Continued neurogenesis is not a pre-requisite for regeneration of a topographic retino-tectal projection.

Electrophysiological recording demonstrated that visuo-tectal projections are topographically organised after optic nerve regeneration in aged Xenopus laevis. 3H-thymidine autoradiography confirmed previous reports [Taylor, Lack, & Easter, Eur. Journal of Neuroscience 1 (1989) 626-638] that cell division had already ceased at the retinal ciliary margin. The results demonstrate that, contrary to a previous suggestion [Holder & Clarke, Trends in Neuroscience 11 (1988) 94-99], continued neurogenesis is not a pre-requisite for the re-establishment of appropriate connections with target cells.

Modulation of cell proliferation in the embryonic retina of zebrafish (Danio rerio).

We describe light-microscopically the development of the embryonic zebrafish eye with particular attention to cell number, cell proliferation, and cell death. The period from 16 to 36 hr post fertilization (hpf) comprises two phases; during the first (16-27 hpf) the optic vesicle becomes the eye cup, and during the second (27-36 hpf) the eye cup begins to differentiate into the neural retina and pigmented epithelium. All cells in the eye primordium are proliferative prior to 28 hpf, and the length of the cell cycle has been estimated to be 10 hr at 24-28 hpf (Nawrocki, 1985). Our cell counts are consistent with that estimate at that age, but not at earlier ages. A 10-hr cell cycle predicts that the cell number should increase by 7% per hr, but during 16-24 hpf the cell number increased by only 1.5% per hr. Despite the low rate of increase, all cells labeled with bromo-deoxyuridine, so all were proliferative. We considered three possible explanations for the nearly-constant cell number in the first phase: proliferation balanced by cell emigration from the eye, proliferation balanced by cell death, and low proliferation caused by a transient prolongation of the cell cycle. We excluded the first two, and found direct support for the third. Previous examinations of the cell cycle length in vertebrate central nervous system have concluded that it increases monotonically, in contrast to the modulation that we have shown. Modulation of the cell cycle length is well-known in flies, but it is generally effected by a prolonged arrest at one phase, in contrast to the general deceleration that we have shown.

The morphogenesis of the zebrafish eye, including a fate map of the optic vesicle.

We have examined the morphogenesis of the zebrafish eye, from the flat optic vesicle at 16 hours post fertilization (hpf) to the functional hemispheric eye at 72 hpf. We have produced three-dimensional reconstructions from semithin sections, measured volumes and areas, and produced a fate map by labeling clusters of cells at 14-15 hpf and finding them in the 24 hpf eye cup. Both volume and area increased sevenfold, with different schedules. Initially (16-33 hpf), area increased but volume remained constant; later (33-72 hpf) both increased. When the volume remained constant, the presumptive pigmented epithelium (PE) shrank and the presumptive neural retina (NR) enlarged. The fate map revealed that during 14-24 hpf cells changed layers, moving from the PE into the NR, probably through involution around the margin of the eye. The transformation of the flat epithelial layers of the vesicle into their cup-shaped counterparts in the eye was also accompanied by cellular rearrangements; most cells in a cluster labeled in the vesicle remained neighbors in the eye cup, but occasionally they were separated widely. This description of normal zebrafish eye development provides explanations for some mutant phenotypes and for the effects of altered retinoic acid.

Neurogenesis in the visual system of embryonic and adult zebrafish (Danio rerio). off.

The zebrafish has recently assumed a central position in the study of vertebrate development. Numerous studies of other fish have shown that their central nervous systems, and especially their visual systems, continue to add new neurons throughout life, which is probably related to their abilities to regenerate axons and whole nervous tissue. Retinal neurogenesis had not been examined in adult zebrafish, and two reports concluded that the optic tectum ceased neurogenesis early in life, so the question arose whether the zebrafish was anomalous in this regard. We labeled embryonic (24- and 48-h postfertilization) and adult zebrafish with the thymidine analog, bromo-deoxyuridine, and, after short and long survivals, examined the retina and brain for labeled cells. They were abundant in both the optic tectum and the retina. Although the rate of retinal growth slows considerably between embryonic and adult stages, the patterns of neurogenesis in both the embryo and the adult are similar to those described in other fish, so these "fish-specific" features of general interest can justifiably be studied in zebrafish.

Retinal neurogenesis: the formation of the initial central patch of postmitotic cells.

We have investigated the relationship between the birthdate and the onset of differentiation of neurons in the embryonic zebrafish neural retina. Birthdates were established by a single injection of bromodeoxyuridine into embryos of closely spaced ages. Differentiation was revealed in the same embryos with a neuron-specific antibody, zn12. The first bromodeoxyuridine-negative (postmitotic) cells occupied the ganglion cell layer of ventronasal retina, where they formed a small cluster of 10 cells or less that included the first zn12-positive cells (neurons). New cells were recruited to both populations (bromodeoxyuridine-negative and zn12-positive) along the same front, similar to the unfolding of a fan, to produce a circular central patch of hundreds of cells in the ganglion cell layer about 9 h later. Thus the formation of this central patch, previously considered as the start of retinal neurogenesis, was actually a secondary event, with a developmental history of its own. The first neurons outside the ganglion cell layer also appeared in ventronasal retina, indicating that the ventronasal region was the site of initiation of all retinal neurogenesis. Within a column (a small cluster of neuroepithelial cells), postmitotic cells appeared first in the ganglion cell layer, then the inner nuclear layer, and then the outer nuclear layer, so cell birthday and cell fate were correlated within a column. The terminal mitoses occurred in three bursts separated by two 10-h intervals during which proliferation continued without terminal mitoses.

The development of eye movements in the zebrafish (Danio rerio).

We investigated the development of oculomotor activity in zebrafish embryos and larvae of ages 48-96 hrs postfertilization (hpf). The optokinetic response (OKR: smooth tracking movements evoked by a rotating striped drum) improved steadily after its onset at 73 hpf, and by 96 hpf had a achieved a gain (eye velocity/drum velocity) of 0.9, comparable to adult performance. Reset movements (the fast phase of optokinetic nystagmus) developed over 75-81 hpf. The vestibuloocular reflex (VOR: compensatory eye movements evoked by passive rotation of the head) developed over 74-81 hpf, and the associated reset movements, over 76-81 hpf. The VOR was qualitatively normal in dark-reared fish, which excludes an essential role for visual experience in its early development. Spontaneous saccadic movements (the fast shift of eye position) appeared between 81 and 96 hpf, and at 96 hpf had maximum velocities that were comparable to adults. These results are compared to, and found to be incompatible with, two earlier ideas of motor development: behavioral "differentiation" and "encephalization."

An assessment of rat photoreceptor sensitivity to mitochondrial blockade.

PURPOSE: To report results of functional, biochemical and structural studies of photoreceptor mitochondria in isolated rat retinas under conditions of mitochondrial inhibition. METHODS: Dark-adapted rat retinas were incubated in a modified Ringer's bicarbonate medium under aerobic and anaerobic conditions. Several different procedures were used to inhibit mitochondrial function; N2, 0.01 mM antimycin A, and 1 and 10 mM potassium cyanide (KCN). Measurements were made of lactic acid production, retinal adenosine triphosphate (ATP) content, and receptor potentials. Morphology of the inner segment mitochondria was examined by electron microscopy. RESULTS: In the presence of N2, 0.01 mM antimycin, or 1 mM KCN, lactic acid production was linear throughout the 60- minute period; and the rate was similar for each condition. Retinal ATP content and the amplitude of the receptor potential were also maintained at high levels after short-term incubations with either N2, antimycin A, or 1 mM KCN. In contrast, use of 10 mM KCN produced an entirely different set of results. These effects were studied both at the alkaline pH (8.9) found when this concentration of KCN was simply added to bicarbonate-buffered media and at the normal pH (after readjustment) of 7.4. With 10 mM KCN (pH 8.9), retinal lactate production was severely depressed, retinal ATP content was nearly depleted within 5 to 10 minutes, and the amplitude of the receptor potential rapidly declined to a low level. The deleterious effects of 10 mM KCN on these parameters were lessened to varying degrees when pH was readjusted to 7.4. Electron microscopic observations of rat rod inner segments indicated generally excellent survival of these organelles after incubation with either N2, antimycin A, or 1 mM KCN in comparison with their appearance under oxygenated conditions. However, the inner segments were significantly disrupted after incubation of retinas with 10 mM KCN. CONCLUSIONS: Findings suggest that the loss of the receptor potential and depletion of ATP observed with minutes after exposing isolated rat retinas to media containing 10 mM KCN results from the inhibition of both respiration and glycolysis by this high concentration of KCN. In contrast, when conditions are chosen so that only respiration is impaired (as with N2, antimycin A, or 1 mM KCN) photoreceptor cells are resistant to short-term episodes of mitochondrial inhibition, principally because the upregulation of glycolysis generates sufficient ATP to compensate reasonably well for the loss in mitochondrially produced ATP.

Pax-6 functions in boundary formation and axon guidance in the embryonic mouse forebrain.

The Pax-6 gene encodes a transcription factor that is expressed in regionally restricted patterns in the developing brain and eye. Here we describe Pax-6 expression in the early forebrain (prosencephalon) on embryonic day 9.5 (E9.5) to E10.5 using both whole-mount in situ hybridization and antibody labeling. We find close correlations between Pax-6+ domains and initial neural patterning, and identify corresponding defects in embryos homozygous for the Pax-6 allele, Small eye (Sey). Pax-6 expression defines the prosencephalon-mesencephalon boundary, and mutant embryos lack this morphological boundary. Markers of the caudal prosencephalon are lost (Pax-6, Lim-1, Gsh-1) and a marker for mesencephalon is expanded rostrally into the prosencephalon (Dbx). We conclude that the caudal prosencephalon (prosomere 1) is at least partially transformed to a mesencephalic fate. This transformation results in a specific deficit of posterior commissure axons. Sey/Sey embryos also exhibit an axon pathfinding defect specific to the first longitudinal tract in the prosencephalon (tpoc, tract of the postoptic commissure). In wild type, tpoc axons fan out upon coming in contact with a superficial patch of Pax-6+ neuron cell bodies. In the mutant, the tpoc axons have normal initial projections, but make dramatic errors where they contact the neuron cell bodies, and fail to pioneer this first tract. Thus Pax-6 is required for local navigational information used by axons passing through its domain of expression. We conclude that Pax-6 plays multiple roles in forebrain patterning, including boundary formation, regional patterning, neuron specification and axon guidance.

The development of vision in the zebrafish (Danio rerio).

We studied the development and maturation of the visual system by determining when zebrafish begin to see and to move their eyes. This information was correlated with the time courses of the development of the retina, the retinofugal projection, the retinal image, and the extraocular muscles, to obtain an integrated picture of early visual development. Two visual behaviors were monitored over 48-96 hr postfertilization (hpf). The startle response (body twitch) was evoked by an abrupt decrease in light intensity. The optokinetic response (tracking eye movements) was evoked by rotation of a striped drum. Visually evoked startle developed over 68-79 hpf, more than 20 hr after the onset of a touch-evoked startle. It was not seen in eyeless fish, excluding a role for nonretinal light senses. Tracking eye movements developed over 73-80 hpf. They were always in the direction of drum rotation, even when the fish had been light deprived from blastula stage, ruling out a "trial and error" period of learning to track the drum. The image formed by the ocular lens was examined in intact fish made transparent by suppressing the formation of melanin. The eye was initially far sighted and gradually improved, so that by 72 hpf the image plane coincided with the photoreceptor layer. The extraocular muscles assumed their adult configuration between 66 and 72 hpf. Thus, the retinal image and functional extraocular muscles appeared nearly simultaneously with the onset of tracking eye movements and probably represent the last events in the construction of this behavior.

Early deletion of neuromeres in Wnt-1-/- mutant mice: evaluation by morphological and molecular markers.

The Wnt-1 gene is required for the development of midbrain and cerebellum; previous work showed that knockout of Wnt-1 causes the loss of most molecular markers of these structures in early embryos and deletion of these structures by birth. However, neither the extent of early neuronal defects nor any possible alterations in structures adjacent to presumptive midbrain and cerebellum were examined. By using a neuron-specific antibody and fluorescent axon tracers, we show that central and peripheral neuronal development are altered in mutants during initial axonogenesis on embryonic day 9.5. The absence of neuronal landmarks, including oculomotor and trochlear nerves and cerebellar plate, suggests that both mesencephalon and rhombomere 1 (r1) are delected, with the remaining neural tube fused to form a new border between the caudalmost portion of the prosencephalon (prosomere 1, or p1) and r2. Central axons accurately traverse this novel border by forming normal longitudinal tracts into the rhombencephalon, implying that the cues that direct these axons are aligned across neuromeres and are not affected by the delection. The presence of intact p1 and r2 is further supported by the retention of markers for these two neuromers, including a marker of p1, the Sim-2 gene, and an r2-specific lacZ transgene in mutant embryos. In addition, alterations in the Sim-2 expression domain in ventral prosencephalon, rostral to p1, provide novel evidence for Wnt-1 function in this region.

Initial organization of neurons and tracts in the embryonic mouse fore- and midbrain.

We investigated the potential role of rostral-caudal and dorsal-ventral subdivisions of the early rostral brain by relating these subdivisions to the early patterning of neuron cell bodies and their axon projections. The earliest neurons were mapped using the lipophilic axon tracers diI and diO on embryos fixed on embryonic days 9.5-10.5 (E9.5-E10.5); neuromeric boundaries were marked by diO. The tracts were small in number, were organized orthogonally (2 dorsal-ventral and 4 rostral-caudal), and originated from groups of cell bodies which we term "sources." Two parallel longitudinal axon systems, one dorsal (the tract of the postoptic commissure and the mesencephalic tract of the trigeminal nerve) and one ventral (the mammillotegmental tract and the medial longitudinal fasciculus), projected caudally from the prosencephalon into the rhombencephalon. We argue that the dorsal longitudinal pathway marked the boundary between the alar and basal plates along the entire neuraxis. The dorsal-ventral axons coursed circumferentially and either crossed the midline (forming the posterior and ventral tegmental commissures) or turned caudally without crossing the midline. The dorsal-ventral axons were not generally restricted to the interneuromeric boundaries, as others have suggested. Earlier, all neighboring neurons projected their axons together; later, nearby neurons projected into different pathways. Some tracts originated in single neuromeres, while other tracts had origins in two or more neuromeres. The dorsal longitudinal axons altered course at several of the borders, but the ventral longitudinal axons did not. In summary, the early subdivisions appeared to influence some, but not all, aspects of tract formation.

Expression of glial fibrillary acidic protein and its relation to tract formation in embryonic zebrafish (Danio rerio).

To address possible roles of glial cells during axon outgrowth in the vertebrate central nervous system, we investigated the appearance and distribution of the glial-specific intermediate filament, glial fibrillary acidic protein (GFAP), during early embryogenesis of the zebrafish (Danio rerio). Immunopositive cells first appear at 15 hours, which is at the time of, or slightly before, the first axon outgrowth in the brain. Immunopositive processes are not initially present in a pattern that prefigures the location of the first tracts but rather are distributed widely as endfeet adjacent to the pia, overlying most of the surface of the brain with the exception of the dorsal and ventral midline. The first evidence for a specific association of immunopositive cells with the developing tracts is observed at 24 hours in the hindbrain, where immunopositive processes border axons in the medial longitudinal fasciculus. By 48 hours, immunopositive processes have disappeared from most of the subpial lamina and are found exclusively in association with tracts and commissures in three forms: endfeet, radially oriented processes, and tangentially oriented processes parallel to axons. This last form is particularly prominent in the transverse plane of the hindbrain, where they define the boundaries between rhombomeres. These results suggest that glial cells contribute to the development and organization of the central nervous system by supporting early axon outgrowth in the subpial lamina and by forming boundaries around tracts and between neuromeres. The results are discussed in relation to previous results on neuron-glia interactions and possible roles of glial cells in axonal guidance.

The first retinal axons and their microenvironment in zebrafish: cryptic pioneers and the pretract.

The initial development of the optic tract was studied with light and electron microscopy in the zebrafish (Danio rerio). Intraocular injections of the fluorescent marker, 1,1'-dioctadecyl-3,3,3',3' tetramethylindocarbocyanine perchlorate (dil), labeled retinal axons and growth cones anterogradely, and injections of dil into the optic chiasm labeled retinal ganglion cells retrogradely. Labeled tissue was photoconverted and examined electron microscopically. The ventronasal retinal quadrant produced the first growth cones. They were the first growth cones in the optic stalk. The leading retinal growth cones, typically 4-10 in number, advanced alongside the tract of the postoptic commissure but rarely sent filopodia into it and never wrapped its axons. Instead, the retinal growth cones followed a pretract, a subpial region that was morphologically distinct from its surroundings and extended out in front of the leading growth cones, presaging the optic tract. Thus, the retinal growth cones, previously thought to be followers of preexisting axons, are actually cryptic pioneers whose proximity to the earlier axons masks their pioneering nature. We suggest that cryptic pioneers and pretracts are probably common elsewhere in the nervous system.

Cone photoreceptor regeneration in adult fish retina: phenotypic determination and mosaic pattern formation.

The retina of anamniotes (fish and amphibia), unlike the CNS of most vertebrates, can regenerate neurons following injury. Using the highly ordered mosaic of single and double cones in the retina of the adult green sunfish (Lepomis cyanellus) as our model system, we examined the events that followed the surgical excision of a small patch of central retina. After surgery there was a transient elevation in the number, and a change in the distribution, of proliferative cells within the retina. The wound was filled in two ways: a proliferative regeneration of new retina and a nonproliferative movement of the wound boundaries toward the center of the lesion. The nonproliferative movement stretched the surrounding, intact retina. In stretched retina the basic pattern of the cone mosaic was maintained, but it was augmented by new cones, even though cones are not normally generated in intact central retina. The stretch itself likely triggered the anomalous cone production. The new and preexisting cones in stretched retina had their morphological phenotypes influenced by mutual contact, often resulting in atypical morphologies (triple and quadruple cones). In the center of the lesioned area, the regenerated cone mosaic was disordered, had a higher than normal cone density, and contained atypical morphologies. The presence of outer segments and synaptic pedicles suggested that the new cones in regenerated and stretched retina were functional. We interpret these results to mean (1) a stretch-induced decrease in cell density can trigger a compensatory, adaptive neurogenesis, (2) cone morphological phenotypes in fish retina are plastic throughout life, and are influenced by cone-cone contacts, (3) the mechanisms that spatially regulate cone production during normal growth are disrupted regeneration.

Development of the retinofugal projections in the embryonic and larval zebrafish (Brachydanio rerio).

Studies of the projection from the vertebrate retina have contributed significantly to current concepts of neural development. The zebrafish has recently become a favored system for the study of development in general and neural development in particular. Although the development of both the optic nerve and the retinotectal projection of the zebrafish has been described, the retinofugal projection in its entirety has not. This paper describes it and also addresses the issue of projectional exuberance: i.e., transient projections to targets that are not innervated in the adult. The retinofugal projection of embryonic and larval zebrafish (32 hours to 7 days post-fertilization) was labeled by intraocular injection of DiI (1,1'-dioctadecyl-3,3,3',3',tetramethylindocarbocyanine perchlorate) and then studied in wholemounts and sections. The first optic axons crossed the chiasm at 32 hours post-fertilization and projected in a straight line to reach the tectum at about 44 hours. At 48 hours, a few optic axons deviated along either the tract of the posterior commissure or the tract of the postoptic commissure. By 72 hours (about the time of hatching) optic axons arborized in ten distinct regions, termed arborization fields. At 6-7 days post-fertilization, the same ten arborization fields (nine contralateral, one bilateral) were evident. Most of the arborization fields were located in the superficial neuropil and were not associated with morphologically identifiable clusters of somata. On the basis of various landmarks, the ten arborization fields are identified as precursors of retinorecipient nuclei previously described in other adult cypriniform fishes. The development was characterized by the nearly complete absence of any transient projections. Thus, the idea that axonal outgrowth is initially exuberant and trimmed back later is not supported by these results.

Graded-index model of a fish double cone exhibits differential polarization sensitivity.

The close apposition of the inner segments of the two cones that combine to form a double cone causes the pair of cone inner segments to guide light as a unitary structure whose transverse sections are roughly elliptical. Electron micrographs of the photoreceptors of a green sunfish (Lepomis cyanellus) retina provide evidence that the refractive index in the ellipsoid region of the inner segments of the double cones is higher in the center than at the perimeter. The hypothesis that the shape and refractive-index gradient could confer differential polarization sensitivity on double cones is examined with a two-dimensional waveguide model of a double-cone inner segment. The model has a dielectric constant that varies parabolically along the narrowest (x) dimension, leading to the index profile: n(x) = nmax[1-(x/x0)2]1/2, where nmax is the peak value of the index and x0 is a parameter specifying the rate at which the index decreases with increasing magnitude of x. A quantity, the polarization contrast, is introduced as a measure of the differential polarization sensitivity of adjacent receptors in the square mosaic of double cones in the sunfish retina. Polarization contrast is proportional to the relative difference in power absorbed by two double cones oriented with their shortest axes orthogonal to each other and stimulated by a field of uniform polarization. Polarization contrast is computed as a function of wavelength for appropriate values of nmax and x0. For normally incident light polarized parallel to one of the two axes of the double cones' cross sections, the polarization contrast is generally between 1% and 5% for wavelengths ranging from 550 to 750 nm. Over most of those wavelengths the polarization contrast of the graded-index-model double cone is approximately five times as large as that of a homogeneous-slab model of the same size and average refractive index. Additional benefits of a graded index, optical isolation of adjacent photoreceptors and antireflection at the photoreceptor entrance, are also discussed.

The cone photoreceptor mosaic of the green sunfish, Lepomis cyanellus.

Recent empirical and theoretical evidence has implicated the geometrical birefringence of the double cones of the green sunfish (Lepomis cyanellus) as the biophysical basis of this vertebrate's sensitivity to polarized light. Because of the intimate link between the organization of the cone-photoreceptor mosaic and the psychophysical details of polarization sensitivity, we have examined the structural features of the green sunfish cone-photoreceptor mosaic, in particular the orientation of the elliptical cross sections of the double cones. Our primary observations are that (1) the arrangement of the cone-photoreceptor mosaic is constant across the retina (with two regional exceptions), with double cones arranged in a rhombic mosaic and aligned roughly +/- 45 deg to the nearest retinal margin; (2) the double-cone/single-cone ratio is everywhere the same; (3) cone density is inhomogeneous across the retina, with the highest densities in the temporal hemiretina. These results are discussed as they relate to the animal's retinal growth and visual mechanisms, particularly the sensitivity to polarized light.

Initial tract formation in the mouse brain.

Mouse embryos from embryonic days 8.5-10.5 (E8.5-E10.5) were fixed and labeled with an antibody to neuron-specific class III beta-tubulin (Moody et al., 1987; Lee et al., 1990a,b) to reveal the first neurons, axons, and tracts in the brain. They were studied in whole-mounts and in light microscopic sections. Some conclusions were checked by labeling tracts in older embryos (E11.5 and E12.5) with the lipophilic dye 1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine. The first immunoreactive cells appeared at E8.5, prior to neural tube closure, in the neural plate immediately caudal to the optic vesicle. Cells along the dorsal midline of the mesencephalon issued the first axons, on E9.0; the cells were the mesencephalic nucleus of the trigeminal nerve, and the axons formed its descending tract. The tract reached the level of the trigeminal ganglion by E10.0 but did not enter the ganglion until after E12.5. On E9.5, the number of labeled cells and axons in the alar plate of the presumptive diencephalon and mesencephalon had increased substantially, and many of the rostral ones coursed into the basal plate to enter longitudinal tracts there. Two tracts originated from cells in the basal plate: the tract of the postoptic commissure (from the base of the optic stalk to the level of the cephalic flexure) and the medial longitudinal fasciculus (from the level of the cephalic flexure caudally through the mid and hind-brains). By E10.0, a small mammillotegmental tract paralleled the tract of the postoptic commissure, but immunolabeling was so widespread that discrete tracts were impossible to discern in the presumptive diencephalon and mesencephalon. The more rostral regions remained lightly labeled. In the cerebral vesicle, the presumptive cerebral cortex, the first immunoreactive cells appeared at E10.0; they had multiple processes oriented parallel to the pia, and were identified as the Cajal-Retzius cells. By E10.5, no tracts had formed in the cerebral vesicle. All the tracts formed by E10.0 were superficial, in the subpial lamina. Those that can be identified in the adult brain are very deep structures. These results are compared with previous descriptions of the embryonic brains of amphibians, fish, birds, and other mammals, including humans.