EphrinB2a in the zebrafish retinotectal system

Members of the Eph-B family of receptors tyrosine kinase and their transmembrane ligands have been implicated in dorsoventral patterning of the vertebrate retinotectal projection. In the zebrafish retinotectal system, however, ephrinB2a is expressed strongly in the posterior tectum, in tectal neurons that form physical contacts with retinal ganglion cell (RGC) axons. In the gnarled mutant, where tectal neurons form ectopically in the pretectum, RGC axons stall before entering the tectum, or else are misrouted or branch aberrantly in the tectal neuropil. Ectopic expression of ephrinB2a in the anterior midbrain of wild-type embryos, with the aid of baculovirus, also inhibits RGC axon entry into the tectum. In vitro, zebrafish RGC axons are repelled by stripes of purified ephrinB2a. It is proposed that ephrinB2a may signal a subpopulation of RGC axons that they have reached their target neurons in the tectum.     

Requirement for the zebrafish mid-hindbrain boundary in midbrain polarisation, mapping and confinement of the retinotectal projection.

The organizer at the midbrain-hindbrain boundary (MHB organizer) has been proposed to induce and polarize the midbrain during development. We investigate the requirement for the MHB organizer in acerebellar mutants, which lack a MHB and cerebellum, but retain a tectum, and are mutant for fgf8, a candidate inducer and polarizer. We examine the retinotectal projection in the mutants to assay polarity in the tectum. In mutant tecta, retinal ganglion cell (RGC) axons form overlapping termination fields, especially in the ventral tectum, and along both the anterior-posterior and dorsal-ventral axis of the tectum, consistent with a MHB requirement in generating midbrain polarity. However, polarity is not completely lost in the mutant tecta, in spite of the absence of the MHB. Moreover, graded expression of the ephrin family ligand Ephrin-A5b is eliminated, whereas Ephrin-A2 and Ephrin-A5a expression is leveled in acerebellar mutant tecta, showing that ephrins are differentially affected by the absence of the MHB. Some RGC axons overshoot beyond the mutant tectum, suggesting that the MHB also serves a barrier function for axonal growth. By transplanting whole eye primordia, we show that mapping defects and overshooting largely, but not exclusively, depend on tectal, but not retinal genotype, and thus demonstrate an independent function for Fgf8 in retinal development. The MHB organizer, possibly via Fgf8 itself, is thus required for midbrain polarisation and for restricting axonal growth, but other cell populations may also influence midbrain polarity.     

Axon sorting in the optic tract requires HSPG synthesis by ext2 (dackel) and extl3 (boxer)

Retinal ganglion cell (RGC) axons are topographically ordered in the optic tract according to their retinal origin. In zebrafish dackel (dak) and boxer (box) mutants, some dorsal RGC axons missort in the optic tract but innervate the tectum topographically. Molecular cloning reveals that dak and box encode ext2 and extl3, glycosyltransferases implicated in heparan sulfate (HS) biosynthesis. Both genes are required for HS synthesis, as shown by biochemical and immunohistochemical analysis, and are expressed maternally and then ubiquitously, likely playing permissive roles. Missorting in box can be rescued by overexpression of extl3. dak;box double mutants show synthetic pathfinding phenotypes that phenocopy robo2 mutants, suggesting that Robo2 function requires HS in vivo; however, tract sorting does not require Robo function, since it is normal in robo2 null mutants. This genetic evidence that heparan sulfate proteoglycan function is required for optic tract sorting provides clues to begin understanding the underlying molecular mechanisms.     

A GFP-based genetic screen reveals mutations that disrupt the architecture of the zebrafish retinotectal projection

The retinotectal projection is a premier model system for the investigation of molecular mechanisms that underlie axon pathfinding and map formation. Other important features, such as the laminar targeting of retinal axons, the control of axon fasciculation and the intrinsic organization of the tectal neuropil, have been less accessible to investigation. In order to visualize these processes in vivo, we generated a transgenic zebrafish line expressing membrane-targeted GFP under control of the brn3c promoter/enhancer. The GFP reporter labels a distinct subset of retinal ganglion cells (RGCs), which project mainly into one of the four retinorecipient layers of the tectum and into a small subset of the extratectal arborization fields. In this transgenic line, we carried out an ENU-mutagenesis screen by scoring live zebrafish larvae for anatomical phenotypes. Thirteen recessive mutations in 12 genes were discovered. In one mutant, ddl, the majority of RGCs fail to differentiate. Three of the mutations, vrt, late and tard, delay the orderly ingrowth of retinal axons into the tectum. Two alleles of drg disrupt the layer-specific targeting of retinal axons. Three genes, fuzz, beyo and brek, are required for confinement of the tectal neuropil. Fasciculation within the optic tract and adhesion within the tectal neuropil are regulated by vrt, coma, bluk, clew and blin. The mutated genes are predicted to encode molecules essential for building the intricate neural architecture of the visual system.     

Rotation of the tectal primordium reveals plasticity of target recognition in retinotectal projection

Retinotectal projection is precisely organized in a retinotopic manner. In normal projection, temporal retinal axons project to the rostral part of the tectum, and nasal axons to the caudal part of the tectum. The two-dimensional relationship between the retina and the tectum offers a useful experimental system for analysis of neuronal target recognition. We carried out rotation of the tectal primordium in birds at an early stage of development, around the 10-somite stage, to achieve a better understanding of the characteristics of target recognition, especially the rostrocaudal specificity of the tectum. Our results showed that temporal retinal axons projected to the rostral part of the rotated tectum, which was originally caudal, and that nasal axons projected to the caudal part of the rotated tectum, which was originally rostral. Therefore, the tectum that had been rotated at the 10-somite stage received normal topographic projection from the retinal ganglion cells. Rostrocaudal specificity of the tectum for target recognition is not determined by the 10-somite stage and is acquired through interactions between the tectal primordium and its surrounding structures.     

Genetic dissection of the formation of the forebrain in Medaka, Oryzias latipes

The forebrain, consisting of the telencephalon and diencephalon, is essential for processing sensory information. To genetically dissect formation of the forebrain in vertebrates, we carried out a systematic screen for mutations affecting morphogenesis of the forebrain in Medaka. Thirty-three mutations defining 25 genes affecting the morphological development of the forebrain were grouped into two classes. Class 1 mutants commonly showing a decrease in forebrain size, were further divided into subclasses 1A to 1D. Class 1A mutation (1 gene) caused an early defect evidenced by the lack of bf1 expression, Class 1B mutations (6 genes) patterning defects revealed by the aberrant expression of regional marker genes, Class 1C mutation (1 gene) a defect in a later stage, and Class 1D (3 genes) a midline defect analogous to the zebrafish one-eyed pinhead mutation. Class 2 mutations caused morphological abnormalities in the forebrain without considerably affecting its size, Class 2A mutations (6 genes) caused abnormalities in the development of the ventricle, Class 2B mutations (2 genes) severely affected the anterior commissure, and Class 2C (6 genes) mutations resulted in a unique forebrain morphology. Many of these mutants showed the compromised sonic hedgehog expression in the zona-limitans-intrathalamica (zli), arguing for the importance of this structure as a secondary signaling center. These mutants should provide important clues to the elucidation of the molecular mechanisms underlying forebrain development, and shed new light on phylogenically conserved and divergent functions in the developmental process.     

Dynamic responses of Xenopus retinal ganglion cell axon growth cones to netrin-1 as they innervate their in vivo target

Netrin-1 influences retinal ganglion cell (RGC) axon pathfinding and also participates in the branching and synaptic differentiation of mature RGC axons at their target. To investigate whether netrin also serves as an early target recognition signal in the brain, we examined the dynamic behavior of Xenopus RGC axons soon after they innervate the optic tectum. Time-lapse confocal microscopy imaging of RGC axons expressing enhanced yellow fluorescent protein demonstrated that netrin-1 is involved in early axon branching, as recombinant netrin-1 halted further advancement of growth cones into the tectum and induced back branching. RGC growth cones exhibited differential responses to netrin-1 that depended on the degree of differentiation of the axon and the developmental stage of the tadpole. Netrin-1 decreased the total number of branches on newly arrived RGC growth cones at the target, but increased the dynamic branching of more mature arbors at the later developmental stage. To further explore the response of axonal growth cones to netrin, Xenopus RGC axons were followed in culture by time-lapse imaging. Exposure to netrin-1 rapidly increased the forward advancement of the axon and decreased the size and expanse of the growth cone, while also inducing back branching. Taken together, the differential in vivo and in vitro responses to netrin-1 suggest that netrin alone is not sufficient to induce the cessation of growth cone advancement in the absence of a target but can independently modulate axon branching. Collectively, our findings reveal a novel role for netrin on RGC axon branch initiation as growth cones innervate their target.     

Topographic targeting and pathfinding errors of retinal axons following overexpression of ephrinA ligands on retinal ganglion cell axons

In the retinotectal projection, the Eph receptor tyrosine kinase ligands ephrinA2 and ephrinA5 are differentially expressed not only in the tectum, but also in a high-nasal-to-low-temporal pattern in the retina. Recently, we have shown that retrovirally driven overexpression of ephrinA2 on retinal axons leads to topographic targeting errors of temporal axons in that they overshoot their normal termination zones in the rostral tectum and project onto the mid- and caudal tectum. The behavior of nasal axons, however, was only marginally affected. Here, we show that overexpression of ephrinA5 affects the topographic targeting behavior of both temporal and nasal axons. These data reinforce the idea that differential ligand expression on retinal axons contributes to topographic targeting in the retinotectal projection. Additionally, we found that ectopic expression of ephrinA2 and ephrinA5 frequently leads to pathfinding errors at the chiasm, resulting in an increased stable ipsilateral projection.     

Perturbations of MicroRNA Function in Mouse Dicer Mutants Produce Retinal Defects and Lead to Aberrant Axon Pathfinding at the Optic Chiasm

Background

During development axons encounter a variety of choice points where they have to make appropriate pathfinding decisions. The optic chiasm is a major decision point for retinal ganglion cell (RGC) axons en route to their target in order to ensure the correct wiring of the visual system. MicroRNAs (miRNAs) belong to the class of small non-coding RNA molecules and have been identified as important regulators of a variety of processes during embryonic development. However, their involvement in axon guidance decisions is less clear.

Methodology/Principal Findings

We report here that the early loss of Dicer, an essential protein for the maturation of miRNAs, in all cells of the forming retina and optic chiasm leads to severe phenotypes of RGC axon pathfinding at the midline. Using a conditional deletion approach in mice, we find in homozygous Dicer mutants a marked increase of ipsilateral projections, RGC axons extending outside the optic chiasm, the formation of a secondary optic tract and a substantial number of RGC axons projecting aberrantly into the contralateral eye. In addition, the mutant mice display a microphthalmia phenotype.

Conclusions

Our work demonstrates an important role of Dicer controlling the extension of RGC axons to the brain proper. It indicates that miRNAs are essential regulatory elements for mechanisms that ensure correct axon guidance decisions at the midline and thus have a central function in the establishment of circuitry during the development of the nervous system.     

Trying to understand axonal regeneration in the CNS of fish.

In contrast to the situation in mammals and birds, neurons in the central nervous system (CNS) of fish--such as the retinal ganglion cells--are capable of regenerating their axons and restoring vision. Special properties of the glial cells and the neurons of the fish visual pathway appear to contribute to the success of axonal regeneration. The fish oligodendrocytes lack the axon growth inhibiting molecules that interfere with axonal extension in mammals. Instead, fish optic nerve oligodendrocytes support--at least in vitro--axonal elongation of fish as well as that of rat retinal axons. Moreover, the fish retinal ganglion cells re-express upon injury a set of growth-associated cell surface molecules and equip the regenerating axons throughout their path and up into their target, the tectum opticum with these molecules. This may indicate that the injured fish ganglion cells reactivate the cellular machinery necessary for axonal regrowth and pathfinding. Furthermore, the target itself provides positional marker molecules even in adult fish. These marker molecules are required to guide the regenerating axons back to their retinotopic home territory within the tectum.     

Role of cell adhesion molecule DM-GRASP in growth and orientation of retinal ganglion cell axons

The cell adhesion molecule (CAM) DM-GRASP was investigated with respect to a role for axonal growth and navigation in the developing visual system. Expression analysis reveals that DM-GRASP's presence is highly spatiotemporally regulated in the chick embryo retina. It is restricted to the optic fiber layer (OFL) and shows an expression maximum in a phase when the highest number of retinal ganglion cell (RGC) axons extend. In the developing retina, axons grow between the DM-GRASP-displaying OFL and the Laminin-rich basal lamina. We show that DM-GRASP enhances RGC axon extension and growth cone size on Laminin substrate in vitro. Preference assays reveal that DM-GRASP-containing lanes guide RGC axons, partially depending on NgCAM in the axonal membrane. Inhibition of DM-GRASP in organ-cultured eyes perturbs orientation of RGC axons at the optic fissure. Instead of leaving the retina, RGC axons cross the optic fissure and grow onto the opposite side of the retina. RGC axon extension per se and navigation from the peripheral retina towards the optic fissure, however, is not affected. Our results demonstrate a role of DM-GRASP for axonal pathfinding in an early phase of the formation of the higher vertebrate central nervous system.     

Tissue-specific neuro-glia interactions determine neurite differentiation in ganglion cells

Guided formation and extension of axons versus dendrites is considered crucial for structuring the nervous system. In the chick visual system, retinal ganglion cells (RGCs) extend their axons into the tectum opticum, but not into glial somata containing retina layers. We addressed the question whether the different glia of retina and tectum opticum differentially affect axon growth. Glial cells were purified from retina and tectum opticum by complement-mediated cytolysis of non-glial cells. RGCs were purified by enzymatic delayering from flat mounted retina. RGCs were seeded onto retinal versus tectal glia monolayers. Subsequent neuritic differentiation was analysed by immunofluorescence microscopy and scanning electron microscopy. Qualitative and quantitative evaluation revealed that retinal glia somata inhibited axons. Time-lapse video recording indicated that axonal inhibition was based on the collapse of lamellipodia- and filopodia-rich growth cones of axons. In contrast to retinal glia, tectal glia supported axonal extension. Notably, retinal glia were not inhibitory for neurons in general, because in control experiments axon extension of dorsal root ganglia was not hampered. Therefore, the axon inhibition by retinal glia was neuron type-specific. In summary, the data demonstrate that homotopic (retinal) glia somata inhibit axonal outgrowth of RGCs, whereas heterotopic (tectal) glia of the synaptic target area support RGC axon extension. The data underscore the pivotal role of glia in structuring the developing nervous system.     

Mutations affecting the formation of posterior lateral line system in Medaka, Oryzias latipes

We performed a systematic screen for mutations affecting the trajectory of axons visualized by immunohistochemical staining of Medaka embryos with anti-acetylated tubulin antibody. Among the mutations identified, yanagi (yan) and kazura (kaz) mutations caused specific defects in projection of the posterior lateral line (PLL) nerve. In yan and kaz mutant embryos, the PLL nerve main bundle was misrouted ventrally and dorsally or anteriorly. Medaka semaphorin3A, sdf1, and cxcr4 cDNA fragments were cloned to allow analysis of these mutants. There were no changes in semaphorin3A or sdf1 expression in mutant embryos, suggesting that the tissues expressing semaphorin3A or sdf1 that are involved in PLL nerve guidance are present in these mutant embryos. Double staining revealed that the mislocated PLL primordium and growth cone of the ectopically projected PLL nerve were always colocalized in both yan and kaz mutant embryos, suggesting that migration of PLL primordia and PLL nerve growth cones are not uncoupled in these mutants. Although homozygous yan larvae showed incomplete migration of the PLL primordium along the anteroposterior axis, ventral proneuromast migration was complete, suggesting that ventral migration of the proneuromast does not require the signaling affected in yan mutants. In addition to the PLL system, the distribution of primordial germ cells (PGCs) was also affected in both yan and kaz mutant embryos, indicating that yan and kaz genes are required for the migration of both PLL primordia and PGCs. Genetic linkage analysis indicated that kaz is linked to cxcr4, but yan is not linked to sdf1 or cxcr4. These mutations will provide genetic clues to investigate the molecular mechanism underlying formation of the PLL system.     

Mutations affecting thymus organogenesis in Medaka, Oryzias latipes

The thymus is an organ for T lymphocyte maturation and is indispensable for the establishment of a highly developed immune system in vertebrates. In order to genetically dissect thymus organogenesis, we carried out a large-scale mutagenesis screening for Medaka mutations affecting recombination activating gene 1 (rag1) expression in the developing thymus. We identified 24 mutations, defining at least 13 genes, which led to a marked reduction of rag1 expression in the thymus. As thymus development depends on pharyngeal arches, we classified those mutations into three classes according to the defects in the pharyngeal arches. Class 1 mutants had no or slight morphological abnormalities in the pharyngeal arches, implying that the mutations may include defects in such thymus-specific events as lymphocyte development and thymic epithelial cell maturation. Class 2 mutants had abnormally shaped pharyngeal arches. Class 3 mutants showed severely attenuated pharyngeal arch development. In Class 2 and Class 3 mutants, the defects in thymus development may be due to abnormal pharyngeal arch development. Those mutations are expected to be useful for identifying the molecular mechanisms underlying thymus organogenesis.     

Trying to understand axonal regeneration in the CNS of fish

In contrast to the situation in mammals and birds, neurons in the central nervous system (CNS) of fish—such as the retinal ganglion cells—are capable of regenerating their axons and restoring vision. Special properties of the glial cells and the neurons of the fish visual pathway appear to contribute to the success of axonal regeneration. The fish oligodendrocytes lack the axon growth inhibiting molecules that interfere with axonal extension in mammals. Instead, fish optic nerve oligodendrocytes support—at least in vitro—axonal elongation of fish as well as that of rat retinal axons. Moreover, the fish retinal ganglion cells re-express upon injury a set of growth associated cell surface molecules and equip the regenerating axons throughout their path and up into their target, the tectum opticum with these molecules. This may indicate that the injured fish ganglion cells reactivate the cellular machinery necessary for axonal regrowth and pathfinding. Furthermore, the target itself provides positional marker molecules even in adult fish. These marker molecules are required to guide the regenerating axons back to their retinotopic home territory within the tectum. © 1992 John Wiley & Sons, Inc.     

Slit1 and Slit2 cooperate to prevent premature midline crossing of retinal axons in the mouse visual system.

During development, retinal ganglion cell (RGC) axons either cross or avoid the midline at the optic chiasm. In Drosophila, the Slit protein regulates midline axon crossing through repulsion. To determine the role of Slit proteins in RGC axon guidance, we disrupted Slit1 and Slit2, two of three known mouse Slit genes. Mice defective in either gene alone exhibited few RGC axon guidance defects, but in double mutant mice a large additional chiasm developed anterior to the true chiasm, many retinal axons projected into the contralateral optic nerve, and some extended ectopically-dorsal and lateral to the chiasm. Our results indicate that Slit proteins repel retinal axons in vivo and cooperate to establish a corridor through which the axons are channeled, thereby helping define the site in the ventral diencephalon where the optic chiasm forms.     

The cell adhesion molecule NrCAM is crucial for growth cone behaviour and pathfinding of retinal ganglion cell axons

We investigated the role of the cell adhesion molecule NrCAM for axonal growth and pathfinding in the developing retina. Analysis of the distribution pattern of NrCAM in chick embryo retina sections and flat-mounts shows its presence during extension of retinal ganglion cell (RGC) axons; NrCAM is selectively present on RGC axons and is absent from the soma. Single cell cultures show an enrichment of NrCAM in the distal axon and growth cone. When offered as a substrate in addition to Laminin, NrCAM promotes RGC axon extension and the formation of growth cone protrusions. In substrate stripe assays, mimicking the NrCAM-displaying optic fibre layer and the Laminin-rich basal lamina, RGC axons preferentially grow on NrCAM lanes. The three-dimensional analysis of RGC growth cones in retina flat-mounts reveals that they are enlarged and form more protrusions extending away from the correct pathway under conditions of NrCAM-inhibition. Time-lapse analyses show that these growth cones pause longer to explore their environment, proceed for shorter time spans, and retract more often than under control conditions; in addition, they often deviate from the correct pathway towards the optic fissure. Inhibition of NrCAM in organ-cultured intact eyes causes RGC axons to misroute at the optic fissure; instead of diving into the optic nerve head, these axons cross onto the opposite side of the retina. Our results demonstrate a crucial role for NrCAM in the navigation of RGC axons in the developing retina towards the optic fissure, and also for pathfinding into the optic nerve.     

Shared and distinct functions of RAGS and ELF-1 in guiding retinal axons.

Two ligands for Eph-related receptor tyrosine kinases, RAGS and ELF-1, have been implicated in the control of development of the retinotectal projection. Both molecules are expressed in overlapping gradients in the tectum, the target area of retinal ganglion cell axons. In two in vitro assays ELF-1 is shown to have a repellent axon guidance function for temporal, but apparently not for nasal axons. RAGS on the other hand is repellent for both types of axons, though to different degrees. Thus, RAGS and ELF-1 share some and differ in other properties. The biological activities of these molecules correlate with the strength of interaction with their receptors expressed on RGC axons. The meaning of these findings for guidance of retinal axons in the tectum is discussed.