Cytogenesis in the monkey retina

Spatiotemporal pattern of ontogenetic expression of calbindin-28/kD in the retinorecipient layers of rat superior colliculus

Using an antibody against calbindin-28kD, we have studied the spatial pattern of expression of this protein in the superior colliculi (SC) of four strains of mature laboratory rats. In all four strains, calbindin-expressing cells (CECs) formed horizontally oriented tiers in the retinorecipient and intermediate gray layers but were diffusely distributed throughout the deep layers. Ontogenetically, calbindin-28kD was expressed for the first time in the retinorecipient layers at postconceptional day 20 (PCD 20), by cells located in the rostrolateral region where the first born retinal ganglion cells (RGCs) are represented. Although on the day of birth (PCD 22/23), the CECs were distributed more widely, they were still absent in the most medial part of the SC, that is, the region where the latest born RGCs are represented. The spatial distribution of CECs became adultlike only by PCD 29, that is, at the end of the period of the naturally occurring death of the RGCs. Monocular eye enucleations on PCD 23 prevented the expression of calbindin in the medial fifth of the retinorecipient layers of the contralateral SC, while the unilateral removal of the visual cortices had no discernable effect on the numbers and distribution of the CECs in either SC. Thus, the spatiotemporal pattern of ontogenetic expression of calbindin-28kD in the retinorecipient layers of SC reflects the spatiotemporal pattern of generation of the RGCs, and the retinal input appears to induce neuronal expression of calbindin-28kD in these layers.

Flk-1, a receptor for vascular endothelial growth factor (VEGF), is expressed by retinal progenitor cells

Throughout development of the vertebrate retina, progenitor cells are multipotential, producing a variety of distinctive cell types. Little is known of the molecular mechanisms directing the determination of cell fate. We have examined retinal progenitor cells for expression of receptor tyrosine kinases in an attempt to define receptors that could allow a progenitor to respond to its environment. We found that the receptor tyrosine kinase Flk-1, previously shown to be expressed in endothelial cells, is also expressed in neural progenitor cells of the mouse retina. Flk-1 RNA expression in the retinal progenitors commences with the onset of neuronal differentiation and persists throughout retinal neurogenesis. Flk-1 RNA and protein levels in the retina vary temporally during development, as shown by in situ hybridization and Western blot analysis. Patterns of beta-galactosidase expression in mice containing the lacZ gene in place of the Flk-1 gene are consistent with Flk-1 being expressed in retinal progenitors. In addition, we show that the ligand of Flk-1, vascular endothelial growth factor (VEGF), is expressed in the developing retina by differentiated cells and that a chimeric ligand of VEGF fused to alkaline phosphatase binds to proliferating retinal progenitors. Furthermore, the neural retina-derived Flk-1 protein kinase is activated by VEGF in vitro. Thus, the Flk-1 receptor protein kinase is expressed on the surface of neural progenitors in mouse retina and may play a critical role in neurogenesis as well as in vasculogenesis.

Retinotopy of the human retinal nerve fibre layer and optic nerve head

The organisation of the primate nerve fibre layer and optic nerve head with respect to eccentricity or the positioning of central and peripheral axons remains controversial. Crystals of the carbocyanine dyes DiI (1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate), or DiA (4-[4-didecylaminostryryl]-N-methylpridiniumiodide) were used to trace retinal ganglion cell axons within the nerve fibre layer, optic nerve head, and optic nerve. The present study demonstrated that peripheral retinal axons were scattered throughout the vitreal-scleral depth of the nerve fibre layer. This scattered distribution was maintained as the fibres passed through the optic nerve head and into the optic nerve. Axons of the arcuate bundles showed a bias towards the scleral portions of the nerve fibre layer and a variable degree of fibre scatter across the nerve fibre layer which was not as evident in labelling from other retinal regions. There was a rough topographic representation within the optic nerve head according to retinal circumference such that both peripheral and central fibres were mixed within a wedge extending from the periphery to the centre of the nerve. Foveal fibres occupied a large proportion of the temporal aspect of the optic nerve head and nerve, whereas fibres from areas temporal to the fovea appeared to be displaced to more superior and inferior regions. Consistent with the scleral bias seen in the retina, arcuate fibres maintained a peripheral position as they passed through the optic nerve head and occupied a more peripheral position in the nerve. The present results suggest that any degree of order present within the optic nerve is not an active process; optic axons are not instructed to establish a retinotopic order within the initial portions of the visual pathway.

Parvalbumin, calbindin, and calretinin mark distinct pathways during development of monkey dorsal lateral geniculate nucleus

Immunocyochemical labeling was applied to follow the developmental changes in the calcium-binding proteins parvalbumin (PV), calbindin D28k (CaB), and calretinin (CaR) during fetal and infant development of Macaca monkey dorsal lateral geniculate nucleus (LGN). For all three proteins, LGN cell body and retinal ganglion cell (RGC) axon labeling patterns changed temporally and spatially over development, and many of these were LGN laminar specific. CaR+ and CaB+ cells were present at the youngest age studied, fetal day 55 (F55). After lamination of the LGN occurred between F90 and F115, CaR+ and CaB+ neurons were specific markers for the S, intercalated, and interlaminar layers. Double label immunocytochemistry showed that all CaR+ cells contained CaB, and none contained GABA. CaR+ cell bodies decreased in number soon after birth so that adult LGN contained only a very small number of CaR+ cells. These patterns and cell counts indicated that a downregulation of CaR had occurred in the CaB+ population. Although CaB+ cell density in S and interlaminar zones declined in the adult, cell counts indicated that this is due to dilution of a stable population into a much larger nucleus during development. PV+ cells appeared at F85 only within the putative magnocellular (M) and parvocellular (P) layers, and PV remained a marker for these layers throughout development. Fetal PV cells also contained GABA, indicating that they were LGN interneurons. After birth, GABA-/PV+ cell numbers increased dramatically throughout the whole nucleus so that by the end of the first year, P and M layers were filled with PV+ cells. Their number and size indicated that these were the LGN projection neurons. Beginning at F66, bundles of PV+ axons occupied the anterior-middle LGN and filled the optic tract. Up to F101, PV+ synaptic terminals were restricted to Players, but after F132 labeling in M layers was heavier than in P layers. Axonal labeling for CaR began at F125. Prenatally CaR+ terminals were present mainly in P layers, whereas by postnatal 9 weeks labeling in M layers much exceeded P layers. Axonal labeling for CaB was present at F132, but CaB+ terminals were observed only after birth with labeling always heavier in M than P layers. By postnatal 9 weeks, PV, CaR, and CaB were colocalized in the same axons and terminals. These experiments indicated that during development and in the adult LGN, both CaR and CaB were markers for the LGN neurons in the S and intercalated pathway. CaR was present transiently while CaB persisted into adulthood. PV was a M and P layer marker first for interneurons and later for projection cells. The complex temporal developmental patterns found in this study suggested that viewing PV, CaB, and CaR simply as calcium-buffering proteins severely underestimates their functional roles during visual system maturation.

Comparison of topographical patterns of ganglion and photoreceptor cell differentiation in the retina of the zebrafish, Danio rerio

Earlier studies suggested retinal differentiation in the zebrafish commences ventrally rather than centrally as is the case in other vertebrates. Here we describe the topographical spread of cell differentiation for ganglion cells, double cones and rods in the zebrafish retina between 36 and 72 hours postfertilization (hpf), by using immunohistochemical markers in retinal wholemounts. Staining for all three cell types commenced within the ventral retina on the nasal side of the optic nerve and choroid fissure, at 38 hpf for ganglion cells and 50 hpf for double cones and rods. Within 3 to 4 hours, the staining of ganglion cells and double cones spread in a continuous wave-like fashion into the nasal region of the ventral retina. After this time, the staining patterns for ganglion cells and double cones progressed dorsally into the central and temporal retina. Finally, stained somata of ganglion cells were observed within the temporal-ventral region by approximately 48 hpf, more than 8 hours later than the first ganglion cells within the nasal retina. The topographical spread of double cone staining was slightly less orderly. After staining had extended into the nasal retina between 50 and 54 hpf, a small group of stained double cones often appeared at the temporal edge of the choroid fissure by 56 hpf, simultaneously with initial staining observed dorsal and temporal to the optic nerve. The topographical spread of rod staining in the ventral retina was more symmetrical. After rod staining appeared near the nasal edge of the choroid fissure at 50 hpf, rods accumulated within a localized patch nasal to the fissure. Approximately 5 hours after initial rod staining, scattered rod staining appeared on the temporal side of the choroid fissure (approximately 55-57 hpf). Rods increased rapidly within the ventral retina, and a dense symmetrical patch extended out from the choroid fissure into the nasal and temporal regions of the ventral retina by 70 hpf. A scattered pattern of rod staining also occurred within the dorsal retina at this time.

Retinal pigment epithelium and photoreceptor maturation in a wallaby, the quokka

Cell generation and the early stages of maturation of the retinal pigment epithelium (RPE) and photoreceptors were examined in a marsupial, the quokka, Setonix brachyurus. Results are presented for animals aged up to postnatal day (P)250. RPE cell generation was studied by analysis of cell number from wholemounted retinae and by tritiated thymidine (3HThy) autoradiography in sectioned material. For 3HThy autoradiography, quokkas aged P1-P200 were injected with 3HThy and killed either 6-20 hours later (pulse-kill) or at P100 or P250 (pulse-leave). The extent of pigmentation of the RPE sheet was examined from sections of embryonic and early postnatal stages. Retinae from animals aged P5 to P160 were also examined at the electron microscope. By P100, RPE cell number is within the range found in adults. New RPE cells are generated in a peripheral band which moves outwards as cells leave the cell cycle in more central locations. RPE cells thus complete their last cell division in a centre-to-periphery wave centred about the optic nerve head. At any given retinal location, RPE cells complete their last cell division earlier than the overlying layers of the neural retina. Cells of the RPE rapidly develop a mature morphology. For example, melanin granules are observed at P5 and Verhoeff's membrane (the terminal bar complex) is evident by P25. By contrast, photoreceptor development in this species is protracted; cone inner segments are observed by P40, whilst the first rod inner segments are observed at P60. Despite being generated earlier, morphological maturation of the cones appears retarded and prolonged compared with that of the rods. The last stages of RPE cell maturation occur late in development, in synchrony with the generation of rods.

Temporal and spatial pattern of MASH-1 expression in the developing rat retina demonstrates progenitor cell heterogeneity

The temporal and spatial pattern of mammalian achaete-scute homolog 1 (MASH-1) expression in the developing rat retina was examined in an effort to correlate achaete-scute homolog expression with the generation of particular cell classes. The expression of MASH-1 was restricted to the latter portion of retinal neurogenesis and was most closely correlated with the appearance of bipolar cells and Müller glia, two cell classes that are generated late in retinogenesis. We also examined the proliferative nature of the MASH-1 -expressing cell type to confirm that MASH-1 is expressed by progenitor cells and to determine the proportion of the proliferating population that expresses MASH-1. MASH-1 was expressed by only 10-30% of the total proliferating population, depending on the age examined. Thus, MASH-1 expression provides a molecular marker of heterogeneity among retinal progenitor cells and may play a role in the commitment and/or differentiation of one or more of the late-appearing retinal phenotypes.

Normal and pathological mechanisms in retinal vascular development

Angiogenesis is a complex biologic process that occurs normally in development and in turnover and remodeling of mature vascular networks. Pathological angiogenesis and neovascularization occur in association with retinal and ocular ischemic diseases, in retinopathy of prematurity and other developmental disorders, and in tumor growth and metastasis. We describe current understanding of cellular and molecular mechanisms of retinal vascular development, highlighting aspects that relate to eye diseases, that provide sites of therapeutic intervention in ophthalmology and that are potential avenues for research.

Numerical relationship between neurons in the lateral geniculate nucleus and primary visual cortex in macaque monkeys

We examined the numerical correlation between total populations of neurons in the lateral geniculate nucleus (LGN) and the primary visual cortex (area 17 of Brodmann) in ten cerebral hemispheres of five normal rhesus monkeys using an unbiased three-dimensional counting method. There were 1.4 +/- 0.2 million and 341 +/- 54 million neurons in the LGN and area 17, respectively. In each animal, a larger LGN on one side was in register with a larger area 17 of the cortex on the same side. Furthermore, asymmetry in the number of neurons in both the LGN and area 17 favored the right side. However, because of small variations across subjects, correlation between the total neuron number in LGN and area 17 was weak (r2 = 0.29). These results suggest that the final numbers of neurons in these visual centers may be established independently or by multiple factors controlling elimination of initially overproduced neurons.

Comparison of photoreceptor spatial density and ganglion cell morphology in the retina of human, macaque monkey, cat, and the marmoset Callithrix jacchus

We studied the relationship between the morphology of ganglion cells and the spatial density of photoreceptors in the retina of two Old World primates, human and macaque monkey; the diurnal New World marmoset Callithrix jacchus; and the cat. Ganglion cells in macaque and marmoset were labelled by intracellular injection with Neurobiotin or by DiI diffusion labelling in fixed tissue. Cone photoreceptor densities were measured from the same retinas. Supplemental data for macaque and data for human and cat were taken from published studies. For the primates studied, the central retina is characterised by a constant numerical convergence of cones to ganglion cells. Midget ganglion cells derive their input, via a midget bipolar cell, from a single cone. Parasol cells derive their input from 40-140 cones. Outside the central retina, the convergence increases with eccentricity. The convergence to beta cells in the cat retina is very close to that for parasol cells in primate retina. The convergence of rod photoreceptors to ganglion cells is similar in human, macaque, and marmoset, with parasol cells receiving input from 10-15 times more rods than midget cells. The low convergence of cones to midget cells in human and macaque retinas is associated with distinctive dendritic "clusters" in midget cells' dendritic fields. Convergence in marmoset is higher, and the clusters are absent. We conclude that the complementary changes in photoreceptor density and ganglion cell morphology should be considered when forming linking hypotheses between dendritic field, receptive field, and psychophysical properties of primate vision.

Topographical regulation of cone and rod opsin genes: parallel, position dependent levels of transcription

RNase protection assays were used to follow rhodopsin and red cone opsin mRNA levels during bovine fetal development as a function of retinal position. Following induction, an equivalent radial gradient of rod and cone opsin mRNA is present in the fetal retina. This gradient is maintained in the adult retina even though no corresponding gradient in rod or cone cell density is present. Since the mRNA expression gradient does not progress radially, position dependent levels of photoreceptor-specific transcription is suggested.

Quantitative analysis of synaptogenesis in the inner plexiform layer of macaque monkey fovea

Synaptogenesis has been tracked by using quantitative electron microscopic methods in the inner plexiform layer (IPL) of the developing Macaca monkey fovea from fetal day (Fd) 55 to Fd132. Vesicle-containing profiles were classified according to whether (1) they contained a ribbon indicating that they originated from a bipolar cell, or (2) the profile formed a junction. Group 2 was further subdivided by morphological characteristics into (2a) amacrine, (2b) bipolar, or (2c) unknown profiles. Ribbon-containing bipolar profiles are clearly identifiable at Fd55 when they occur at a density of 0.9/100 microns2. Bipolar synapses increase rapidly to 4.7/100 microns2 by Fd88, similar to their density at Fd132. Identifiable amacrine profiles forming a junction are rare at Fd55-68. By Fd88, amacrine synaptic density has jumped to 6.7/100 microns2 and continues to increase to 9.5/100 microns2 at Fd132. These quantitative data strongly suggest that, at the Macaca fovea, bipolar synaptogenesis both begins and ends before amacrine synaptogenesis. The large number of immature amacrine synaptic profiles and densities at Fd132 suggests that amacrine synapses continue to form after Fd132. This study confirms that cone-dominated monkey fovea has a different sequence of synaptogenesis than the rod-dominated peripheral retina (Nishimura and Rakic, [1985] J. Comp. Neurol 241:420-434). The data support the concept that synaptic developmental sequence is determined by the type of photoreceptor which dominates a particular retinal region or species. Bipolar ribbon synapses are observed in the outer half of the IPL at Fd55, are present in the inner IPL at Fd60, and then, with increasing age, are found throughout the IPL. This pattern strongly suggests that vertical OFF bipolar pathways form earlier than ON pathways in the IPL. In contrast, amacrine profiles are found throughout the IPL at the youngest ages, with an adult-like banding pattern present by Fd132.

Synapses between cones and diffuse bipolar cells of a primate retina

The photoreceptor synapses of three representative cells of the six types of diffuse bipolar cell of the rhesus macaque monkey's retina are described at 3.5-4.0 mm eccentricity. Bipolar cell DB3 was found to be postsynaptic to 11 cones at 155 basal synapses; about 70% of these were triad associated. Bipolar cell DB4 as postsynaptic to eight cones at 52 ribbon synapses; in addition it was found also to make an average of two or three basal (non-ribbon) synapses per cone (total 23). The DB5 bipolar cell type had 57 invaginating synapses with seven cones. It too had basal synapses, but only two with each of three cones. The diffuse invaginating bipolar cell described by Mariani (1981) is identified as a member of the DB5 category. Dendrites of cone bipolar cell types which have axons ending in the a-layer of the inner plexiform layer make only basal synapses with the cone pedicle. Those so far investigated are the flat midget bipolar cell and the DB2 and DB3 flat diffuse bipolar cells. All bipolar cells whose axons terminate in the b-layer of the inner plexiform layer are postsynaptic at the ribbon synapses of the cone pedicles. They now appear to fall into two groups. Those whose dendrites are exclusively postsynaptic at the ribbons; these are the blue cone and invaginating midget bipolar cells. And the diffuse bipolar cell DB4, that has both ribbon and basal synapses in a ratio of about 2.3:1. It is uncertain into which category cell DB5 should be placed; its basal synapses are so few the cell could be anomalous. It now seems that at least one primate bipolar cell type may be like those of other vertebrates in having, as defined ultrastructurally, two different kinds of synaptic connection with its cones. The results are discussed in the context of a brief review of the photoreceptor synapses of other mammalian bipolar cells.

Ontogeny of the opioid growth factor, [Met5]-enkephalin, and its binding activity in the rat retina

The endogenous opioid peptide [Met5]-enkephalin is a tonically active opioid growth factor (OGF) with an inhibitory action on DNA synthesis in the developing rat retina. In this study, the ontogeny of the spatial and temporal expression of OGF and its binding activity was examined. OGF-like immunoreactivity was detected in the retina at gestation day (E) 20, but not at E18, and was localized to ganglion cell and neuroblast layers; immunochemical reaction was no longer seen in the retina by postnatal day 6. Native OGF was further identified and characterized by high-performance liquid chromatography (HPLC) studies and immunodot assays, which revealed that [Met5]-enkephalin was present in the neonatal, but not adult, rat retina. OGF binding activity was detected as early as E18 using [125I]-[Met5]-enkephalin and in vitro receptor autoradiography. Little OGF binding activity was noted for prenatal retinas, but appreciable activity was observed from birth to postnatal day 4; no OGF binding could be detected after postnatal day 5 or in the adult. These results reveal the transient appearance of the OGF, [Met5]-enkephalin, and its receptor binding activity in the developing mammalian retina, and show that their ontogeny coincides with the timetable of DNA synthesis of retinal neuroblasts.

Involvement of Notch-1 in mammalian retinal neurogenesis: association of Notch-1 activity with both immature and terminally differentiated cells

The Notch pathway is thought to define an evolutionarily conserved signaling mechanism that regulates the differentiation of immature cells through cells interactions. We have examined the expression of the Notch-1 receptor, the central element of this pathway, in the developing rat retina, where cell-fate choices depend upon a series of local cell interactions. Notch-1 immunoreactivity is associated with differentiating cells at different stages of retinal neurogenesis, suggesting that Notch-1 may play a role in the successive cell-fate determination which governs retinal development. In addition, the Notch-1 immunoreactivity is detected in nuclei of postmitotic, differentiated neurons of the adult retina. Our observations raise the possibility that besides its role in the differentiation of immature cell populations Notch-1 activity may also be involved in the maintenance of the differentiated state.

Shifting relationships between photoreceptors and pigment epithelial cells in monkey retina: implications for the development of retinal topography

This study examines the spatiotemporal relationships between retinal pigment epithelium (RPE) and photoreceptors (PR) during development of Macaca nemestrina retina. Our aim was to learn more about the developmental dynamics of these two important cell populations, particularly whether development changes in RPE cell densities mimic those of PR at selected retinal points. Twelve eyes ranging in age from 100 fetal days (Fd) to adulthood were flatmounted; the retinal perimeters were traced; and then sample punches were taken of the RPE and neural retina at the fovea, optic disc, mid- and far-nasal periphery, and far temporal, inferior and superior periphery. The two tissues were gently separated and the RPE cells and photoreceptors from the same region of the punch were counted using Nomarski contrast interference optics. We found that the total number of cones remains stable around 4 million between Fd100 and adulthood, but RPE number increases from 1.6 million at Fd100 to 2.56 million in adulthood. At the fovea, the core:RPE ratio increases from 5.4:1 at Fd100 to 28:1 by adulthood. In the temporal periphery by contrast, the cone:RPE ratio declines from 2.2:1 at Fd100-110 to less than 1:1 in the adult. In the vicinity of the optic disc, the ratio of (cones+rods); RPE remains around 35:1 throughout development, but in the retinal periphery it decreases to the adult value of 22:1. These changing ratios indicate that photoreceptors and RPE cells are redistributed independently during development, and that these two cellular sheets slide over one another to achieve their final distribution. This situation suggests that the forces or factors causing foveation are intrinsic to the neural retina.

Linked regularities in the development and evolution of mammalian brains

Analysis of data collected on 131 species of primates, bats, and insectivores showed that the sizes of brain components, from medulla to forebrain, are highly predictable from absolute brain size by a nonlinear function. The order of neurogenesis was found to be highly conserved across a wide range of mammals and to correlate with the relative enlargement of structures as brain size increases, with disproportionately large growth occurring in late-generated structures. Because the order of neurogenesis is conserved, the most likely brain alteration resulting from selection for any behavioral ability may be a coordinated enlargement of the entire nonolfactory brain.

Origin of microglia in the quail retina: central-to-peripheral and vitreal-to-scleral migration of microglial precursors during development

The origin, migration, and differentiation of microglial precursors in the avascular quail retina during embryonic and posthatching development were examined in this study. Microglial precursors and developing microglia were immunocytochemically labeled with QH1 antibody in retinal whole mounts and sections. The retina was free of QH1+ macrophages at embryonic day 5 (E5). Ameboid QH1+ macrophages from the pecten entered the retina from E7 on. These macrophages spread from central to peripheral areas in the retina by migrating on the endfeet of the Müller cells and reached the periphery of the retina at E12. While earlier macrophages were migrating along the inner limiting membrane, other macrophages continued to enter the retina from the pecten until hatching (E16). From E9 on, macrophages were seen to colonize progressively more scleral retinal layers as development advanced. Macrophages first appeared in the ganglion cell layer at E9, in the inner plexiform layer at E12, and in the outer plexiform layer at E14. Therefore, it seems that macrophages first migrated tangentially along the inner retinal surface and then migrated from vitreal to scleral levels to gain access to the plexiform layers, where they differentiated into ramified microglia. Macrophages appeared to differentiate shortly after arrival in the plexiform layers, as poorly ramified QH1+ cells were seen as early as E12 in the inner plexiform layer and at E14 in the outer plexiform layer. Radial migration of macrophages toward the outer plexiform layer continued until posthatching day 3, after which retinal microglia showed an adult distribution pattern. We also observed numerous vitreal macrophages intimately adhered to the surface of the pecten during embryonic development, when macrophages migrated into the retina. These vitreal macrophages were not seen from hatching onwards, when no further macrophages entered the retina.

Distinct temporal patterns of expression of sodium channel-like immunoreactivity during the prenatal development of the monkey and cat retina

Polyclonal and monoclonal antibodies prepared against the alpha-subunit of the voltage-gated sodium channel (alpha NaCh) were used to examine the distribution of sodium channel-like immunoreactivity during the prenatal development of the cat and rhesus monkey (Macaca mulatta) retina. At all prenatal ages studied, beginning on embryonic day 29 (E29) in the cat and E52 in the monkey, both antibodies labelled optic axons. With the polyclonal antibodies, the appearance of positive cells largely mirrored the onset of their morphological maturation. Immunoreactivity appeared first in the somata of ganglion cells, and subsequently the inner plexiform layer could be distinguished by its intense immunolabelling. A few weeks later horizontal cells displayed immunolabelling that extended to their dendrites in the developing outer plexiform layer. This was followed by immunoreactive cones, with bipolar cells labelled only postnatally. By contrast, with the monoclonal antibody some cells were found to be immunoreactive while their somata were still in the ventricular layer (E33 in cat and E52 in monkey). Many of these cells appeared to migrate to the outer portion of the prospective inner nuclear layer, where they gradually acquired the morphological appearance of bipolar cells. Transient expression of immunolabelling with monoclonal sodium channel antibody was found in the cones of the cat and cones and rods of the monkey. These results indicate that different types of alpha NaCh-like proteins are expressed in the mammalian retina at distinct developmental periods. Their presence at very early stages during development suggests that these proteins could play a specific role in the commitment and/or differentiation of specific retinal cell types.

Early appearance and transient expression of putative amino acid neurotransmitters and related molecules in the developing rabbit retina: an immunocytochemical study

We have studied, by immunocytochemistry, the ontogeny of GABA, glycine, glutamate, glutamine, and taurine-containing cells in the rabbit retina. Amacrine cells show GABA immunoreactivity by embryonic day 25 (E25) and throughout postnatal life. By contrast, ganglion cells and horizontal cells are only transiently GABA-immunoreactive (-IR); few appear GABA-IR by the third postnatal week. At maturity, glycine is present in amacrine cells and in some bipolar cells. During development, putative ganglion cells transiently contained glycine between E25 and postnatal day 3 (P3), whereas immunolabelling in presumed amacrine cells and bipolar cells persists after birth. Ganglion cells, bipolar cells, photoreceptors, and some amacrine cells are glutamate-IR in the adult retina. Glutamate immunoreactivity first appears in the somata and processes of cytoblastic cells by E20 and is prominent by E25. Surprisingly, ganglion cells are not strongly glutamate-IR until just before eye-opening, at postnatal day 10 (P10), coincident with the appearance of glutamine in their somata and in Müller glial cells. Bipolar cells are glutamate-IR before they or Müller cells contain high levels of glutamine (at P10). Glutamate immunoreactivity in photoreceptors is progressively restricted to the inner segments by eye-opening. At no stage are presumed horizontal cells glutamate-IR or glutamine-IR, but some amacrine cells show glutamate- and glutamine-IR by P10. Taurine is localized to photoreceptors and Müller glial in the adult retina. Some cytoblasts are taurine-IR at E20; with ensuing development, taurine labelling becomes restricted primarily to Müller cells and photoreceptors; some putative bipolar cells may also be labelled. However, for a few days around birth, cells resembling horizontal cells, also show taurine immunoreactivity. The early appearance and often transient expression of these amino acids in retinal cells suggests that these neuroactive molecules may be involved in the structural and functional development of the retina.

Horizontal cells in the rabbit retina: differentiation of subtypes at neonatal and postnatal stages

We are investigating the differentiation of the major subtypes of horizontal cell in the rabbit retina in order to learn more about developmental controls responsible for the variety of neuronal phenotypes. Immunohistochemistry with anti-neurofilament and anti-calbindin-D antibodies, followed by epoxy resin embedding, has facilitated study of these neurons. In the mature rabbit retina, axonless (A-type) horizontal cells reacted strongly in procedures using either antibody; short axon (B-type) somas did not show a reaction with anti-neurofilament antibodies and stained moderately using anti-calbindin antibodies. In the immature neonatal retina the somas of all the horizontal cells seemed to be similar with regard to general morphology, but two populations could be distinguished on the basis of immunostaining. Some, identified as A-type horizontal cells (by comparison with mature retina), were stained using either antibody. Interspersed among these were similar cells with no detectable immunoreactivity, identified as B-type horizontal cells. By the end of the first postnatal week, faint reactivity to anti-calbindin-D was present in the somas of B-type horizontal cells; they stained moderately throughout the rest of the period studied. Thus differences in immunostaining indicate that the two horizontal cell subpopulations are established early in the rabbit, though some other distinguishing characteristics emerge only gradually as the retina matures. These results suggest that in mammals the determination of phenotypic subtype occurs early, possibly at the time that the cell is specified as a horizontal neuron, or shortly thereafter.

Radial asymmetry in the topography of retinoblastoma. Clues to the cell of origin

Retinoblastoma is a malignancy of the human developing retina. In situ as well as in vitro studies have attributed tumoral histogenesis either to a primitive retinoblast with neuronal and glial differentiation potentials, or to a photosensory progenitor cell. Here it is shown in vivo that the retinal topography of 457 retinoblastoma and retinoma foci is radially asymmetrical. Tumor density appears to mimic the horizontal visual streak characteristic of red/green cone cell distribution. Such a non-random distribution seems to invalidate the hypothesis of a primitive multipotential neuroblast as the unique source of retinoblastoma and may support the view that retinoblastoma evolves along the cone cell lineage.

Retinoic acid promotes differentiation of photoreceptors in vitro

The results of several recent studies have demonstrated that cell commitment and differentiation in the developing vertebrate retina are influenced by cell-cell interactions within the microenvironment. Retinoic acid has been shown to influence cell fates during development of the nervous system, and retinoic acid has been detected in the embryonic retina. To determine whether retinoic acid mediates the differentiation of specific neuronal phenotypes during retinal histogenesis, we treated dissociated cell cultures of embryonic and neonatal rat retina with varying concentrations of all-trans or 9-cis retinoic acid and analyzed the effects on cell fate using neuron and photoreceptor-specific antibodies. Addition of exogenous retinoic acid caused a dose-dependent, specific increase in the number of cells that developed as photoreceptors in culture throughout the period of retinal neurogenesis. In the same cultures, retinoic acid also caused a dose-dependent decrease in the number of cells that developed as amacrine cells. Also, results of double-labeled immunohistochemical studies using bromodeoxyuridine demonstrated that the primary effect of retinoic acid was to influence progenitor cells to develop as newly generated rod photoreceptors. Since retinoic acid and at least one of the retinoic acid receptors (RAR alpha) have been localized to the developing neural retina, these results suggest that retinoic acid may play a role in the normal development of photoreceptor cells in vivo.

An array of early differentiating cones precedes the emergence of the photoreceptor mosaic in the fetal monkey retina

We previously have demonstrated that approximately 10% of cones in the fetal monkey retina precociously express the red/green opsin. These data suggested the possibility that a subset of cones differentiates prior to their nascent cone neighbors. To further assess this early cone differentiation in the fetal monkey retina, we used monoclonal antibodies proven to be important developmental markers of photoreceptor phenotypes and synaptogenesis (XAP-1, specific to photoreceptor membranes; SV2, specific to synaptic vesicle protein). Although these two antibodies recognize functionally distinct antigens, our analyses revealed that both identify a subset of precociously immunoreactive cones. Further, XAP-1- and SV2-positive cones are distributed in the same pattern as precocious red/green-sensitive cones in immature regions of the fetal monkey retina. These results support the hypothesis that the primate retina possesses a spatially organized protomap that may induce the emergence of the photoreceptor mosaic and trigger the formation of color-specific pathways that include horizontal, bipolar, and retinal ganglion cells.

Development of the rabbit retina. V. The question of 'columnar units'

A qualitative and quantitative description of the columnar units in the mammalian retina, and a discussion of their ontogeny and putative functions is given. Columnar arrangements of cells exist in the developing retina which can be observed by means of scanning electron microscopy. In the adult retina, each Müller cell ensheaths a columnar group of neuronal cells. Counting the number of cells in radial H/E stained sections at various developmental stages reveals a constant ratio of neuronal cells per Müller cell, independent of the developmental stage (after postnatal day 9), and independent of the retinal topography. Such groups of cells always consist of one Müller cell, 11 rod photoreceptor cells, about 2 bipolar cells, and 1 to 2 amacrine cells. Retinal ganglion cells, cone photoreceptor cells, and horizontal cells are more sparsely distributed in the retina than these units; since they are known to arise earlier in the ontogenesis than other cell types they are considered to exist independently of the columnar units. It is suggested that the units arise by migration of groups of preneurons along a common Müller (precursor) cell; these preneurons and the corresponding Müller cell may be clonally related. In the adult retina, such columns might constitute metabolic and functional units.

Organization of pioneer retinal axons within the optic tract of the rhesus monkey

Retinal ganglion cell axons must make a decision at the embryonic optic chiasm to grow into the appropriate optic tract. To gain insight into the cues that play a role in sorting out the crossed from the uncrossed optic axons, we investigated the sequence of their initial ingrowth in rhesus monkey embryos. Two carbocyanine dyes, 1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate and 4-(4-dihexadecylaminostyryl)-N-methylpyridinium iodide, were placed, respectively, into the left and right retinas to identify the course of uncrossed and crossed retinal axons through the optic chiasm and tract. Our results show that at embryonic day 36 the most advanced retinal projections are uncrossed. At this age the leading crossed axons are just reaching the chiasmatic midline, whereas the uncrossed fibers have already entered the optic tract. This indicates that the pathfinding of these pioneer uncrossed fibers does not require the presence of retinal axons from the opposite eye. At subsequent stages of development (embryonic days 40 and 42) there is a clear partial segregation of the uncrossed and crossed retinal axons within the optic tract: the uncrossed-component course is in the deeper portion of the optic tract, whereas the crossed component lies in a more superficial region. Thus, the spatial organization of retinal axons within the primordial optic tract reflects the sequential addition of the uncrossed and crossed retinal fibers. The orderly and sequential ingrowth of these pioneer retinal axons indicates that specific chiasmatic cues are expressed early in development and that such pioneer fibers may serve as guides for the later-arriving retinal fibers.

Birthdates of neurons in the retinal ganglion cell layer of the ferret

The present study determined the temporal and spatial patterns of genesis for neurons of different sizes in the retinal ganglion cell layer of the ferret. Fetal ferrets were exposed to tritiated thymidine on embryonic days E-22 through E-36. One to 3 months after birth, they were perfused and their retinae dissected, and autoradiographs were prepared from resin-embedded sections throughout the entire flattened retinal ganglion cell layer. Soma size differences in conjunction with separate retrograde labeling and calbindin immunocytochemical studies were used as criteria for identifying different retinal ganglion cell subtypes in juvenile and adult ferrets. Neurons of different sizes in the ganglion cell layer were generated at different stages during development. Medium sized cells were generated primarily between E-22 and E-26; the largest cells were generated between E-24 and E-29; small cells were generated between E-26 and E-32; and very small cells were generated between E-29 and E-36. The former three groups were interpreted to be three subtypes of retinal ganglion cells, while the latter group was interpreted to be displaced amacrine cells. This temporal order of the genesis of ganglion cell classes is consistent with the spatial ordering of their fibers in the mature optic chiasm and tract, and it is consistent with the developmental change in decussation pattern recently shown in the optic pathway of embryonic ferrets. The spatial pattern of genesis suggests that ganglion cells of a particular class are added to the ganglion cell layer in a centroperipheral fashion initiated in the dorsocentral retina nasal to the area centralis. No evidence was found for a wave of ganglion cell addition that proceeded in a spiralling pattern around the area centralis, as has been reported in the cat.

Light and electron microscopic analysis of synaptic development in Macaca monkey retina as detected by immunocytochemical labeling for the synaptic vesicle protein, SV2

The development of synapses has been followed in Macaca monkey fetal and infant retina using immunocytochemical labeling for the transmembrane synaptic vesicle glycoprotein, SV2. Electron microscopy (EM) was used to verify the presence of morphological synapses at selected ages. EM immunocytochemical labeling in adult retina showed that all synaptic types contained SV2 in inner (IPL) and outer (OPL) plexiform layers. In fetal retina, SV2 expression and the appearance of morphological synapses were closely related in time, demonstrating that SV2 is a reliable marker for synaptogenesis. SV2 expression appears along a foveal to peripheral gradient. Both SV2 and synapses appear in the foveal IPL at Fd50-55, and reach the retinal edge by Fd90-103. Cone ribbon synapses and SV2 labeling are not present in the foveal OPL until Fd60. Photoreceptors in the far periphery contain SV2 by Fd119-125. This pattern demonstrates an "inner to outer" direction of synaptogenesis. Cones show SV2 labeling before rods at the same retinal eccentricity. In the cone-dominated fovea, SV2 labeling and bipolar cell ribbon-containing terminals are present at Fd55 when amacrine cell conventional terminals are very scarce, indicating that bipolar synapses precede amacrine synapses in monkey foveal IPL. SV2 labeling and bipolar terminals appear first in the outer IPL which contains "OFF" ganglion and bipolar processes in the adult, suggesting that "OFF" midget bipolar cells may form the first synapses. Both SV2 immunocytochemical labeling and EM morphology find that monkey retina follows a generalized inner before outer, and cone before rod synaptic developmental pattern, similar to that in other mammals. The cone-dominated fovea initiates synaptogenesis, and shows a different synaptic sequence from rod-dominated peripheral retina.

Central retinal area is not the site where ganglion cells are generated first

The development of retinal ganglion cells (RGC) was studied in the chick from stage 18 to adulthood. Our main objectives were to identify the retinal site where the first RGCs differentiate, to locate this site relative to the optically defined central retinal area, and to map the spatial arrangement of the RGC field at different stages in development. The eyes of the experimental animals were fixed and serially sectioned. The borders of RGC fields were determined from the presence of either ganglion cell perikarya or ganglion cell axons. In seven cases between stages 21 and 26, the borders of the RGC fields were confirmed electron microscopically. The serial sections together with the RGC fields were then reconstructed in three dimensions. The reconstructed retinae were projected onto a plane by using the radially equidistant polar azimuthal projection. First, RGCs appear dorsal to the apex of the optic fissure. Ganglion cell development then initially spreads out symmetrically with respect to the optic fissure. However, from stage 29 on, the nasal half of the retina expands much more than the temporal half. This asymmetrical growth entails that the optic fissure is eventually located in the temporal half of the retina in the mature animal. The RGC fields of the embryonic stages were superimposed on the retina of a visually active animal according to their real size and position. It turned out that the central retinal area was at least 2 mm away from the site where the first RGCs were generated. It is not before stage 28 that the prospective central retinal area is included into the expanding ganglion cell field. The fact that RGCs at the central retinal area are generated 2.5 days later than first RGCs near the apex of the optic fissure has important implications for the formation of the retinotectal projection.

Early axon outgrowth of retinal ganglion cells in the fetal rhesus macaque

Employing retinal explants and retrograde transport techniques, we studied the formation of the arcuate fascicles by examining the growth of the central retina, the emergence of the adult fiber layer pattern, and the projections of retinal ganglion cells in the central and peripheral retina. Sixty days prior to foveal pit formation, the distance from the incipient fovea to the optic disk was equal to the adult, even though the retinal area was only 8% of the adult. Arcuate fibers, at this age, were observed to avoid the incipient fovea, with no fascicles and few axons projecting over this region. A small population of 15.2% of the ganglion cells located within 2 mm of the incipient fovea possessed an axon with an aberrant trajectory that wound around and projected 50 to several hundred microns away from the optic disk, compared to only 3% at other retinal locations. The incidence of disorder decreased with increasing fetal age, establishing mature values in late fetal periods. These findings suggest that the area of the central retina does not increase after embryonic day 60 and that guidance factors are present that allow outgrowing ganglion cell axons to distinguish and avoid that portion of the retina that will become the fovea.

What do retinal müller (glial) cells do for their neuronal 'small siblings'?

Müller (radial glial) cells are the predominant glia of the vertebrate retina. They arise, together with rod photoreceptor cells, bipolar cells, and a subset of amacrine cells, from common precursor cells during a late proliferative phase. One Müller cell and a species-specific number of such neurons seem to form a columnar unit within the retinal tissue. In contrast, 'extracolumnar neurons' (ganglion cells, cone photoreceptor cells, horizontal cells, and another subset of amacrine cells) are born and start differentiation before most Müller cells are generated. It may be essential for such neurons to develop metabolic capacities sufficient to support their own survival, whereas late-born ('columnar') neurons seem to depend on a nursing function of their 'sisterly' Müller cell. Thus, out of the cell types within a retinal column it is exclusively the Müller cell that possesses the enzymes for glycogen metabolism. We present evidence that Müller cells express functional insulin receptors. Furthermore, isolated Müller cells rapidly hydrolyse glycogen when they are exposed to an elevated extracellular K+ ion concentration, a signal that is involved in the regulation of neuronal-glial metabolic cooperation in the brain. Müller cells are also thought to be essential for rapid and effective retinal K+ homeostasis. We present patch-clamp measurements on Müller cells of various vertebrate species that all demonstrate inwardly rectifying K+ channels; this type of channel is well-suited to mediate spatial buffering currents. A mathematical model is presented that allows estimation of Müller cell-mediated K+ currents. A simulation analysis shows that these currents greatly limit lateral spread of excitation beyond the borders of light-stimulated retinal columns, and thus help to maintain visual acuity.

Morphogenesis of retinal ganglion cells during formation of the fovea in the Rhesus macaque

The morphology of retinal ganglion cells within the central retina during formation of the fovea was examined in retinal explants with horseradish-peroxidase histochemistry. A foveal depression was first apparent in retinal wholemounts at embryonic day 112 (E112; gestational term is approximately 165 days). At earlier fetal ages, the site of the future fovea was identified by several criteria that included peak density of ganglion cells, lack of blood vessels in the inner retinal layers, arcuate fiber bundles, and the absence of rod outer segments in the photoreceptor layer. Prior to E112, the terminal dendritic arbor of retinal ganglion cells within the central retina extended into the inner plexiform layer and were located directly beneath their somas of origin or at most were slightly displaced from it. For example, at E90 the mean horizontal displacement of the geometric center of the dendritic arbor from the somas of cells within 600 microns of the estimated center of the future fovea was 4.1 microns (S.D. 2.7, range 1.0-10.0, n = 97). Following formation of the foveal depression the dendritic arbors of cells were significantly displaced from their somas. For example, at E138 the mean displacement was 41.2 microns (S.D. 12.2, range 12.0-56.0, n = 97). The displacement of the dendritic arbor which occurred during this period was not accounted for by areal growth of the dendritic arbor, the somas, or the retina, but was produced by the lengthening of the primary dendritic trunk. Moreover, no significant displacement was observed within the remaining 1.5-6.5 mm of the central retina. These observations provide evidence supporting early speculations that the formation of the foveal pit occurs, in part, by the radial migration of ganglion cells from the center of the fovea during its formation. Our analyses suggest that this migration occurs by the lengthening of the primary dendrite presumably by the addition of membrane. This migration is in a direction opposite to the inward movement of photoreceptors that occurs during late fetal and early postnatal periods (Packer et al., 1990, Journal of Comparative Neurology 298, 472-493).

Biphasic retinal neurogenesis in the brush-tailed possum, Trichosurus vulpecula: further evidence for the mechanisms involved in formation of ganglion cell density gradients

We investigated cell generation in the retina of the brush-tailed possum (Trichosurus vulpecula) by using tritiated (3H)-thymidine labelling of newly generated cells. Animals aged between postnatal day (P) 5 and 85 each received a single injection of 3H-thymidine. Following autoradiographic processing, maps of labelled cells were constructed from retinal sections. Retinal cell generation takes place in two phases, the first is concluding in the retinal periphery at P53 as the second is seen to commence in midtemporal retina. In the first phase, cells in central retina are generated earlier than those in peripheral regions. In the second phase, cells complete their final division in midtemporal retina first and in the periphery last. Cells generated in the first phase comprise virtually all cells in the ganglion cell layer, amacrine cells, horizontal cells, and cones. Ganglion cells are produced at a slightly earlier stage than displaced amacrine cells, horizontal cells, or cones. Amacrine cells in the inner nuclear layer are the final cells produced in the first phase. When ganglion cells and amacrine cells are pooled, their combined rate of production matches that of the other cell types. These data indicate that the ratio of displaced amacrine cells: horizontal cells: cones: combined ganglion cells and amacrine cells does not change throughout development. However, the ratio of ganglion cells:macrines changes steadily as development proceeds to favour amacrine cells. In the second phase, sparse numbers of nonganglion cells in the ganglion cell layer and large numbers of bipolar and Müller cells are produced along with all rods. The two phases in the possum are similar to those seen in the wallaby, the quokka. However, fewer cells are added in central retina in the possum than in the quokka and cell addition continues for a more extended period in the periphery in the possum. We suggest that this difference in cell addition could account for the development of a more pronounced visual streak of retinal ganglion cells in the possum than in the quokka. A comparison of possum retinal cell generation with that of other marsupials adds support for the "homochrony theory."

Synaptic connections of the narrow-field, bistratified rod amacrine cell (AII) in the rabbit retina

The synaptic connections of the narrow-field, bistratified rod amacrine cell (AII) in the inner plexiform layer (IPL) of the rabbit retina were reconstructed from electron micrographs of continuous series of thin sections. The AII amacrine cell receives a large synaptic input from the axonal endings of rod bipolar cells in the most vitreal region of the IPL (sublamina b, S5) and a smaller input from axonal endings of cone bipolar cells in the scleral region of the IPL (sublamina a, S1-S2). Amacrine input, localized at multiple levels in the IPL, equals the total number of synapses received from bipolar cells. The axonal endings of cone bipolar cells represent the major target for the chemical output of the AII amacrine cell: these synapses are established by the lobular appendages in sublamina a (S1-S2). Ganglion cell dendrites represent only 4% of the output of the AII amacrine and most of them are also postsynaptic to the cone bipolars which receive AII input. The AII amacrine is not presynaptic to other amacrine cells. Finally, the AII amacrine makes gap junctions with the axonal arborizations of cone bipolars that stratify in sublamina b (S3-S4) as well as with other AII amacrine cells in S5. Therefore, in the rabbit retina 1) the rod pathway consists of five neurons arranged in series: rod-->rod bipolar-->AII amacrine-->cone bipolar-->ganglion cell; 2) it seems unlikely that a class of ganglion cells exists that is exclusively devoted to scotopic functions. In ventral, midperipheral retina, about nine rod bipolar cells converge onto a single AII amacrine, but one of them establishes a much higher proportion of synaptic contacts than the rest. Conversely, each rod bipolar cell diverges onto four AII amacrine cells, but one of them receives the largest fraction of synapses. Thus, within the pattern of convergence and divergence suggested by population studies, preferential synaptic pathways are established.

Lineage versus environment in embryonic retina: a revisionist perspective

The idea that microenvironmental cues act alone late in development to determine a cell's phenotype has dominated recent discussion of, retinal development, and has successfully displaced the notion of any role for cell lineage in the process of cell determination. We argue that there is, in fact, evidence favoring a degree of lineage restriction during the development of the vertebrate retina. We propose that environmental factors modulate a process of progressive lineage restriction. In this model, progenitor cells are viewed as having unequal potential, and their progeny are viewed as being committed to one of the major retinal cell classes before the stage at which they become postmitotic.

Cellular interactions determine neuronal phenotypes in rodent retinal cultures

Progenitor cells isolated from early rat embryo retinas differentiate into phenotypes normally generated early in retinal development (e.g., ganglion cells), whereas progenitors isolated from postnatal retinas differentiate into later-generated retinal cell types (e.g., rod photoreceptors; Reh and Kljavin, J. Neurosci. 9:4179-4189; 1989; Adler and Hatlee, 1989; Science 243:391-393; Sparrow, Hicks, and Barnstable, 1990, Dev. Brain Res. 51:69-84). To determine whether this change in committment is intrinsic to the progenitor cells, or alternatively can be modified by interactions with their developing environment, I co-cultured mouse and rat retinal cells, from different developmental stages, and identified the resulting phenotypes with species-specific and cell class-specific antibodies. I found that the phenotypes into which mouse neuroepithelial cells differentiate depends on the phenotypes of the rat cells that surround them. Retinal precursor cells from embryonic day (E) 10-12 will adopt the rod photoreceptor phenotype only when close to cells expressing this phenotype. By contrast, when the E10-12 retinal progenitor cells are cultured with cells from the cerebral cortex, they differentiate primarily into large multipolar neurons, similar in their morphology and antigen expression to retinal ganglion cells. These results indicate that interactions among the cells of the developing retina are important in the determination of cell fate.

Genesis of neurons in the retinal ganglion cell layer of the monkey

We have analyzed the genesis of various neuronal classes and subclasses in the ganglion cell layer of the primate retina. Neurons were classified according to their size and the time of their origin was determined by pulse labeling with 3H-thymidine administered to female monkeys 38 to 70 days pregnant. All offspring were sacrificed postnatally, and their retinas processed for autoradiography. The somata of cells in the retinal ganglion cell layer generated on embryonic day (E) 38 ranged from 9 to 14 microns in diameter. Between E40 and E56, the minimum soma diameter remained around 8-9 microns, while the maximum gradually increased to 22 microns. As a consequence, the means of the distributions of labeled cells also increased with age, from 11.8 microns diameter for cells generated on E38 to 14.6 microns diameter at E56. Over this period the percentage of labeled cells in the 10.5-16.5 microns and greater than 16.5 microns diameter range gradually increased. The proportion of the labeled cells in the less than 10.5 microns diameter range decreased from E38 to E45, but subsequently increased rapidly. At the end of neurogenesis in the retinal ganglion cell layer, around E70, most labeled cells were considerably smaller (7-9 microns) than those generated earlier. Our results indicate that within the ganglion cell layer of the macaque, neurons of small caliber are generated first, followed successively by medium sized cells. Large, putative P alpha cells are generated late. The production between E56 and E70 of cells with the smallest somata suggests that the last-generated neurons in the ganglion cell layer are predominantly displaced amacrine cells. Within the same sector of retina, different classes of neurons in the ganglion cell layer of the rhesus monkey appear to have a sequential schedule of production.

Two cellular inductions involved in photoreceptor determination in the Xenopus retina

Cellular determination in the Xenopus retina is not a strict consequence of cell lineage or cell birthdate. This suggests that a retinal cell gets its fate by either local cellular interactions, diffusible factors, or an indeterminate stochastic mechanism. We have performed an in vitro experiment in which cellular contact is controlled to test the first possibility directly. We use these experiments to demonstrate that two cellular inductions are involved in photoreceptor determination in vitro and that these inductions also occur during development in the retina in vivo. The first interaction is responsible for biasing cells toward either a generic photoreceptor or a cone fate, while the second directs cells toward a rod cell fate.

The development of parafoveal and mid-peripheral human retina

The morphological development of parafoveal retina (1-1.5 mm from the foveal center) and the mid-peripheral (4 mm from the foveal center) human retina has been studied from fetal (F) 26 weeks to adulthood. At both retinal points, all layers and neuronal types are present at F26 weeks. In parafovea at F26 weeks photoreceptors have only a rudimentary inner segment and no outer segments. Short outer segments are present on both rods and cones at F36 weeks. By postnatal (P) 5-8 days the inner retina is relatively mature. Photoreceptors have elongated basal axons which cause the photoreceptor layer to become much thicker than in prenatal retina. At birth cone inner segments are untapered, but rod inner segments have already reached their adult width of 2 microns. Both rod and cone inner and outer segments are 30-50% of adult length. By 13 months both inner and outer retina are mature appearing, with the photoreceptors accounting for half the retinal thickness due to the elongation of the fibers of Henle. Cone outer segments elongate up to P5 years and rod outer segments to P13 years. At mid-peripheral or rod-ring retina outer segments are present on rods at F26 weeks and on cones at F36 weeks. At birth the inner retina is adultlike. The outer plexiform layer becomes thicker up to P45 months due to the elongation of fibers of Henle. At birth both rod and cone mid-peripheral inner segments are slightly longer and outer segments are 50% longer than in parafoveal retina. By P5 years mid-peripheral rod outer segments are slightly longer than in parafoveal retina, and this changes little thereafter. This anatomical study has found that the photoreceptors in peripheral rod-ring retina develop earlier than those in more central retina, and in turn parafoveal photoreceptors develop well in advance of foveal cones. This suggests that human neonates may utilize more peripheral retinal regions for some aspects of visual function before foveal cone vision becomes dominant.

Development of catecholaminergic, indoleamine-accumulating and NADPH-diaphorase amacrine cells in rabbit retinae

We have investigated the ontogeny of four classes of amacrine cells in the rabbit retina. In particular, the distribution, number, soma diameter, dendritic field diameter, and pattern of dendritic stratification were studied in catecholaminergic (CA) and indoleamine-accumulating (IA) amacrines and in two classes of nicotinamide adenine dinucleotide phosphate (NADPH) diaphorase amacrine cells. The first CA and IA cells are observed on the 27th postconceptional day (27PCD) and the first NADPH-diaphorase cells on 28PCD. These first cells are concentrated in the central part of the visual streak, and at subsequent ages, cells in this part of the streak have larger somata and more mature dendritic fields than those elsewhere, supporting the notion that the peak density region is a developmentally advanced part of the retina. Throughout development, amacrine cells of all classes are concentrated in the visual streak, with their density reaching minima at the superior and inferior retinal margins. As their total number increases, the difference in cell density between the streak and the periphery decreases, presumably because proportionately more cells are added at the periphery. Their total number peaks around 42PCD, followed by a decline of 12-31% to adult values. Once the peak number of cells has been reached, the difference in cell density between the streak and periphery begins to increase. The rate of this increase is closely correlated with the increase in retinal area. This redistribution of amacrine cells, as well as a greater expansion of their dendritic fields in peripheral retina, is almost certainly the product of nonuniform retinal expansion.

Epi-polarization and incident light microscopy readily resolve an autoradiographic or heavy metal label from an obscuring background or second label

Difficulty encountered in resolving grains of exposed photographic emulsion in autoradiographs of the densely melanized retinal pigment epithelium was solved by using epi-polarized or incident light microscopy. The apparatus used included a metallurgical illuminator specifically designed for epi-polarization microscopy or, as a less expensive but only slightly less effective alternative, a modified fluorescence illuminator. The black melanin granules absorb incident light (as they do in vivo) while the silver grains reflect it producing a "darkfield-like" representation. Brightfield and darkfield-like images can be alternated easily and quickly, or both can be viewed simultaneously. Epi-polarization microscopy has wider application in resolving a reflective label over any opaque background staining or dark second label.

A morphological comparison of foveal development in man and monkey

The fovea can first be identified in both monkey and human retina at 26-30% gestation as a region containing all adult retinal layers and only cone photoreceptors. A shallow foveal pit and cone outer segments appear by 63-65% gestation in both species. Prenatal development continues rapidly in the monkey, so that by birth a single layer of inner retinal neurons are present in the fovea, cones are three cells deep, inner segments are elongated, and outer segments are up to 50% of inner segment length. In contrast, human fovea does not reach a similar stage until several months after birth. The fovea is adult-like in monkey at 12 weeks and in human at 11-15 months, although human will mature further up to four to five years. This study shows that human fovea is less mature at birth than monkey but develops rapidly in infancy, suggesting that it may be even more susceptible to postnatal environmental influences than the commonly-used monkey model.

Identity of cells produced by two stages of cytogenesis in the postnatal cat retina

Cytogenesis in the postnatal cat retina was studied with the aid of 3H-thymidine autoradiography to identify the cell classes generated. Cells proliferate in two stages, which are separate spatially and temporally. Previous studies have shown that during Stage 1, cytogenesis occurs at high density at the ventricular surface of the retina, whereas Stage 2 occurs at low density in the inner retinal layers. At the ages studied, the progeny of Stage 1 cytogenesis are distributed in an annulus toward the margin of the retina, and those of Stage 2 occur central to the annulus, indicating that Stage 2 follows Stage 1. Cell genesis in Stage 1 appears to cease by P16; genesis in Stage 2 persists until between P21 and P30. The same cell classes (amacrine cells, bipolar cells, Müller cells, and rod photoreceptors) are generated during both Stages 1 and 2, but there are significant changes in their proportions both within and between stages. The proportion of the Stage 1 mitoses that form bipolar cells increases from 31% at P0 to 62% at P14. A corresponding decrease is observed in the proportion of rods (from 60% at P0 to 32% at P14). The proportion of cells generated during Stage 2 that become rods increases from 39% at P0 to 70% at P21, whereas the proportion of bipolar cells decreases from 50% at P0 to 23% at P21. Müller cells form a relatively constant proportion (8 to 15%) of the cells generated during both Stage 1 and 2. Thus at the end of Stage 1, mostly bipolar cells are generated; at the end of Stage 2, mostly rods are generated.

The human retina has a cone-enriched rim

Video-enhanced imaging of retinal wholemounts reveals an abrupt change in the composition of the photoreceptor mosaic at the edge of the human retina. Cone densities rise threefold and rod densities fall tenfold in a 1-mm-wide peripheral band. Antibodies directed against cones confirm the identification of the major subtypes of photoreceptors within this peripheral band. The cone-enriched rim is most highly developed along the nasal retinal margin, an area where the extreme lateral periphery of the visual field is imaged. This rim of cones may function as part of a rapid-acting alert mechanism under conditions of moderate and bright illumination.