homologue of D. melanogaster elav (Good 1995; Kim et al., 1996). The differentiation of particular types of neurons occurs in spatially and temporally restricted patterns within the CNS. Primary neurons are distributed in three bilateral longitudinal columns, medial, intermediate, and lateral, which give rise to motoneurons, interneurons, and sensory neurons, respectively (Myers et al., 1986; Kimmel and Westerfield 1990). At the three-somite stage, huc expression reveals nascent neurons in the midbrain, hindbrain and the trigeminal ganglia. As neurulation proceeds, cells maintain their relative positions within the neural plate, medial neural plate cells occupy ventral neural tube and lateral neural plate cells occupy the dorsal neural tube. Thus changing the position of the motoneurons, interneurons, and sensory neurons from their lateromedial position at the neural plate stage to their characteristic DV position within the neural tube. The DV pattern of neurons in the neural tube is like the pattern of neurons of other vertebrates, so that different types of neurons are located at distinct positions within the neural tube. Patterning the nervous system The nervous system is patterned along both the AP axis and the DV axis. Along the AP axis, there are four primary components of the CNS, the forebrain, midbrain, hindbrain and spinal cord. As stated earlier, the initial neural 6

22 induction in vertebrates induces anterior neural tissue, namely forebrain. The forebrain, like all other regions of the CNS is derived from the simple neuroepithelial cell sheet of the neural plate. Before neural tube formation, the anterior CNS undergoes morphogenetic movements that transform the simple neural plate into the highly complex derivatives of the forebrain (Loosli et al., 2001). The most complex region of the CNS is the forebrain or anterior neural plate. The forebrain region comprises the telencephalon, optic vesicles, hypothalamus, ventral and dorsal thalamus, and the pretectum. The telencephalon in mammals is the region that includes the cerebral cortex, olfactory bulb, and the basal ganglia. Within the ventral region of the forebrain, Nodal and Hedgehog signaling pathways appear crucial to its development in zebrafish. Mutations that affect Hh signaling result in a reduction of hypothalamus tissue, which is ventral tissue, after initial expression of a hypothalamic marker is seen (Rohr and Concha, 2000; Varga et al., 2001). Markers for ventral sub-pallial telencephalon are reduced or absent whereas dorsal pallial markers are unaffected or expanded in zebrafish embryos carrying mutations in the Hh signaling pathway (Rohr et al., 2001; Varga et al., 2001). Establishment of the dorsal forebrain seems to require Bmp signaling, although only a few studies have been performed to address this issue. Like other components of the brain, formation of the zebrafish midbrain and hindbrain result from inductive signals that start during gastrulation (Woo and Fraser, 1995). Caudal to the forebrain lays the midbrain or mesencephalon, 7

23 followed by the isthmus rhombencephali, a region considered the most rostral portion of the hindbrain or rhombencephalon. The midbrain includes the tectum dorsally and the tegmentum ventrally. The tectum serves as a major relay center for sensory information from the eyes, ears, and lateral line organs, whereas the tegmentum contains motor nuclei that are involved in directing eye movements. The isthmus contains dorsally the cerebellar primordium (Müller et al., 1996). Later during early somitogenesis the midbrain and isthmus fold up and form the midbrain-hindbrain boundary (MHB). The repression of the transcription factors otx2 and gbx are important for positioning the MHB. The MHB is suggested to be required for AP polarization of the midbrain because when this tissue is grafted to caudal forebrain the surrounding tissue adopts a midbrain character. Early in development, Wnt8 signaling has been shown to be important for positioning the midbrain and hindbrain. Later, the graded expression of Ephrin A2 and Ephrin A5 ligands are required for AP pattern within the midbrain (Picker et al., 1999). Fgf8 functions at the end of gastrulation in the early hindbrain primordium (Reifers et al., 2998, Furthauer et al., 2001; Raible and Brand, 2001). In addition, the expression of two targets of Fgf8 signaling, erm and pea3 suggests that Fgfs function during early hindbrain segmentation (Nusslein-Volhard 2001). Once the hindbrain primordium is induced, it becomes divided into seven segmental rhombomeres along the AP axis. These rhombomeres can be clearly visualized at about the 18-somite stage in zebrafish (Hanneman et al., 1988). Each rhombomere segment contains specific subtypes of neurons that are identifiable based on the segment they occupy and their axon projection (Fig. 8

24 1.3). The hindbrain is also divided along the DV axis into roof plate, alar plate, basal plate and floor plate which is similar to the spinal cord. In zebrafish, the Sonic hedgehog (Shh) family of secreted proteins is involved in patterning the ventral hindbrain. Inactivation of both shh and tiggy winkle hedgehog (twh) results in a complete absence of ventrally located branchiomotor neurons (Bingham et al., 2001; Etheridge et al., 2001). As has been shown for dorsal derivatives of other components of the nervous system, the Bmp pathway regulates the production of dorsal hindbrain derivatives. Patterning the spinal cord The spinal cord is the most posterior part of the nervous system and more simply organized than the other brain structures. It is patterned along both the AP and DV axis. One aspect of patterning along the AP axis is revealed by the segmental arrangement of the cell bodies of early-developing primary motoneurons that innervate trunk muscles (Myers 1985; Eisen et al., 1986; Westerfiled et al., 1986). Perhaps the most apparent patterning within the developing spinal cord of zebrafish is the organization along the DV axis, which is the same as other vertebrates, with sensory neurons present dorsally, interneurons present in the middle region of the spinal cord and motoneurons present ventrally. Specialized non-neuronal cells occupying the dorsal midline 9

25 of the spinal cord and contacting the lumen of the central canal are roof plate cells, whose function is not known (Altman and Bayer 1984). Floor plate cells contact the lumen of the central canal ventrally. Patterning along the DV axis of the spinal cord Dorsal and intermediate spinal cord. The spinal cord contains sensory neurons dorsally and interneurons in the midregion. In addition to these cell types another population of cells, the neural crest, occupy the dorsolateral aspect of the neural tube. Neural crest cells arise as bilateral domains of the neuroectoderm that later converge on the midline and form a distinct population of mesenchymal cells (LeDourain and Kalcheim, 1999; Luo et al., 2001). One class of sensory neurons present in the zebrafish dorsal spinal cord is the Rohon- Beard (RB) cells that are specified at the neural plate stage (Korzh et al., 1993). RB cells are large primary neurons that occupy a continuous row on either side of the dorsal spinal cord midline (Bernhardt et al., 1990). These RB extends their axons longitudinally within the dorsal spinal cord, one rostrally and one caudally, as well as a peripheral dendrite that branches extensively and innervates the skin. The midregion of the spinal cord contains many types of interneurons, whose functions are not fully known, but several of them may play inhibitory roles (Bernhardt et al., 1992). Bmps regulate the DV patterning of the spinal cord. Zebrafish homozygous for mutations in Bmp pathway genes, like swirl/bmp2b, 10

26 snailhouse/bmp7 and somitabun/smad5 (Nguyen et al, 1998; 2000), shows that Bmp signaling is necessary to establish neural crest and dorsal sensory neurons. When embryos with varying levels of Bmp activity are analyzed, it is revealed that Bmps have graded effects throughout the mediolateral axis of the neural plate (Barth et al., 1999). In embryos where Bmp signaling is severely depleted, there is a loss of dorsal sensory neurons (RBs) as well as interneurons in the midregion of the spinal cord. Mild depletion of Bmp signaling as in somitabun/smad5 mutants leads to a lateral displacement of RBs and interneurons. Thus, Bmp signaling is critical to the development of the dorsal and intermediate regions of the spinal cord. Ventral spinal cord. Floor plate cells and motoneurons occupy the ventral domain of the vertebrate spinal cord. Floor plate cells are non-neuronal cells in the ventral midline of the spinal cord. The floor plate consists of two cell types: medial and lateral cells (Kuwada et al., 1990; Odenthal et al., 1998, 2000; Schauerte et al., 1998). These cells secrete proteins that affect fate decisions and axon trajectories of adjacent neurons (Colamarino and Tessier-Lavigne 1995; Tanabe and Jessell 1996). Motoneurons are located laterally to the floor plate cells in the ventral spinal cord. They are born during the two distinct waves of neurogenesis; primary motoneurons undergo their final S phase between 9 and 16 h, and secondary motoneurons have their last S phases beginning after 14 h and continuing until after 25 h (Myers et al., 1986). Patterning of the ventral spinal cord requires Bmp, Shh, and Nodal signaling. In addition to being important for dosal spinal cord patterning, Bmp 11

27 signaling is also important for patterning the ventral spinal cord. Severe depletion of Bmp leads to an expansion of the floor plate (Barth et al., 1999; Nguyen et al., 2000). This suggests that Bmp signaling suppresses the formation of ventral cell types. Shh is produced by notochord and floor plate cells and influences the patterning of the ventral neural tube in vertebrates (reviewed by Tanabe and Jessell, 1996). Early in development Shh is expressed by the notochord as neuroectoderm is specified; later floor plates cells also express Shh. Targeted deletion of Shh in mice and blocking antibody experiments demonstrate that Shh is necessary for specification of ventral cell types including floor plate and motoneurons (reviewed by Tanabe and Jessell, 1996). In zebrafish, mutations have revealed that Hedgehog signaling disrupts ventral spinal cord development. Embryos mutant for sonic you (syu), which encodes Shh (Schauerte et al., 1998), Smoothened, encoded by smo, a Hedgehog co-receptor (Chen et al., 2001; Varga et al., 2001), or you-too (yot) which encodes Gli2, a downstream factor in the Hedgehog pathway (Karlstrom et al., 1999), lack lateral floor plate cells. The specification of medial floor plate appears to require Nodal signaling. In mutants lacking molecules involved in this pathway, cyclops (cyc) (Rebagaliati et al., 1998b; Sampath et al., 1998), one-eyed pinhead (oep)(zhang et al., 1998b; Gritsman et al., 1999) or schmalspur (sur) (Pogoda et al., 2000; Sirotkin et al., 2000b) lack medial floorplate, but contain lateral floor plate (Hatta et al., 1991b; Brand et al., 1996a; Schier et al., 1996, 1997). 12

28 The syu mutant zebrafish embryos have motoneurons, unlike Shh mutant mouse embryos (Schauerte et al., 1998). However the midline expression of the two other hedgehog molecules echidna hegehog (ehh; Currie and Ingham, 1996) and twhh (Ekker et al., 1995) might be sufficient to induce motoneuron development in the absence of syu. In support of this, smo mutant embryos have a deficit of primary motoneurons and no secondary motoneurons (Chen et al., 2001a; Lewis and Eisen 2001; Varga et al., 2001). Embryos lacking both notochord and floor plate (the sources of Hedgehog signals) develop a small number of anterior primary motoneurons but no secondary motoneurons (Beattie et al., 1997). These embryos transiently express shh during gastrulation; and thus it has been hypothesized that brief exposure to Shh may induce neuroectodermal cells to develop as primary motoneurons whereas extended exposure may be required to induce secondary motoneuron development. The identity of primary motoneurons is refined by the expression of genes encoding LIM-class homeodomain proteins. All of the primary motoneurons express lim3. MiP (middle primary) and RoP (rostral primary) express islet 1 (isl1), while CaP (caudal primary) and VaP (variable primary, a transient motoneuron), express islet 2 (isl2, Appel et al., 1995). The position a primary motoneuron occupies determines which lim genes it expresses and its axon projection. When potential MiP neurons are transplanted to a CaP position, two hours before axongensis, the transplanted cells express isl2 and project axons into the ventral development (Eisen et al., 1991; Appel et al., 1995). Furthermore in embryos injected with isl2 antisense morpholinos, CaP and VaP axons fail to 13

29 extend into the ventral myotome, but usually projected caudally within the CNS (Segawa et al., 2001), suggesting that isl2 is required for the development of a subset of primary motoneurons Thus, the primary motoneurons are not only defined by the expression of the LIM-class of transcription factors, but need them to establish their final identity. Proneural and neurogenic gene activity Proneural and neurogenic genes were initially identified in Drosophila (reviewed in Campos-Ortega, 1995). Early AP and DV patterning mechanisms in Drosophila define proneural clusters in the neuroectoderm, where cells express proneural genes. The genes are bhlh protein-encoding genes of the acheatescute class. Each cell within the proneural cluster has the potential to take on the neural fate. Through the process of lateral inhibition, one cell within the proneural cluster inhibits neighboring cells so only one cell develops as a neuroblast within each proneural cluster (Skeath and Carroll, 1992). The other cells in the proneural cluster will take on an epidermal fate. A membrane bound receptor, Delta and the transmembrane receptor, Notch, mediate the process of lateral inhibition. Proneural genes promote the expression of Delta, whereas the activation of Notch inhibits the expression of proneural genes. Through the interaction of Delta and Notch, proteolytic events are initiated that produce an intracellular fragment (Notch intra ) that together with the Suppressor of Hairless [Su (H)] protein, promotes the expression of genes of the Enhancer of split [E 14

30 (spl)] complex. The E(spl) proteins then inhibit expression of proneural genes, thus these cells with a high level of Notch activity becomes epidermoblasts, whereas those cells with a low levels of Notch activity becomes neuroblasts. Mutations of Delta, Notch, or any of the genes in the Notch-Delta signaling pathway in Drosophila result in neurogenic phenotypes where excess neuroblasts are made at the expense of epidermoblasts. The genes that are involved in the production of neuronal cell types in flies are also conserved in vertebrates including zebrafish. During vertebrate neurogenesis, all the cells within the neuroectoderm are destined for neural development; vertebrates are unlike Drosophila in this respect. Furthermore the cells within the vertebrate proneural domains that have been singled out to become neurons stop dividing while the other cells continue dividing. These cells will become neurons during the next wave of neurogenesis. Functional studies reveal that mechanisms similar to those in Drosophila are involved in selecting cells that become neurons within proneural domains in vertebrates. By the end of gastrulation, groups of cells that will give rise to the longitudinal columns (medial, intermediate, and lateral) of early developing primary neurons begin to express the proneural gene neurogenin1 (ngn1). Two of these domains have been well studied: the lateral domain that generates neural crest and RB neurons, and the more medial domain that generates primary and secondary motoneurons. Within these early proneural domains and later stages of neurogenesis, newly specified neurons transiently express three Delta homologues, deltaa (dla), deltab (dlb) and deltad (dld)(dornseifer et al., 1997; 15

31 Haddon et al., 1998; Appel and Eisen, 1998; Appel et al., 2001). Subsets of neural precursors also express delta genes; these could be cells fated to exit the cell cycle and differentiate as neurons (Appel et al., 2001). Neural plate cells and neural precursors of 24 h embryos, uniformly express several Notch homologues (Bierkamp and Campos-Ortega 1993; Westin and Lardelli, 1997; Appel et al., 2001). Overexpression of notch and delta genes results in an inhibition of neurogenesis. Furthermore the activation of Notch1a leads to expression of her4, a Hairy Enhancer of split related gene. When Notch signaling is disrupted by expression of dominant negative forms of Delta protein or by mutations in dla, excess cells within proneural domains express elevated levels of ngn1 and develop as neurons (Dornseifer et al., 1997; Haddon et al., 1998; Appel and Eisen, 1998; Appel et al., 2001). Experiments that block Delta-Notch signaling in the lateral neural plate result in a larger number of RB neurons, showing that surrounding cells can become RBs, thus Delta-Notch mediated lateral inhibition prevents cells from becoming that fate (Appel and Eisen, 1998; Haddon et al., 1998; Cornell and Eisen, 2000). In the medial neural plate, Delta-Notch signaling prevents cells surrounding prospective primary motoneurons from taking that fate and they become secondary motoneurons instead (Appel and Eisen, 1998; Haddon et al., 1998). All of these data demonstrate the role of Delta-Notch signaling in regulating neurogenesis in the zebrafish neural plate. 16

32 Other roles for neurogenic genes during vertebrate development In addition to being involved in vertebrate nervous system development, neurogenic genes are also involved in other aspects of development. One of these developmental processes is somitogenesis. In mice the Delta homologue Dll, is expressed in the paraxial mesoderm and the posterior halves of somites. Mice homozygous mutant for Dll1, lack epithelial somites and have somite segments without cranio-caudal (anterior-posterior) polarity (Hrabe de Angelis et al., 1997). Furthermore, Notch1 mutant mice have irregularly shaped somites (Conlon et al., 1995). In zebrafish notch1a, notch1b, notch5, notch6, deltac and delta D and downstream targets of the pathway, her1, her4, and her7 are expressed in the presomitic mesoderm and formed somites. All of these genes are expressed in distinct bands in the formed somites. deltac, notch5, notch1a, and notch1b are expressed in the posterior half of presumptive and formed somites and deltad and notch6 in the anterior half of the somite. Overexpression of RNA encoding a constitutively active form of Notch in zebrafish results in a loss of segmentation (Takke and Campos-Ortega, 1999). This also results in the transcriptional activation of her1 and her4 throughout the regions in which they are normally expressed in discrete bands (Takke and Campos-Ortega, 1999; Takke et al., 1999). In addition, overexpression of deltad, her1, and her4 results in the loss of somite boundaries (Dornseifer et al., 1997). The after eight (aei) mutation has been identified as deltad (van Eeden et al. 1996, Holley et al., 2000). These mutants form paraxial mesoderm, but somite boundaries are abnormal in these mutants. This mutation 17

33 also results in a perturbation in the expression of other components of the pathway, her1 and deltac. Taken together these results demonstrate the importance of the Delta-Notch signaling pathway in somite formation. Production of functional neural circuits In order for the nervous system to be a properly functioning unit, the cells that compose it must make precise connections with their targets. The earliest connections in embryonic zebrafish are made by the primary neurons. The connection between some of these cells establishes functional circuits that are first manifested in the embryo, but can be seen throughout the life of the adult. One of the functional circuits that has been identified and well studied in zebrafish embryos and larvae is the escape response circuit, which controls escape behavior (startle response or C-start), an avoidance behavior that includes a characteristic C-bend of the muscle of the trunk and tail (Yasagril and Diamond, 1968; Eaton and Hackett, 1984; Eaton et al., 1988). This circuit is composed of sensory nerves, hindbrain reticulospinal interneurons (Fig. 1.3), and spinal neurons. The escape response in teleost fish and amphibians is mediated by the Mauthner reticulospinal interneuron, which receives pre-synaptic sensory input from cranial nerves onto its lateral dendrite and post-synaptically contacts spinal motoneurons and interneurons. The single bilaterally located pair of Mauthner neurons arise at about 7.5 h (Mendelson, 1986a). The Mauthner cell body is easily identifiable, because of its 18

34 large soma size and presence in rhombomere 4 (Mendelson, 1986b). The Mauthner axon appears at h, projects across the midline then caudally into the trunk of the embryo, where it reaches the rostral spinal cord at h. At about the same time, the neuron takes on a bipolar shape with the presence of the lateral dendrite opposite of the axons (Kimmel et al., 1990). The trigeminal, acoustic-vestibular, and lateral line sensory nerves innervate the Mauthner lateral dendrite. The trigeminal ganglion cells are the first sensory neurons to develop in the head of zebrafish. About 20 trigeminal neurons arise together in the ganglion on each side of the head during the first day of development. The ganglion projects axons at about h in the lateral longitudinal fascicle in the hindbrain. The growth cones of 1-2 axons in the trigeminal ganglion are the earliest to arrive at the Mauthner cell and contact its pial-most surface at about 18 h before the Mauthner lateral dendrites have formed (Kimmel et al., 1990). At about 19 h, a bundle of trigeminal axons are present at this location, where they occupy a prominent indentation in the Mauthner cell surface. This indentation is the same one that is present later at the base of the newly forming lateral dendrite. The inputs from the acousticvestibular ganglion can first be seen at 22 h. The central axons grow into the brain and appear to contact the newly forming lateral dendrite at 23 h at a site distal to the site of trigeminal input. The axons of lateral line neurons arrive at the Mauthner cell last. The lateral line develops during embryogenesis in a ganglion just caudal to the otic vesicle (Metcalfe et al., 1985). The inputs from the lateral line appear to arrive at the Mauthner neuron at about 25 h. Contacts were 19

35 observed between the growth cones of lateral line axons and the tip of the growing lateral dendrite of the Mauthner neuron. The Mauthner axon descends into the spinal cord where it makes monosynaptic contacts onto ipsilateral spinal motoneurons and interneurons. Mauthner axon collaterals contact the ventral processes of primary motoneurons that innervate and excite the fast muscle fibers in the myotomes (Fetcho, 1986; Diamond, 1971; Fetcho and Faber, 1988; Celio et al., 1979; Myers, 1985). Thus, the escape response is initiated by the Mauthner neuron and the excitation of the muscle to cause the C-bend away from a potential threat is carried out by primary motoneurons in the trunk and tail of the fish. Connecting the components of neural circuits In order to achieve a properly functioning neural circuit, all of the components must make the correct connections with their targets. To reach their targets, axon growth cones respond to cues in the environment. Motor axon projections and the molecules that influence them have been studied in vertebrates. Chick, mouse, and culture experiments implicate diffusible chemoattractants in guiding motor axons. Hepatocyte growth factor/scatter factor, a diffusible ligand for the c-met receptor tyrosine kinase, is released by limb mesenchyme and appears to be the major chemoattractant for limb motoneurons (Ebens et al., 1996). Chemorepellants are also involved in the guidance of motor axons in vertebrates. Some of the molecules that act as 20

36 chemporepellants during motor axon pathfinding are the Semaphorins, which are a large family of secreted and transmembrane molecules, some of which utilize the Neuropilin receptors (He et al., 1997; Kolodkin et al., 1997; Kolodkin, 1996). Semaphorins are thought to guide motor axons away from the midline so that they exit the CNS properly (Varela-Echavarria et al., 1997). After exiting the spinal cord, motor axon extension is defined to a specific region of the myotome. Numerous experiments demonstrate that motor axons seem to be inhibited from entering the caudal sclerotome (reviewed by Eisen, 1994). The caudal sclerotome expresses molecules that act as chemorepellants that may mediate this inhibition. Collagen IX, a chondroitin sulfate proteoglycan (Ring et al., 1996), Ephrins, a large class of chemorepellants (Wang et al., 1997; Donoghue, et al., 1996), and T- cadherin, a homophilic cell adhesion molecule (Fredette et al., 1996) are all expressed by the caudal sclerotome and shown to be important for inhibiting motor axon pathfinding. The escape response activates primary motoneurons that innervate fish axial muscle. The proper innervation of the muscle by these axons is essential to the escape behavior of the animal, thus reaching their targets is vital. Primary motoneurons are born early during development, at about 12 h (Myers et al., 1986) and start to extend their axons at about 18 h along the myotome. There are three individually identifiable primary motoneurons, CaP (caudal primary), MiP (middle primary), and RoP (rostral primary) (Myers et al., 1986, Westerfield et al., 1986, Eisen et al., 1986). The axons of these motoneurons have very stereotypical axon projections and pathways (Fig. 1.2;Myers et al., 1986). The 21

37 CaP axon grows ventrally along the medial surface of the anterior domain of the myotome, and finally innervates the ventral myotome. The MiP axon projects ventrally, first following the same pathway as the CaP axon, but then forms a dorsal collateral and project into the dorsal myotome, while the RoP axon projects in the middle part of the myotome. All of these axons are confined to the anterior domain of the myotome and never project into the posterior domain. As is the case for other vertebrates, the posterior domain of the somite, which gives rise to myotome, in zebrafish appears to be inhibitory to the axons of the primary motoneurons. Chemorepellants have been shown to be restricted to the posterior domain of the somite in zebrafish. The expression of semaphorin 3A2 is restricted to the posterior domain, and its overexpression results in a loss or truncation of CaP axons. In addition, the expression of ephrin-b2 is restricted to the posterior domain of the somite. Thus, molecules important for motor axon pathfinding in other vertebrates are also conserved in zebrafish and they ensure that the axons are in the proper areas to appropriately innervate their targets. The nervous system requires the production of proper cell types, at the correct times, in the right proportions, and the proper connections between those cell types to produce an accurately functioning system. The use of mutagenesis screens in the vertebrate zebrafish model to understand the development of the nervous system allows researchers the opportunity to identify molecules that could not otherwise be found by using other vertebrate systems. The work discussed previously shows clearly that the isolation of mutations that affect a 22

38 specific aspect of development can lend insights into nervous system development. The work presented in this document uses a mutation that was isolated in a mutagenesis screen. This mutation affects multiple aspects of zebrafish development and therefore provided the opportunity to analyze the production of neurons during nervous system development (types and numbers), what effect those cell types had on the activity of a functional neural circuit, and how the formation of those circuits are affected by the environment. 23

39 Figure 1.1: Schematic representation of a cross-section of zebrafish during different stages of neurulation. As neurulation proceeds, the sensory neurons (RB), interneurons, and motoneurons change from their lateromedial distribution to the dorsol-ventral distribution at the neural plate stage. RB, Rohon-Beard sensory neurons;in, interneurons; and MN, motoneurons; FP, floor plate; RP, roof plate 24

40 Neural plate FP IN MN RB Neural keel Neural tube RP RB IN MN FP 25

41 Figure 1.2: Schematic representation of primary motoneurons Lateral view of the organization of primary motoneurons as they migrate along the myotome, CaP in red, MiP in blue and RoP in green (A). Cross-section through trunk revealing CaP and MiP (B). 26

42 Figure 1.3: Schematic representation of subsets of hindbrain reticulospinal neurons. This is a dorsal view of some of the neurons present in each rhomobomere. The neurons are present in specific rhombomeres and have specific axonal projections. Note the large Mauthner neuron present in rhombomere four; this neuron is involved in the escape behavior. 27

43 CHAPTER 2 ZEBRAFISH DEADLY SEVEN FUNCTIONS IN NEUROGENESIS Introduction Mutagenesis screens in zebrafish have identified mutants that affect both neurogenesis and somitogenesis (Jiang et al., 1996; Schier et al., 1996; van Eeden et al., 1996). Two mutants, mindbomb (mib) and after eight (aei, encoding deltad; Holley et al., 2000), show varying degrees of nervous system hyperplasia. mib has an over-abundance of early developing neurons, including Mauthner cells, whereas aei/deltad has a more subtle neurogenic phenotype only reported to affect primary sensory neurons (Schier et al., 1996; Jiang et al., 1996; Holley et al., 2000). Furthermore, mutations in both of these genes cause defects in somite boundary formation (van Eeden et al., 1996; Jiang et al., 1996, 2000; Durbin et al., 2000; Holley et al., 2000). Mutations in two other genes, beamter (bea) and deadly seven (des) have somite defects similar to mib and aei/deltad (van Eeden et al., 1996; Durbin et al., 2000; Jiang et al., 2000), but nervous system abnormalities have not been described for these mutants. In frogs, fish, and mice, perturbing genes in the Notch-Delta signaling pathway leads to defects both in neurogenesis and somitogenesis. In Xenopus 28

44 laevis, the Delta homologue, X-Delta-1 is expressed in the nervous system and interfering with its activity results in an overproduction of early developing neurons (Chitnis et al., 1995). Overexpressing RNA encoding X-Delta-1 or a constitutively activated form of Xenopus Notch, has the opposite effect causing a decrease in the number of early developing neurons (Chitnis et al., 1995). Overexpression of a dominant-negative form of another Delta homolog, X-Delta- 2, normally expressed in the presomitic mesoderm, leads to a defect in somitogenesis (Jen et al., 1997). Overexpressing zebrafish delta genes also causes neurogenic defects (Haddon et al., 1998; Appel et al., 1998) and overexpression of deltad results in defects in somite boundary formation (Dornseifer et al., 1997; Takke et al., 1999). Consistent with this, aei/ deltad mutants have a neurogenic phenotype and defects in somitogenesis (van Eeden et al., 1996; Holley et al., 2000). Both mouse Notch1 and a mouse Delta homologue, Dll1, are expressed in paraxial mesoderm. Disruption of either of these genes, as well as genes required for Notch glycosylation and processing, results in abnormal somite formation (Conlon et al., 1995; Oka et al., 1995; de Angelis et al., 1997; Wong et al., 1997; Evard et al., 1998; Kusumi et al., 1998; Zhang and Gridley, 1998). A neurogenic defect is also observed when Notch1, its downstream effector RBP- JK, or Presenilin-1, a gene involved in Notch processing, are mutated in mice (de la Pompa et al., 1997; Handler et al., 2000; Donoviel et al., 1999). Thus, there is overwhelming evidence that disrupting the Notch-Delta signaling pathway perturbs both neurogenesis and somitogenesis in vertebrates. We have isolated a mutation with a defect in somite formation closely resembling that of known zebrafish and mouse Notch-Delta pathway mutants. 29

45 In this report we demonstrate that the mutation we isolated is an allele of the zebrafish mutant des (des b420 ). We expand on the previously described motor axon defect (van Eeden et al., 1996) by analyzing individual primary motor axons. By creating genetic mosaics, we show that the motor axon defect in des b420 mutants is non-cell autonomous and is likely due to aberrant somite and myotome formation. We also find that the normally segmental pattern of neural crest cell migration is lost and neural crest-derived dorsal root ganglion (DRG) neurons are misplaced and fail to coalesce into ganglia. Furthermore, analysis of the primary nervous system of des b420 mutants reveals a restricted neurogenic defect that only affects a subset of neuronal cell types, including the identified, hindbrain interneuron, the Mauthner cell. We show that overexpressing RNA encoding an activated form of Xenopus Notch (Notch-ICD; Coffman et al., 1993; Chitnis et al., 1995) rescues the Mauthner cell phenotype suggesting that des acts in the Notch-Delta signaling pathway at or upstream of the level of the Notch receptor. The specificity of the des neurogenic phenotype, with respect to the subsets of neurons affected, implies that different Notch-Delta pathway components may exhibit temporal or regional specificity during zebrafish development. 30

46 Materials and Methods Fish Adult zebrafish and embryos were maintained essentially as described in Westerfield (1995) and staged by hours (h) or days (d) postfertilization at approximately 28.5 C as in Kimmel (1995). Mutant embryos (*AB background) were collected from pairwise matings of heterozygous adults and identified based on somite morphology. In situ hybridization and immunohistochemistry Staged embryos were processed for whole mount in situ hybridization as described by Thisse et al. (1993). Antisense digoxigenin islet1, islet2 (Korzh et al., 1993; Inoue et al., 1994; Appel et al., 1995), mesp-a (Sawada et al., 2000; Durbin et al., 2000), crestin (Rubenstein et al., 2000) and dopachrome tautomerase/tyrosinaserelated protein 2 (dct, Kelsh and Eisen, 2000) riboprobes were synthesized from plasmids linearized with EcoR1 and transcribed with T7 polymerase. valentino/kriesler (val) was synthesized from a plasmid linearized with Pst1 and transcribed with Sp6 (Moens et al., 1998). myod was synthesized from a plasmid linearized with Xba1 and transcribed with T7 (Weinberg et al., 1996) and her1 was synthesized from a plasmid linearized with XhoI and transcribed with T3 (Muller et al., 1996). For znp1 (Melancon et al., 1997), monoclonal antibody 16A11 (anti-hu; (Henion et al., 1996), acetylated tubulin (Sigma), anti-islet 1 (Developmental 31

47 Studies Hybridoma Bank; Korzh et al., 1993), zrf-1 (Trevarrow et al., 1990) and 3A10 (Hatta, 1992) immunohistochemistry, embryos were fixed in 4% paraformaldehyde overnight at 4 C, then washed in phosphate buffered saline (PBS) and preincubated in PBS with 0.5%Triton X-100, 1% bovine serum albumin, 1% dimethysulfoxide, and 2.5% goat serum (PBDT). Antibodies were diluted in PBDT and incubated overnight at 4 C. For anti-neurofilament (RMO44) antibody labeling, 48 h embryos were fixed in 2% trichloroacetic acid and processed as described by Pöpperl et al., For visualization under transmitted light, the Clonal PAP system (Sternberger Monoclonals Inc.) with 3',3'-diaminobenzidine (DAB) as substrate was used whereas anti-mouse Oregon Green (Molecular Probes) was used for fluorescent detection. For cross-sectional analysis of DRG and enteric neurons, embryos were fixed at 5 d, embedded in 1.5% agar/5% sucrose and sectioned on a cryostat at 16 µm. Anti-Hu was added for 2 hrs at room temperature followed by incubation in anti-mouse Oregon Green for 1 h at room temperature. Embryos and sections were analyzed using a Zeiss Axioplan microscope and photographed with Kodak Ektachrome 64T film or digitally imaged using a Photometrics SPOT camera. BrdU incorporation and in situ hybridization Embryos collected from pairwise matings of heterozygous des b420 fish were dechorinated at 6-14 h with 2mg/ml pronase (Sigma) and incubated in 10mM BrdU (Roche) /10%DMSO in embryo medium (Westerfield, 1995). Embryos were soaked for 45 minutes in BrdU starting at 7 h, 9 h, 11 h, and 14 h. After 32

48 incorporation, embryos were washed once quickly followed by two 15 minute washes in embryo medium. To process for val RNA in situ hybridization, BrdUtreated embryos remained in embryo medium until 19 h. Afterward they were fixed for hrs in 4% paraformaldehyde at room temperature followed by in situ hybridization with a digoxigenin val riboprobe. Following the color reaction with NBT/BCIP and washing in PBSt (1XPBS and 0.5% tween), embryos were incubated in 2N HCl for 1 hr at 37 C then processed for BrdU detection using an anti-brdu (Roche; 1/100) followed with a rhodamine-conjugated secondary antibody. Following antibody labeling, embryos were embedded in 1.5% agar/5% sucrose and sectioned on a cryostat at 16 µm. Single cell labels and transplants Individual motoneurons were labeled with rhodamine dextran (3 X 10 3 MW; Molecular Probes) as described (Eisen et al. 1989; Beattie et al. 2000). Embryos were mounted in 1.2% agar on a microslide. After labeling, embryos were removed from the agar and placed in embryo medium (Westerfield, 1995) containing 50 units penicillin and 5 µg streptomycin at 28.5 C. Labeled cells were visualized using a Zeiss Axioskop. Images were captured with a Photometrics SPOT camera and were colorized using Photoshop (Adobe). CaP transplants were performed essentially as described (Eisen (1991; Beattie et al. 2000). Briefly, 16 h donor embryos labeled at the 1-4 cell stage with rhodamine dextran (10 X 10 3 MW; Molecular Probes) were mounted side by side with unlabeled host embryos in 1.2% agar on a microslide. Individual CaP 33

49 motoneurons were transplanted from labeled donors to unlabeled host spinal hemisegments from which the native CaP and VaP (Eisen et al., 1990) motoneurons had been removed. Transplanted cells were visualized using a Zeiss Universal Compound Microscope equipped with a Dark Invader low light level camera. Images were captured using AxoVideo (Axon Instruments) and colorized using Photoshop (Adobe). RNA injections and analysis of injected embryos To make Xenopus Notch-ICD myc-tagged RNA, plasmid DNA was linearized with Not1 (Chitnis et al., 1995). Transcription was performed using Sp6 mmessage mmachine Kit (Ambion). After transcription, RNA was phenol/chloroform extracted and concentrated using Microcon YM-50 microconcentrator filter devices (Amicon). RNA quality was assayed by formaldehyde gel electrophoresis. The RNA was diluted in 1% phenol red and water to 0.8 pg/pl and was pressure injected into 1-2 cell stage embryos. The amount of injected RNA was approximately pg. All Notch-ICD injected embryos, except those that were severely deformed and/or dying, were fixed at 28 h in 4% formaldehyde and processed for antibody labeling using a 3A10 monoclonal antibody (Developmental Studies Hybridoma Bank) and a polyclonal anti-c-myc antibody (Santa Cruz Biotechnology). Antibody labeling was visualized using an anti-mouse Oregon Green (Molecular Probes) and anti-rabbit Cy3 (Jackson ImmunoResearch Laboratories Inc.) under 34

50 an Axioplan compound microscope. Only embryos with myc expression in the head and hindbrain were analyzed for 3A10 staining. Results In a mutagenesis screen designed to identify genes involved in motor axon guidance (Beattie et al., 1999), we isolated a mutation, b420, that had abnormal motor axons (see below) and that also caused defects in somite formation similar to those described for fused-somite-type mutants; a class including fused somites (fss), bea, des, aei, and mib (van Eeden et al., 1996, 1998). To determine whether b420 was an allele of any of these mutations, we performed complementation tests by pairwise matings between heterozygous b420 and heterozygous mib a52b, bea tm98, aei tr233, or des tp37 embryos. In crosses between heterozygous b420 and des tp37, approximately 25% of the embryos displayed the mutant phenotype indicating that these were mutations in the same gene (2 crosses; 403 embryos). b420 crossed with other mutants yielded only wild-type embryos. Consistent with zebrafish nomenclature (Mullins, 1995), we named b420, des b420. Unlike the other ten alleles isolated (van Eeden et al., 1996), and another allele described here, des b638, des b420 is a late embryonic lethal; larvae fail to form swim bladders and die at 9-10 days post fertilization (d). 35

51 Somitogenesis is disrupted in des b420 mutants While the first approximately 6 somites appeared normal at 18 h in des b420 mutants, the remainder formed incomplete somite boundaries (Fig. 2.1, Table 2.1). The number of correctly formed somites increased slightly over time with somite boundaries present for the first 8 or so somites at 48 h, suggesting that 2-3 somite boundaries may recover (Table 2.1). To examine the des b420 somite defect in more detail, we analyzed expression of a number of genes by RNA in situ hybridization. All of these genes were expressed in patterns consistent with those previously described for des mutants (Fig. 2.2; van Eeden et al., 1996; Durbin et al., 2000; Holley et al., 2000; Jiang et al., 2000; Sawada et al., 2000). In particular, the normal bands of her1 expression in the unsegmented mesoderm (Muller et al., 1996) were not distinct (Fig. 2.2 A, D). myod, which is normally expressed in the posterior region of developing myotomes (Weinberg et al., 1996), was expressed throughout myotomes (Fig. 2.2 B, E). However, myod expression was normal in rostral myotomes of des b420 mutants where somite boundaries were present (arrow in Fig. 2.2 E). Genes expressed in the anterior portion of somites were also perturbed. For example, mespa, normally expressed in the anterior region of presumptive somites (Durbin et al., 2000; Sawada et al., 2000) lacked this restricted expression pattern in des b420 mutants (Fig. 2.2 C, F). These data, as discussed previously (Durbin et al., 2000; Sawada et al., 2000), indicate that cells with anterior and posterior fates are not segregated within des mutant somites, but are mixed together resulting in no clear distinction between anterior and posterior somite and myotome regions. 36

52 Spinal motor axons are disorganized in des b420 mutants des b420 was originally identified by its motor axon defect (Beattie et al., 1999). znp1, which labels a subset of axons, antibody, labeling at 24 and 48 h revealed that primary motor axons and motor nerves branched aberrantly and lacked their stereotyped morphology in segments posterior to somite 7 (Fig. 2.3). In the rostral trunk where somite and myotome formation was normal in des b420, motor axons were also normal (data not shown). To characterize axonal morphology in more detail, we labeled individual caudal and middle primary motoneurons (CaP and MiP) in spinal cord hemisegments corresponding to myotomes 8-10 where somite formation was aberrant, at 22 h with vital fluorescent dye and visualized their axonal projections. We found that axons of both ventrally (CaP) and dorsally (MiP) projecting primary motor neurons had abnormalities including excessive branching, short axons and bifurcating axons (Fig. 2.4). To determine whether the defect in motor axons was cell-autonomous, we created genetic mosaics between wild-type and des b420 mutant embryos. Single CaP motoneurons were transplanted from labeled donor embryos (either wild type or mutant) into unlabeled host embryos (either wild type or mutant) before axogenesis and followed over time in living embryos to determine axonal morphology. In control transplants, consisting of wild-type CaPs transplanted into wild-type host embryos, the transplanted CaP always exhibited a normal morphology (Fig. 2.5 A). When wild-type CaP motoneurons were transplanted into des b420 mutant hosts, however, they had an abnormal axonal morphology (Fig. 2.5 B). CaP motoneurons from des b420 mutants transplanted into wild-type 37

53 host embryos exhibited normal CaP axonal morphology (Fig. 2.5 C). These experiments demonstrate that des function is not required in motoneurons for the development of normal axonal trajectories. Previous studies have shown that disrupting somitogenesis alters formation of motor axons (Eisen and Pike, 1991) and that anterioposterior myotome patterning may affect motor axon guidance (Bernhardt et al., 1998; Roos et al., 1999). Therefore, the most likely cause of the motor axon phenotype in des b420 mutants is disrupted anterioposterior myotome patterning. Trunk neural crest migration is selectively disrupted in des b420 mutants In higher vertebrates, mesoderm patterning influences neural crest migration. In chick embryos, for example, neural crest cells migrate exclusively along the rostral sclerotome (Keynes and Stern, 1984; Rickmann et al., 1985; Bronner-Fraser, 1986; Teillet et al., 1987; Loring and Erickson, 1987). Ablation of the entire somite or experimental manipulations that result in rostral-only or caudal-only sclerotome result in disrupted neural crest migration and dorsal root ganglion (DRG) formation (Kalcheim and Teillet, 1989). In zebrafish, neural crest cells migrate ventromedially (referred to as the medial pathway) between somite and notochord and ventrolaterally between myotome and skin (referred to as the lateral pathway) (Raible et al., 1992). Neural crest cells migrating along the medial pathway give rise to DRG neurons, sympathetic neurons, enteric neurons, glial cells, and pigment cells, whereas those migrating on the lateral pathway, only give rise to pigment cells (Raible and Eisen, 1994). 38

54 Since somite and myotome patterning are disrupted in des mutants, we asked whether neural crest migration was also affected. crestin mrna is expressed in migrating neural crest cells (Rubinstein et al., 2000), but is much more strongly expressed in cells migrating along the medial pathway; pigment cells migrating along the lateral pathway begin to differentiate as they migrate (Raible et al., 1992) and concomitantly downregulate crestin expression (Luo et al., 2001). Analysis of crestin expression in des b420 mutants revealed that in the first approximately 6-7 somites, the pattern of crestin expression along the medial pathway was comparable to that seen in wild-type embryos (data not shown). crestin-expressing cells in segments posterior to somite 7, however, lacked the striking segmental pattern observed in wild-type embryos (Fig. 2.6 A, B). We also examined the gene, dct, which is expressed in differentiating melanocytes as they migrate along both the medial and lateral pathways (Kelsh et al, 2000). In wild-type embryos, dct-expressing cells on the medial pathway are segmentally patterned; however, this pattern is more subtle than that observed with crestin (Fig. 2.6 A, E; Lou et al., 2000; Kelsh et al., 2000). We found that the pattern of dct-expressing cells was similar between wild-type and mutant embryos at 28 h on the medial pathway, although the segmental arrangement of dct-expressing cells in mutants was slightly less regular (Fig. 2.6 E, F). We also found that dctexpressing cells along the lateral pathway in des b420 mutants looked comparable to wild-type dct-expressing cells along the lateral pathway (data not shown). To determine if neural crest derivatives were affected, we examined mutants at 3-4 d. Approximately, 35% of DRG neurons failed to coalesce into ganglia resulting in misplaced neurons (Fig. 2.6 C, D; see also Fig. 2.11). We did 39

55 not see any defect in cervical sympathetic ganglion or enteric neurons. These latter cell populations originate from the vagal/anterior trunk crest and migrate along the anterior 1-5 somites (Epstein et al., 1994; Raible et al., 1992) which are unaffected in des b420 mutants. When we analyzed the pattern of all three pigment cell types, melanophores, iridophores, and xanthophores at 3-5 days in mutant embryos, they looked indistinguishable from the pattern in wild-type embryos (Fig. 2.6 G, H and data not shown). Due to the lack of molecular markers for glial cells in zebrafish, we cannot determine whether this population is affected in des b420 mutants. These data suggest that there are non-pigment neural crest cell precursors along the medial pathway in des b420 mutants whose segmental pattern of migration is disrupted. Since neural crest cells that give rise to DRG neurons migrate along the medial pathway, it is possible that this migration defect leads to the failure of the DRG to form correctly at 3 d. Neural crest migration along the lateral pathway appears unaffected. des b420 mutants have a neurogenic phenotype Two other fused somite-type zebrafish mutants, mib and aei/deltad have somite defects similar to des, but also exhibit neurogenic phenotypes (Jiang et al., 1996; Schier et al., 1996; Holley et al., 2000). mib mutants exhibit a strong neurogenic phenotype, resulting in supernumerary primary neurons throughout the CNS and periphery and a concomitant reduction in secondary motoneurons, melanocytes, and eye and hindbrain radial glial cells (Schier et al., 1996; Jiang et al., 1996). A thorough characterization of the neurogenic phenotype in aei/deltad 40

56 mutants has not been reported; however, they do have increased numbers of spinal primary sensory neurons (Holley et al., 2000). To determine whether des b420 mutants also had a neurogenic defect, we examined numerous cell types within the nervous system. Primary motoneurons: We used RNA in situ hybridization with anti-sense islet1 and islet2 probes to examine spinal motoneurons (Appel et al., 1995; Inoue et al., 1994). At 20 h we found a subtle, but statistically significant increase in the number of motoneurons expressing islet2 in des b420 mutants (Fig. 2.8). islet2 expression is usually limited to 1-2 ventral cells per spinal hemisegment, corresponding to CaP and VaP motoneurons (Appel et al., 1995). In wild-type embryos, only 4% of the hemisegments had more than two cells in this location whereas in des b420 mutants approximately 20% of the hemisegments contained more than two ventral, islet2-expressing cells. This result was corroborated using islet1, which is expressed in MiP and RoP motoneurons at this stage of development (Appel et al., 1995). Again, we saw a small but statistically significant increase in islet1-expressing cells in the ventral spinal cord of des b420 mutants (Fig. 2.8). Reticulospinal neurons: The embryonic zebrafish hindbrain contains a set of individually identified, primary reticulospinal interneurons that have characteristic cell body positions within rhombomeres 1-7 and stereotyped axonal projections that descend within the spinal cord (Metcalfe et al., 1986; Mendelson, 1986). Previous experiments demonstrated that perturbing Delta function in zebrafish by expression of wild-type or dominant negative forms of X-Delta-1 altered the number of Mauthner neurons, a conspicuous reticulospinal 41

57 neuron present in hindbrain rhombomere 4 (Haddon et al., 1998), suggesting that Notch-Delta signaling determines Mauthner cell number. We visualized the primary reticulospinal neurons in des b420 mutants both by retrograde labeling (data not shown) and using a neurofilament monoclonal antibody (RMO44; Lee et al., 1987). We observed a dramatic increase in Mauthner cells present in rhombomere 4 and a significant increase in RoL2 cells present in rhombomere 2 and MiD3cm cells present in rhombomere 6 (Fig. 2.9). Other reticulospinal neurons, such as those present in odd-numbered rhombomeres 3, 5 and 7, and the T-interneurons in the caudal hindbrain were unaffected in des b420 (Table 2.2 and data not shown). We also examined reticulospinal neurons in two other des alleles; tp37 (Jiang et al., 1996 and b638 (this report). We found that all alleles had an increase in Mauthner cells, RoL2 cells and Mid3cm cells. b420 mutant embryos, however, had the greatest increase in Mauthner and Mid3cm cells, while tp37 mutant embryos were the least affected (Table 2.3). Indeed, the difference in Mauthner cell number between these two alleles was statistically significant (P< 0.001; Table 2.3). Therefore, b420 may be a stronger allele compared to tp37 and b638. We examined other cells in des b420 mutants and found them to be unaffected (Table 2.2). Many of these cell types are increased or decreased in number in mib mutants (Schier et al., 1996; Jiang et al., 1996). Therefore, des mutant embryos exhibit a restricted neurogenic phenotype with only particular populations of neurons affected. 42

58 Reciprocal effects on primary and secondary sensory neurons in des mutants Rohon-Beard (RB) cells are primary sensory neurons present in the dorsal spinal cord (Lamborghini, 1980; Metcalfe et al., 1990) which, like the neural crest cells that differentiate into DRG neurons, are derived from the lateral edge of the neural plate. It has been suggested that RB cells and neural crest form an equivalence group and their fates are determined by Notch-Delta signaling during development (Cornell and Eisen, 2000). To determine whether des b420 mutants had a defect in this process, we examined RB and DRG neurons in mutant embryos. RNA in situ hybridization with an islet2 antisense probe at 20 h revealed that RB cells were decreased in des b420 mutants (Fig A, B,Fig. 2.12). To ensure that this reduction was not allele specific, we examined RB cell number in two other alleles, tp37 and b638, and found a similar reduction in RB cell number (see Fig.2.12 legend for cell counts). Using the anti-hu antibody (Henion et al., 1996), we examined DRG neurons at 72 h. We found that in addition to the previously described defect in DRG ganglion formation, there was an increase in the number of DRG neurons in mutant embryos compared to wild types (Fig C, D, Fig. 2.12). To determine if this increase was specific for DRG neurons, we examined other trunk neural crest derivatives. We found no difference in the number of trunk melanocytes, iridophores, or enteric neurons (Table 2.2). There was also no obvious difference in the number of xanthophores, and cervical sympathetic ganglion neurons; however, these cells cannot be readily counted. Thus, the only trunk neural crest derivative that appears to be increased in des b420 mutants is DRG neurons. 43

59 Supernumerary Mauthner cells are born at the same stage in des b420 mutants and wild types There are at least two scenarios for how supernumerary neurons could arise during development; a defect in cell-cycle regulation or a defect in lateral inhibition. If excess Mauthner cells in des b420 mutants were produced by a defect in cell-cycle regulation, we would predict that cells would be generated over a protracted period of time. Alternatively, if the des mutant phenotype were caused by a defect in lateral inhibition, we would predict that all of the Mauthner cells might be born at the same time, since lateral inhibition occurs within a cluster of cells with equivalent developmental potential (see Greenwald, 1998). Thus, a disruption in lateral inhibition would result in the entire cluster of cells acquiring the same fate at presumably the same time. To differentiate between these two possibilities, we examined the birthdate of Mauthner cells in des b420 mutants using BrdU incorporation followed by RNA in situ hybridization with an anti-sense probe to val (Moens et al., 1998) to identify Mauthner cells. Wild-type Mauthner cells are born at approximately 7.5 h (Mendelson, 1986); the cell cycle at this time is 1-3 hrs (Kimmel et al., 1994). We performed BrdU incorporation at 7 h, 9 h, 11 h, and 14 h for 45 minutes followed by val RNA in situ hybridization at 19 h. If a Mauthner cell precursor were still dividing at any of these times, then we would expect the Mauthner cells it generated to be both BrdU and val positive. A Mauthner cell that has completed its last S-phase will only be val positive. In wild-type embryos, valexpressing Mauthner cells incorporated BrdU at 7 h and 9 h and did not incorporate BrdU at 11 h and 14 h (Fig. 2.13). All Mauthner cells examined in 44

60 des b420 mutants incorporated BrdU at 7 h (Fig. 13). At 9 h, 78% were both val and BrdU positive and at 11 h and 14 h, all val positive cells were BrdU negative. Thus, Mauthner cells in des b420 mutants were not generated over a protracted time period, but underwent their final cell division at a time coincident with the production of Mauthner cells in wild-type embryos. These data are consistent with the hypothesis that supernumerary Mauthner cells in des b420 mutants result from a defect in lateral inhibition. Activated Notch rescues the Mauthner cell phenotype in des b420 mutants To further investigate whether the neurogenic defect in des b420 mutants is caused by a disruption in lateral inhibition, we asked if an activated form of Notch could rescue the mutant phenotype. When a constitutively active form of Xenopus Notch lacking its extracellular domain was expressed in Xenopus embryos, it inhibited primary neurogenesis (Notch ICD or Notch E; Coffman et al., 1993; Chitnis et al., 1995). Injecting RNA encoding Notch E into zebrafish embryos also blocked primary neurogenesis (Haddon et al., 1998). We focused our analysis on Mauthner cells, as they are the most dramatically affected neuronal cell-type in des b420 mutants. In wild-type embryos, there is one Mauthner cell on both the right and left sides of rhombomere 4. Even though we analyzed embryos that expressed the myc-tagged Notch-ICD RNA throughout the brain, we analyzed each side separately since exogenous RNA may be distributed differently between the two sides. When embryos from heterozygous des b420 matings were mock-injected with phenol red (a tracer for 45

61 injections) at the one- to two-cell stage, the expected ratio of mutants and wildtypes was observed with approximately 75% of the sides possessing one Mauthner cell and the remaining 25% of the sides having between 4-9 Mauthner cells (Fig 2.14, Table 2.4). When embryos from wild-type matings were injected with RNA encoding Notch-ICD, we found that 34.8% of the sides had zero Mauthner cells whereas 65.2% of the sides had the expected single Mauthner cell (Fig. 2.14, Table 2.4). When embryos from heterozygous des b420 matings were injected with RNA encoding Notch-ICD, only 5.5% of the sides had more than 4 Mauthner cells while 19.6% had 2 or 3 Mauthner cells (compare to 0% of the sides having this number in mock-injected embryos from heterozygous des b420 matings). Moreover, 12.4% of the sides had no Mauthner cells (Table 2.4). Although we were not able to determine the genotype of injected embryos, approximately 17% had 2 or more Mauthner cells on one side and zero or 1 Mauthner cell on the other side (Fig. 14 D). Since wild-type embryos never have more than one, whereas mutants always have more than 3 Mauthner cells per side (Fig A, B), these embryos are likely homozygous mutants that have a decreased number of Mauthner cells due to the presence of activated Notch. Therefore, using an activated form of Notch, we were able to decrease the number of Mauthner cells in des b420 mutant embryos supporting the idea that des acts in the Notch-Delta signaling pathway, at the level of or upstream of, the Notch receptor. 46

62 Discussion Here we describe the characterization of a mutant allele of the zebrafish des gene. In agreement with other work (van Eeden et al., 1996; Durbin et al., 2000; Sawada et al., 2000), our studies show that des is critical for anterioposterior patterning within the somite and myotome. Our genetic mosaic experiments support the idea that this mispatterning affects motor axon outgrowth. The normal segmented pattern of neural crest migration is aberrant for some populations of neural crest cells migrating along the medial pathway. Moreover, DRG are disorganized suggestive of a correlation between these two defects. Our finding that des mutants have a restricted neurogenic phenotype suggests that des functions to regulate the number of neurons in specific regions of the central and peripheral nervous system. Spatial localization of des expression or temporal control of des expression during development could generate this specificity. To investigate whether des functions in lateral inhibition via the Notch-Delta pathway, we determined when the supernumerary Mauthner cells in des mutants were born and whether activated Notch could rescue the Mauthner cell phenotype. Data from both of these experiments support the hypothesis that des functions in subsets of cells where it is involved in Notch- Delta signaling during Mauthner cell fate specification. Although des b420 is lethal, all the other alleles of des are viable (van Eeden et al., 1996), including the b638 allele also used in this study. Since all of these mutations, except the b638 allele, were induced by ENU mutagenesis of spermatogonial stem cells, they are likely point mutations or small deletions 47

63 (Knapik, 2000). b638 was induced by ENU mutagenesis of post-meiotic sperm (Riley and Grunwald, 1995). Since the phenotypes of the different alleles are similar, it is possible that another gene is affected in des b420 which causes the lethality. Motor axon outgrowth and neural crest migration in des b420 mutants In mouse and chick, anterioposterior patterning of the sclerotomal compartment of the somite functions to pattern motor nerves and neural crest (Keynes and Stern, 1984; Rickmann et al., 1985; Bronner-Fraser, 1986; Teillet et al., 1987; Loring and Erickson, 1987; Krull and Koblar, 2000). In contrast, zebrafish sclerotome does not appear important for motor nerves or neural crest migration, rather the myotome appears to function in this capacity (Morin-Kensicki and Eisen, 1997). Myotome is the largest component of the zebrafish somite (Kimmel et al., 1995) and gene expression patterns have revealed anterioposterior patterning within this tissue (reviewed in Stickney et al., 2000). Motor axons extend along developing myotomes during the first day of development and Bernhardt et al., (1998) have shown that the CaP axon extends along the anterior region of the medial myotome immediately adjacent to the border between anterior and posterior myotome. Because somitic anterioposterior polarity is severely disturbed in des mutants, we were not surprised to see a pathfinding defect. We extend this observation using genetic mosaic analysis to demonstrate that the motor axon defect in CaP is non-cell autonomous. Because motor axons extend directly out of the spinal cord and onto the myotome, it is likely that the 48

64 motor axon defect is caused by the lack of distinction between anterior and posterior myotome regions. However, we have not ruled out the possibility that other tissues are involved. Zebrafish neural crest cells also migrate along the somite and developing myotome (Raible et al., 1992). The neural crest defect in des b420 mutants is intriguing as certain populations of neural crest are affected while others are not. For example, streams of crestin-expressing cells migrating along the medial pathway lack their segmental pattern. DRG neurons, which are derived from neural crest cells that migrate along the medial pathway (Raible and Eisen, 1994), are misplaced in mutant embryos and often fail to coalesce into ganglia. However, dct-expressing cells migrating along the medial pathway apparently retain their segmental migration pattern and their derivatives, melanocytes, are unaffected in mutant embryos. One interpretation of these data is that neural crest cells migrating along the medial pathway that give rise to DRG neurons are affected, either cell autonomously or non-cell autonomously, in des b420 mutants whereas cells that give rise to melanocytes are unaffected. This hypothesis is supported by data showing that there are populations of fate-restricted neural crest cells (Raible and Eisen, 1994; Henion and Weston, 1997; Ma et al., 1998) and some zebrafish trunk neural crest cells become lineage-restricted before they reach their final location (Raible and Eisen, 1994). Neural crest specified to become DRG neurons may be more sensitive to the disruption of somitic mesoderm anterioposterior patterning than neural crest cells specified to become 49

65 melanocytes. Even without evoking fate restriction, it is possible that there is differential sensitivity to disrupted anterioposterior pattern by unique progeny of multipotent stem cells (see Anderson, 1997). des regulates neuronal cell number Unlike the zebrafish mutant mib, which contains a dramatic excess of neurons throughout the nervous system (Jiang et al., 1996; Schier et al., 1996), des b420 has a different neurogenic profile. In the hindbrain, only the number of Mauthner cells is increased dramatically. Two other hindbrain reticulospinal neurons, RoL2 and MiD3cm show a modest increase in number, whereas other reticulospinal neurons appear unaffected. In the des b420 mutant trunk, there is also an increase in the number of primary motoneurons and DRG neurons and a decrease in the number of RB neurons. Our finding that only certain subpopulations of neurons are regulated by des is intriguing. If des is acting in the Notch-Delta signaling pathway, we speculate that it only acts in a subpopulation of cells. Analysis of zebrafish notch and delta genes reveal both overlapping and unique patterns of expression throughout the nervous system and mesoderm (Bierkamp and Campos-Ortega, 1993; Dornseifer et al., 1997; Westin and Lardelli, 1997; Appel and Eisen, 1998; Haddon et al., 1998; Smithers et al., 2000). Thus, mutations in these genes or in their specific downstream components could cause a restricted neurogenic phenotype like that seen in des mutants. Eventual cloning of the des gene and localization of its product will clarify how des function is spatiotemporally regulated. 50

66 des regulates the number of trunk sensory neurons The relationship between RB cells and neural crest has recently been investigated by examining deltaa mutant embryos. Cornell and Eisen (2000) found that in deltaa mutants there was supernumerary RB cells and a concomitant decrease in trunk neural crest cells and derivatives, including DRG neurons and melanocytes. These results parallel what was found when a dominant negative form of delta, that apparently reduces function of all Delta proteins, was expressed in embryos (Haddon et al., 1998; Appel and Eisen, 1998). Moreover, aei/deltad mutants also have supernumerary RB cells (Holley et al., 2000). These findings indicate that the function of DeltaA and DeltaD is to limit the number of cells that take on the RB cell fate. We also found an effect of des on the number of RB cells and DRG neurons in mutants; but surprisingly, it was in the opposite direction. Throughout the dorsal spinal cord we found a sporadic decrease in RB cells; a curious finding because other neurogenic mutants have increased numbers of RB cells (Holley et al., 2000; Cornell and Eisen, 2000). Nevertheless, the molecular regulation of fate decisions in the putative RB-neural crest equivalence group is largely unknown. Given the role of Notch-Delta signaling in other cellular contexts, it is possible that if des is a member of this signaling pathway, it could play no role in fate specification in the RB-neural crest equivalence group. This would be consistent with the fact that the number of neural crest cells is not qualitatively decreased in des mutants. An alternate explanation for the observed des phenotype may be more similar to the role of Notch signaling in the maintenance of neural precursors in mammalian cerebral cortex. In this scenario, the reduced number 51

67 of RB neurons observed in des mutants may result from a reduction in RB precursors due to a lack of des function. Raible and Eisen (1996) showed that interactions, such as lateral inhibition, among neural crest cells influence cell fate. Moreover, in chick, Morrison et al., (2000) showed that activated Notch causes a decrease in the number of DRG neurons and an increase in glial cells whereas inhibiting Notch function enhanced neurogenesis. Thus, because there is precedent for Notch- Delta signaling functioning in cell fate determination among neural crest cells, it is not surprising that decreasing des, which acts in the Notch-Delta pathway, also affects cell fate determination within the neural crest. Thus alterations in Notch signaling are likely to result in a change in the neural crest cell population that could be reflected in the number of different derivatives. The increase in DRG neurons in des mutants that display normal melanogenesis, is consistent with previous studies of Notch signaling in neural crest development. Therefore, it is possible that the RB-DRG phenotype results from the normal role of des during two different events, first the development of RB neurons and second during lateral inhibition activity regulating neuron and glia fate decisions amoung neural crest derived cells. The number of Mauthner cells is controlled by des-mediated lateral inhibition Previous studies suggest that within the hindbrain, more than one cell can take on the Mauthner cell fate (reviewed in Kimmel and Model, 1978) supporting the idea that lateral inhibition within an equivalence group regulates the number 52

68 of Mauthner cells. More recent experiments demonstrate a direct role for Notch- Delta mediated lateral inhibition in Mauthner cell generation. Haddon et al. (1998) blocked Notch-Delta signaling by over-expressing a construct encoding a dominant negative form of delta (X-Delta1 dn ) and found extra Mauthner cells in the hindbrain. Conversely, over-expressing full-length delta (X-Delta1) resulted in a reduction in Mauthner cell number. In this study, we find that injecting RNA encoding an activated form of Notch into embryos at the 1-2 cell stage, causes a decrease in Mauthner cells in both wild-type embryos and embryos from heterozygous des b420 matings. Our interpretation of these results is that des is a gene that functions in lateral inhibition either epistatic to, or at the level of, Notch and Delta. Cells expressing Notch-ICD fail to take on the primary, Mauthner cell fate and instead take on the secondary, non-mauthner cell fate. At this time, we do not know what these cells become. We were never able to completely eliminate Mauthner cells in embryos from des b420 matings; however, in 17% of wild-type embryos we did eliminate both Mauthner cells. When Xenopus X-Delta1 was overexpressed, 27% of injected wild-type zebrafish lacked both Mauthner cells (Haddon et al., 1998). The discrepancy in these numbers may be due to the effectiveness of the gene in suppressing Mauthner cell-fate specification or the concentration of RNA. We did not analyze embryos that were overtly aberrant suggestive of overly high concentrations of injected RNA. Analysis of these embryos would most likely have increased the number of fish observed with no Mauthner cells. We also 53

69 suspect that it may take a higher concentration of Notch to decrease the greater number of Mauthner cells seen in mutant embryos compared to wild-type embryos. Multiple Notch-Delta signaling pathways In vertebrate species, there are multiple Notch and Delta proteins as well as upstream processing proteins and downstream effector molecules (Artavanis- Tsakonas et al., 1999). The sheer number of these proteins suggests that specific interactions and pathways exist that function during particular times and/or regions within the developing nervous system. mib mutants have a very dramatic neurogenic phenotype throughout the nervous system suggesting that Mib controls a common aspect of lateral inhibition. Des, in contrast, is a protein required for the generation of particular neuronal types within the hindbrain and spinal cord. This specificity may be due to a spatial localization of des function or temporal windows in the requirement for Notch-Delta signaling. It is intriguing that the reticulospinal neurons affected in des b420 mutants reside within even numbered rhombomeres. In general, neurogenesis in even numbered rhombomeres precedes neurogenesis in odd numbered rhombomeres in vertebrates (Lumsden and Keynes, 1989). Thus, the specificity we observe may reflect timing as opposed to spatial requirements. It may be that the complexity of the nervous system calls for molecular diversity in order to produce a wide array of neurons. Des appears to function in this process by mediating lateral inhibition within particular cell populations. 54

70 Figure 2.1: Somite boundary formation is aberrant in des b420 mutants. Lateral views of live 18 h wild-type (A, B, C) and mutant (D, E, F) embryos at the level of anterior trunk (segments 2-5; A, D), mid-trunk (segments 8-12; B, E) and tail (segments 14-18; C, F). Black arrowheads denote normal somite boundaries and white arrowheads denote incomplete somite boundaries. In this and all subsequent lateral views, anterior is to the left and dorsal is to the top. Scale bar, 25 µm. 55

71 Age (h) (n=10) Abnormal somites Beginning at somite number ± ± ± ± ± 0.6 Table 2.1: Analysis of aberrant somites in des b420 mutants Somites were analyzed along the entire rostrocaudal axis with the most rostral somite being number 1. A somite was considered normal if both the anterior and posterior boundaries were complete. n= 10 embryos analyzed for each time point. Numbers are represented at mean ± 95% confidence interval. 56

72 Figure 2.2: Somite and myotome gene expression is perturbed in des b420 mutants. Dorsal views (anterior to the top) of whole mount RNA in situ hybridization of her1 (A, D; 13 h), myod (B, E; 15 h) and mespa (C, F; 13 h) in wildtype (A, B, C) and mutant (D, E, F) embryos. Black arrowheads denote bands of gene expression in wild-type embryos and the corresponding region in mutants. Arrow in E designates normal myod expression in anterior somites. Scale bar, 150 µm (A,C, D, F), 50 µm (B, E). 57

73 Figure 2.3: Motor nerves are disorganized in des b420 mutants. Lateral views of whole mount embryos labeled with the znp1 monoclonal antibody at 24 h (A, C) and 48 h (B, D) in wild-type (A, B) and mutant (C, D) embryos. Arrowheads point to one set of axons in each panel. Scale bar, 50 µm. 58

74 Figure 2.4: CaP and MiP motor axons display abnormalities in des b420 mutants. Lateral view of pseudocolor images of live CaP (A, B) and MiP (C, D) motoneurons labeled with rhodamine dextran and visualized at approximately 26 h in wild-type (A, C) and des b420 mutant (B, D) embryos. Cell bodies reside in the spinal cord and white lines denote the dorsal and ventral aspects of the notochord (A, B) and the dorsal aspect of the notochord (C, D). Scale bar, 10 µm. 59

75 Figure 2.5: des b420 is non-cell autonomous for CaP motoneurons. Single CaP motoneurons were removed from rhodamine dextran labeled donors at approximately 16 h, a time before axogenesis, and transplanted into stagematched, unlabeled host embryos from which native CaP and VaP motoneurons had been removed. (A) Wild-type CaP transplanted into wild-type host (n=4). (B) Wild-type CaP transplanted into a des b420 mutant host (n=4). (C) des b420 mutant CaP transplanted into wild-type host (n=4). White lines denote dorsal and ventral aspects of the notochord. Scale bar, 10 µm. 60

76 Figure 2.6: Migrating neural crest and neural crest derivatives are differentially affected in des b420 mutants. Lateral view of mid-trunk (segments 9-14) neural crest on the medial pathway as revealed by crestin expression (arrowheads) at 24 h in (A) wild-type and (B) mutant embryos. Cross section view of Hu-positive DRG neurons (arrowheads) in 72 h (C) wild-type and (D) mutant embryos. Lateral view of mid-trunk (segments 9-14) dct-expressing cells in (E) wild-type and (F) mutant embryos along the medial pathway. Pigment cells positioned in the lateral stripe in 3 d (G) wild-type and (H) mutant embryos. Scale bar, 50 µm (A, B); 30 µm (C-F); 75 µm (G, H). 61

77 62

78 Figure 2.7: des b420 mutants have excess primary motoneurons. CaP and VaP motoneurons (arrowheads) as revealed by islet2 RNA in situ hybridization at 20 h in a (A) wild-type and (B) mutant embryo. 63

79 Figure 2.8: Number of primary motoneurons in des b420 mutant embryos. Cell counts of islet2 and islet1 expression at 20 h in spinal hemisegments 5-9; n=10 embryos for each point. For these and all subsequent histograms the mean ± 95% confidence interval is plotted. Significance was determined by Student t-test with *P = , **P< Scale bar, 25 µm. 64

80 Figure 2.9: Supernumerary reticulospinal neurons are present in des b420 mutants. Dorsal view of a confocal image (anterior to the top) of RMO44 labeled (A) wild type and (B) des b420 mutant embryo. Somata of affected cells are labeled in A and arrowheads point to their corresponding axons. Note the increased number of axons in the mutant embryo. Performed in collaboration with Dr. Cecilia Moens. 65

81 Figure 2.10: Number of supernumerary reticulospinal neurons present in des b420 mutants. Cell counts of affected reticulospinal neurons. Cells on both sides of the midline were analyzed for wild type (n=14) and mutant (n=14) embryos. See Figure 7 legend for details regarding statistics. Mth, Mauthner cell. Scale bar, 20 µm 66

82 Cell Type Age Probe wt b420 Trigeminal ganglion neurons 26 h anti-islet1 antibody 28.5±1.8 28±1.5 Ear sensory hair cells $ 28 h anti-acetylated tubulin antibody 8.0± ±0.6 Statoacoustic neurons 24 h anti-islet1 antibody 16.6± ±0.9 Hindbrain radial glial # 38 h Zrf1 antibody 6.0± ±0.0 Hindbrain T-cells 48 h anti-neurofilament antibody 4.2± ±0.5 Enteric neurons* 5 d anti-hu antibody 7.5± ±0.6 Trunk 24 h dct mrna 47.0± ±5.4 Trunk iridophores 4 d NA 40.5± ±4.1 Table 2.2: Cell types unaffected in des b420 mutants Cell types were analyzed at specific stages (Age) using either RNA in situ hybridization or immunohistochemistry (Probe). Under these conditions, we found no significant difference between the number of cells in mutant and in wild-type embryos. Cells were counted in 10 embryos/larvae except for hindbrain T-cells (n=14) and statoacoustic neurons (n=7). $ Ear sensory hair cells were counted in both the anterior and posterior maculae. # Hindbrain radial glial fiber bundles are present in pairs between rhombomeres. The number of individual glial fibers in each bundle was not quantitated. * Hu-positive neurons encircling the gut were counted in 5 cross sections per larvae from the mid-trunk Trunk melanocytes were counted on one side of each embryo along the entire length of the trunk. Trunk iridophores were counted on one side of each embryo along the entire length of the trunk using indirect light. 67

83 Embryos (n) Mauthner RoL2 Mid3cm wt (19) 2.0 ± ± ± 0.3 b420 (14) 9.6 ± ± ± 0.6 b638 (16) 8.6 ± ± ± 0.5 tp37 (24) 7.0 ± ± ± 0.5 Table 2.3: Two other des alleles have supernumerary reticulospinal neurons. Embryos were labeled at 48 h with the anti-neurofilament antibody, RMO44, and axons originating from cells on both sides of the midline were counted. Numbers are reported as mean ± 95% confidence interval. Differences in the mean cell counts of Mauthner cells, RoL2 cells and Mid3cm cells of the three alleles compared to wild-type embryos were statistically significant (P<0.01- P<0.001). The difference in the means between the number of Mauthner cells in b420 and tp37 was also statistically significant (P<0.001). All other differences between the three alleles were not statistically significant. sd=standard deviation. 68

84 Figure 2.11: Analysis of trunk sensory neurons in des b420 mutants. Dorsal view (anterior to the left) of a 20 h (A) wild-type and (B) mutant embryo showing islet2 expressing RB neurons. Arrowheads denote the absence of RB neurons in the spinal cord of des b420 mutants. All three des alleles examined had a similar decrease in RB cells (b420=2.5±0.2, b638 =2.6±0.2, and tp37=2.3±0.1 compared to wild type=3.1±0.2). Lateral view of 72 h (C) wild-type and (D) mutant embryo showing Hu-expressing DRG neurons (arrowheads). (E) islet2 expressing RB neurons and Hu expressing DRG neurons were counted in hemisegments 5-9; n= 10 embryos for RB counts and n=6 embryos for DRG counts. See Figure 7 legend for details regarding statistics. Scale bar, 20 µm 69

85 Figure 2.12: Number of sensory neurons in des b420 mutants. islet2 expressing RB neurons and Hu expressing DRG neurons were counted in hemisegments 5-9; n= 10 embryos for RB counts and n=6 embryos for DRG counts. See Figure 7 legend for details regarding statistics. Scale bar, 20 µm 70

86 Figure 2.13: Supernumerary Mauthner cells in des b420 mutants are formed at the same time as Mauthner cells in wild-type embryos. Numerical representation of Mauthner cell birthdays in wild-type and des b420 mutants. At 7 h, all Mauthner cell precursors examined in wild-type (n=13) and des b420 (n=64) were undergoing DNA synthesis. At 9 h, all wild-type (n=16) and 78% of mutant (n=46) Mauthner cell precursors were still dividing. At 11 h and 14 h, all Mauthner cells in both wild-type (n=16, n=15) and mutant (n=65, n=87) embryos had undergone their last S-phase. val-expressing cells (n) in 10 embryos were examined and the mean ± 95% confidence interval were plotted. Due to cross sectional analysis, not all Mauthner cells present in each embryo were analyzed. 71

87 Figure 2.14: Notch rescues the des b420 Mauthner cell defect. Embryos from wild-type or heterozygous des b420 matings were injected at the 1-2 cell stage with RNA encoding Notch-ICD-myc (Coffman et al., 1993; Chitnis et al., 1995). Embryos were fixed at 28 h and processed for 3A10 antibody labeling to visualize Mauthner cells. (A) Mock-injected wild-type embryos have one Mauthner cell on either side of hindbrain rhombomere 4. (B) Approximately 25% of mock-injected embryos from a heterozygous des b420 mating have supernumerary Mauthner cells. Injecting RNA encoding Notch-ICD decreases the number of Mauthner cells in (C) wild-type and (D) des b420 mutant embryos. 72

88 Cross used to obtain embryos Construct injected PERCENTAGE OF EMBRYOS WITH THE INDICATED NUMBER OF MAUTHNER CELLS PER SIDE Wild-type (n = 30) Wild-type (n=46) b420 het (n=78) b420 het (n=194) Phenol red Notch- ICD phenol red Notch- ICD Table 2.4: Affect of Notch-ICD on Mauthner cell number. Wild-type embryos or embryos from des b420 heterozyogous matings (b420 het) were injected at the 1-2 cell-stage with control solution containing 0.2% phenol red or solution containing mrna encoding a myc-taggedxenopus Notch-ICD (Coffman et al., 1993; Chitnis et al., 1995). Mauthner cells were visualized at 28 h by 3A10 antibody labeling and reported as percent Mauthner cells per side. Numbers for the b420 het cross Notch-ICD are from 3 separate injection experiments. n = number of sides counted (two sides per embryo). 73

89 CHAPTER 3 MUTATIONS IN DEADLY SEVEN REVEAL DEVELOPMENTAL PLASTICITY IN THE ESCAPE RESPONSE CIRCUIT Introduction Establishing functional neural circuits involves numerous developmental processes including the generation of the correct cell number and cell type and the establishment of the appropriate connections between cells. Mutations that alter these processes have the potential to disrupt the relationship between cell types in a circuit and can compromise downstream behavior unless the animal can adapt to such changes. Zebrafish offer several advantages as a model system for studying the impact of mutations on circuitry and behavior. In aquatic vertebrates, such as fish and premetamorphic amphibians, relatively simple neural circuits control fundamental motor behaviors. These circuits often have relatively few cells, and in some cases it is possible to identify individual cells and link them to a behavior. This is particularly true in the zebrafish hindbrain where there is an array of morphologically distinct reticulospinal neurons with characteristic cell body locations and stereotyped axon projections (Metcalfe et al., 1986). The largest of these, the Mauthner cell, is present bilaterally in 74

90 rhombomere segment 4 and sends its axon across the midline where it descends to make monosynaptic connections with contralateral inter- and motoneurons along the length of the spinal cord (Faber and Korn, 1978). Experiments in both zebrafish and goldfish have shown that activation of the Mauthner cell initiates a coordinated bend away from the direction of the stimulus, allowing the animal to swim away from a perceived threat (Zottoli, 1977). Two homologous reticulospinal neurons are also involved in the escape circuitry (Lui and Fetcho, 1999); however, activation of the Mauthner cell is sufficient to trigger the behavior (Nissanov et al., 1990; Zottoli, 1977; Eaton et al. 1981) demonstrating a one to one correlation between Mauthner cell firing and the escape response. Mutations in the des/notch1a gene (van Eeden et al., 1996; Holley et al., 2002) result in supernumerary Mauthner cells (Gray et al., 2001). We took advantage of this defect to ask how the nervous system responds to specific alterations in a defined neural circuit. We find that all Mauthner cells present in des b420 mutants are active during an escape response. Quantitation of kinematics of the escape movements reveals little difference between wild-type and mutant larvae. Analysis of Mauthner axons in two des alleles, however, reveals a dramatic decrease in the number of axon collaterals compared to wild-type Mauthner axons. In addition, the number of Mauthner cells is inversely proportional to the number of axon collaterals per Mauthner cell. Mauthner axons extending down the same side of the spinal cord largely have collaterals in different positions suggesting that contacts are being formed on non-overlapping populations of cells. These data suggest that the formation of the escape response circuit is highly adaptive and can compensate for supernumerary 75

91 Mauthner cells by regulating the number of synapses formed on target cells in the spinal cord thus maintaining a normal escape response. Materials and Methods Fish Care and Identification Mutant embryos and larvae were obtained from a laboratory breeding stock of heterozygous adult fish. Homozygote mutants were obtained from natural matings of heterozygous adults and were identified by their somitic defect that results in irregular muscle patterns. Both the b420 and tp37allele carry ENU-induced mutations (van Eeden et al., 1996; Gray et al., 2001). Larval zebrafish refer to fish at 3-6 days post fertilization (d). Whole mount antibody labeling Whole mount antibody labeling with 3A10 antibody (1/10; Developmental Studies Hybridoma Bank) and acetylated tubulin (1/400; Sigma) were processed as described in Gray et al. (2001). Retrograde Labeling of calcium green dextran Larval zebrafish were anesthetized with 0.02% 3-aminobenzoic acid ethyl ester (MS222, tricaine). The targeted cells were retrogradely labeled by pressure injection via a glass microelectrode of a 50% solution of calcium green dextran 76

92 (10,000 MW; Molecular Probes) in 10% hanks into the caudal spinal cord (Fetcho and O'Malley, 1995). Injections were targeted to the ventral cord to selectively label Mauthner cells and their homologs without disrupting more dorsal sensory pathways. After injection, fish were allowed to recover in 10% Hanks. Carefully done, such injections do not alter the behavior of the fish and therefore are unlikely to have altered the normal activity in the labeled cells (Liu and Fetcho, 1999). Calcium Imaging Calcium imaging was performed as previously described (Fetcho and O Malley, 1995; O Malley et al. 1996). Nine to ten hours after labeling, fish were briefly anesthetized in tricaine, placed on a cover glass in a Petri dish, embedded on their backs in a thin layer of 1.2% agar and screened under confocal microscopy. Confocal images were obtained by looking into the head of the intact fish using a Zeiss inverted microscope with a 63X-water objective and a Zeiss laser-scanning confocal imaging system. Mauthner cells were identified by their highly characteristic morphology and position (Metcalfe et al, 1986). Escapes were elicited by a small tap from a polished glass probe attached to a piezoelectric crystal. The stimulus (tap) strength could be controlled by the amount of voltage applied to the crystal. Taps given to the head were aimed at the ear of the animal, ipsilateral to the group of cells under observation. Taps given to the tail were aimed rostral of the injection site and ipsilateral to the observed cells. 77

93 The fluorescence intensities of the cells were monitored by collecting a baseline series of images of a group of labeled Mauthner cells at 300 msec intervals before delivering the stimulus. To avoid false positives due to the cells moving to a brighter plane of focus, we started with the brightest plane of focus, optimizing for one cell at a time/per trial. While the optimized cell increased in fluorescence intensity, in most cases, the surrounding Mauthner cells could be seen to brighten as well. High-speed recording of behavior High-speed recording was performed as previously described in Liu and Fetcho (1999), except that a different escape-inducing stimulus was used. Test larvae were placed into individual Petri dishes (3.5 cm) and assigned a number designation. Escape responses were recorded under a high-speed camera that captures images digitally at 1,000 frames/seconds connected to a dissecting microscope. Escape responses were elicited by a small tap to either the head or tail of the fish delivered by a small, polished glass probe. A successful escape trial was recorded from the time the probe visually contacts the fish to when the fish swims to the edge of the field of view. Fish were tapped on the left side of the head for the first trial, then on the right side of the head for the second trial, then the left side of the tail for the third trial etc. Stimuli to the head were directed at the ear whereas stimuli to the tail were directed caudal to the anal pore. All the fish in the group were tested for the same quadrant before moving on to the next trial/quadrant resulting in each 78

94 fish resting for an average of twenty minutes between trails and each fish was not presented with a stimulus in the same quadrant for well over an hour. Such delays should prevent habituation, fatigue etc. On occasion, a fish failed to respond to the stimulus, gave a premature response, or turned slightly on its side. Only responses that occurred after the stimulus was deployed and in which the fish remained upright throughout the initial bend were kept digitally. Since des b420 homozygotes do not develop a swim bladder, we chose to collect our trials at 3 d when neither wild-type larvae nor des b420 larvae have swim bladders. Data Analysis Analyses of the recorded movement data were done as previously described in Liu and Fetcho (1999). Several kinematic parameters of the initial turn of the escape response were selected for analysis: latency to its initiation (time from the contact of glass probe to the beginning of movement), the maximum angle of the turn, its peak angular velocity, and its duration (time from beginning of movement to maximum angle). Movements were then analyzed for these kinematic parameters using a specialized program written in Labview. The analysis was automated; the image of the fish was thresholded and the silhouette of the fish was used to determine the location of the rostral midline. The midline from each successive frame was plotted to give a representation of the animal s movements. The program calculated the angle between the position of the midline in successive frames and its original position 79

95 and also provided other kinematic data such as the angular velocity. These parameters were then statistically compared using analysis of variance (SuperAnova). Single Mauthner axon labels Mauthner axons of 3 d larvae, anesthetized in tricaine, were retrogradely labeled by pressure injection into the spinal cord using a glass electrode filled with 5% lysinated rhodamine dextran (10 K MW; all flourescent dextrans used in this study were obtained from Molecular Probes) at the mid-trunk level. After labeling, larvae were allowed to recover for 24 hours. Because of the difficulty in resolving multiple fluorescently labeled axons in des b420 mutants, only larvae with one labeled Mauthner axon were scored. Collaterals were imaged and counted under confocal microscopy (63X) over a three somite length corresponding to three spinal cord hemisegments. The total number of Mauthner cells present on the corresponding side of the hindbrain was quantitated on a Zeiss Axioplan compound microscope under transmitted light (40X). Dual Color Mauthner axon labels Mauthner cells of 4 d larvae, anesthetized in tricaine, were retrogradely labeled by pressure injection into the spinal cord using a glass electrode filled with 10% fluorescein dextran (3K MW). Following a recovery of 48 hours, fish were screened for labeled Mauthner cells. These fish were anesthetized again, 80

96 placed on a cover glass in a petri dish, and embedded upright (dorsal side up) in a layer of 1.2% agar. Mauthner cells were targeted for electorporation under FITC fluorescence. For electroporation, thin-walled filamented glass microelectrodes backfilled with either 10% Rhodamine Dextran (3 K MW) or 10% Alexa 647 (10 K MW) were used to electroporate single Mauthner cells (see Haas et al., 2002). Once the electrode was in position next to a Mauthner cell body, 2-3 trains of pulses (train width: 3 sec, pulse duration: 1 msec, pulse period: 10 msec, frequency: 100 pulses/sec, amplitude: 10V) were applied using an A-M systems isolated pulse stimulator (model 2100) and a custom built electroporation apparatus. The number of pulse trains (usually 2 or 3) was determined visually by monitoring the extent of cell filling under fluorescence. Once a single Mauthner cell was filled with dye (either Rhodamine or Far Red), the positive pressure was reapplied to the electrode as it was moved out of the head. A second microelectrode, this time filled with the other dye (ie Far Red if Rhodamine was used first), was positioned in the head adjacent to a second fluorescein labeled Mauthner cell located within the same cluster of cells and electroporated using the method described above. Following electroporation, the fish were re-embedded on their sides in 1.2% agar, in tricaine, and imaged on a Zeiss 510 confocal laser scanning microscope using a Zeiss 63X water objective to image collaterals. 81

97 Results The escape response circuit in des b420 mutants The neural circuit driving the escape response in zebrafish is mediated by a defined set of neurons. Sensory input from the trigeminal (V) cranial nerve, the acoustic (VIII) cranial nerve, and lateral line contact the Mauthner cell lateral dendrite (Kimmel et al., 1981; Kimmel et al.,1990). Mauthner cells extend axons down the spinal cord and contact primary motoneurons and interneurons (Fetcho and Farber, 1988). We have previously shown that des b420 mutants have a restricted neurogenic phenotype (Gray et al., 2001). Wild-type animals possess a single pair of Mauthner cells located in hindbrain rhombomere 4 (Fig. 3.1 A). In des b420 mutants, however, the Mauthner cell is variably increased in number with between 3-8 cells present on each side representing a 4.8-fold increase in Mauthner cell number (Fig. 3.1 B; Gray et al., 2001). There is also an approximately 19% increase in primary motoneurons in des b420 mutants whereas all of the other cells in the escape response circuit appeared unaffected (Gray et al., 2001). Mauthner axons in des mutants cross the midline then extend posteriorly within the contralateral side of the spinal cord consistent with what is seen in wild-type larva (Fig. 3.1). Thus, the escape response circuit in des b420 mutants is characterized as having a dramatic increase in Mauthner cell number, without major alterations in cell body location or axon projections, with a small or no increase in the other cell types participating in the circuit. 82

98 Supernumerary Mauthner cells are active during the escape response The presence of excess Mauthner cells in des b420 mutants prompted us to ask whether all of these cells participated in the escape response. Taking advantage of the transparency of zebrafish larvae, we used optical imaging of calcium-dependent fluorescent indicators to monitor the activity of several cells simultaneously (O'Malley et al., 1996). This technique was used to determine the activity patterns of supernumerary Mauthner cells during escape responses elicited by a tap with a glass probe, deflected by a piezo-electric crystal. While some taps were administered to the head, we concentrated on escapes elicited by taps to the tail because the Mauthner cell is known to play a larger role in these responses (O Malley et al, 1996; Lui and Fetcho, 1999). Data were collected from 21 escape trials (17 to tail and 4 to head) from 6 fish. Figure 2 shows an example of a calcium imaging trial. In this case, four cells were labeled on the side being imaged (Fig. 3.2 A, arrow) and the change in fluorescence intensities was quantified (Fig. 3.2 B, Fig. 3.3). Because of potential movement artifacts, each calcium imaging trial began at the brightest plane of focus for the cell of interest (Fetcho and O Mallley, 1995). Since the supernumerary Mauthner cells do not all lie in the same plane of focus, the focal plane was altered in consecutive trials within the same animal, each time optimizing for a particular cell (cell 3 in Fig. 3.2 B). In most cases, even the surrounding cells outside of the focal plane gave robust responses. In every case, the optimized cell displayed a significant increase in fluorescence. All labeled 83

99 Mauthner cells examined were activated during escape. These data suggest that all of the supernumerary Mauthner cells in des b420 mutants are integrated into the escape circuit. The observation that all of the Mauthner cells responded to the stimulus might be a consequence of a strong stimulus that was suprathreshold for all cells, or it might be a result of mutual excitation among the cells via gap junctions or other synaptic interactions. To attempt to distinguish between these two possibilities, we repeated the calcium imaging experiments while lowering the stimulus intensity to threshold, which is the intensity at which the fish escapes 50% of the time when presented with the stimulus. If Mauthner cells were independently activated by the sensory inputs, then we might expect that near threshold for the behavior, some cells would fire and not others because of slight differences in the level of sensory input or threshold of the cells that would bring some cells to threshold and not others. The threshold stimulus intensity was determined for a mutant fish by varying the tap strength to find a strength at which the fish escaped to about 50% of the taps. Once the threshold was determined for a particular fish, the responses of the Mauthner cells to a stimulus at threshold were examined in that fish (n=3 fish). As before, all labeled Mauthner cells were activated during these escapes to a threshold stimulus. These data suggest that all the Mauthner cells are firing together even at threshold, supporting the hypothesis that some interactions between the cells (e.g. via gap junctions or chemical synapses) assure that they function together. Mauthner cells are contacted presynaptically on the Mauthner cell lateral dendrite by the trigeminal, acoustic, and the lateral line nerves (Kimmel et al., 84

100 1984; Kimmel et al., 1990). We used confocal microscopy to examine the spatial relationship between these nerves and Mauthner cell lateral dendrites in des b420 mutants. Mauthner cell bodies in des b42 mutants are compact and assume the position of wild-type Mauthner cells. We found that all of the Mauthner cell lateral dendrites in des b420 mutants spatially overlapped with the three sensory nerves, with the trigeminal and lateral line nerves the most easily observed (Fig. 3.4). Due to the small axon diameter of these nerves, we were unable to visualize singular axon collaterals, however the close apposition of the presynaptic nerves and the postsynaptic lateral dendrite shows that all Mauthner cells in des b42 mutants could potentially receive sensory input. This lends support to the hypothesis that all of the Mauthner cells receive some direct input during an escape response. Escape performance is similar in wild-type and des b420 mutant larvae The Mauthner cell is known to play a major role in escape performance, particularly in tail-elicited escapes (Liu and Fetcho, 1999). The calcium imaging data revealed that all Mauthner cells were active during an escape response suggesting that des b420 mutants might exhibit abnormal escape responses. One prediction is that the escape response in mutants would be exaggerated in terms of the speed of the escape (angular velocity) due to the extra Mauthner cells. Moreover, because there are multiple Mauthner cells present in mutants as compared to the single Mauthner cell in wild-type this might alter the time to respond to the stimuli (latency to response). 85

101 To compare the escape performance of des b420 mutants to wild-type animals, we captured escape responses with a high-speed (1,000 frames/second) digital camera (Fig. 3.5) and compared several kinematic parameters (Table 3.1). Data were collected from two groups: 20 trials (5 each to left head, right head, left tail, right tail) from 6 des b420 mutants and 5 wild-type animals, and 20 trials from 6 des b420 mutants and 4 wild-type siblings. The trials were analyzed via a computer program written to automatically extract several parameters for each frame (Liu and Fetcho, 1999). Surprisingly, performance measurements that assess the function of Mauthner cells in the escape response were largely unaffected in des b420 mutant larvae compared to wild-type larvae (Table 3.1). In particular, the peak angular velocity was unaffected (p>0.05) suggesting that performance was not increased by the presence of extra Mauthner cells. The latency to respond was also not significantly different between mutant and wild-type larvae for both head and tail elicited responses (p>0.05). Previous Mauthner cell lesion studies showed no effect on turn angle after killing the Mauthner cell and its hindbrain homologues (Liu and Fetcho, 1999). Turn angle was significantly different between mutant and wild-type larvae for the head elicited response (p<0.001), but was not significantly different for the tail response in which the Mauthner cell is normally activated independently of its hindbrain homologues (p>0.05). This suggests that some of the other defects in des mutants could be affecting this parameter in head elicited responses. The peak turn duration was statistically longer in mutant compared to wild-type larvae (p<0.001) in tail-elicited responses, but not head. Thus, 6 of 8 comparisons of the movements showed no differences 86

102 between mutant and wild-type larvae and these unchanged parameters included two (latency and peak angular velocity) in which the Mauthner cell is thought to play a key role based upon previous lesion experiments. Although des mutants have defects in somite and myotome formation (van Eeden et al., 1996; Gray et al., 2001, Holley et al., 2002), this does not appear to dramatically affect the escape response. Mauthner axons in des mutants have fewer axon collaterals Despite the presence of supernumerary Mauthner cells contributing to the circuit, the escape response in des b420 mutants was relatively unaffected. One explanation for this finding is that each Mauthner cell could be playing a correspondingly smaller role in mediating the escape behavior. Mauthner cells drive the escape response by synapsing directly onto contralateral interneurons and motoneurons in the spinal cord via axon collaterals (Fetcho and Faber, 1988). To determine the relationship between Mauthner cells and downstream spinal neurons, we quantitated the number of axon collaterals present on individual Mauthner axons in wild-type and mutant larvae over a defined length of axon corresponding to three spinal cord hemisegments. Single Mauthner axons in wild-types and des mutants were back labeled from the spinal cord in 3 d larvae and analyzed under confocal microscopy at 4 d. Collaterals were positively identified by their knob-like appearance as previously reported (Gahtan and O Malley, 2003). Axon collateral counts at 4 d revealed that individual Mauthner axons in des b420 mutants had a dramatic decrease in axon collaterals compared to 87

103 individual wild-type Mauthner axons (Fig 3.6 A, B; Table 3.2). Mauthner cells in des b420 mutants had an approximately 78% decrease in the number of axon collaterals compared to wild-type Mauthner axons (Fig 3.6; Table 3.2). These data suggest that there is a relationship between the number of axon collaterals and the number of Mauthner cells; that is, when more Mauthner cells are present, there are fewer axon collaterals per Mauthner axon. To examine the relationship between Mauthner cell number and the number of axon collaterals per Mauthner axon, we analyzed des tp37 mutants (van Eeden et al., 1996), which have a less dramatic increase in Mauthner cells (Gray et al., 2001). Whereas des b420 mutants have an approximately 4.8-fold increase in Mauthner cells when compared to wild-type embryos, des tp37 mutants have an approximately 3.6-fold increase in Mauthner cells (Gray et al., 2001; Table 3.2). Quantitation of axon collaterals in des tp37 mutants revealed an approximately 56% decrease in the number of collaterals per Mauthner axon, a less dramatic decrease than that seen in the b420 mutant allele, but still considerably fewer collaterals than seen on wild-type Mauthner axons (Table 3.2). When there are a greater number of Mauthner cells, each Mauthner axon has fewer collaterals suggesting that the Mauthner axons are dividing up their synaptic targets in the spinal cord. If this were the case, we would expect Mauthner axons on the same side of the spinal cord to have non-overlapping axon collaterals. To test this, two different fluorophore-conjugated dextrans were electroporated into two separate Mauthner cells on the same side of the hindbrain in des b420 mutants. The positions of the labeled collaterals from the two different Mauthner cells were analyzed under confocal microscopy. We found 88

104 that of 61 axon collaterals analyzed, 93% did not overlap with collaterals from the other labeled axon (Fig. 3.7; Table 3.3). In the small number of cases where we did see overlap (2 pair), the collaterals might be contacting the same postsynaptic targets, although even in these cases they could be contacting different, but adjacent postsynaptic neurons. These data suggest that supernumerary Mauthner axons in mutant larvae are forming synapses with largely non-overlapping populations of spinal neurons and dividing up their target territory. Taken together, these data are consistent with the hypothesis that each Mauthner cell in des mutants exerts less of an effect on downstream spinal neurons and participation in the escape response is distributed among the population of Mauthner cells thus maintaining a largely normal escape response. Discussion The escape response in zebrafish is initiated by Mauthner cells, a single pair of large reticulospinal neurons. A mutation in the zebrafish gene des/notch1a results in a dramatic increase in the number of Mauthner cells without an equivalent increase in other neurons within the neural circuit that drives the escape response (Gray et al., 2001). This mutation, therefore, offers the unique opportunity to ask how the developing organism adapts to the presence of excess neurons and if/whether all of the cells are incorporated into the developing neural circuitry. Our analysis reveals that excess neurons are incorporated into the circuitry, but in a way that maintains the behavioral output. The more Mauthner cells present, the fewer collaterals there are per 89

105 Mauthner axon. Moreover, Mauthner axons extending caudally in the same region of the spinal cord have very few collaterals that form in the same locations strongly suggesting that Mauthner cells in des mutants are dividing up their targets. These data reveal that there is plasticity in the formation of this vertebrate neural circuit and the animal can adapt to the presence of excess cells in a manner that preserves the downstream behavior. Mauthner cell and the escape response Both experimental manipulations and genetic perturbations have been used to study the affect of cell loss on the escape response circuit. Ablating Mauthner cells in wild-type animals caused a serious diminution of the tailelicited escape response while the head-elicited escape appeared unchanged suggesting that the Mauthner cell is the principal interneuron mediating the tailelicited response (Liu and Fetcho, 1999). Additionally, in the zebrafish space cadet mutant spiral fibers fail to contact Mauthner cells resulting in aberrant escape behavior revealing a critical role for these cells in modulating the escape response (Lorent et al., 2001). Thus, the Mauthner cell, and modulation of the Mauthner cell, is essential for the tail elicited escape behavior and Mauthner cell loss cannot be compensated for by other reticulospinal neurons involved in the escape response (i.e. MiD3cm and MiD2cm). However, our work indicates that an increase in the number of Mauthner cells neither strongly enhances nor diminishes the escape response. 90

106 Fish and premetamorphic amphibians have only a single, large, fastconducting Mauthner cell that initiates the escape response. A similar response in mammals, however, is primarily mediated by a population of giant neurons in the caudal pontine reticular-formation in the brainstem (Lingenhohl and Friauf, 1992; 1994; Yeomans and Frankland, 1995). The question in terms of the construction of neuronal circuitry is whether the use of a single, large cell, as opposed to several smaller cells, is advantageous for this crucial behavior. We have focused primarily on the motor performance, the output side of the escape circuit, in mutant fish. Our data indicate that altering the number of cells has no dramatic affect on the motor output in the behavior. We have not explored the sensory side in as much detail because it is more difficult to approach. The presence of extra cells might generate severe problems with the ability to properly integrate different sensory modalities (vision, somatosensation, lateral line, audition) to produce an appropriately timed motor response. Such deficits in integration of different modalities on the sensory side might explain why fish with extra Mauthner cells have not been observed in natural populations. Mauthner axon collaterals Our data show that there is a relationship between the number of Mauthner cells and the number of collaterals per Mauthner axon. It remains to be determined how the number and distribution of Mauthner axon collaterals is regulated. It may be that a Mauthner cell is sensitive to the presence of other Mauthner cells. Whether this regulation occurs at the level of the Mauthner cell 91

107 body or at the level of the axon is unclear. Alternatively, target cells in the spinal cord may mediate regulation. It is possible that once a spinal neuron is innervated by a Mauthner axon, it is bypassed by other Mauthner axons. Based on our data, we would predict that individual Mauthner cells in des mutants would have decreased output due to their decreased numbers of axon collaterals as compared to wild types. One way to test this idea would be to eliminate Mauthner cells and examine performance changes. We tried to laser ablate all but one Mauthner cell in des b420 mutants to get the biggest possible behavioral effect. Unfortunately, the tight juxtaposition of the Mauthner cell bodies makes controlled removal of the cells using dye phototoxicity very difficult and we were unsuccessful in performing this experiment. Our data show that all Mauthner cells in des b420 mutants are activated during an elicited escape response (see Fig. 3.2). It is possible, however, that different Mauthner cells in mutants contribute unequally to the response. For example, some Mauthner axons could have many collaterals and thus a stronger input onto spinal neurons whereas other Mauthner axons could have fewer collaterals and less of an impact on spinal neurons. The totality of these responses would result in a normal escape response. The axon collateral data, however, does not support this idea. While there was variability in the number of axon collaterals counted, there was no clear subgrouping of axons having the wild-type number of collaterals and other axons have substantially fewer collaterals. The one exception to this was an example in des tp37 where there were only two Mauthner cells and one of them had 12 collaterals over the three spinal cord hemisegment distance (Table 3.2). This number of collaterals was the 92

108 smallest number seen in wild-type larvae (range: collaterals/ 3 hemisegments) and raises the possibility that when only two Mauthner cells are present, the number of collaterals approximates what is seen in wild-type larvae where there is only one Mauthner cell. However, as soon as there are three Mauthner cells present, the number of collaterals decreases significantly (range: 4-8 collaterals/ 3 hemisegments). Thus, there appears to be an overall decrease in axon collateral number on each axon, rather than a large decrease in collaterals on some axons. Evolution of the escape response Analysis of des mutants may yield insight into the evolution of the escape response. As mentioned above, a startle response in mammals analogous to the fish escape response is mediated by a population of neurons in the caudal pontine reticular-formation that contact spinal interneurons and motoneurons (Lingenhohl and Friauf, 1992; 1994; Yeomans and Frankland, 1996). Although individual axons have not been traced in mammals, it is thought that these cells distribute the activation of the stimulus to interneurons and motoneurons throughout the spinal cord. This is supported by the finding that neurotoxic lesions of the caudal pontine nucleus decreased the startle response (Koch et al., 1992). This scenario is similar to what we see in des mutants in that Mauthner cells divide-up targets in the spinal cord. des mutants reveal that an evolutionary switch between one neuron and a population of neurons could be easily achieved by altering the number of target neurons contacted by each pre- 93

109 synaptic cell as a way to maintain the integrity of the behavior. As a result, the animal can adapt to genetic changes thus ensuring integrity of the behavior and survival. Such plasticity may have been important for the evolution of motor behaviors controlled by larger numbers of neurons. 94

110 Figure 3.1: des mutants have supernumerary Mauthner cells. Mauthner cells in a 30 h (A) wild-type and (B) des b420 mutant embryo as revealed by 3A10 antibody labeling and confocal microscopy. Dorsal views with anterior to the top. Scale bar, 40 µm. 95

111 Figure 3.2: Mauthner cells in des mutants are active during an escape response. A calcium imaging trial of a des b420 mutant. (A) Mauthner cells in a mutant larvae were backfilled with the fluorescent dye, calcium green dextran at 4 d and imaged on confocal microscopy with a 63X objective. (B) For quantification purposes, Mauthner cells on the right side of the hindbrain (arrow in A) were numbered 1 through 4. Fluorescence intensities were presented in pseudo-color with red being the brightest. In this trial, the plane of focus was optimized for cell 3. However, as is typically the case, all four cells were noticeably brighter after the escape, when compared to the baseline of frames collected before the stimulus. The distortion in frame number six indicates movement of the fish in response to a tap on the tail. Frames are ordered from left to right with 578 msec between frames. Scale bar, 30 µm. Performed in collaboration with Dr Joseph Fetcho and Dr. Katherine Liu at State University of New York, Stony Brook. 96

112 Figure 3.3: Quantification of the increase in fluorescence intensities. The fluorescence intensities for each cell (1-4) were plotted for each frame. The y-axis shows the normalized intensities ( F/F). The x-axis shows the time in seconds. Performed in collaboration with Dr Joseph Fetcho and Dr. Katherine Liu at State University of New York, Stony Brook. 97

113 Figure 3.4: Mauthner cell lateral dendrites in des mutants are in close apposition to sensory nerves. 72 h wild-type (A,C) and mutant (B,D) embryos were labeled with 3A10 antibody and analyzed by confocal microscopy. The trigeminal (Tg) and lateral line (Ll) nerves extend perpendicular to the Mauthner cell (Mth) and are closely juxtaposed to Mauthner cell lateral dendrites (arrow). A and B are the left sides of merged sections of wild-type and mutant embryos; C and D are single optical sections through the wild-type and mutant embryo shown in A and B. All images are dorsal views with anterior to the top. Scale bar 35 µm 98

114 Figure 3.5: des b420 mutant larvae exhibited largely normal escape responses. A tap with a small, polished glass probe elicited escapes. In this trial, the tap was delivered to the left side of the head (lower left-hand corner). Images were captured at 1,000 frames/sec. Every third frame (msec) is shown, starting at the initiation of the response (frame 0). Following the start of movement, peak turn angle is reached at 15 msec (15). For an example of a wild-type escape response see Liu and Fetcho (1999). Scale bar, 550 µm. Performed in collaboration with Dr Joseph Fetcho and Dr. Katherine Liu at State University of New York, Stony Brook. 99

115 *latency to response (msec) peak angular velocity ( /msec) peak turn duration (msec) peak turn angle ( ) wild type heads 6.13± ± ± ±3.30 tails 11.0± ± ± ±3.32 des b420 heads 6.81± ± ± ±4.30 tails 10.92± ± ± ±3.11 Table 3.1: Escape response performance in wild-type and des b420 mutant larvae Kinematic parameters were analyzed and compared using analysis of variance. Turn duration was defined as the time from initiation of response to peak turn angle. Data are reported as mean ± standard error. *3 d homozygous mutants (n=6) and wild-type siblings (n=4). 5 d homozygous mutants (n=6) and wild-type (n=5). Performed in collaboration with Dr Joseph Fetcho and Dr. Katherine Liu at State University of New York, Stony Brook. 100

116 Figure 3.6. The number of collaterals per Mauthner axon is dramatically decreased in des mutants. 4 d wild-type (A), des b420 (B) and des tp37 larvae were backlabeled from the caudal spinal cord with lysinated rhodamine dextran. The number of axon collaterals was quantitated over a three spinal cord hemisegment region in the midtrunk and the number of Mauthner cells were counted under transmitted light (40X) (C). Arrow heads denote axon collaterals. Scale bar, 10 µm. 101

117 wt (n=10) dest p37 (n=10) des b420 (n=10) Mauthner Mauthner Mauthner collaterals collaterals cells cells cells Collaterals ± ± ± ± ± ± 1.6 Table 3.2: Axon collaterals in wild-type and des mutant larvae Axons in 3 d wild-type (wt), des t[p37 mutant and des b420 mutant larvae were backfilled with lysinated rhodamine dextran at the mid-trunk level. Larvae were fixed at 4 d and axon collaterals were counted over three spinal cord hemisegments under confocal microscopy. The number of Mauthner cells present on the side where the analyzed axon(s) originated were counted under transmitted light. Data are reported as mean ± 95% confidence interval. The decrease in axon collaterals observed in both des t[p37 and des b420 mutant were statistically significant compared to wild-types (p < ). In addition, the difference in axon collaterals between des t[p37 and des b420 mutants was also significant (p < 0.005). 102

118 Figure 3.7: Collaterals from distinct Mauthner axons do not form in the same location. Two Mauthner cell bodies in des b420 mutants were electroporated with different fluorophores and their axon collaterals imaged at 6 d. One axon is shown in red (A) the other in green (B). The merged image (C) shows that the collaterals from the two axons do not form in the same location. Arrowheads denote collaterals. Scale bar, 10 µm. Performed in collaboration with Dr Joseph Fetcho and Dr. Katherine Liu at State University of New York, Stony Brook. 103

119 Fish Collaterals overlapping collaterals (pair) Red Green A B C D E Table 3.3: Mauthner axon collaterals from distinct axons rarely form in the same position. Two Mauthner cell bodies from 4 d des b420 larvae were electroporated with Rhodamine (R) or Far Red dextran (G) and their axons were analyzed over a three hemisegment length for collaterals at 6 d. Collaterals were considered overlapping if they where in the same position throughout the z-series. In Fish D, collaterals were examined along two distinct axon regions. Performed in collaboration with Dr Joseph Fetcho and Dr. Katherine Liu at State University of New York, Stony Brook. 104

120 CHAPTER 4 MOTOR AXON BEHAVIOR IN ABNORMALLY PATTERNED SOMITES Introduction Nerve segmentation in the vertebrate embryo corresponds to the segmentation of the paraxial mesoderm. The segmentation of the unsegmented presomitic mesoderm (PSM) results in the formation of somites, which form in an anterior to posterior pattern. The nerves of the trunk and components of the peripheral nervous system (PNS), which also mature in an anterior to posterior pattern, are aligned segmentally with respect to the somites. Many studies have revealed that the somites may be responsible for the segmental arrangement of the PNS and peripheral nerves. The peripheral nerves are restricted to the anterior half of each avian somite (Keynes and Stern, 1984; Rickman et al., 1985). In avian and amphibian embryos, somite ablation results in the loss of segmented PNS components such as peripheral nerves to the limbs, thus, demonstrating the importance of the somite on the patterning of these structures (Lewis et al., 1981; Tosney, 1987, 1988; Lehmann, 1927; Detwiler, 1934). Other studies performed in avian embryos in which the early somites were reversed 180, revealed that components of the PNS and peripheral axons still 105

121 associated with the original anterior half of each somite although it was located posteriorly after the rotation (Aoyama and Asamoto, 1988). Furthermore, when the manipulation of the somites resulted in regions composed only of anterior half somites, segmentation of the PNS is lost, although axons still extended into the periphery (Stern and Keynes, 1987). In addition, in posterior half somite domains, the peripheral axons did not migrate into the periphery (Stern and Keynes, 1987). These data suggest that the anterior half of the somite is permissive for PNS and nerve segmentation, while the posterior half may be non-permissive. In zebrafish, the primary motoneurons extend their axons out of the spinal cord at a midsegmental location onto a common pathway along the medial portion of the myotome, which is derived from the somite (Bernhardt et al., 1998; Eisen et al., 1986; Myers et al., 1986). There are three types of easily identifiable motoneurons that supply the first innervations of the myotome in embryonic zebrafish. These neurons are arranged in specific locations in the spinal cord and each innervates a different part of the myotome. The Caudal Primary (CaP), Middle Primary (MiP), and the Rostral Primary (RoP) innervate the ventral, dorsal and middle myotome respectively (Eisen et al., 1986). The pathway for these motor axons is pioneered by CaP. CaP axons extend out of the spinal cord along the medial aspect of the anterior myotome immediately adjacent to the boundary with the posterior domain (Fig 4.1). Unlike chick and mouse where sclerotome functions to pattern motor axons, in zebrafish the myotome appears to function in this role. When motor axons extend out of the spinal cord, the somite consists primarily of myotome, with a small component of ventrally 106

122 located sclerotomal cells. Furthermore, ablation of the sclerotome does not disrupt segmental patterning of motor axons (Morin-Kensicki and Eisen, 1997). Like other vertebrate species, zebrafish somites and myotomes are subdivided into anterior and posterior domains as revealed by gene expression and cellular morphology (Bernhardt et al., 1998; Stickney et al., 2000). CaP axons are restricted to the anterior domain of the myotome and never enter the posterior myotome (Bernhardt et al., 1998, Eisen 1991, 1994). Previous studies in zebrafish have implicated this division of the somite/myotome into anterior and posterior domains as an important event for motor axon pathfinding (Bernhardt et al., 1998; Roos et al., 1999). Studies performed in avian and mammals suggest that inhibitory signals are predominately provided to motor axons by the posterior sclerotome (Keynes and Stern, 1988; Fredette et al., 1996; Wang and Anderson, 1997; reviewed by Flanagan and Veanderhaegehen, 1998;). Moreover, surgical formation of regions of posterior character in avian somites results in a failure of axons to extend normally (Stern and Keynes, 1997). Since zebrafish somites and myotomes are patterned along the anterior-posterior axis and motor axons extending along myotomes obey domain restrictions, this suggests that motor axons may be guided by anterior-posterior restricted signals. Earlier experiments implicated semaphorin 3A2 (sema 3A2) in the guidance of zebrafish motor axons. sema 3A2 expression is restricted to the posterior domain of the somite. Furthermore, its overexpression resulted in a loss or a truncation of CaP motor axons, suggesting that this molecule is an inhibitory signal for motor axons (Bernhardt et al., 1998; Roos et al., 1999). This work proposed that the position of the growing CaP axon 107

123 might be controlled by the expression of inhibitory molecules in the posterior somite (Roos et al 1999). Together with data from avian and mammalian embryos, this led to the hypothesis that the posterior domain of the somite/myotome was inhibitory for motor axons. In order to get a better understanding of the effect that the domains of the somite/myotome have on CaP motor axon behavior, we have analyzed the CaP motor axon in mutants with defective anterior-posterior patterning of the somite/myotome. deadly seven/notch1a (des) is necessary for proper somite formation in zebrafish (van Eeden et al., 1996; Gray et al., 2001; Holley et al., 2002). Mutations in des cause abnormal somite formation as revealed by morphological examination and gene expression. Defective somite boundaries are seen in the somites posterior to somites 6-7at 18 h in des mutant embryos (Fig. 4.2). Proper boundaries are seen in the first approximately 8 somites at 48 h suggesting a rescue of about 1-2 somite boundaries as development proceeds (Gray et al., 2001). The analysis of the somites using RNA in situ hybridization reveals a patterning defect within the affected somites of des mutant embryos (Fig. 4.3; Gray et al., 2001; van Eeden et al., 1996; Durbin et al., 2000; Holley et al., 2000; Jiang et al., 2000; Sawada et al., 2000). Expression of myod, normally restricted to the posterior domain of each somite (Weinberg et al., 1996), is abnormal in somites after somite 6 in mutant embryos. The expression of myod is not restricted to the posterior domain of the somite, but is expressed throughout the aberrant somites in des mutants. The expression of a marker of anteriorly restricted cells in the newly formed somites, fgf8, is aberrant. fgf8, normally expressed in anterior somite cells, is no longer restricted to the anterior domain is 108

124 also seen throughout the somite. The expression of these markers shows that cells with anterior and posterior character are not separated and therefore demonstrates a loss of somite polarity in the somites caudal to somite 7 in the mutant embryos. Notch1 has been shown to be important for the oscillation in gene expression of hairy enhance-of-split related 1 (her1) and deltac within the PSM in zebrafish. The oscillations create stripes of gene expression that move through the cells within the PSM in waves traveling rostral to caudal. When the oscillation of these molecules is stabilized in the rostral PSM somite boundary formation can take place. The disruption of oscillation results in abnormal gene expression within the somites and abnormal boundary formation. Another mutation, fused somites/tbx24 (fss) also exhibits defects in somitogenesis and myogenesis. There are no morphologically distinct somite boundaries present in the mutant embryos at 18 h and the somitic mesoderm is not patterned along the anteroposterior axis (Fig 4.2;Durbin et al., 2000, van Eeden et al., 1996, Nakaido et al., 2002). myod expression is present throughout the myotome in these mutant embryos and there is a loss of fgf8 expression (Fig 3.3; Durbin et al., 2000). Thus, somitic cells in fss mutants display only posterior identity. In this study we analyze the role of somite/myotome patterning in zebrafish motor axon outgrowth, by analyzing ventral motor axon (CaP) pathfinding during zebrafish development in des and fss mutant embryos. We show that in both des and fss mutants, motor axons are truncated and excessively branched. However, motor axons can extend along posteriorized myotomes indicating that this region of the myotome can support axon outgrowth. By 109

125 anteriorizing somites, we find similar motor axon defects as those seen when the somites/myotomes are posteriorized. These results suggest that neither anterior nor posterior domain is sufficient to support the stereotyped migration of motor axons. We propose that it is the juxtaposition of these two domains that set up a region of wild-type myotome that is sufficient to support stereotyped motor axon outgrowth. Materials and Methods Fish Adult zebrafish were maintained as described by Westerfield (1995). The staging was by hours (h) or days (d) post-fertilization at approximately 28.5 C as in Kimmel (1995). The des and fss mutant embryos were collected from pairwise crosses of heterozygous adults and identified based on somite morphology. In situ Hybridization and Immunohistochemistry Embryos were collected and staged at various times of development and processed for RNA in situ hybridization as described by Thisse et al. (1993). Antisense digoxigenin myod riboprobe was synthesized from a plasmid linearized with XbaI and transcribed with T7 RNA polymerase (Weinberg et al., 1996). fgf8 riboprobe was synthesized from a plasmid linearized with Xho1 and 110

126 transcribed with T7 RNA polymerase. papc riboprobe was synthesized from a plasmid linearized with Apa1 and transcribed with T3 RNA polymerase. For immunohistochemistry with the monoclonal antibody znp1 (Trevarrow et al., 1990) embryos were fixed in 4% paraformaldehyde overnight at 4 C, then washed in PBS and preincubated in phosphate buffered saline with 0.5% Triton X-100, 1% bovine serum albumin, 1% dimethylsulfoxide, and 2.5% goat serum (PBDT) (Melancon et al., 1997). The antibody was diluted in PBDT and incubated overnight at 4 C. For visualization under transmitted light the Clonal PAP system (Sternberger Monoclonals Inc.) with 3,3 -diaminobenzidine (DAB) as substrate. For fluorescent detection embryos were incubated with antimouse Oregon Green (Molecular Probes). Blastula Transplants Embryos were collected from pairwise crosses of AB* and des b420 heterozygous fish. The AB* embryos were aligned in 1.2% agar plates and injected with the vital dye, 3,000 MW lysinated rhodamine dextran using a MPPI-2 pressure injector. Transplants were performed as described by Ho and Kane (1990) using unlabelled host embryos from a des b420 heterozygous cross. Host embryos were allowed to develop to the desired age then subjected to immunohistochemistry as described using znp1 primary monoclonal antibody and a fluorescent Oregon Green (Molecular Probes) secondary antibody. For cross-sectional analysis of transplanted cells and primary motoneurons, embryos were embedded in 1.5% agar/5% sucrose and sectioned on a cryostat at 16 µm. 111

127 Embryos and sections were analyzed with a Zeis Axioplan microscope and digitally imaged by using a Photometrics SPOT camera. Construction and overexpression of muscle actin promoter-epha4 GFP construct A αp-xepha4egfp construct was generated using the αp (α-actin promoter) plasmid vector (Higashijima et al., 1997) and a pcs2-xepha4egfp or pcs2-epha4egfp construct (Durbin et al., 2000). The αp and pcs2-epha4 plasmids were cut with EcoRI and NotI, the appropriate fragments were purified and ligated. The construct was linearized with Apa1, purified and injected into 1-2 cell wt embryos obtained from AB* crosses. Approximately ng of DNA was injected into each embryo using a MPPI-2 microinjector. The embryos were allowed to develop to h and fixed for antibody processing with znp1, according to above protocol. Results Motor axons show defective pathfinding in des b420 mutant embryos Previous work suggested that the patterning of the somite/myotome is critical for proper pathfinding of the CaP motor axon in zebrafish (Roos et al., 1999, Bernhardt et al., 1998). We examined CaP axons of des embryos at 26 h. znp1 antibody staining showed that primary motor axons at 26 h in des mutant 112

128 embryos were aberrantly branched and highly disorganized in myotomes of the mid and caudal trunk, posterior to myotome 7 (Fig 4.4). CaP motor axons extending along unaffected myotomes 1-6, are normal, they displayed the stereotyped wild-type projection (Gray et al., 2001). To further characterize the axonal defect in mutant embryos, we examined CaP motor axons by quantitating the number of branches and by analyzing growth cone position. Examination of CaP motor axon branches over a five-segment length revealed that 70% of motor axons in des mutant embryos had two or more branches as compared to 18% in wild-type embryos (Fig. 4.5). Examination of CaP growth cones at 26 h showed that 25% of growth cones were in areas dorsal to the ventral myotome at this stage of development, however in wild-type embryos all growth cones had reached the ventral myotome (Fig. 4.6). These results suggest that anterioposterior patterning within the ventral myotome is critical for stereotyped ventral motor axon outgrowth. Wild-type muscle cells can rescue the motor axon pathfinding defect in des b420 mutant embryos Experiments using des b420 mutants have shown that the CaP motor axon defect was non-cell autonomous for CaP and likely due to the environment through which the axons extend (Gray et al., 2001). CaP axons extend out of the spinal root and ventrally along the medial surface of the myotome between the myotome and notochord. To identify the environmental component responsible for the stereotyped ventral axon projections, blastula transplants were performed 113

129 between labeled wild-type embryos and unlabeled embryos from a des b420 heterozygous cross. When wild-type cells transplanted into mutant embryos developed as muscle, axons in that region were completely rescued (Fig 4.7). Analysis of cross sections confirmed that wild-type muscle cells located in the medial myotome could rescue the axons. In Figure 4.7, the axon on the right, where there are transplanted muscle cells, displays the wild-type axon projection; however, the axon on the left without any transplanted muscle cells display an abnormal projection, its axon actually goes out of the plane of the image and then reemerges in the most ventral area of the somite. In embryos without transplanted wild-type muscle cells or in hemisegments without any wild-type muscle cells no rescue of the axonal defect was observed. These data demonstrate that the axonal defects observed in mutant embryos is due to the somite/myotome defect exhibited by des b420 mutants. fused somites mutant embryos have defective somites and disorganized motor axons In des somites, cells of anterior character, suggested to be permissive for axon outgrowth, are intermingled with cells of posterior character, which are suggested to be inhibitory for axon outgrowth. Our characterization of the motor axon defect in des mutant embryos and rescue of that defect with wildtype muscle suggests that patterning of the somite/myotome, by domain restriction of anterior and posterior cells, is required for stereotyped axon pathfinding in zebrafish. The somites/myotomes of fss mutants only contain cells of posterior character, suggesting that these myotomes would be inhibitory 114

130 for motor axon outgrowth. Furthermore, sema3a2, a molecule suggested to be inhibitory to motor axons is expressed throughout the somite/myotome, in fss mutants (Fig 4.8). van Eeden et al., (1996) showed that motor axon pathfinding in fss mutants was abnormal. We have characterized this defect in more detail by analyzing motor axon branching and growth cone position. At 26 h, axons in fss mutants are highly branched and disorganized (Fig 4.4). All of the axons examined over a five-segment length in fss embryos have two or more branches when compared to 18 % in wt and 70 % in des mutant embryos (Fig. 4.5). Seventy percent of CaP growth cones in fss mutant embryos failed to extend into the distal ventral myotome at 26 h, while all CaP growth cones in wild-type embryos had extended into this area (Fig. 4.6). The defects displayed by CaP axons in fss mutants were similar to those seen in des mutants, but more severe. Thus, ventral motor axons can extend along posteriorized myotomes, suggesting that the posterior myotome is not inhibitory. However, the axons were highly disorganized and truncated indicating that posterior myotome lacks the cues necessary for stereotyped ventral motor axon pathfinding. Motor axons pathfind abnormally along anteriorized somites/myotomes Zebrafish motor axons in wild-type embryos extend along the anterior myotome suggesting that this region of the myotome contains the cues necessary for stereotyped motor axon pathfinding. To test this idea, we generated embryos with anteriorized myotomes by overexpression of the EphA4 receptor that is 115

131 expressed in the anterior domain of zebrafish somites during the segmentation period and has been implicated to play a role in somite formation (Durbin et al., 1997). EphA4 was ectopically expressed throughout the myotome using an α- actin promotor to drive expression. The α-actin-epha4gfp construct was injected into 1-2 cell stage embryos and used to drive myotome cells towards anterior identity. Expression of EphA4 in myotomes was revealed by GFP expression and was found to be mosaic throughout the myotome. We analyzed motor axons extending along myotomes with at least 3-4 GFP expressing cells (Fig 4.9). In these embryos axons displayed abnormalities similar to those found in des and fss mutant embryos. The axons appear to be slightly less branched, 85 % of axons had two or more branches as compared to 100 % of the fss mutant axons. The axons were slightly more branched as compared to des mutant embryos where 70 % of axons had two or more branches. Furthermore, there were a small number of truncations observed. These data suggest that neither anterior nor posterior myotome contains the cues necessary for stereotyped ventral motor axon outgrowth. Based on this data we suggest that the juxtaposition of anterior and posterior myotome is needed to set up the necessary guidance cues. As CaP axons and later developing secondary motor axons extend along the anterior myotome adjacent to the boundary between anterior and posterior myotome, it implicates this boundary as a critical component for establishing the ventral axon pathway. 116

132 Discussion Zebrafish motor axons extend along a narrow region of the anterior myotome immediately adjacent to the posterior myotome. Analysis of mutants lacking anterior-posterior somite/myotome domains indicate that this patterning is essential for stereotyped motor axon outgrowth. Myotomes in des mutants lack most anterior character and CaP axons extending along these myotomes are short and highly branched. Moreover, in fss mutants, which lack all anterior character, similar axon defects are seen and they are even more severe. Axonal defects were rescued when wild-type muscle was transplanted into des mutant embryos supporting the idea that the axonal defects are due to mispatterned myotomes. The finding that motor axons extend along fss myotomes indicates that the posterior myotome is not inhibitory for motor axons. In addition, we have shown that motor axon outgrowth along anteriorized myotomes exhibit the same defects as seen on posteriorized myotomes. Taken together, these data support the idea that the juxtaposition of anterior and posterior myotome is critical for stereotyped motor axon outgrowth. This juxtaposition of anterior and posterior somite/myotome cells may signal the formation of a region permissive for motor axons and may control abnormal branching and pathfinding. Motor axon pathfinding in somite patterning mutants We had previously shown through genetic mosaic analysis that the CaP motor axon defect in des mutant embryos was non-cell autonomous for the 117

133 motoneuron and likely due to the environment (Gray et al., 2000). Since the primary motor axons extend along the myotome, it was hypothesized that this was the likely cause for the defect in the mutant embryos. Here we show through transplantation experiments that indeed muscle cells from wild-type animals can rescue the axon defect in des mutant embryos. This allowed us to determine how myotome patterning affected the morphology and behavior of CaP motor axons. It was revealed that CaP axons in des and fss mutant embryos displayed many of the same characteristics. These axons had more branches than wild-type axons, and in fss mutant embryos the axons had even more branches than axons in des mutant embryos. Therefore, fss myotomes are even less conducive for stereotyped axonal outgrowth than the des myotome. Although the posteriorized myotomes in fss mutants are not inhibitory for motor axons, these data do demonstrate that this environment has different aeffects on axons than des mutant myotomes. Both fss and des somites/myotomes lack anterior-posterior patterning, des myotomes do contain cells of both anterior and posterior character, however, fss myotomes are completely posteriorized. This could explain the differences in the axon branching defects. The presence of some cells with anterior character, although dispersed throughout the somite, may be enough to account for the decreased number of branches in the des mutant compared to fss mutants. These few anteriorly fated cells may produce signals, which in the presence of posteriorly fated cells, affects the branching morphology of primary motor axons. 118

134 A properly patterned somite is required for guidance molecules to be expressed in the correct domains Patterning within the somite establishes pattern within the myotome and thus affects axon outgrowth. Anterior and posterior somite domains can be distinguished morphologically and based on gene expression (Stickney et. al, 2000). The posterior domain in zebrafish has been hypothesized to be an inhibitory or repulsive domain for motor axons. The expression of guidance cues is restricted early during development in many organisms, into the anterior or posterior domains of the somite/myotome or sclerotome. One of the earliest molecules that shows a restriction in expression to one domain in zebrafish is semaphorin 3A2 (sema3a2). sema3a2 is expressed by somites early in development and has been implicated as an axon guidance cue (Roos et al 1999). In newly formed somites (12-14 h) it is expressed in the posterior domain. The anterior domain of the somite is devoid of any expression at this stage (Bernhardt et al., 1998; Roos et al., 1999). One interesting aspect of sema3a2 RNA expression is that it is restricted to the medial portion of the anterior domain of the somite a little later in development at h. Previous studies in which a construct encoding a his-tagged sema3a2 was overexpressed in embryos resulted in missing or truncated motor axons suggesting that sema3a2 was acting as a repulsive cue (Roos et al., 1999). When we assay the expression of sema3a2 in fss mutant embryos at 14h, we find that the mrna is expressed throughout the somite. Although the expression of this molecule is not restricted to any domain in fss mutant embryos, the axon defects are not the same as those observed in embryos that had sema 3A2 overexpressed. In embryos overexpressing sema3a2 119

135 axons were truncated or absent, we never saw absent motor axons in fss mutants. So even though all the cells of the somite/myotome in fss mutants express sema3a2, axons are still able to extend out of the spinal cord and along the myotome, albeit abnormally. This result does not suggest that sema3a2 is not a guidance cue that may be inhibitory to motor axons, but its true activity may only be seen in the context of a somite with some degree of anterior-posterior patterning. The ephrins and Eph receptors have also been implicated in controlling aspects of motor axon pathfinding in vertebrate species. These molecules are expressed at the appropriate times and places during development to mediate the patterning of motor axons. Data from rat indicates that ephrin-b ligands and EphB receptors are involved in repulsion of motor axons from improper domains of the somite. Specifically in rat, ephrin-b2 is expressed in the caudal half of the sclerotome and the receptor EphB2 in the rostral half of the sclerotome (Wang and Anderson, 1997). Through in vitro analysis it was shown that ephrin-b2 is likely a repulsive cue for motor axons (Wang and Anderson, 1997). When presented with stripes of ephrin-b1 or ephrin-b2, axons were biased towards stripes devoid of these proteins (Wang and Anderson, 1997). One interesting aspect of in vitro analysis of the ephrins is that when axons are presented with a continuous field of either ephrin, it did not inhibit axon guidance. Thus suggesting that an environment consisting of domains of permissive and nonpermissive molecules is needed for proper axon guidance. Examination of these receptors and ligands in zebrafish has led to the hypothesis that these molecules may play the same roles in guidance of motor 120

136 axons in zebrafish. The ephrin-b2 ligand and the EphA4 receptor are expressed during embryonic zebrafish development. Expression of ephrin-b2 can be found in the posterior domains of formed somites. The receptor EphA4 is expressed in the anterior domain of the somites throughout somitogenesis (Durbin et al., 1998). There is a loss in the polarity of the expression domains of these molecules when we examine their expression in des mutant embryos (data not shown). Thus far in zebrafish, these molecules have only been studied as far their involvement in the development of the somites is concerned. However, based on the conservation of their expression domains it is not unlikely that they may also be involved in setting up domains that are important for axon outgrowth. Degree of motor axon rescue may be dependent on the position of cells within the somite. Transplanted wild-type muscle cells rescue motor axons in des mutant embryos. Although transplanted muscle cells could be seen in nearly every myotome, only subsets of those myotomes showed rescued CaP axons. Only muscle cells located medially could rescue these axons. This was not very surprising since it is known that axons grow along the medial portion of the myotome in zebrafish. Furthermore, data from mouse suggests that there is clonal separation of precursors of medial and lateral myotomes in its somite (Trinquet and Nicolas, 2002). In avian somites, it has been recognized that the medial region of the somite nearest each developing PNS structure, including the peripheral nerves, is sufficient for proper segmentation (Tosney 1987, 1988). The cells that make up the developing myotomes are not only patterned along the 121

137 anterior-posterior axis, but there may also be patterning along the medio-lateral axis as evidenced by cycling of the expression of genes known to play roles in anterior-posterior patterning. It has been suggested that this medial-lateral regionalization of the myotome may occur in the presomitic mesoderm before segmentation has occurred (Freitas et al., 2001). It may be possible that myotome cells located medially or laterally have different gene expression profiles. Therefore it is not only important for the muscle cells to be of anterior character, they must also be located medially in order to have an effect on the projection of the axons. The fact that the axons in des mutant embryos are not as severely affected as their fss counterparts may be due to some cells of anterior identity being present in the correct medio-lateral positions in the mispatterned somites. Juxataposed anterior and posterior cells Since zebrafish spinal motor axons grow along a region of the anterior somite/myotome at the interface of anterior and posterior domains, and are abnormal in the absence of such an interface, shows that this region is critical for proper motor axon pathfinding. Each domain of the somite is defined by the expression of numerous genes, whose effects on motor axon pathfinding are not completely clear. A restriction of gene expression to the region between anterior and posterior cells has not been shown in zebrafish, although our data suggest that this region may be of different character than either the anterior or posterior domains. Work performed on the AP patterning of the Drosophila wing imaginal disc, shows that there is a compartment boundary that divides the anterior and 122

138 posterior domains of the wing. Hedgehog (Hh) synthesized by posterior cells diffuses into the anterior compartment forming a short-range activity gradient that activates the localized expression of a number of target genes including patched and decapentaplegic (Chen and Struhl, 1996). The short range activity of Hh is responsible for patterning the central region of the Drosophila wing (Mullor et al., 1997; Strigini and Cohen, 1997). Furthermore, this activity is mediated by activation of the transcription factor Collier/Knot (Col/Kn) in the narrow stripe of cells along the anterior-posterior boundary, called the anterior-posterior organizer (Mohler et al., 2000; Vervoort et al., 1999). Therefore, it is possible that proteins in the anterior or posterior somite/myotome domain in zebrafish could control the expression of molecules along the boundary and thus produce a region that is conducive to stereotyped axon projection. 123

139 Figure 4.1: Schematic representation of wild-type CaP axon projection. Lateral view of a trunk myotome segment with a CaP axon extending along the anterior domain of the myotome next to the boundary with posterior myotome. Blue domain is anterior, A; Red domain is posterior,p 124

140 Figure 4.2: des and fss mutants have abnormal somites. Lateral views of somites in live h wild-type (A), fss mutant embryos(b), des mutant embryos (C). Arrow in B indicates no somite boundaries, arrow in A and C indicate normal boundaries, and arrowhead in C indicates abnormal somite boundaries. Anterior is to the left. Scale bar:100 µm. 125

141 Figure 4.3: Gene expression is perturbed in somites of des and fss mutants. Dorsal views of RNA in situ hybridization of fgf8 (A,B,C; 14 h), myod (D,E,F; 16 h), papc (E,F,G; 13 h) in wild-type (A,D,G), des (B,E,H) where arrowheads reveal abnormal expression, fss (C,F,I) embryos,, arrow in C and I indicates a lack of expression, and the arrow in F indicates expression throughout the somitic region. Scale bar is 100 µm. 126

142 Figure 4.4: CaP motor axons are abnormal in des and fss mutants. Lateral views of trunks of 26 h wild-type (A), des mutant (B) and fss mutant(c) embryos as visualized by whole-mount znp1 antibody labeling. Arrows indicate abnormal axons. 127

143 Number of axons wt des fss Number of branches Figure 4.5: Number of branches on CaP motor axons. Branches were counted for 5 axons over a five-segment (segments 7-11) length in 10 embryos. Axon were labeled with the znp1 antibody at 26 h in whole-mount and branches were counted using Zeis Axioplan microscope. 128

144 60 50 Number of axons wt des fss 10 0 SR HM VN MVM VM Area of myotome Figure 4.6: Position of growth cones of CaP motor axons at 26 h. Growth cones for 5 axons in each fish were examined. The positions are various locations along the wild-type path of the growth cone. SR, spinal root; HM, horizontal myoseptum; VN, ventral edge of the notochord; MVM; middle region of the ventral muscle; VM, ventral muscle. Axons were visualized by antibody staining with znp1. 129

145 Figure 4.7: Wild-type muscle cells rescues CaP motor axons in des mutant embryos. Wild-type cells (red) transplanted to des mutants. znp1 antibody staining of CaP motor axons (A), merged view of transplanted muscle cells and znp1 stained CaP, motor axons (B). Cross-sections of des embryos with transplanted muscle cells. znp1 antibody staining (C), transplanted cells (D), and (C) and (D) merged in (E). Arrows indicate rescued motor axons. 130

146 Figure 4.8: sema3a2 is expressed throughout the somitic domain in fss mutants. sema3a2 expression is restricted to the posterior domain of the somite in wild-type (A). Expression is not restricted in the somite region of fss mutants. Scale bar is 50 µm 131

147 Figure 4.9: Overexpression of α-actinepha4egfp results in abnormal CaP motor axons. znp1 antibody staining of CaP motor axons (A). α-actinepha4egfp expressing muscle cells (B). (A) and (B) merged (C). Arrow in A indicates an abnormal axon and arrowhead in B indicates GFP expressing muscle cells. (n=50 axons) 132