Patterning of retinotectal connections in the vertebrate visual system

Calcium Imaging in the Zebrafish

The zebrafish (Danio rerio) has emerged as a widely used model system during the last four decades. The fact that the zebrafish larva is transparent enables sophisticated in vivo imaging, including calcium imaging of intracellular transients in many different tissues. While being a vertebrate, the reduced complexity of its nervous system and small size make it possible to follow large-scale activity in the whole brain. Its genome is sequenced and many genetic and molecular tools have been developed that simplify the study of gene function in health and disease. Since the mid 90's, the development and neuronal function of the embryonic, larval, and later, adult zebrafish have been studied using calcium imaging methods. This updated chapter is reviewing the advances in methods and research findings of zebrafish calcium imaging during the last decade. The choice of calcium indicator depends on the desired number of cells to study and cell accessibility. Synthetic calcium indicators, conjugated to dextrans and acetoxymethyl (AM) esters, are still used to label specific neuronal cell types in the hindbrain and the olfactory system. However, genetically encoded calcium indicators, such as aequorin and the GCaMP family of indicators, expressed in various tissues by the use of cell-specific promoters, are now the choice for most applications, including brain-wide imaging. Calcium imaging in the zebrafish has contributed greatly to our understanding of basic biological principles during development and adulthood, and the function of disease-related genes in a vertebrate system.

Topographic wiring of the retinotectal connection in zebrafish

The zebrafish retinotectal projection provides an attractive model system for studying many aspects of topographic map formation and maintenance. Visual connections initially start to form between 3 and 5 days postfertilization, and remain plastic throughout the life of the fish. Zebrafish are easily manipulated surgically, genetically, and chemically, and a variety of molecular tools exist to enable visualization and control of various aspects of map development. Here, we review zebrafish retinotectal map formation, focusing particularly on the detailed structure and dynamics of the connections, the molecules that are important in map creation, and how activity regulates the maintenance of the map.

Imaging axon pathfinding in zebrafish in vivo

Axon pathfinding in the developing animal involves a highly dynamic process in which the axonal growth cone makes continuous decisions as it navigates toward its target. Changes occurring in the growth cone with respect to retracting from or extending into complex new territories can occur in minutes. Thus, the advent of strategies to visualize axon path-finding in vivo in a live intact animal is crucial for a better understanding of how the growth cone makes such rapid decisions in response to multiple cues. Combining these strategies with loss-of-function and/or gain-of-function techniques, one can gain some insight as to which molecules are crucial to particular growth cone behaviors at specific choice points during navigation. The major advantage of using zebrafish lies in the accessibility of major axon tracts for live microscopy, as their embryonic development occurs ex utero. Furthermore, the robust embryos remain healthy during immobilization and allow for good imaging for long periods. This protocol describes the method for stabilizing and preparing live zebrafish embryos for imaging labeled axonal tracts at high spatial and temporal resolution for up to 72 h. It has been used for retinotectal axon pathfinding, but can be adapted to visualize other axon tracts of interest.

Imaging axon pathfinding in Xenopus in vivo

Axon pathfinding in the developing animal involves a highly dynamic process in which the axonal growth cone makes continuous decisions as it navigates toward its target. Changes occurring in the growth cone with respect to retracting from or extending into complex new territories can occur in minutes. Thus, the advent of strategies to visualize axon path-finding in vivo in a live intact animal is crucial for a better understanding of how the growth cone makes such rapid decisions in response to multiple cues. Combining these strategies with loss-of-function and/or gain-of-function techniques allows one to gain some insight as to which molecules are crucial to particular growth cone behaviors at specific choice points during navigation. The main advantage of using Xenopus lies in the accessibility of major axon tracts for live microscopy, as their embryonic development occurs ex utero. Furthermore, the robust embryos remain healthy during immobilization and allow for good imaging for long periods. This protocol describes the methods for stabilizing and preparing live Xenopus embryos for imaging labeled axonal tracts at high spatial and temporal resolution for up to 72 h. This approach can been used to investigate how the knockdown of certain gene functions can affect the speed of navigation through the well-studied Xenopus retinotectal pathway. It can be adapted to visualize other axon tracts of interest.

Calcium imaging in the zebrafish

The zebrafish (Danio rerio) has emerged as a new model system during the last three decades. The fact that the zebrafish larva is transparent enables sophisticated in vivo imaging. While being the vertebrate, the reduced complexity of its nervous system and small size make it possible to follow large-scale activity in the whole brain. Its genome is sequenced and many genetic and molecular tools have been developed that simplify the study of gene function. Since the mid 1990s, the embryonic development and neuronal function of the larval, and later, adult zebrafish have been studied using calcium imaging methods. The choice of calcium indicator depends on the desired number of cells to study and cell accessibility. Dextran indicators have been used to label cells in the developing embryo from dye injection into the one-cell stage. Dextrans have also been useful for retrograde labeling of spinal cord neurons and cells in the olfactory system. Acetoxymethyl (AM) esters permit labeling of larger areas of tissue such as the tectum, a region responsible for visual processing. Genetically encoded calcium indicators have been expressed in various tissues by the use of cell-specific promoters. These studies have contributed greatly to our understanding of basic biological principles during development and adulthood, and of the function of disease-related genes in a vertebrate system.

Analysis of the retina in the zebrafish model

The zebrafish is one of the leading models for the analysis of the vertebrate visual system. A wide assortment of molecular, genetic, and cell biological approaches is available to study zebrafish visual system development and function. As new techniques become available, genetic analysis and imaging continue to be the strengths of the zebrafish model. In particular, recent developments in the use of transposons and zinc finger nucleases to produce new generations of mutant strains enhance both forward and reverse genetic analysis. Similarly, the imaging of developmental and physiological processes benefits from a wide assortment of fluorescent proteins and the ways to express them in the embryo. The zebrafish is also highly attractive for high-throughput screening of small molecules, a promising strategy to search for compounds with therapeutic potential. Here we discuss experimental approaches used in the zebrafish model to study morphogenetic transformations, cell fate decisions, and the differentiation of fine morphological features that ultimately lead to the formation of the functional vertebrate visual system.

Functional imaging reveals rapid development of visual response properties in the zebrafish tectum

The visual pathway from the retina to the optic tectum in fish and frogs has long been studied as a model for neural circuit formation. Although morphological aspects, such as axonal and dendritic arborization, have been well characterized, less is known about how this translates into functional properties of tectal neurons during development. We developed a system to provide controlled visual stimuli to larval zebrafish, while performing two-photon imaging of tectal neurons loaded with a fluorescent calcium indicator, allowing us to determine visual response properties in intact fish. In relatively mature larvae, we describe receptive field sizes, visual topography, and direction and size selectivity. We also characterize the onset and development of visual responses, beginning when retinal axons first arborize in the tectum. Surprisingly, most of these properties are established soon after dendrite growth and synaptogenesis begin and do not require patterned visual experience or a protracted period of refinement.

Approaches to study neurogenesis in the zebrafish retina

Similar to other vertebrate species, the zebrafish retina is simpler than other regions of the central nervous system (CNS). Relative simplicity, rapid development, and accessibility to genetic analysis make the zebrafish retina an excellent model system for the studies of neurogenesis in the vertebrate CNS. Numerous genetic screens have led to isolation of an impressive collection of mutations affecting the retina and the retinotectal projection in zebrafish. Mutant phenotypes are being studied using a rich variety of markers: antibodies, RNA probes, retrograde and anterograde tracers, as well as transgenic lines. Particularly impressive progress has been made in the characterization of the zebrafish genome. Consequently, positional and candidate cloning of mutant genes are now fairly easy to accomplish in zebrafish. Many mutant genes have, in fact, already been cloned and their analysis has provided important insights into the gene circuitry that regulates retinal neurogenesis. Genetic screens for visual system defects will continue in the future and progressively more sophisticated screening approaches will make it possible to detect a variety of subtle mutant phenotypes in retinal development. The remarkable evolutionary conservation of the vertebrate eye provides the basis for the use of the zebrafish retina as a model of human disorders. Some of the genetic defects of the zebrafish retina indeed resemble human retinopathies. As new techniques are being introduced and improved at a rapid pace, the zebrafish will continue to be an important organism for the studies of the vertebrate visual system.

Developmental remodeling of the retinogeniculate synapse

Anatomical rearrangement of retinogeniculate connections contributes to the refinement of synaptic circuits in the developing visual system, but the underlying changes in synaptic function are unclear. Here, we study such changes in mouse brain slices. Each geniculate cell receives a surprisingly large number of retinal inputs (>20) well after eye-specific zones are formed. All but one to three of these inputs are eliminated over a 3-week period spanning eye opening. Remaining inputs are strengthened approximately 50-fold, in part through an increase in quantal size, but primarily through an increase in the number of release sites. Changes in release probability do not contribute significantly. Thus, a redistribution of release sites from many inputs to few inputs at this late developmental stage contributes to the precise receptive fields of thalamic relay neurons.

Morphogenesis of the optic tectum in the medaka (Oryzias latipes): a morphological and molecular study, with special emphasis on cell proliferation

We analyzed the medaka optic tectum (OT) morphogenesis by using 5-bromo-2'-deoxyuridine (BrdU) immunohistochemistry (with a new method we developed for pulse-labeling embryos) and in situ hybridization with three probes, two for recently cloned homeobox genes (Ol-Prx3 [Paired-Related-Homeobox3] and Ol-Gsh1 [Genetic-Screen-Homeobox1]) and one for Ol-tailless. The tectal anlage first appears as a sheet of proliferating cells expressing Ol-Gsh1 and Ol-tailless but not Ol-Prx3. Cells subsequently cease to proliferate in a superficial and rostral zone and begin to express Ol-Prx3. When tectal lamination begins, the proliferative zone (mpz) becomes restricted to a crescent at the OT medial, caudal, and lateral margin. This mpz functions throughout the fish's entire life. It produces cells that are added at the OT's edge as radial rows, spanning every layer of the OT. The cells of the mpz continue to express Ol-tailless in the adult, whereas Ol-Gsh1 expression is turned off. When superficial layers form, Ol-Prx3 expression becomes restricted to the underlying deep layer, where it persists in the adult. Ol-Prx3 seems to be a marker for the differentiation of a subset of deep cells and allows analysis of tectal lamination, whereas Ol-tailless and Ol-Gsh1 could be involved in the control of tectal cell proliferation. This study constitutes a first step toward molecular approach to OT development in anamniotes. We compare and discuss the expression patterns of the homologs of the genes studied, and more generally the morphogenetic patterns of the medaka tectum, with those encountered in other cortical structures and in other vertebrate groups.

Development of the retina

As in other vertebrate species, the zebrafish retina is simpler than other regions of the central nervous system. This relative simplicity along with rapid development, and accessibility to genetic analysis make the zebrafish retina an excellent model system for studies of neurogenesis in the vertebrate CNS. Several genetic screens have led to the isolation of an impressive collection of mutants affecting the retina and the retinotectal projections in zebrafish. A variety of techniques and markers are available to study the isolated mutants. These include several antigen- and transcript-detection methods, retrograde and anterograde labeling of neurons, blastomere transplantations, H3 labeling, and others. As past genetic screens have achieved a rather low level of saturation, the current collection of mutants can only grow in the future. Morphological and behavioral criteria have been successfully applied in zebrafish to search for defects in spinal development. In future genetic screens, progressively more sophisticated screening approaches will make it possible to detect very subtle changes in the retinal development. The remarkable evolutionary conservation of the vertebrate eye provides the basis for using the zebrafish as a model system for the detection and analysis of genetic defects potentially related to human eye disorders. Some of the genetic defects of the zebrafish retina indeed resemble human retinopathies. As the genetic analysis of the vertebrate visual system is far from being complete and new techniques are being introduced at a rapid pace, the zebrafish embryo will become increasingly useful as a model for studies of the vertebrate retina.

The ephrins and Eph receptors in neural development

The Eph receptors are the largest known family of receptor tyrosine kinases. Initially all of them were identified as orphan receptors without known ligands, and their specific functions were not well understood. During the past few years, a corresponding family of ligands has been identified, called the ephrins, and specific functions have now been identified in neural development. The ephrins and Eph receptors are implicated as positional labels that may guide the development of neural topographic maps. They have also been implicated in pathway selection by axons, the guidance of cell migration, and the establishment of regional pattern in the nervous system. The ligands are anchored to cell surfaces, and most of the functions so far identified can be interpreted as precise guidance of cell or axon movement. This large family of ligands and receptors may make a major contribution to the accurate spatial patterning of connections and cell position in the nervous system.

Cellular localization of ephrin-A2, ephrin-A5, and other functional guidance cues underlies retinotopic development across species

Avian retinotectal and rodent retinocollicular systems are general model systems used to examine developmental processes that underpin topographically organized neuronal circuits. The two systems rely on guidance components to establish their precise retinotopic maps, but many cellular events differ during their development. For example, compared with the chick, a generally less restricted outgrowth pattern is observed when retinae innervate their targets in rodents. Cellular or molecular distributions of guidance components may account for such differences in retinotopic development across species. Candidate repellent molecules, such as ephrin-A2 and ephrin-A5, have been cloned in both chick and rodents; however, it has not yet been shown in rodents that living cells express sufficient amounts of any repellent components to deter outgrowth. We used a coculture assay that gives cellular resolution of retinotarget interactions and demonstrate that living, caudal superior colliculus cells selectively prevent extension of axons from temporal regions of the retinae. Time-lapse video microscopy revealed the cellular localization of permissive and repulsive guidance components in rodents, which differed from that in chick. To analyze the potential molecular basis for these differences, we investigated the function and localization of ephrin-A2 and -A5. Cells transfected with ephrin-A2 and -A5 selectively repelled retinal axons. Ephrin-A2 and -A5 RNA expression patterns differed across cell populations and between species, suggesting molecular mechanisms and key cellular interactions that may underlie fundamental differences in the development of retinotectal and retinocollicular maps.

Organization of the medial pulvinar nucleus in the macaque

BACKGROUND

The medial pulvinar appears to subserve the integration of associative cortical information and projects to visuomotor-related cortex. In contrast to the other pulvinar subdivisions, the medial pulvinar is a polymodal structure. Therefore, we studied the structural organization of the medial pulvinar to determine how it differs from the surrounding unimodal nuclei.

METHODS

Nissl-stained sections were examined to determine the boundaries of, and the distribution of neuronal sizes within, the medial pulvinar. In addition, Golgi-impregnated neurons were examined and drawn for analysis. Only rhesus monkey specimens were used, and the material had been prepared previously for other studies.

RESULTS

Projection neurons have round to oval somata and moderate numbers of primary dendrites that extend for short distances before branching into many secondary branches. Two variations of projection neurons (P1 and P2) were distinguished on the basis of the diameters of their dendritic tree. Both varieties have short dendrites that radiate in all directions. They differ in that P2 cells have longer second tier dendrites than P1 cells. Three types of local circuit neurons, tufted, radiating and varicose, were distinguished on the basis of their dendritic morphology. Four types of afferent fibers were identified. Type 1 afferents form cone-shape terminal arbors. Type 2 afferents are similar to those reported for retinal or cortical terminals. Type 3 afferents are of medium thickness and of an unknown origin. Type 4 afferents are thin and have small varicosities consistent with previously described cortical afferents. Afferent fibers are predominantly oriented along the mediolateral axis of the nucleus. We observed putative contacts between some afferents and local circuit neurons and between local circuit neurons and projection neurons.

CONCLUSIONS

Medial pulvinar neurons are generally smaller and rounder than those found in the adjacent pulvinar nuclei. These results provide additional evidence for structural distinctions between thalamic nuclei having different functions. However, the observed differences are subtle. In addition, the data in this report provide morphological evidence that cortical signals are likely to be integrated by means of the circuitry located within the nucleus.

Absence of topography in precociously innervated tecta

The retinotectal map in Xenopus forms very early: retinal axons are topographically ordered along the dorsoventral axis of the tectum by stage 39, as they first arrive. To test whether topographic cues are present even earlier, we forced retinal axons to innervate the tectum prematurely by transplanting stage 28 eye primordia into stage 20 hosts, then assayed dorsoventral topography using focal injections of lipophilic dye into dorsal and ventral retina at donor stages 39–40. Unoperated and isochronic control projections showed normal dorsoventral ordering both in the optic tract and in the tectum. In contrast, projections from heterochronically transplanted eyes were ordered in the tract, but spread out upon entering the tectum and did not show significant dorsoventral ordering. Individual axons entering the tectum precociously often made abnormally abrupt and topographically incorrect turns. Thus, the topographical cues normally expressed in the tectum at stage 39 appear to be absent a few hours earlier. However, this lack of cues is only temporary, since heterochronic transplants allowed to survive to donor stages 45–46 showed normal topography. The absence of tectal topography at a stage when retinal axons can navigate to the young tectum strongly suggests that the molecules that provide tectal topographical signals are distinct from those used for pathfinding in the diencephalon and target recognition at the tectum.

Target preference of embryonic retina cells and retinal cell lines is cell-autonomous, position-specific, early determined and heritable

Retinal ganglion cells (RGCs) form the topographic connection between retina and optic tectum in the developing avian embryo. In vitro, neurons with the morphological traits and marker expression of RGCs were found both in single-cell cultures from embryonic day (E) 6 chick retina and in retinal cell lines derived from E3.5 quail retina. Rapid and substantial differentiation of RGC-like cells could be induced in the lines by addition of fibroblast growth factor aFGF or bFGF. RGC-like cells were examined with respect to their target discrimination properties as single cells in the stripe carpet assay. In this assay system, alternating stripes of membrane vesicles prepared from the anterior and posterior tectum are offered to growing axonal processes as a substrate. Temporal RGC-like cells, both primary cells prepared from the temporal retina and immortalized cells of those retinal lines derived from the temporal retina, avoid stripes of membrane vesicles from posterior tectum; they prefer to grow on membrane vesicles from the anterior tectum, which is their in vivo target. Nasal RGC-like cells did not exhibit a target preference, in accordance with previous findings. Together the experiments show that target preference of RGCs is a cell-autonomous and heritable mechanism that is determined early and is position-dependent.

Connectional topography in the zebrafish olfactory system: random positions but regular spacing of sensory neurons projecting to an individual glomerulus

It is unknown how neuronal connections are specified in the olfactory system. To define rules of connectivity in this system, we investigated whether the projection of sensory neurons from the olfactory epithelium to the olfactory bulb is topographically ordered. By backtracking with 1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate (DiI), we find that neurons projecting into a single identified glomerulus are widely dispersed over the olfactory epithelium. Their positions in the sensory surface do not predict their glomerulus specificity and are probably random. A statistical analysis reveals that neurons connected to the same glomerulus are spaced at distances of several cell diameters from each other. The convergence of projections to one point in the target area from neurons that are widely and evenly distributed in the sensory surface constitutes an unusual type of connectional topography that contrasts with the precise topological (neighborhood-preserving) maps found in other sensory systems. It may maximize the probability to detect odorants that activate a single glomerular unit.

Control of topographic retinal axon branching by inhibitory membrane-bound molecules

Retinotopic map development in nonmammalian vertebrates appears to be controlled by molecules that guide or restrict retinal axons to correct locations in their targets. However, the retinotopic map in the superior colliculus (SC) of the rat is developed instead by a topographic bias in collateral branching and arborization. Temporal retinal axons extending across alternating membranes from the topographically correct rostral SC or the incorrect caudal SC of embryonic rats preferentially branch on rostral membranes. Branching preference is due to an inhibitory phosphatidylinositol-linked molecule in the caudal SC. Thus, position-encoding membrane-bound molecules may establish retinotopic maps in mammals by regulating axon branching, not by directing axon growth.

The molecular basis of retinotectal topography

Over 50 years have passed since Roger Sperry formulated a simple model of how visual space, as seen by the retina, can be projected onto the brain in a two-dimensional, topographic map during development. Sperry posited a set of two orthogonal gradients in the retina that gives each cell a positional identity. He further suggested that these molecules could be used to match up with complementary gradients in the target field of the retinal projection, the tectum. While some investigators hold that the existence of such molecules may not be necessary to establish retinotectal maps, recent work has identified several cell surface proteins whose distributions are of the type predicted by Sperry. An unexpected twist comes from culture assays demonstrating that inhibitory activities on tectal membranes can guide the growth of processes from retinal neurons. Moreover, the expression patterns of several enzymes and three transcription factors suggest that these proteins are candidates for regulatory agents in the determination of cell position in the retina. In addition, results from perturbation experiments support the candidacy of two of the enzymes, and a new mutant screen has uncovered several as yet unidentified genes that are required for establishment of the proper retinotectal map. A number of these results were presented at a recent meeting on neurospecificity held in Cargese, Corsica and sponsored by NATO and NSF.

Regenerating retinal fibers display error-free homing along undamaged normal fibers

After crushing one optic nerve in a bony fish, retinal fibers regenerate to both tecta. Anterograde labelling indicates that the ipsilaterally regenerating fibers have a rather straight growth, apparently along the undamaged fibers of the contralateral retina. In contrast, the contralaterally regenerating fibers deviate widely from a straight course. Retrograde labelling shows a mirror-symmetric distribution of regenerated ipsilateral and resident contralateral ganglion cells in a comparable annulus. In contrast, ganglion cells in the regenerated contralateral retina show no topological order after comparable small Dil applications to the ventrolateral tectum. These data suggest that regenerating fibers can orient on the undisturbed, contralateral fibers.

Responses of retinal axons in vivo and in vitro to position-encoding molecules in the embryonic superior colliculus

We show that rat retinal ganglion cell axons exhibit no topographic specificity in growth along the rostral-caudal axis of the embryonic superior colliculus (SC). Position-related, morphological differences are not found between temporal and nasal axon growth cones. However, embryonic retinal axons respond in vitro to a position-dependent molecular property of SC membranes. In vivo, regional specificity in side branching is the earliest indication that axons make topographic distinctions along the rostral-caudal SC axis. Our contrasting in vivo and in vitro results indicate that molecules encoding rostral-caudal position in the SC neither guide nor restrict retinal axon growth, but may promote the development of topographic connections by controlling specificity in the extension or stabilization of branches.