Expression of Chick Semaphorin 5B During Neuronal Development by A R T H U R T E R R E N C E L E G G B . S c , The University of British Columbia, 1998 A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S OF T H E D E G R E E OF M A S T E R OF S C I E N C E in T H E F A C U L T Y OF G R A D U A T E S T U D I E S (Department of Anatomy and Cel l Biology) We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y OF B R I T I S H C O L U M B I A Apr i l 2003 © Arthur Terrence Legg, 2003 U B C Rare Books and Special Collections - Thesis Authorisation Form Page 1 o f 1 In presenting t h i s t h e s i s i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree that the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r reference and study. I further agree that permission f o r extensive copying of t h i s t hesis f o r s c h o l a r l y purposes may be granted by the head of my department or by his or her representatives. I t i s understood that copying or p u b l i c a t i o n of t h i s t hesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. The U n i v e r s i t y of B r i t i s h Columbia Vancouver, Canada http://www. library. ubc. ca/spcoll/thesauth. html 4/22/2003 Abstract During development, the formation of a functional nervous system requires precise pathfinding of axons to their targets. Growth cones at the leading edge of these axons use the information supplied by a variety of cues in the environment to navigate their course. The semaphorins, comprising a large family o f neuronal guidance cues, were first identified in the grasshopper limb bud and were shown to be important for proper T i l axon pathfinding. Most of the studies from invertebrates and vertebrates have demonstrated an inhibitory role but more recent studies have shown them to have a b i -functional nature, also acting as attractive cues. The transmembrane semaphorin, chick semaphorin 5B ( c S E M A 5 B ) , is unique from other semaphorin in that it contains both an inhibitory S E M A domain and a region of thrombospondin type-1-like repeats that have been associated with neuronal outgrowth. To determine whether this semaphorin plays a role in axon guidance in the developing chick nervous system, its expression was analyzed using a variety of techniques including in situ hybridization, R T - P C R , and immmunocytochemistry. Expression of c S E M A 5 B is first clearly identified at E5 in the spinal cord, D R G s , retina, and in a variety of neuroepithelia associated with the tectum, ventricular regions, and the olfactory system. Expression within the spinal cord is dynamic being first broadly expressed in both dorsal and ventral regions at E5 with the ventral expression persisting through E l 1. A t this later stage the expression is associated with large-diameter cells in the lateral motor column. Expression within the retina occurs along the retinal ganglion cell layer and is relatively uniform in distribution and is maintained from E5 through E10, while the expression in the tectum appears to occur in a gradient with highest levels in the anterior region. Results from these studies, along with i i some important in vitro studies performed by others in our lab, have suggested that c S E M A 5 B is important for the neuronal pathfinding within the spinal cord, retina, and developing tectum. Table of Contents Abstract i i List o f Figures v i List of Tables v i i List o f Abbreviations v i i i Acknowledgements x i I N T R O D U C T I O N 1 A x o n guidance 1 Semaphorins 1 Genetic Analysis of Semaphorins 5 Semaphorin Receptors 7 Genetic Analysis of Semaphorin Receptors 10 Semaphorin Signaling 12 Transmembrane Semaphorins 17 Class 5 Semaphorins 18 Hypothesis 20 M A T E R I A L S A N D M E T H O D S 22 R N A Work 22 Probe sequence 22 Synthesis of DIG-labeled R N A probe 22 In Situ Hybridization studies 23 Tissue preparation 23 Pre-hybridization 24 In Situ Hybridization 24 Post-hybridization. 25 Niss l stained sections 25 Northern Analysis 26 R T - P C R studies 26 Affinity Purification of c S E M A 5 B Antibody : 27 Purification of GST-fusion protein 27 Affinity column preparation 28 Antibody purification procedure 29 Antibody Staining 29 Antibody staining o f dissociated cultures 30 Western Blots 30 Cytoskeletal extractions 31 R E S U L T S 33 Structure of c S E M A 5 B 33 Construction o f an R N A probe 33 Northern Studies 38 R T - P C R Studies 41 In Situ Hybridization Studies 42 c S E M A 5 B expression in the Spinal Cord 42 c S E M A 5 B expression in D R G 45 c S E M A 5 B expression in the Retina 45 iv c S E M A 5 B expression in the Tectum 49 Epithelial expression 52 Immunohistological Studies 57 Westerns 57 c S E M A 5 B distribution in the Spinal Cord 60 Distribution of c S E M A 5 B in D R G s 68 c S E M A 5 B distribution in the retina 68 D I S C U S S I O N 72 Role of S E M A 5B in developing spinal cord 73 Functional Studies 79 Role of S E M A 5 B in retina 81 Functional Studies 82 Role of c S E M A 5 B in tectum 83 Role of S E M A 5 B in the neuroepithelium 85 Future directions 88 R E F E R E N C E S 89 v List of Figures Figure 1: Semaphorins and Semaphorin Receptors 3 Figure 2: Semaphorin Signaling 13 Figure 3: Semaphorin 5B Structure and amino acid sequence 34 Figure 4: Analysis and Alignment of the c S E M A 5 B R N A Probe 36 Figure 5: Northern and R T - P C R Results 39 Figure 6: In situ hybridization studies in the spinal cord 43 Figure 7: In situ hybridization studies in the retina 47 Figure 8: In situ hybridization studies in the tectum 50 Figure 9: In situ hybridization studies in ventricular regions 53 Figure 10: In situ hybridization studies in nasal regions 55 Figure 11: Antibody production and western blots 61 Figure 12: Semaphorin 5B distribution in Spinal Cord 63 Figure 13: Dissociated Spinal Cultures 66 Figure 14: Semaphorin 5B distribution in the Retina 69 Figure 15: Models for c S E M A 5 B functions in axon guidance 75 Figure 16: A Model for c S E M A 5 B function in migration 86 vi List of Tables Table I: Summary of c S E M A 5 B expression in all regions 58 vn List of Abbreviations bp basepairs B C I P 5-bromo-4-chloro-3-indolyl-phosphate C N S central nervous system Cy3 indocarbocyanine C U B complement-binding D H dorsal horn D R G dorsal root ganglion E C M extracellular matrix E N external nuclear layer E P ependymal layer D E P C diethyl pyrocarbonate F ITC fluorescein isothiocyanate G C L retinal ganglion cell layer GPI glycosylphosphatidylinositol G S T glutathione S-transferase I N inner nuclear layer kb kilobase k D kilodalton mg milligram M g C b magnesium chloride MgS04 magnesium sulphate M I C A L flavoprotein monooxy-genase ml milliliter m M millimolar N B T 4-nitro blue tetrazolium chloride N G F nerve growth factor Npn neuropilin N T neurotrophin O F L optic fiber layer O N optic nerve O N H optic nerve head O T K off-track P B S phosphate buffered saline P C R polymerase chain reaction P L M C presumptive lateral motor column P M S F phenylmethylsulfonyl fluoride P N S peripheral nervous system R G C retinal ganglion cell R T - P C R reverse-transcriptase polymerase chain reaction SDS sodium dodecyl sulfate S G C stratum griseum centrale SO stratum opticum ng microgram ul microliter u M micromolar ventral horn white matter Acknowledgements I would firstly like to thank Dr. Timothy P. O'Connor for allowing me to conduct my research in his lab, for his excellent guidance throughout the entire duration of this degree, his passion for science, and for his friendship in and out of the lab. I would like to thank my committee members, Dr. W . Tetzlaff and Dr. C . Roskelley for their suggestions and scientific advise. I would like to thank Dr. Jennifer Bonner for her scientific guidance, her friendship while sharing an office. I would like to thank my fellow graduate students in Anatomy as they made this whole process a lot of fun. I would like to thank my parents who have supported me in whatever endeavors I have gone through. Finally, I would like to thank my Lord Jesus Christ who gives me strength and guidance always. x i Introduction: Axon guidance: The generation of a functional nervous system requires precise pathfinding o f axons to their appropriate targets. How neurons accomplish this feat while making relatively few errors is unclear, however, the precise temporal and spatial distribution of guidance factors is a requirement. Guidance cues evoke stereotyped responses from neurons including attraction or repulsion, and can act locally i f bound to cell membranes and the E C M , or at a distance i f they are secreted. The growth cone, a highly dynamic, actin-rich sensory structure found at the leading edge of the axon, is responsible for interpreting these signals and translating them into a number of growth cone behaviors including outgrowth, turning, fasciculation, retraction, and stalling (Suter and Forscher 2000). When first describing a growth cone over a century ago, Ramon y Cajal proposed that this structure responds to guidance factors to direct axonal growth. This has proven to be true and in the subsequent century, numerous guidance cues have been identified, and their effects on growth cone behaviors have been characterized. Even though a number of families o f guidance cues have been identified, the precise mechanisms underlying growth cone behavior and response to these guidance cues are not resolved. Semaphorins: To date the major families of guidance proteins identified include the Ephrins, Slits, Netrins, and Semaphorins (Grunwald and K l e i n , 2002). These guidance cues can localize to the membrane, as is the case with ephrins and some members of the semaphorin family, or they may be secreted proteins like netrins, slits, as wel l as some of the semaphorins. Different families are typically associated with a particular response, 1 either attractive or inhibitory, but many exhibit bi-functional guidance properties. O f these guidance cues, semaphorins comprise the largest family. With more than 20 different members identified to date, the semaphorins are subdivided into eight different classes based on domain structure and species o f origin (figure 1; Semaphorin Nomenclature Committee 1999). Semaphorins comprise a diverse family of secreted and membrane associated proteins characterized by the presence of a conserved semaphorin (Sema) domain, 500 amino acids in size. The semaphorins may be bound to the membrane through either a transmembrane domain (classes 1, 4, 5, 6) or through a glycophosphatidyl inositol linkage (class 7), or they may be secreted as is the case with classes 2, 3, and V (figure 1). The first identified semaphorin, grasshopper Sema-la (previously named fasciclin IV), was initially implicated in axon guidance using antibody perturbation experiments (Kolodkin et al., 1992, 1993). Subsequent purification of a secreted semaphorin, chick Collapsin-1 (Sema3A), demonstrated the chemorepulsive nature o f this semaphorin as acute addition led to rapid actin depolymerization and growth cone collapse of D R G neurons in vitro (Luo et al., 1993). Following this initial characterization, a number of studies went on to demonstrate the repulsive nature of semaphorins for a variety of neuronal populations (Matthes et al., 1995; Tanelian et al., 1997; Bagnard et al., 1998; Chedotal et al., 1998; Shoji et al., 1998; Winberg et al., 1998b; Y u et al., 1998; de Castro et al., 1999; Miyazaki et al., 1999; Rabacchi et al., 1999; Roos et al., 1999, Steup et al., 1999; Y u et al., 2000; Tamamaki et al., 2003). In addition to functioning as repulsive guidance cues, semaphorins have also been found to function as attractive cues (Wong et al., 1997; Bagnard et al., 1998; Song et al., 1998; de Castro et al., 1999; Wong et al., 1999; Bagnard et al., 2000; Fujioka et al., 2003). For 2 Figure 1 : Semaphorins and Semaphorin Receptors Genes encoding semaphorins are highly conserved from invertebrates to humans. There are at least 20 known semaphorins that can be divided into 8 subclasses based on structural similarities. Classes 1, 2, and 5 are invertebrate semaphorins, classes 3 through 7 are vertebrate semaphorins, and class V are viral associated semaphorins. Common to all semaphorins is a 500-amino acid length semaphorin domain at their amino termini (orange), which is highly conserved between invertebrates and vertebrates. The carboxyl regions of the semaphorins are more variable. Classes 2 through 4, 7, and a subset of V contain an Ig domain (blue/white), while class 5 members contain a thrombospondin type-1 repeat domain (purple). The semaphorins are further distinguished biochemically as being secreted, membrane glycosylphosphatidylinositol (GPI)-anchored, or transmembrane molecules. A schematic representing the two classes of semaphorin receptors, the neuropilins and plexins, depicts the various domain structures associated with each. There are two classes of neuropilins, both of which have a very small intracellular domain, which interact with the secreted class 3 semaphorins and the plexins, and are only found in vertebrates. The plexin family has four classes (A-D) in vertebrates, with known interactions with semaphorins from classes 3, 4, and 7. There are two classes ( A & B ) in invertebrates that are known to interact with semaphorins from class 1. Known interactions between the receptors and semaphorins are shown with lines (red). 3 example, while Sema3A is repulsive for cortical axons, another semaphorin, Sema3C acts as an attractive guidance cue for these same neurons (Bagnard et al., 1998). Similarly, in the developing grasshopper limb, the T i l neurons turn away from their normal pathway to contact cells ectopically expressing Sema-la, suggesting an attractive response (Wong et al., 1999). In addition, the activation of the c G M P pathway in Xenopus spinal neurons can convert Sema3A repulsion to attraction (Song et al., 1998). Additional examples o f both repulsive and attractive roles for these proteins in vitro have been demonstrated (Raper, 2000). Genetic Analysis of Semaphorins: Along with many in vitro studies describing the function of semaphorins, a number of studies have described a role for semaphorins in vivo. Observations on a number of mutants from classes 1 and 2 in invertebrates, and from classes 3 and 6 in vertebrates, have led to many conclusions about the in vivo function of these semaphorins (Behar et al., 1996; Taniguchi et al., 1997; Shoji et al., 1998; Catalano et al., 1998; Winberg et al., 1998; Y u et al., 1998, 2000; Roy et al., 2000; White et al., 2000; Bahri et al., 2001; Leighton et al., 2001; Ginzburg et al., 2002). There are three semaphorins (Ce-sema-la or smp-1, Ce-sema-lb or smp-2, Ce-sema-2a or mab-20) in C. elegans and mutants for each have been described (Roy et al., 2000; Ginzburg et al., 2002). While the majority of defects associated with Ce-sema-2a include errors in morphogenesis and epithelial cell-to-cell contacts, these embryos also display errors in axon guidance and cell migration (Roy et al., 2000). Mutants exhibit a number o f fasciculation and pioneer guidance defects in at least one of the D A and D B motor neurons, as well as errors in neuroblast cell migration (Roy et al., 2000). A mutation in either the Ce-sema-la or Ce-5 sema-lb also results in a number of defects in epidermal cell morphogenesis and cell-to-cell contacts (Ginzburg et al., 2002). In addition to errors in morphogenesis, Ce-sema-la and Ce-sema-lb mutants exhibit mild defects in the guidance o f P L M L and P L M R axons, including premature stopping and unusual branching (Ginzburg et al., 2002). In Drosophila, null mutations in D-Sema-la result in premature stalling of motor axons and failure o f these axons to defasciculate (Yu et al., 1998). In addition, ectopic expression of D-Sema-la in muscle cells results in motor axons avoiding these targets ( Y u et al., 1998). Expression o f a truncated form of D-Sema-la in these flies is able to rescue this phenotype, suggesting that the cytoplasmic domain of this transmembrane semaphorin is not required for this function (Raper, 2000). Recent studies on D-Sema-la mutants have demonstrated that this semaphorin is also required for proper axon guidance and synapse formation in the adult Giant Fibre system (Godenschwege et al., 2002). Rescue experiments in these mutants further demonstrated that D-Sema-la is required both pre-and post-synaptically (Godenschwege et al., 2002). Ectopic expression of D-Sema-2a in muscles prevents motoraxons from correctly pathfinding to these targets (Winberg et al., 1998). The converse is true in D-Sema-2a mutants, where many motorneurons are observed to incorrectly invade muscle fibers that normally express D-Sema-2a (Winberg etal., 1998). In addition to genetic analysis of semaphorins in invertebrates, vertebrate mutants have also been described. M i c e lacking Sema3 A exhibit a number o f defects in sensory and motor axon patterning in the developing spinal cord (Behar et al., 1996; Taniguchi et al., 1997), however, the overall morphology of the C N S in these mice is relatively normal (Catalano et al., 1998, Ulupinar et al., 1999). This may be due to functional redundancy 6 of different semaphorins. For example, sympathetic neurons are repulsed by multiple semaphorins from the class 3 family in vitro (Adams et al., 1997; Takahashi et al., 1998). Homozygous mutant mice lacking Sema6A, the mouse homologue o f invertebrate D -Sema-la, display a number of aberrant thalamocortical projections (Leighton et al., 2001). These neurons normally express Sema6A, which suggests that this guidance defect is acting cell-autonomously and that this semaphorin serves as a guidance receptor in these neurons (Leighton et al., 2001). Thus, while the studies on the Sema3A knock-out mouse are not conclusive, other in vivo studies from invertebrates and mice support the in vitro findings and demonstrate that a number of different semaphorin classes normally function in axon guidance during development. Semaphorin receptors: Two families of transmembrane proteins, the neuropilins (Npn) and the plexins have been identified as receptors for semaphorins. The first receptors discovered belong to the neuropilin family o f proteins, Npn-1 and Npn-2. They were first identified as semaphorin receptors in an expression screen which exploited their high affinity for semaphorin-3A (Kolodkin et al., 1997; He and Tessier-Lavigne, 1997; Fujisawa and Kitsukawa, 1998; Kolodk in 1998). Subsequent analysis demonstrated that this interaction was conserved for other class three members, however, different class members were shown to interact with different combinations of neuropilins for signaling (Nakamura et al., 2000). Specifically, Sema3A was shown to signal through a receptor complex consisting of Npn-1, while Sema3F signals through a complex consisting of Npn-2, and 7 Sema3C signals through a hetero-dimeric complex consisting of both Npn-1 and Npn-2 (Nakamura et al., 2000). Npn-1 was originally characterized as a cell surface molecule with a restricted expression pattern in the optic tectum of Xenopus (Takagi et al., 1991). When over expressed in embryonic mice, Npn-1 leads to axon defasciculation and sprouting (Kitsukawa et al., 1995). The Npns are structurally characterized by the presence of two extracellular complement-binding ( C U B ) domains, followed by the presence o f two coagulation factor V / V I H homology domains, a M A M domain, and a small cytoplasmic region (figure 1; Chen et al., 1997). The C U B domain was shown to be the interacting domain for Sema3A from analysis of various Npn-1 deletion mutants (Nakamura et al., 1998). Many observations apart from the biochemical interaction studies have demonstrated that Npn-1 is required for functional Sema3A receptors. Antibodies to Npn-1 block Sema3A induced collapse o f neurons in vitro (He and Tessier-Lavigne, 1997; Kolodkin et al., 1997). In addition, D R G neurons isolated form Npn-1 null mice do not undergo growth cone collapse in the presence of semaphorin 3 A (Kitsukawa et al., 1997). The small cytoplasmic tail of Npn-1 interacts with a S E M C A P protein, neuropilin interacting protein (NIP) which is thought to induce clustering, however, it is the extracellular region of Npn-1 which is thought to mediate its function in semaphorin binding (figure 2C; Cai and Reed, 1999). Based on its structure and some key observations, it appears as though at least one additional transmembrane protein is required for Sema3A activity (Nakamura et al., 1998). Presently, an invertebrate equivalent for neuropilin has not been found. 8 Another family of transmembrane proteins, the plexins bind specifically to semaphorins either alone, or in combination with the neuropilins (Comeau et al., 1998; Winberg et al., 1998). There are nine known plexins, which subdivide into four families based on structure; p lex in-Al-4 , B l - 3 , C l , and D l (Tamagnone et al., 1999). Plexins are transmembrane proteins characterized by the presence of an extracellular semaphorin domain and cysteine-rich region, as well as a conserved plexin-specific cytoplasmic domain (Tamagnone et al., 1999). There are 3 known classes of semaphorins which bind directly to plexins. In Drosophila, sema-la and sema-lb bind p l ex in -A l directly, while in vertebrates Sema4B binds p l ex in -B l , and Sema7A binds p lex in-Cl (figure 1; Winberg et al., 1998; Lange et al., 1998; X u et al., 1998). Neuropilins and plexins interact to form a receptor complex for the secreted class 3 semaphorins. Initial results demonstrated that plexins form stable complexes with Npn-1 and Npn-2 when ectopically expressed in COS-7 cells (Takahashi et al., 1999). When co-expressed in COS-7 cells, neuropilins and plexins induce collapse in the presence of Sema3A, while expression of either one alone does not (Takahashi et al., 1999). Over-expression of a truncated p l ex in -A l lacking the intracellular domain inhibits Sema3A induced collapse in cultured sensory neurons (Takahashi et al., 1999). In addition to the Npns and the plexins, another family o f receptor molecules shown to contribute to semaphorin receptor complexes includes L l , a member o f the immunoglobulin superfamily o f adhesion molecules. Analysis from L l -deficient mice demonstrated that cortical neurons in these mice were unresponsive to Sema3A but not to Sema3B or Sema3E (Castellani et al., 2000). Further analysis demonstrated that the extracellular domains of L l and Npn-1 directly associate to form a receptor for Sema-3A 9 (Castellani et al., 2000). Interestingly, a soluble form of L I , with an attached Fc domain, can convert Sema3A-induced axonal repulsion into attraction in vitro (Castellani et al., 2000, 2002). Studies on plexin associated proteins from Drosophila have implicated the re-named Dtrk receptor off-track ( O T K ) , as a binding partner required for Plexin A l signaling (Winberg et al., 2001). This was supported from experiments which demonstrated that otk mutants displayed similar phenotypes to loss-of-function mutations of either Semala or PlexA (Winberg et al., 2001). Additionally, otk loss-of-function mutations interact genetically with Semala and P lexA, and reduced levels of otk suppress Sema-la gain-of-function phenotypes (Winberg et al., 2001). Finally, recent work by Giordano et al. 2002 has demonstrated that Plexin B I (the Sema4D Receptor) and Met (the Scatter Factor 1/ Hepatocyte Growth Factor Receptor) associate in a complex when co-expressed in M L P 2 9 epithelial cells. Binding of Sema4D to Plexin B I stimulates the tyrosine kinase activity o f Met, resulting in tyrosine phosphorylation of both receptors (Giordano et al., 2002). Cells that lack Met expression do not respond to Sema4D (Giordano et al., 2002). This same cells w i l l respond to Sema4D is they are then forced to express Met (Giordano et al., 2002). These results suggest that the Met receptor may also contribute to a receptor complex for various Sema4D. Genetic Analysis of Semaphorin Receptors: A number of mutants have been described for both the neuropilin and the plexin families o f receptors. There are two plexins (plx-1 ,plx-2), in C. elegans, but no obvious neuropilin genes (Fujii et al., 2002). P lexinA (plx-1) null-mutants exhibit a number of 10 morphological defects including the displacement o f r a y l , one of nine male-specific genital sensilla (simple sense organ) in the tale, and a partial loss of seam cells which lead to a discontinuous outer alae, a cuticular structure running longitudinally along the lateral surface of the body wal l (Fujii et al., 2002). Interestingly, suppression of Ce-Sema-la and Ce-Sema-lb displayed a similar phenotype toplx-1, whereas Ce-Sema-2a mutants exhibit a distinct phenotype (Fujii et al., 2002). In Drosophila, PlexA loss-of-function mutant phenotypes phenocopy those of Semala including a failure o f ISNb growth cones to defasciculate from one another at any or all three of the ISNb choice points (Winberg et al., 1998). Also , PlexA and Semala loss-of-function mutations interact genetically, while PlexA loss-of-function suppresses Semala gain-of-function phenotypes (Winberg et al., 1998). Interestingly, over-expression of Plex A in all neurons leads to axon guidance defects in all parts of the motor projection and within the C N S (Winberg et al., 1998). Mutations in PlexB also result in axon guidance phenotypes (Hu et al., 2001). The ISNb neurons in flies overexpressing PlexB fail to innervate their normal muscle targets (Hu et al., 2001). In mice lacking neuropilin-1, the phenotype is similar to that of Sema3A but it is much more severe, and mice arrest developmentally (Kitsukawa et al., 1997). These animals display severe abnormalities in cranial and spinal neurons, along with aberrant efferent projections within the PNS including abnormal limb innervation (Kitsukawa et al., 1997). M i c e with mutations in neuropilins-2 also display a number of neuronal defects within the C N S and P N S , including severe defasciculation of oculomotor neurons (Giger et al., 2000, Mar in et al., 2001). Mice that lack plexin-A3 have a number of axon guidance defects including an inability o f the ophthalmic branches of the trigeminal nerve 11 to fasciculate (Cheng et al, 2001). These mice also display a number o f guidance defects associated with the hippocampus including aberrant dentate granule cell and mossy fibre projections (Cheng et al., 2001). Semaphorin Signaling: Recent studies have begun to identify the downstream signaling targets involved in transmitting the semaphorin signal (reviewed in Castellani and Rougon, 2002). A number of these studies have implemented the Rho GTPases in semaphorin signaling. Dominant negative R a c l was shown to inhibit Sema3A-induced collapse o f D R G neurons, while a constitutively active R a c l partially mimicked the Sema3A growth cone collapse (Jin and Strittmatter, 1997). Both plexins and neuropilin signaling cascades result in the activation or inhibition of specific Rho GTPases including Rho A , R a c l , R n d l , and RhoD (figure 2). In Droshophila, P lexB was shown to bind activated R a c l , sequestering it from p21 activated kinase ( P A K ) its downstream effector (figure 2 A ; Driessens et al., 2001, H u et al., 2001). In addition to an interaction with activated R a c l , P lexB also binds R h o A creating a scenario were signaling through PlexB leads to collapse by simultaneously inhibiting the R a c l pathway, while leading to activation o f the R h o A pathway (figure 2 A ; Driessens et al., 2001, H u et al., 2001). This model was also supported from in vitro studies on interactions between Sema4D and plexin B I , which demonstrated that clustering of this plexin leads to an interaction with active Rac-1, that the collapse induced by Sema4D is characteristic of Rho activation, and that there was again an inhibition of P A K activity (figure 2B; V i k i s et al., 2000; Rohm et al., 2000; Driessens et al., 2000). Again, R a c l and R h o A activities are regulated through distinct 12 Figure 2: Semaphorin Signaling A series of schematic representations o f signal transduction pathways involved in semaphorin activity is shown. In (A), the invertebrate Semal A interacts with Plexin A l and O T K to recruit kinases including enabled. In (B), the vertebrate class 4 transmembrane semaphorins signal through the class B plexins associated with M E T . This activation leads to a recruitment o f the Rho family of small GTPases that signal to either inhibit, or induce collapse. The guanine exchange factor L A R G interacts with plexin B l and activates the Rho A pathway. In (C), the secreted semaphorins, signal through a clustered neuropilin/plexin complex to recruit R n d l , which activates P A K leading to collapse. Recruitment o f RhoD to the same binding region inhibits this R n d l activity. Signaling through plexin A l also leads to activation of a number of other pathways involving C R M P s , Fyn, Cdk5, and a pool of G S K - 3 . Receptor interactions with L l are thought to modulate semaphorin signaling through changes in c G M P levels. In (D) bi-directional signaling of transmembrane semaphorins may occur through a number o f possible candidates. 13 ScmalA B Recruitment Of Other Kinases 1 / 1 Increased 1 motility Decreased motility j / Actio Turnover / Decreased Motility/ Actin Turnover PLEXIN SIGNALLING (transmembrane semaphorins) Local Protein Translation Tubulin Dynamics Reorganization of Cytoskeleton NEUROPILIN/PLEXIN SIGNALING (secreted semaphorins) SEMAPHORIN BI-DIRECTIONAL SIGNALING 14 binding sites where Plexin B l directly interacts with activated R a c l (figure 2B; H u et al., 2001; V i k i s et al., 2002). Recent work by Aurandt et al. (2002) has implicated the activation of Rho by a Rho-specific nucleotide exchange factor L A R G , in response to stimulation by Sema4D (figure 2B). The L A R G protein was shown to interact directly with the C-terminal region of Plexin B l through its P D Z binding domain (Aurandt et al., 2002; Perrot et al., 2002). Sema3A signals through clustering o f a neuropilin/plexin A l complex (figure 2C). There appear to be differences in the downstream effectors o f Plexin A l compared to plexin B l (figure 2B, C) . While R a c l activity is important for signaling through plexin A l , it does not appear to interact with this plexin (figure 2C; Kuhn et al., 1999; Rohm et al., 2000). While R a c l does not interact with plexin A l , both the Rho-like GTPases R n d l and RhoD were shown to bind to the intracellular region of plexin A l , and actually compete for the same binding region (figure 2C; Rohm et al., 2000; Zanata et al., 2002). Interestingly, while R n d l activation by plexin A l results in cytoskeletal collapse, binding of RhoD has an antagonist effect, blocking R n d l activity (figure 2C; Zanata et al., 2002). Plexin and neuropilin receptor complexes also induce activation of L L M kinase and subsequent inhibition of cofilin, resulting in a reduction in actin turnover and retraction of the growth cone (figure 2C; Aizawa et al., 2001). The Collapsin response mediator proteins ( C R M P s ) are another family of proteins which were first identified by their possible involvement in the Sema3A-induced collapse o f D R G growth cones (Goshima et al., 1995). Further evidence came from a study demonstrating that an an t i -CRMP antibody blocks collapse induced by Sema3 A (Wang and Stittmatter, 1997). A recent study examining the role of C R M P s in Sema3A 15 signaling demonstrated that they are linked to Plexin A l through activation of Fes/Fps (Fes) tyrosine kinase which then phosphorylates plexin A l and C R M P in a Fes /CRMP/CRAM(CRMP-assoc ia ted molecule) complex (figure 2C; Mitsui et al., 2002). The activation of Fes appears to be required for Sema3A-induced collapse of COS7 cells expressing P l e x A l , Npn-1, and Fes (Matsui et al., 2001). Signaling through Sema3A also appears to activate a pool of glycogen synthetase kinase 3 (GSK3) , a seronine/threonine kinase implicated in a variety of growth factor signaling cascades (figure 2C; Eickholt et al., 2002). This activation occurs at the leading edge of the growth cone (Eickholt et al., 2002). Other downstream effectors of semaphorin signaling include the flavoprotein monooxy-genases ( M I C A L S ) , which have been shown to interact directly with P l exA in Drosophila (figure 2A; Terman et al., 2002). The M I C A L family o f proteins have multiple domains which are known to be important for interactions with actin, intermediate filaments, and cytoskeletal-associated adaptor proteins, and may thus mediate the cytoskeletal alterations characteristic o f semaphorin signaling (Terman et al., 2002). A n additional kinase dependent pathway in Sema3A signaling involves the src kinase Fyn and the Thr/Ser kinase cyclin-dependent kinase 5 (Cdk5: figure 2C; Sasaki et al., 2002; Pasterkamp and Kolodkin, 2003). Evidence from Xenopus neurons suggests that Sema3A signaling also requires local protein synthesis in the growth cone, as the inhibition of translation in the axon prevents Sema3 A induced collapse and turning (figure 2C; Campbell and Holt, 2001). In addition to functioning as ligands, transmembrane semaphorins also contain intracellular regions which have been shown to interact with a variety of targets that may be involved in various signaling cascades. In Drosophila, Sema-la is involved in 16 synapse formation and may signal bi-directionally via Enabled (figure 2D; Godenschwege et al., 2002). Other possible interacting proteins include the synapse associated protein PSD-95, S E M C A P - 1 , and E V L (ena vasp regulator) protein (figure 2D; Inagaki et al., 2001; Wang et al., 1999; Klostermann et a l , 2000). While the list o f possible candidates involved in semaphorin signaling is constantly increasing, a clear picture as to how the semaphorins are able to pass on guidance information to the cytoskeleton is unclear. Even more unclear is whether the transmembrane semaphorins are able to signaling bi-directionally, and whether this is crucial for their role in axon guidance. The studies on bi-directional Sema-la signaling in Drosophila would suggest that the answer to the last two questions is yes (Godenschwege et al., 2002). Transmembrane Semaphorins: The majority o f research on semaphorins in vertebrates has focused on characterizing the role of the class 3 secreted semaphorins during neuronal development. Recently, more work has begun to focus on the role of transmembrane semaphorins in development, however little is known about their function. Most of the studies on vertebrate transmembrane semaphorins have looked at the role o f the class four semaphorin Sema4D (CD100), a semaphorin important in the immune system (Hall et al., 1996; reviewed in Suzuki et al., 2003). Recent experiments on the vertebrate semaphorin Sema6A, a transmembrane semaphorin homologous to the insect semaphorin class 1, have demonstrated its ability to repel sympathetic and D R G neurons in vitro ( X u et al., 2000). Analyses from mice lacking Sema6A transcript suggest that it is also important for proper development o f the thalamocortical projection in vivo (Leighton et al., 2001). Based on research showing that a transmembrane semaphorin, semaphorin l a has the 17 potential to be attractive for Til neurons in the developing grasshopper limb, our lab focused on determining whether transmembrane semaphorins might play a similar attractive role in vertebrates (Wong et al., 1999). Using a series of degenerate primers to the grasshopper semaphorin 1 a semaphorin domain, an embryonic chick cDNA library was screened to determine if other semaphorins with a conserved semaphorin domain might convey a similar function. One of the candidate genes that was identified corresponded to a class 5 transmembrane semaphorin, chick Semaphorin 5B (cSEMA5B). Class 5 Semaphorins: The class 5 family members are distinguished from other semaphorins by the presence of a unique protein domain containing seven thrombospondin type-1 repeats on their extracellular region, located 3' of the semaphorin domain. This domain is of particular interest as previous work has show that thrombospondin type-1 repeats can induce outgrowth of rat cortical neurons in culture, and is responsible for the attachment of various cell types, both neuronal and non-neuronal, when grown on a substrate containing thrombospondin (Neugebauer et al., 1991; O'Shea et al., 1991 Osterhout et al., 1992; Adams and Tucker, 2000). This class of semaphorins was first described in mice, and contains two known family members in vertebrates, Sema5A and Sema5B (Adams et al., 1996). These two family members are 58% identical and 72% similar to each other, with Sema 5 A spanning 1077 amino acids, and Sema5B spanning 1093 amino acids (figure 4; Adams et al., 1996). The semaphorin domains in these mice are 64% identical to one another and are most similar to invertebrate semaphorin 1 domains (-60-66% similar). In mice, Sema5 A and Sema5B are differentially expressed in embryonic and 18 adult tissues. Though expressed in similar regions, Sema5A is expressed primarily in mesodermal cells, while Sema5B is exclusively in the neuroepithelium (Adams et al., 1996). In addition, northern blot analysis on mouse tissue shows that the highest expression levels of Sema5B were found in brain tissue expressed as a single transcript, 5.9kb in size (Adams et al., 1996). The expression of semaphorin 5B has also been examined in the forebrain of rats over a development period ranging from E15-P7 (Skaliora et al., 1998). S E M A 5 B is broadly expressed in all neuroepithelia throughout the brain at each of the ages examined. In particular, epithelial regions, such as the ventricular zone, which normally corresponds to high areas of neuronal proliferation, show strong expression (Skaliora et al., 1998). S E M A 5 B is distributed uniformly and is strongly expressed in cortical ventricular and subventricular zones at postnatal day 0 (Skaliora et al., 1998). This wide spread epithelial expression suggests that S E M A 5 B might play a general role in these proliferative tissues, possibly with cell migration (Skaliora et al., 1998). Interestingly in the rat, S E M A 5 B is expressed in discrete patches within the striatum and may be involved in the correct patterning o f cells within this region o f the brain. A t P7, S E M A 5 B is highly expressed in the rat visual cortex (Skaliora et al., 1998). In Drosophila, only one member of this class of semaphorins exists (Khare et al., 2000; Bahri et al., 2001). It is most similar to Sema5B but is not a homologue o f either vertebrate class 5 members, and was thus given the designation of Semaphorin 5C, (DSema 5C; Khare et al., 2000; Bahri et al., 2001). In flies, DSema 5C is found at segment boundaries in regions giving rise to muscle attachments (Bahri et al., 2001). DSema 5C exists as two different protein isoforms that share a similar expression pattern. 19 The DSema 5C mutant flies are homozygous viable and display no obvious embryonic phenotypes (Bahri et al., 2001). A s no detailed analysis was performed on the nervous system, a clear role for this semaphorin in Drosophila has not been established. A possible role for class V semaphorins in axon guidance was suggested from studies examining the well known neurological disorder in humans, Cris-du-chat syndrome (Simmons et al., 1998). These studies reveal that the S E M A 5 A gene accounts for 10% of the deleted region associated with this disorder. In addition, recent studies on Sema5 A in mouse have demonstrated its ability to repulse retinal axons in vitro (Oster et al., 2003). In these studies, Oster et al. 2003 examined the expression of this semaphorin in the developing retina and found it to be specifically expressed in the optic disk and along the optic nerve. They were able to demonstrate that the inhibitory response o f R G C to Sema5A was maintained in the presence o f L l , laminin, or netrin-1 signaling in vitro (Oster et al., 2003). Hypothesis: While the studies in both the mouse and rat have described the expression of Semaphorin 5B, they both have a number of limitations in that they have not examined the distribution of this semaphorin protein, they have only looked at particular regions or over a small developmental range, and they have not provided much insight as to the role o f the class 5 semaphorins. In order to determine a functional role for this semaphorin, its expression was examined in the developing chick from E2 through until E l 7 . The examination of c S E M A 5 B expression was completed using a series o f in situ hybridization studies, and with antibodies raised against the N-terminal region of the protein. I hypothesize that c S E M A 5 B is expressed during in regions associated with 20 neuronal outgrowth and guidance, including the spinal cord, retina, and tectum. Based on its expression along with the results from functional in vitro studies performed by others in our lab, I propose that c S E M A 5 B is an inhibitory guidance cue that contributes to proper axon outgrowth and guidance within these structures. In addition, I hypothesize that c S E M A 5 B is expressed along many neuroepithelia and that it plays an inhibitory role in these highly proliferative regions, to facilitate there migration from these epithlial layers towards target tissue in the developing C N S . 21 Methods RNA Work: Probe sequence: A cDNA sequence corresponding to the C-terminal cytoplasmic region of cSEMA5B was produced from the full-length cSEMA5B (obtained from an E10 chick cDNA library, Clontech) sequence cloned into the pLitmus29 vector. The following primers were generated to isolate this sequence from pLitmus29: Primerl (forward): 5 A G C T G T C T G A T C C C C A T G A G 3 Primer 2 (reverse): 5 'GAAAAGAACATGCACAAACCC 3 ' The corresponding PCR product was subcloned into the vector bluescript29-KS, containing a T3 and T7 promoter sequence for the production of RNA. Synthesis of DIG-labeled R N A probe: To synthesize DIG-labeled antisense and sense R N A probes, the following reagents (Roche) were mixed in the following order at room temperature to produce a 20ul reaction mixture: 9.5jul of sterile DEPC treated dH 20, 4 pi of 5x transcription buffer, 2/tl 0.1 M DTT, 2/d of nucleotide mix (pH 8.0), 1/d of Linearized bluescript-cSEMA5B plasmid (1 pg/ul), 0.5 pi Placental ribonuclease inhibitor (40 U/ul), ljul of T7 or T3 R N A polymerase (10-20 U/ul). The resulting mixture was incubated at 37° C, for 2 hours. After the reaction was completed, a 1 pi aliquot was removed and separated on a 1.5% agarose/0.5XTBE gel containing 0.5 pg/ml EtBr in order to estimate the amount of probe synthesized. To remove any remaining plasmid 2 pi of ribonuclease-free DNase 1 (10 U/ul) was added to the reaction mixture and it was incubated at 37° C for 15 min. An R N A isolation column (GIBCO) was used to isolate the resulting R N A probe. As an 22 alternative method, R N A was isolated with the addition of a volume of 100 u.1 D E P C treated dH 2 0 and 10 ul of 4 M L i C l (DEPC-treated) to the reaction. The R N A was precipitated with the addition of 300 ul o f ethanol followed by an incubation on ice for 30 minutes. The resulting mixture was spun in a microfuge for 20min. The pellet was washed with 70% ethanol, and spun down for another 2 minutes. A l l o f the E t O H was removed with an aspirator and pellets were allowed to air-dry briefly for 2-5 minutes. The R N A pellet was re-dissolved in D E P C treated dH 2 0 at -0.1 ug/ul (-100 pi) and stored at -80° C. In situ hybridization Studies Tissue Preparation: Embryonic chicks isolated at various stages, were submerged in a solution containing 4% paraformaldehyde in DEPC-treated P B S , and were washed at 4°C for a period of 8 firs. Chicks at E7 or older were first fixed via pericardial infusion, by inserting a needle into the heart and injecting P B S for 2 minutes to clear the blood followed by 10 minutes of Paraformaldehyde to ensure rapid fixation. These embryos were then sectioned into numerous parts before submersion in 4% paraformaldehyde solution at 4°C for 8 hrs. Following the fixation, the resulting embryos were submerged in a solution containing 30% sucrose in DEPC-treated P B S and were left on a rotator at 4°C overnight for cryo-protection. The embryos were submerged in O.C.T (optimal cutting temperature) compound (Tissue-Tek), and placed in a metal tray preteated overnight with 0.2N N a O H . Trays containing embryos submerged in O.C.T compound were frozen in a small vial o f 2-methylbutane contained in a bath of liquid nitrogen. The 23 tissue was sectioned into lOum thick sections using a microtome (Bright Instruments) at -20°C. The sections were placed on slides and dried in an oven at 40°C overnight. Prehybridization: Slides were washed 2 X for 5minutes in DEPC-treated P B S followed by two washes with DEPC-treated P B S containing lOOmM glycine. Sections were treated with DEPC-treated P B S containing 0.3% Triton® X-100 and followed by two washes with DEPC-treated P B S . The sections were permeabilized for 20 minutes at 37°C with T E buffer ( lOOmM T r i s - H C L , 5 0 m M E D T A , p H 8.0) containing lug /ml Proteinase K (RNase free). Following permeabilization, the sections were fixed with DEPC-treated P B S containing 4% paraformaldehyde at 4°C for 5 minutes and washed twice more with DEPC-treated P B S . Slides were then placed in RNase-free slide containers that contained 0 .1M triethanolamine (TEA) buffer, p H 8.0, containing 0.25% (v/v) acetic anhydride (Sigma), and the slides were allowed to shake for 10 minutes. Slides were washed again 2 X 5min with P B S then the sections were incubated at 42°C for 1 hour in prehybridization buffer D I G Easy Hyb solution (Roche Molecular Biochemicals) containing lmg/ml denatured and sheared salmon sperm and lmg/ml yeast t - R N A . In situ hybridization: A wax pen (PAP) was used to encircle the sections to prevent leakage of hybridization buffer during the hybridization step. The slides were incubated in the presence of 200ul D I G Easy hybridization buffer (Roche) containing 50ng o f D I G -labeled anti-sense or sense R N A probe. The sections were covered with a layer o f parafilm to prevent evaporation and were placed in a humid chamber at 42°C for 24hrs or overnight. 24 Post Hybridization Procedure: Following the hybridization step, the sections were put through a series of washes in a shaking water bath. The slides were washed twice for fifteen minutes in 2 X SSC buffer at 37°C, followed by two washes with I X SSC buffer for 30 minutes at 37°C, and twice more for 45 minutes in a 0.1 X S S C buffer at 37°C. Sections were washed on a shaking platform 2 X for 10 minutes in buffer 1 ( lOOmM Tr i s -HCl (pH 7.5), 150mM NaCl) at room temperature. Sections were covered for 30 minutes with blocking solution (buffer 1 containing 0.1% Triton X-100 and 2% normal sheep serum (Sigma)) at room temperature. The blocking solution was decanted and sections were incubated for 2 hrs in a humid chamber with buffer 1 containing 0.1% Triton X-100, 1% sheep serum, and a 1:500 (weight) dilution of sheep anti-DIG-alkaline phosphatase (Fab fragments; Roche) Sections were washed 2 X 10 minutes in buffer 1 on a shaking platform at room temperature. Sections were incubated for 10 minutes with buffer 2 ( lOOmM Tr i s -HCl (pH 9.5), lOOmM N a C l , 50 m M M g C l 2 ) , and were then covered with the color solution ( lOmL buffer 2 containing 200pl o f nitroblue tetrazolium (NBT) / 5-bromo-4-chloro-3-indolylphosphate (BCfP) stock solution (Boehringer Mannheim)) in a humid chamber for -12-24 hrs until desired color was achieved. The color reaction was stopped with the addition o f buffer 3 ( l O m M tr is -HCl (pH 8.1), I m M E D T A ) followed by a brief wash with distilled H2O. Sections were mounted with CytoSeal 280 (Richard-Allan) and overlaid by a coverslip. Niss l Stained sections: For sections prepared with a nissl-stain, 1 g cresyl violet acetate dissolved in 990 ml o f d d H 2 0 and the mixture was heated to 35°C and stirred for several hours at 35°C. 25 The solution was filtered, allowed to cool at room temperature, and the p H was adjusted to 3.5 with glacial acetic acid. Sections prepared as outlined in tissue preparation above, were then submerged in this cresyl violet solution and stained for 5-15 minutes. The sections were washed with 10 dips in water, followed by 10 dips in 70% ethanol. Sections were dehydrated through ethanol by immersing the slides in 2 changes of 95% ethanol for 2 min each. Northern Analysis Total R N A from spinal cord, tectum, and gut tissues of various chicks ranging in age from E3 to E l 1 was TRIzol extracted (GibcoBRL) . A total of 30u.g of R N A , as measured by the absorbance at 280 nm, was electrophoresed in a 1% denaturing formaldehyde agarose gel for 3 hours at 70Volts. Electrophoresed R N A was then transferred to Hybond-N nylon membrane (Amersham) overnight in 10X SSC transfer buffer. The membrane was baked at 80°C for 2 hrs and incubated in prehybridization buffer (50%) w/v deionized formamide, 5 X Denhardt's solution, 1%SDS, denatured salmon testes D N A ) at 42°C for 2-4 hrs. Membranes were then hybridized in the presence of anti-sense or sense R N A probes overnight at 42°C followed by a series of washes in S S C buffers o f varying concentrations. The membranes were developed using a standard protocol from the D I G Northern Starter K i t (Roche). R T - P C R studies SUPERSCRIPT™ II RNase FT Reverse Transcriptase was used to synthesize first strand c D N A for all R T - P C R studies (Invitrogen). Total R N A from spinal cord, tectum, and gut tissues of various chicks ranging in age from E3 to E l 7 was TRIzol extracted (GibcoBRL) . A total of 1 ug of R N A was incubated in the presence o f 50ng of random 26 primers and 1 pi of l O m M dNTP mix, to a total volume o f 12ul with ddEbO. The mixture was heated to 65°C for 5min followed by a quick chil l on ice. A total of 4pl of 5 X First-Strand Buffer, 2pl 0.1 M D T T , and l u l of RNase inhibitor were added to the mixture. Following incubation at 25°C for 10 minutes, the mixture was incubated at 42°C for 2 minutes and 1 pi of S U P E R S C R I P T II enzyme was added. The mixture was then incubated for 50 minutes at 42°C and the reaction was stopped by heating to 70°C for 15 min. A total of 2pl o f first-strand c D N A mixture was used in the P C R reaction, and the primers used were the same as used to make the R N A probe. Following the P C R reaction, the product was subjected to separation on a 1% agarose gel in the presence o f EtBr to confirm the size (figure 5B). Aff ini ty Purification of cSEMA5B Ant ibody Purification of GST-fusion protein: The p G E X plasmid containing c S E M A 5 B peptide corresponding to the region of antibody recognition and cloned downstream of G S T , was supplied by Dr. Wenyan Wang. E coli D H 5 a cells transformed with p G E X vector containing either the c S E M A 5 B - glutathione-S-transferase fusion protein (GST) or G S T protein alone were used to inoculate 25ml of L B supplemented with 50fig/fil o f ampicillin (amp). These 25ml cultures were grown overnight in the orbitial shaker at 37°C. The resulting cultures were used to inoculate 500ml of L B supplemented with 50/^g//il o f amp. This culture was grown for 1.5-2 hrs at 37°C until reaching an optical density(600) of 0.6-1.0. Protein expression was induced by the addition of I P T G ( O . l m M final) followed by incubation at 37°C for 5 hours. The resulting bacteria were collected by centrifugation at 8000 rpm at 4°C for 10 minutes and the resulting pellet was re-suspended in 9ml of M T - P B S (150mM 27 N a C l , 16 m M N a 2 H P 0 4 , 4 m M NaH 2 P04 (pH 7.3)). The suspension was sonicated six times at 30seconds each, with 30 seconds on ice between each sonication. Following sonication, 1ml of 10% triton-X-100 was added and the suspension was rocked at 4°C for 5 minutes. The resulting lysates were spun down at 8000 rpm for 10 minutes at 4°C and pellets were discarded. The supernatant was added to a 2ml slurry containing 160 mg (80mg/ml) of Glutathione beads (Sigma G4251) pre-incubated for 30 minutes in M T -P B S and washed 4 X with 15ml of M T - P B S . The supernatant and beads were incubated at room temperature for 10 minutes. The mixture was centrifuged at 1000 rpm for 1 minute and the supernatant was removed. The beads were washed 5 X in 15 ml M T - P B S with 1% triton-X-100 followed by 2 washes with M T - P B S (no Triton). The c S E M A 5 B - G S T fusion proteins (or G S T alone) were eluted by three washes with 2ml of elution buffer (50mM Tris HC1 p H 8.0, l O m M Glutathione). Each elution wash was kept separate. 15ul of the resulting elutions were run on a 12% polyacrylamide gel to determine i f elution was successful (figure 4). The OD2go for each elution fraction were measured to determine the peptide concentration ( 1 0 D - 0.5mg/ml). Affinity column preparation: The column was prepared as outlined by the AminoLink® K i t (Pierce Biotechnology). A 500 u.1 solution of c S E M A 5 B peptide at 4mg/ml was diluted 1:3 in coupling buffer and loaded onto an equilibrated AminoLink® column. The resulting slurry was mixed at 4°C for 6hrs in the presence o f AminoLink® Reductant solution to bind the c S E M A 5 B peptide to the column. The remaining active sites in the column were bound using a Tris-based quenching buffer in the presence of AminoLink® Reductant solution. Following a series of washes with 15mls of a I M N a C l buffer (pH 28 7.5), the column was stored at 4°C in P B S buffer (pH 7.3) containing 0.05% sodium azide. Antibody purification procedure: Following equilibration of the column with 10ml of P B S (pH 7.3) containing 0.05% sodium azide, 2.0ml of sera was loaded onto the column and allowed to enter the gel bed. Sample was allowed to circulate though the column overnight at 4°C to allow for maximum binding o f antibodies. The column was then washed with 30ml o f A m m o L m k wash solution. Antibody was eluted with Immunopure® IgG elution buffer and the elution fractions were monitored by absorbance at 280 nm. Fractions containing A b were separated on a 12% polyacrylamide SDS gel and stained with coomassie blue to confirm the correct sizes for heavy and light chain regions (figure 12). Ant ibody Staining Embryo's were fixed in 3.7% paraformaldehyde via pericardial infusion (ie P B S in the heart for ~2-3minutes, followed by 20min of 3.7% para at room temperature). The fixed embryos were then imbedded in O C T compound and sectioned into 1 Oum thick sections using a microtome at -20°C (Bright Instruments). Sections were mounted onto pre-cleaned Superfrost® D/Plus microsope slides (Fischer), and washed 3 X 5 minutes in PBS(A) containing 0.05% Triton X-100. Sections were then washed 2 X 1 0 minutes in PBS(B) containing 0.1% Triton X and 0.005% bovine serum albumin ( B S A ) . To block sections, slides were incubated for 1 hour PBS(B) containing 10% goat serum and 1%BSA. Sections were then washed 2 X 1 0 minutes in P B S ( A ) followed by 2 X 10 minutes in PBS(B) . Sections were incubated overnight at 4°C with primary antibody diluted appropriately in PBS(B) usually 1:1000 ( lOpl in 10ml). Sections were again 29 washed 2 X 10 minutes in P B S ( A ) followed by 2 X 10 minutes in PBS(B) . Section were then incubated for 1 hour at room temperature with a Cy3 conjugated goat anti rabbit fluorescent secondary antibody (Jackson ImmunoResearch Laboratories, West Grove, P A ) diluted at 1:500 in PBS(B) . Sections were washed 2 X 1 5 minutes in P B S ( A ) . Sections were then covered in antifade (Molecular Probes) containing glycerol and P B S , covered with a coverslip and sealed using nail polish. Antibody staining of dissociated cultures Spinal Cords were isolated from E8 chick and subjected to trituration through a small bore pipette in the presence of D M E M - F 1 2 . The resulting suspensions were plated on poly-L-lysine and laminin (lng/ml) coated coverslips in the presence of N G F (20ng/ml), and were allowed to grow for 12 hrs at 37°C in the presence of CO2. After 12hrs cultures were fixed in 3.7% paraformaldehyde and stained as per protocol. Primary mouse anti-NeuN antibody was used at a dilution of 1:100, while an t i - cSEMA5B rabbit antibody was used at a dilution of 1:1000. Pre-absorption o f antibody with fusion protein: A total of 100u.g of purified fusion protein was incubated with lOul o f affinity purified antibody, and lOOul of N e u N antibody in a final volume o f 10ml o f P B S . The next day, cells were stained as per protocol using the pre-absorbed affinity purified antibody and N e u N Antibody. Westerns Blots Approximately 100-200mg of various tissues including spinal cord, tectum, retina, and gut were isolated from chick at various stages in microcentrifuge tubes. These tissues were homogenized using a homogenizing pestle in the presence o f lOOuL of 30 RIP A buffer (150mM N a C l , 5 0 m M Tris p H 7.4, 5 m M E D T A , 5.0% NP-40, 1.0% sodium deoxycholate (DOC) and 0.1 % sodium dodecyl sulphate (SDS), aprotinin, leupeptin, phenylmethyl-sulfonyl floride (PMSF)) . The total volume was brought up to 500(4,1 (200mg/ml) with the addition of RIP A buffer. The Cells were incubated in RTPA buffer on ice for 20 minutes. Supernatants were collected and assayed for protein concentrations according to the manufacturer's instructions (BioRad). A total of 40ptg of protein was loaded and separated on a 12% SDS-poly acrylamide gel electrophoresis ( P A G E ) gels and then electro-transferred to Hybond E C L Nitrocellulose membranes (Amersham Pharmacia, Piscataway, NJ) . Membranes were blocked with 3% milk protein in Tris buffered saline (TBS)-Tween-20 for 1 hour at room temperature. Membranes were probed with of rabbit an t i - cSEMA5B antibody (1:1000) in TBS-Tween-20 overnight at 4°C. Blots were incubated with o f anti-rabbit horse radish peroxidase (1:2000; Jackson ImmunoResearch Laboratories, West Grove, P A ) in 1% milk in TBS-tween-20 for 1 hour at room temperature. Antibody was visualized by enhanced chemiluminescence ( E C L ) according to manufacturer's instructions (Amersham Pharmacia, Piscataway, NJ) . Cytoskeletal Extract ions: For cytoskeletal fractions, the tissue isolation was the same as above, however, homogenized tissue was incubated on ice with constant rotation for 20 min in 500/xl o f cytoskeletal buffer (50mM N a C l , l O m M Pipes p H 6.8, 3 m M M g C l 2 , 300 m M Sucrose, 0.5% triton-X 100, 1.2mM P M S F , 10/ig/ml aprotinin, leupeptin; Hinck et al., 1994) to extract Triton insoluble proteins. The resulting lysates were spun down at 14,000 rpm for 10 minutes at 4°C and the insoluble pellet and supernatant were separated. A total of 100/xl of SDS IP buffer (15mM Tris p H 7.5, 5 m M E D T A , 2 .5mM E G T A , 1% SDS) was 31 added to the pellets and pellets were boiled for 10-15 minutes at 100°C. The resulting suspension was brought up to 400/i.l with the addition of cytoskeletal buffer. Protein assays were performed for both the soluble and insoluble fractions according to the manufacturer's instructions (Biorad) to determine the amount of protein. A total of 25pg o f protein from both fractions was loaded and separated on a 12% P A G E gel as described above. 32 Results The structure of SEMA5B: A n analysis of c S E M A 5 B sequence shows that it is 1092 amino acids in length (figure 3). It is 77% identical and 87% similar to mouse Sema5B based on alignment using N T I Vector Suite 6.0 software. The structural domains of the protein include the semaphorin domain spanning approximately 500 amino acids, followed by seven thrombospondin repeats, a transmembrane region spanning 20 amino acids, and a short cytoplasmic region of 90 amino acids (figure 3A). Sequence analysis suggests that there may be a putative proprotein convertase cleavage sequence, ICKR/KR, within the fifth thrombospondin repeat (figure 3B). A potential cleavage product is supported from westerns using an an t i - cSEMA5B antibody which recognizes two distinct products (figure 1 I D , E) . Precedence for the cleavage o f semaphorins has been established for both the class 3 and 4 semaphorins (Adams et al., 1997; Elhabazi et al., 2001). Adams et al. 1997 demonstrated that furin-dependent proteolytic processing of Sema3a is required for an enhanced repulsive functionality. Similarly, Sema4D is cleaved by a metalloprotease-dependent process and exists as both a transmembrane and diffusible molecule in the immune system (Elhabazi et al., 2001). Construction of an R N A Probe: In mice, the cytoplasmic tails of the class 5 semaphorins consist of 80 (Sema5 A ) or 91 (Sema5B) amino acids. This region shows a lesser degree of similarity between 5 A and 5B, than do other regions o f the protein (Figure 4A) . Indeed, an alignment of Sema5A and Sema5B nucleotide sequences demonstrated that the cytoplasmic region shows the greatest degree o f divergence and thus provided an excellent sequence for the 33 Figure 3: Semaphorin 5B Structure and amino acid sequence A schematic o f c S E M A 5 B displaying the conserved regions between mice and chick. These include the signal sequence (blue), semaphorin domain (orange), thrombospondin repeats (green), and transmembrane domain (red) (A). Also present on the schematic is the representation of an affinity purified A b used for the immuno studies, and a non-radioactive R N A probe used in the in situ and Northern studies. A western blot displays the specificity o f the affinity purified A b , as both a monoclonal H A antibody and the c S E M A 5 B antibody specifically recognize a recombinant c S E M A 5 B protein, containing an H A tag, which is over-expressed in H E K 293 cells. In addition, the amino acid sequence o f c S E M A 5 B is displayed along with the conserved regions as indicated (B). 34 RNAPROBE SEMA5B -1092 AA IN LENGTH -N-TERMTNAL SIGNAL SEQUENCE -500 AA SEMA DOMAIN -SEVEN THROMBOSPONDIN REPEATS 0^ -TRANS MEMBRANE REGION ^ -100 AA CYTOPLAMSIC REGION r — i -EXPRESSED IN THE BRAIN AND SPINAL CORD anti-HA anti-S5b The Amino Acid Sequence for cSEMASB B (l)MWSRLKAISLSLPSLFLL(AJHLSASONVTEYSEAJEHOOCV^KEHPTIAF EDLKPWVSNFTYPGVHDFSOLALDANRNQLIVGARNYLFRLSLHNVSL I O A T A W G S D E D T R R S C Q S K G K T E E E C O N Y V R V L I V l Y G K K V F T C G T N A F S P V C S S R O V G N L S K H D R I N G V A R C P Y D P R H N S T A V 1 T S R G E L Y A A T V I D F S G R D P A 1 Y R SLGNVPPLRTAQYNSKWLNEPNFIAAYDIGLFTYFFFRENAVEHDCGKTVYSRVARV CKNDIGGRFLLEDTWTTFMKARLNCSRAGEIPFYYNELOSTFYLPEQDLIYGVFTTN VNSIAASAVCAFNLSAITOAFNGPFRYOENPRSAWLPTANPIPNFQCGTLSDDSPNEN LTERVLODAORLFLMNDVVOPVTVDPYVTODSIRFSKLVVDIVQGKDTLYHVMYIG TEYGTtLKALSTTNRSLRSCYLEEMOILPDGOREAIKSLOILHSDRSLFVGLNNGVLK IPLERCSMYRTEGECLGARDPYCGWDNKOKRCTTIEDSSNMSLWTONITECPVKNL TTNGRLGPWSP WQPCEHSDGDSTGSCMCRSRSCDSPRPRCGGRSCEGARIEVANC SRNG4 WTPWSSWALCSTSCGIGFQVRQRSCSNPAPRYGGRVCVGQSREERFCNENSPCPLPIFWSSWG PWNKCSVNCGGGIHSRQRTCENGNTCPGCAVEYKTCNPESCPEWRRNTPWTPWMPVNITQN GARQEQRFRYTCRAQLSDPHELQFGRKKTESRFCPNDGSAMCETDSLVDDLLKTGKJSASll iSGGWSFWGAWSSCSRDCELGFRIRKRTCTNPEPKNGGLPCVGSAMEYQDCNPHPCPVKG SWSCWTPWSQCSATCGGGHYQRTRTCTNPAPSSGEDICIGLHTEEALCNTHPCEGGWSEWSE WSLCNEEGIQYRSRYCEVHSPDSSQCVGNSJQYQDCLYNEIPYILPASS1DESTNCGGFSUH U A T G V S C F F G S S I X T F f V I Y V Y C O R C O R O S O E S T V I H P T T P N H L H Y K G N T T P K N E K Y T P M E F K T L N K N N L I P D D R T N F Y P L Q Q T N V Y T T T Y Y P S T L N K Y D Y R P E A S P GRTFTNS)(1092) Signal Sequence Antibody peptide-recognition sequence — Sema domain sequence A A A A A A Tsp type-l-like repeat sequences —AAA Transmembrane sequence AAA R N A probe sequence ( A A A ) Putative cleavage sequence AAA 35 Figure 4: Analysis and alignment of c S E M A 5 B R N A Probe A n alignment of the c S E M A 5 B R N A probe sequence along with c S E M A 5 B , Sema5B, and Sema5A using Vector N T I suite 6.0 (A) . Letters in yellow represent identical amino acids for all three sequences, while letters in light blue represents identical amino acids for c S E M A 5 B and Sema5B. A high divergence between Sema5A and Sema5B in mice over the probe region is indicated by a lack of blue or yellow lettering (A). The probe sequence was amplified from full-length c S E M A 5 B via P C R , and is 350bp in size when run on a 1% agarose gel at 90V (B). A double digest using NotI and Smal releases the probe sequence from the bluescript29 K S vector in the first lane, while a PvuII digest confirms the correct orientation of the probe in the second lane (C). Both the antisense and sense R N A probes are ~350bp in size when run on a 1% agarose gel (D). The amount o f probe is determined to be lOOng/u.1 when compared to a series o f standards of known DIG-labeled R N A s concentrations, using the same color development process employed in the in situ hybridization procedure (E). The R N A isolated from tissue runs as two bands as most of the R N A present is r R N A which runs at ~4.8kb and ~ 2kb in size, corresponding to the 28S and 18S bands (F). The antisense probe recognizes dot blots of total R N A on a nylon membrane following the northern blot process (G). The sense probes also cross react with the total R N A , but to a much lesser degree (G). 36 ChickSEMA5B probe V B Y V Y C Q R C Q R Q S Q E S T V I H P T T Chick SEMA5B PASSIDESTN CGGFSLIHLIATGVSCFFGSSLLTFV1YVYCQRCQRQSQ.ESTVIHPTT Mouse Sema5B PASSVEETTS CGGFNLIHLIVTGVSCFHJSGLLTLAgYgSCQHCQRQSQEST^HPAT Mouse SemaSA V § V f R ^ S V E E K R C G E F N M F H | F H H A v ^ H M s s B G C L L T L W Y T Y C Q R Y Q Q 2 S H B T V I H P V i Consensus PASSIDEST CGGF LIHLIATGVSCFILS LLTLVIYVYCQRCQRQSQESTVIHPTT ChickSEMA5Bprobe p„H L—HYKGNTTPKNEKYTPMEFKTLNKNNLIPDDRTNFYPLQQTNVYTTTYYPSTLNKYDYR Chick SEMA5B p u H L — HYKGNTTPKNEKYTPMEFKTLNKNNL IPDDRTNFYPLQQ7NVYTTTYYPSTLNKYDYR Mouse Sema5B PNHL—HYKGGGTPKNEKYTPMEFKTLNKNNLIPDDRANFYPLQQTNVYTTTYYPSPLNKPSjR Mouse SemaSA PAALNSSITNHINKLDKYDjVEAIKAFNKNNLILB^NK((NPHLTGKTY|NAYF(DLNNYDEY Consensus PNHL HYKGNTTPKNEKYTPMEFKTLNKNNL IPDDRTNFYPLQQTNVYTTTYYPSTLNKYDYR Chick SEMA5B probe p E A S P G R T F T N S - (97AA) ChickSEMA56 P E A S P G R T F T N S - (1092AA) Mouse Sema5B PEASPGQRCFPNS (1093AA) Mouse Sema5A (1077AA) Consensus P E A S P G R T F T N S R N A Standard Antisense -350 bp, Sense Antisense G 37 production of a specific R N A probe to distinguish between the two family members (Adams et al., 1996). Using this information, a chick probe corresponding to the final 300 bp, or the intracellular region of c S E M A 5 B , was used for the in situ hybridization and Northern experiments (figure 4A) . Further sequence analysis demonstrated that this region is not only divergent between the class 5 semaphorins, but is distinct from other known semaphorin sequences. Following amplification of the desired probe sequence using P C R , the resulting product was sub-cloned into a Bluescript29-KS expression vector to produce both sense and anti-sense probes (figure 4B-D) . Initial analysis of the probes was done by spotting total R N A , isolated from 293 cells expressing recombinant c S E M A 5 B , onto a nylon membrane, followed by a Northern blot procedure (figure 4F,G). The presence of strongly hybridized spots on the blot probed with the anti-sense R N A , demonstrated that these probes provided a suitable tool for use in the Northern studies (Figure 4F, G) . In addition, a quantification protocol using dot blots, demonstrated that sufficient quantities o f the DIG-labeled R N A probes were available for in situ hybridization studies (Figure 4E). Northern Studies: In order to determine whether c S E M A 5 B exists as a single transcript in chick, a Northern analysis was performed on spinal cord tissue obtained from an E9 chick. A single band at ~5kb suggests that c S E M A 5 B does exist as a single transcript (Figure 5A). Northern analysis on R N A isolated from 293 cells expressing a full-length recombinant form of c S E M A 5 B also displayed a single band of similar size, confirming the findings from chick tissue (Figure 5A). This single transcript product is consistent with Northern 38 Figure 5: Northern and R T - P C R Results Chick S E M A 5 B is expressed as a single transcript of ~5kb as confirmed from total R N A isolated from cells expressing recombinant c S E M A 5 B and from E9 SC tissue (A). R T - P C R using primers specific to c S E M A 5 B confirms the expression of this semaphorin in spinal cord tissue isolated from chick ranging in age from E3 until E l 7 (B). HEK-293 cells expressing recombinant c S E M A 5 B were used as a positive control. 39 NORTHERN RESULT 293 Cells-CSEMA5B E9 SC 3kb 4.8kb B RT-PCR RESULT sa a. Ok re SC TISSUE 293-E3 E10 E17cSEMA5B 350b^ 40 analyses of other class 5 semaphorins in Drosophila and mice (Adams et al., 1996; Khare et al., 2000). In mice, Northern blot analysis of m R N A isolated from embryos revealed two Sema5A transcripts of 5.5 and 9.4 kb, while Sema5B is expressed as a single transcript o f 5.9kb (Adams et al., 1996). In Drosophila, the class 5 semaphorin DSema5C is expressed as a single transcript ~5kb in size (Khare et al., 2000). A s only a probe to the cytoplasmic region was used in these studies on c S E M A 5 B , it does not rule out the possibility for the existence o f splice variants that lack this intracellular region. R T - P C R studies: A s the Northern studies were technically difficult, and the sensitivity was low, R T - P C R was employed. Using primers specific to c S E M A 5 B , R T - P C R was performed on a variety o f different chick tissues from a wide range of developmental stages. R T -P C R analysis demonstrated expression o f c S E M A 5 B in the neural tube or spinal cord as early as E3 and as late as E l 7 (figure 5B). Expression o f c S E M A 5 B at E3 was not confirmed using Northern blot analysis neither was any message observed in the in situ hybridization studies from chicks at this age. This is likely due to the greater sensitivity of the R T - P C R technique, and that less material is required. The spinal cord expression at E l 7 was also the latest stage examined for all o f the studies presented, and corresponds with a stage when the majority of spinal cord development is complete (Eide and Glover, 1997). Other tissues examined included the tectum, retina, and regions o f the gut (Data not shown). A l l o f these areas were shown to express c S E M A 5 B , and confirmed the findings from both the Northern and in situ hybridization studies. 41 In situ Hybridization Studies: In order to gather more information as to the distribution of c S E M A 5 B message, in situ hybridization was utilized. Based on the results presented above, in situ hybridization studies were performed on chicks ranging in age from E2 until E l l . From these in situ hybridization studies, c S E M A 5 B expression was first clearly observed in a variety of tissues at E5 . This expression was broad and included the gut, epidermis, spinal cord, brain, olfactory and ventricular epithelium, and the retina. While many tissues express c S E M A 5 B , this thesis focuses on its expression in the developing nervous system. c S E M A 5 B expression in the Spinal Cord: Strong c S E M A 5 B expression is observed in the spinal cord at stage 26 (E5; figure 6A). A t this age, the expression in the spinal cord is broad, being present in both dorsal and ventral regions including the mantle layer (figure 6A). Though this expression is broad, it is restricted to the grey matter with no observable message in the surrounding white matter (figure 6A). While the dorsal and ventral horns are not distinguished yet, a number of sensory axons have already reached the dorsal funiculus, an area associated with expression o f c S E M A 5 B (figure 6 A ; Airman and Bayer 1984). There is also expression along the entire dorsal-ventral axis of the medial ependymal (EP; ventricular epithelium) layer, an area associated with a large population of mitotic cells (figure 6A; Bellairs and Osmond, 1998). B y stage 28 (E6) however, the expression of c S E M A 5 B becomes more restricted in appearance, with the highest levels observed in large cells of the ventral horn associated with presumptive lateral and medial motor columns (figure 6B, D) . Again, there is expression in the ependymal layer (figure 6B). 42 Figure 6: In situ hybridization studies in the spinal cord In situ hybridization studies using a probe for c S E M A 5 B clearly confirm its expression in the spinal cord including the mantle layer, D R G s , and the dorsal surface epithelium (SE) at stage 26 (E5; A ) . A t E6, the expression of c S E M A 5 B increases ventrally in presumptive lateral motor columns ( P L M C ) and is present along the ventricular ependymal (EP; epithelial) layer as indicated by the arrow heads (>; B) . A t E9 c S E M A 5 B message is highest in the ventral horn (VH) , with lower levels in dorsal regions including the dorsal horn ( D H ; C) . In contrast, the white matter ( W M ) is negative (C). A t higher magnification, the ventral expression in the S C at E6 is clearly cellular in the presumptive lateral motor column (D). The expression in D R G s at E6 is uniform (E, F) . Scale bar in ( A , B , C ) is 100pm; scale bar in (D,E,F) is 20pm. 43 44 The intensity of c S E M A 5 B expression in the ventral spinal cord continues to increase through E9 (figure 6C). The majority of c S E M A 5 B expression is associated with cells located in the lateral motor column region (figure 6C). While much of the spinal cord patterning has taken place at this stage, a number of primary afferent axons are still invading the dorsal grey matter with many reaching the lateral motor column at E10 (Shiga et al., 2000). Observations from E10 spinal cord sections indicate that this expression of c S E M A 5 B is maintained during these latter axon pathfinding events (data not shown). c S E M A 5 B expression in D R G : Expression of c S E M A 5 B was also apparent within dorsal root ganglia. This expression was evident as early as E5 (Figure 6E, F). Unlike the dynamic expression in the spinal cord, the expression in the D R G s remained uniform and constant throughout development. A s there are a variety of different sensory neurons and glial cells present at all o f the ages examined, characterization of the cell types expressing c S E M A 5 B awaits further analysis. Different populations of cells are localized to different areas within the D R G , and while high magnification on D R G s demonstrates that not all o f the cells express c S E M A 5 B , the expression is uniform and not biased for a particular area associated with a specific cell type (figure 6F; Eide and Glover, 1997). This coincident expression of a semaphorin in both the spinal cord and D R G neurons is similar to the expression of Sema3A in embryonic chick (Shepherd et al., 1996). Expression of c S E M A 5 B in retina: In addition to the spinal cord and D R G s , cells associated with the retina also express c S E M A 5 B . Expression can be clearly observed at stage 26 (E5), in an inner 45 layer of the retina, presumably along the retinal ganglion cell (RGC) layer ( G C L ; Figure 7). The expression is confined to the cell body layer and is not associated with the nerve fibre layer (figure 7C-F). Unlike the uneven distribution of c S E M A 5 B message within the spinal cord, the expression associated with the retina is uniform. Expression is absent in the optic nerve head (ONH) and along the optic nerve ( O N ; Figure 7A). This region shows exclusive expression of Sema5A in mice (Oster et al., 2003). The cells expressing c S E M A 5 B are likely ganglion cells, as the other populations of cells associated with the R G C layer, including some amacrine and Muller glial cells, do not have their nuclei exclusively in this region (Mey and Thanos, 2000). Interestingly retinal axons l ikely receive inunction from the radial glial cells of Mul le r which have been shown to express different guidance cues, for example the netrin receptor D C C (Gad et al., 2000). Retinal ganglion cells encounter other guidance cues, such as laminin, through interactions with the glial endfeet and basal lamina present in the nervefibre layer at the vitreo-retinal border (figure 15B; Stier and Schlosshauer, 1999). Interactions between radial glia and the retinal axon growth cones are assumed to occur within the optic stalk, along the optic tract, and within the optic tectum (Silver, 1984; Thanos and Mey , 2001). A t later stages including E9, this c S E M A 5 B expression becomes more intense and continues to be associated with the inner most cell layer of the retina (Figure 7E). A t these older ages, the distribution of c S E M A 5 B message remains uniform along both the temporal/nasal axis and the rostral/caudal axis. A t these stages, it is still not clear as to what cell-types are expressing c S E M A 5 B , but the expression overlaps the normal distribution of R G C and amacrine cells (Mey and Thanos, 2000). 46 Figure 7: In situ hybridization studies in the retina c S E M A 5 B expression within the retina is relatively uniform in the nasal temporal plane. This expression at stage 26 (E5) is associated with the R G C layer ( G C L ) as indicated from the arrows (A, B , C, D). Notably, c S E M A 5 B expression is absent from the optic fibre layer (OFL) and the optic nerve head (ONH) as indicated with an asterisk (A, C, D). A t a later stage, c S E M A 5 B expression is still associated with the G C L in the E9 retina (E). The sense probes do not have message associated with this layer (F). (G) is an example o f a Nissl-stained retina used in histological analysis of the tissue. The associated retinal layers identified include the optic fibre layer (OFL) , ganglion cell layer ( G C L ) , inner nuclear layer (IN), external nuclear layer (EN), and retinal pigmented epithelium (RPE). Scale bar for (A, B) is 100pm; scale bar for (C-G) is 10pm. 47 c S E M A 5 B expression in the Tectum: A t early stages o f development, the tectum is the most prominent anatomical and functional structure of the avian brain. Results from the in situ hybridization studies demonstrate that c S E M A 5 B is expressed in the tectum at E5 (figure 8), just prior to the arrival of the first retinal fibres at E6, when maximal cell proliferation is occurring in the tectum (Halfter, 1987). Similar to the expression of c S E M A 5 B in the spinal cord, the distribution of message in the tectum is asymmetrical (Figure 8). The highest expression appears to be associated with anterior regions, getting less intense in posterior regions (Figure 8A). Anterior/ventral tectum represents the entry for incoming retinal afferents (figure 16C). In middle regions of the tectum, this expression occurs along the inner most ependymal (EP; neuroepithelium) lining of the ventricle, a layer highly associated with cell proliferation at this stage (figure 8A; M e y and Thanos, 2000). There is also additional expression in a central layer at E6 and E7, which later forms the stratum griseum centrale (SGC; figure 8B, C) . This layer is comprised o f large multipolar neurons, often referred to as ganglion cells, which are generated early in comparison to other neurons in the tectum (figure 8B, C; M e y and Thanos, 2000). While this layer represents the principal tectal efferent neurons, it also acts as a target for incoming retinal afferents growing along the superficial, stratum opticum layer (Mey and Thanos, 2000). Tectal expression was observed at later stages but is not included in this report. The expression in the tectum overlaps a period of maximum cell proliferation as well as innervation from retinal afferents. 49 Figure 8: In situ hybridization studies in the tectum Expression of c S E M A 5 B is present at stage 26 (E5) along the ependymal (EP; epithelial) layer as indicated by arrowheads (>; A - C ) . The anterior expression (<) of c S E M A 5 B is greater than the posterior expression (*) along the E P (A-C) . In sections from the dorsal tectum, there is additional expression of c S E M A 5 B in the presumptive stratum griseum centrale (SGC) layer as indicated (#; B , C) . There is no expression along the stratum opticum (SO) the most superficial layer associated with optical fibres. Scale bar in A is 100pm; scale bar in (B,C) is 10pm. 50 51 Epi the l ia l expression: In all o f the regions examined there is an obvious expression of c S E M A 5 B in the associated neuroepithelia. This expression is consistent with observations from mice and rat, where Semaphorin5B is ubiquitously expressed in most neuroepithelia in the brain and other structures (Adams et al., 1996; Skaliora et al., 1998). Observations from embryonic chick brain show that c S E M A 5 B is expressed in the ventricular ependymal (epithelium) layer at stage 26 (E5; Figure 9). The expression, as seen in the ependymal layer of the spinal cord, is relatively uniform along the anterior-posterior axis. Sections from rostral and caudal regions of the brain again display a similar pattern of expression within the ventricular epithelium (Figure 9A, B) . This expression is consistent with brain sections from rat where S E M A 5 B is found uniformly distributed in most of the epithelia within the brain, including the ventricular zones (Skaliora et al., 1998). High expression of c S E M A 5 B is also observed in the olfactory epithelium in transverse sections from E5 and E6 chick (Figure 10). During embryogenesis many studies have demonstrated that the olfactory epithelium is the site of origin o f several types of cells which migrate along the olfactory nerve (ON) towards the brain (Dryer and Graziadei, 1994; Drapkin and Silverman, 1999; Wray, 2001). The expression of c S E M A 5 B decreases in the surrounding lateral olfactory tissue layers. These layers serve as target regions for a number of cells produced along the olfactory epithelium. Thus, this assortment of migrating cell populations and various neuronal pathfinding events associated with the olfactory epithelium during these stages correlates with expression of C S E M A 5 B . 52 Figure 9: In situ hybridization of ventricular epithelium c S E M A 5 B expression in the ventricles at stage 26 (E5) is highest in the neuroepithelium as indicated by the arrowheads (A-C) . This expression is uniform along anterior-posterior plane (A). A similar, uniform, epithelial-associated expression is seen at lower levels within the brain (B). Scale Bar in (A,B) is 100pm; scale bar in C is 20pm. 53 Figure 10: In situ hybridization studies in nasal regions A t stage 26 (E5), expression o f c S E M A 5 B in the nasal epithelium is relatively uniform (A, B) . A t higher magnification, the c S E M A 5 B distribution is greatest in the inner epithelial layer, decreasing in the outer layers (C, D). Scale bar in (A,B) is 100u,m; scale bar in (C,D) is 10u.m. 55 56 In addition to the tissue regions mentions above, there was expression seen in the surface epithelium located dorsally to the spinal cord. This observation was consistent in both the in situ hybridization studies as well as from antibody studies (figures 6 A & 12 A ) . Again, similar staining patterns for semaphorins have been observed from studies on secreted sema3A (Shepherd et al., 1996). There was also distribution of c S E M A 5 B expression in the epithelium associated with the gut (data not shown). A summary of the expression of c S E M A 5 B from all tissues examined clearly demonstrates the association of this expression with the neuroepithelia in these regions (Table I). Immunohistological studies : Using an affinity purified antibody (see materials and methods) raised against the N-terminal region o f c S E M A 5 B , the distribution o f c S E M A 5 B protein was analyzed (figure 3). Based on the expression of semaphorin from the in situ hybridization studies, chicks were examined from E5 through until E l 1 to determine whether the protein distribution mimicked that of R N A . Initial studies performed using serum from non-affinity purified antibody revealed expression in the spinal cord similar to the R N A distribution from in situ hybridization studies, and seemed to support both the findings from the in situ hybridization studies, as well as the use o f this antibody in further studies (data not shown). Westerns: In order to confirm the specificity of the antibody, a series of western blots from a variety of different tissues were produced and analyzed. Initial blots suggested that the non-affinity purified antibody recognized a number of different antigens on the blot (data not shown). Based on these westerns the antibody underwent an affinity purification 57 Table I: Summary of c S E M A 5 B expression in all regions A summary o f the noted expression of c S E M A 5 B fro the in situ hybridization studies highlights the broad expression of this semaphorin in the developing neuroepithelium as well as within the C N S including the spinal cord and tectum. Abbreviations: N / C , not completed; n/a, not applicable. 58 6 i o ll a CO ii 1 bD 1 > -o 5 "§) co 2 § TJ g O 4> . J > g • 1—t "S g> o G iS ii 3S> CJ o 5 o < Z V pi bD a _o CJ .o 'co co CJ .3 CO co s < £ < & Z Z ~a "a z z 11 co h JS i eggs | | 2 o | ^ is 1» ' a , • a o CO CO CO CO u ^ >^ Vi Vi Oi a a a CO co . > i c j 03 03 — — PL, OH CU pq If a 5 .2 .2 8^ < < > ^ 03 111 •s ft CJ o U O z z 3^ .2 •-3 co a> co 1 Q j l |W B co co co S3 ii ii ii O > » > » > » C (U cu 4> CJ a CJ o o o a c c co co U U CO CO O O ^ ^ ^ ^ CO CO n *o r> W W W 03 .9 t5 t5 ON w in W W o 0 0 —i Pi o 0O ^ VO t^ - ON w w w w 1 cu >