Early emergence of photoreceptor mosaicism in the primate retina revealed by a novel cone-specific monoclonal antibody

Memo1-Mediated Tiling of Radial Glial Cells Facilitates Cerebral Cortical Development

Polarized, non-overlapping, regularly spaced, tiled organization of radial glial cells (RGCs) serves as a framework to generate and organize cortical neuronal columns, layers, and circuitry. Here, we show that mediator of cell motility 1 (Memo1) is a critical determinant of radial glial tiling during neocortical development. Memo1 deletion or knockdown leads to hyperbranching of RGC basal processes and disrupted RGC tiling, resulting in aberrant radial unit assembly and neuronal layering. Memo1 regulates microtubule (MT) stability necessary for RGC tiling. Memo1 deficiency leads to disrupted MT minus-end CAMSAP2 distribution, initiation of aberrant MT branching, and altered polarized trafficking of key basal domain proteins such as GPR56, and thus aberrant RGC tiling. These findings identify Memo1 as a mediator of RGC scaffold tiling, necessary to generate and organize neurons into functional ensembles in the developing cerebral cortex.

Microglia in the primate macula: specializations in microglial distribution and morphology with retinal position and with aging

Microglia, the principal resident immune cell in the retina, play constitutive roles in immune surveillance and synapse maintenance, and are also associated with retinal disease, including those occurring in the macula. Perspectives on retinal microglia function have derived largely from rodent models and how these relate to the macula-bearing primate retina is unclear. In this study, we examined microglial distribution and cellular morphology in the adult rhesus macaque retina, and performed comparative characterizations in three retinal locations along the center-to-periphery axis (parafoveal, macular, and the peripheral retina). We found that microglia density peaked in the parafoveal retina and decreased in the peripheral retina. Individual microglial morphology reflected macular specialization, with macular microglia demonstrating the largest and most complex dendritic arbors relative to other retinal locations. Comparing retinal microglia between young and middle-aged animals, microglial density increased in the macular, but not in the peripheral retina with age, while microglial morphology across all locations remained relatively unchanged. Our findings indicate that microglial distribution and morphology demonstrate regional specialization in the retina, correlating with gradients of other retinal cell types. As microglia are innate immune cells implicated in age-related macular diseases, age-related microglial changes may be related to the increased vulnerability of the aged macula to immune-related neurodegeneration.

Expression of synaptic and phototransduction markers during photoreceptor development in the marmoset monkey Callithrix jacchus

Marmoset photoreceptor development was studied to determine the expression sequence for synaptic, opsin, and phototransduction proteins. All markers appear first in cones within the incipient foveal center or in rods at the foveal edge. Recoverin appears in cones across 70% of the retina at fetal day (Fd) 88, indicating that it is expressed shortly after photoreceptors are generated. Synaptic markers synaptophysin, SV2, glutamate vesicular transporter 1, and CTBP2 label foveal cones at Fd 88 and cones at the retinal edge around birth. Cones and rods have distinctly different patterns of synaptic protein and opsin expression. Synaptic markers are expressed first in cones, with a considerable delay before they appear in rods at the same eccentricity. Cones express synaptic markers 2-3 weeks before they express opsin, but rods express opsin 2-4 weeks before rod synaptic marker labeling is detected. Medium/long-wavelength-selective (M&L) opsin appears in foveal cones and rod opsin in rods around the fovea at Fd 100. Very few cones expressing short-wavelength-selective (S) opsin are found in the Fd 105 fovea. Across peripheral retina, opsin appears first in rods, followed about 1 week later by M&L cone opsin. S cone opsin appears last, and all opsins reach the retinal edge by 1 week after birth. Cone transducin and rod arrestin are expressed concurrently with opsin, but cone arrestin appears slightly later. Marmoset photoreceptor development differs from that in Macaca and humans. It starts relatively late, at 56% gestation, compared with Macaca at 32% gestation. The marmoset opsin expression sequence is also different from that of either Macaca or human.

Gradual morphogenesis of retinal neurons in the peripheral retinal margin of adult monkeys and humans

The adult mammalian retina has for long been considered to lack a neurogenerative capacity. However, retinal stem/progenitor cells, which can originate retinal neurons in vitro, have been recently reported in the ciliary body of adult mammals. Here we explored the possibility of retinal neurogenesis occurring in vivo in adult monkeys and humans. We found the presence of cells expressing molecular markers of neural and retinal progenitors in the nonlaminated retinal margin and ciliary body pars plana of mature primates. By means of immunohistochemistry and electron microscopy we also observed photoreceptors and other retinal cell types in different stages of morphological differentiation along the peripheral retinal margin. These findings allow us to extend to primates the idea of neurogenesis aimed at retinal cell turnover throughout life.

The genesis of retinal architecture: an emerging role for mechanical interactions?

Patterns in nature have always fascinated human beings. They convey the idea of order, organization and optimization, and, to the enquiring mind, the alluring promise that understanding their building rules may uncover the forces that shaped them. In the retina, two patterns are outstanding: the stacking of cells in layers and, within the layers, the prevalent arrangement of neurons of the same type in orderly arrays, often referred to as mosaics for the crystalline-like order that some can display. Layers and mosaics have been essential keys to our present understanding of retinal circuital organization and function. Now, they may also be a precious guide in our exploration of how the retina is built. Here, we will review studies addressing the mechanisms controlling the formation of retinal mosaics and layers, illustrating common themes and unsolved problems. Among the intricacies of the building process, a world of physical forces is making its appearance. Cells are extremely complex to model as "physical entities", and many aspects of cell mechanotransduction are still obscure. Yet, recent experiments, focusing on the mechanical aspects of growth and differentiation, suggest that adopting this viewpoint will open new ways of understanding retinal formation and novel possibilities to approach retinal pathologies and repair.

Molecular mechanisms of vertebrate retina development: Implications for ganglion cell and photoreceptor patterning

Although the neural retina appears as a relatively uniform tissue when viewed from its surface, it is in fact highly patterned along its anterior-posterior and dorso-ventral axes. The question of how and when such patterns arise has been the subject of intensive investigations over several decades. Most studies aimed at understanding retinal pattern formation have used the retinotectal map, the ordered projections of retinal ganglion cells to the brain, as a functional readout of the pattern. However, other cell types are also topographically organized in the retina. The most commonly recognized example of such a topographic cellular organization is the differential distribution of photoreceptor types across the retina. Photoreceptor patterns are highly species-specific and may represent an important adaptation to the visual niche a given species occupies. Nevertheless, few studies have addressed this functional readout of pattern to date and our understanding of its development has remained superficial. Here, we review recent advances in understanding the molecular cascades that control regionalization of the eye anlage, relate these findings to the development of photoreceptor patterns and discuss common and unique strategies involved in both aspects of retinal pattern formation.

Rhodopsin, Violet and Blue Opsin Expressions in the Chick Are Highly Dependent on Tissue and Serum Conditions

The molecular, cellular or tissue environment can influence the expression of genes and thereby regulate processes of tissue formation. Here we determined the tissue and serum dependence of the expression of all photopigments in the chick by a series of distinct retinal cell cultures, analyzed by RT-PCR using specific primers for all four opsins and rhodopsin followed by quantitative scanning of the respective gel bands. For comparison, we first determined expression of all opsins during normal chick retinogenesis, which began with red and violet opsins at E12, shortly followed by blue and green opsins and finally rhodopsin at E14. This period corresponds to the time of synaptogenesis in the inner retina. All cultures were started with 6-day-old dissociated retinal cells. Cells were kept at low or high cell density (called LoDens or HiDens), or they were reaggregated as retinal spheres, whereby all of them were raised at low (2%) or high serum (12%) levels (called LoSer or HiSer). In LoDens/HiSer cultures, expression of all opsins was weak. At HiDens/LoSer red and green opsin expression was strong, while rhodopsin and violet/blue remained low. In HiDens/HiSer cultures the expression of red and green was strong; rhodopsin was almost normal, while violet and green were low. In reaggregates at high serum the expression came closest to a normal retina, but violet and blue opsins were still below normal. At low serum, however, violet and blue were negligible and rhodopsin was low. This in vitro study shows that rhodopsin, followed by violet and blue opsin expressions is highly dependent on serum, cell density and tissue conditions, while red and green opsins are more autonomous.

Mechanisms controlling the formation of retinal mosaics

Most regions of the nervous system derive their power of processing from a modular architecture. The retina is an outstanding example of modular circuit design. Retinal neurons are stacked in layers and within each layer neurons of the same type commonly form orderly arrays, or mosaics. Here we review current knowledge on the mechanisms of retinal mosaic formation, and discuss the hypothesis that retinal mosaics are the building blocks in the assembly of retinal circuitry.

Density-dependent differentiation in nontransformed human retinal progenitor cells in response to basic fibroblast growth factor- and transforming growth factor-alpha

Multipotential retinal precursors give rise to all cell types seen in multilayered retina. The generation of differentiation and diversity of neuronal cell types is determined by both extrinsic regulatory signals and endogenous genetic programs. We have previously reported that cell commitment in human retinal precursor cells (SV-40T) can be modified in response to exogenous growth factors, basic fibroblast growth factor, and transforming growth factor alpha (bFGF and TGFalpha). We report in this study that nontransformed human retinal precursors differentiate into photoreceptors by a cell density-dependent mechanism, and the effects were potentiated by bFGF and TGFalpha alone or in combination. A larger proportion of multipotential precursors plated at a density of 1 x 10(4) cells/cm(2) differentiated into neurons (photoreceptors) compared to cells plated at 3-5 x 10(4)/cm(2) and 1 x 10(5) cells/cm(2) under serum-free conditions and the effects were amplified seven- to eightfold in response to growth factors. Basic fibroblast growth factor (bFGF) and TGFalpha can induce 90% of the cells to assume a photoreceptor phenotype at a lower cell density, compared to only 30 and 25% of the cells acquiring a photoreceptor phenotype at intermediate and higher cell densities. Furthermore, at a lower cell density, 60-70% of the cells incorporate Bromodeoxyuridine (Brdu), suggesting that cells in a cell cycle may make a commitment to a specific fate in response to neurotrophins. Neurons with a photoreceptor phenotype were positive for three different sets of antibodies for rods/cones. Cells also exhibited upregulation of other proteins such as a D4 receptor protein expressed in photoreceptors, protein kinase Calpha (PKCalpha) expressed in rod bipolars and blue cones, and some other neuronal cell types. This was also confirmed by Western blot analysis. Newly derived photoreceptors survive for a few days before significant cell death ensues under serum-free conditions. To summarize, differentiation in precursors is density dependent, and growth factors amplify the effects.

Adeno-Associated Virus Type 5: Transduction Efficiency and Cell-Type Specificity in the Primate Retina

Gene transfer using adeno-associated viruses (AAVs) has been effective for treating inherited retinal diseases in animal models. Further evaluation in primates must be performed prior to clinical application, however, because of the difference between the retina of the primate and those of other animals. Prior work has shown that AAV2 can transduce rod-photoreceptor and RPE cells in the non-human primate retina and that AAV5 is more efficient at transducing photoreceptor cells than AAV2 in the rodent retina. In this study, we evaluated the efficiency of AAV5 in the non-human primate retina after subretinal injections of the vector to distinct anatomic retinal regions (superior, inferior, nasal, macula, temporal). rAAV5 led to a rapid onset of transgene expression (within 2 weeks), with expression persisting up to 10 months. Postoperative electrophysiology studies showed that global retinal function was preserved following gene transfer. Quantitative analysis of gene transfer demonstrated a maximum transduction efficiency of 22% in the injected areas. Evaluation of cell types using confocal microscopy and cone-specific antibodies revealed that AAV5, expressing reporter genes from the cytomegalovirus (CMV) promoter/enhancer, preferentially transduced rods. No significant differences were found in the regional tropism of AAV5 among the five areas injected despite variation in retinal topography. Immunohistochemical studies revealed that the AAV5 receptor, PDGFR-A, is localized to the outer segments of rods but not cones providing a basis for the observed tropism. Our results support the utility of AAV5 for rod photoreceptor degeneration therapies.

Putting neurons in the right places: local interactions in the genesis of retinal architecture

Development of the nervous system can be schematically summarized as (1) making the necessary cells, (2) putting these cells in the right places, and then (3) connecting them appropriately. Each of these steps represents an enormous challenge to our understanding. Focusing on the vertebrate retina, I will consider the question of what defines the right place for a neuron to go. I will illustrate data pointing to the prominent role played by short-range cellular interactions, possibly coordinated by global factors, and will discuss how a few sets of local rules could control cell positioning and proper wiring in retinal circuits.

Dynamic microtubule-dependent interactions position homotypic neurones in regular monolayered arrays during retinal development

In the vertebrate retina cell layers support serial processing, while monolayered arrays of homotypic neurones tile each layer to allow parallel processing. How neurones form layers and arrays is still largely unknown. We show that monolayered retinal arrays are dynamic structures based on dendritic interactions between the array cells. The analysis of three developing retinal arrays shows that these become regular as a net of dendritic processes links neighbouring array cells. Molecular or pharmacological perturbations of microtubules within dendrites lead to a stereotyped and reversible disruption of array organization: array cells lose their regular spacing and the arrangement in a monolayer. This leads to a micro-mechanical explanation of how monolayers of regularly spaced ‘like-cells’ are formed.

Multiple phosphorylated isoforms of NRL are expressed in rod photoreceptors

NRL, a bZIP transcription factor of the Maf subfamily, interacts with the homeodomain protein CRX and synergistically regulates rhodopsin expression. Here we report that six isoforms of NRL (29-35 kDa) are generated by phosphorylation and expressed specifically in the mammalian retina. The anti-NRL antibody also cross-reacts with a cytosolic 45-kDa protein, which is detected in neuronal tissues but is not encoded by the NRL gene. In both human retinal cell cultures and sections of fetal and adult human retina, NRL is present in the nuclei of developing and mature rods but not cones. We propose that NRL regulates rod photoreceptor-specific gene expression and is involved in rod differentiation.

The mammalian photoreceptor mosaic-adaptive design

Unlike in birds and cold-blooded vertebrates' retinas, the photoreceptors of mammalian retinas were long supposed to be morphologically uniform and difficult to distinguish into subtypes. A number of new techniques have now begun to overcome the previous limitations. A hitherto unexpected variability of spectral and morphological subtypes and topographic patterns of distribution in the various retinas are being revealed. We begin to understand the design of the photoreceptor mosaics, the constraints of evolutionary history and the ecological specialization of these mosaics in all the mammalian subgroups. The review discusses current cytological identification of mammalian photoreceptor types and speculates on the likely "bottleneck-scenario" for the origin of the basic design of the mammalian retina. It then provides a brief synopsis of current data on the photoreceptors in the various mammalian orders and derives some trends for phenomena such as rod/cone dualism, spectral range, preservation or loss of double cones and oil droplets, photopigment co-expression and mono- and tri-chromacy. Finally, we attempt to demonstrate that, building on the limits of an ancient rod dominant (probably dichromatic) model, mammalian retinas have developed considerable radiation. Comparing the nonprimate models with the intensively studied primate model should provide us with a deeper understanding of the basic design of the mammalian retina.

Nocturnal tarsier retina has both short and long/medium-wavelength cones in an unusual topography

The evolutionary position of tarsiers with respect to primates is still debated. The type of photoreceptors in the nocturnal Tarsius spectrum retina has been compared with the nocturnal New World monkey Aotus trivulgaris and the Old World monkey Macaca nemestrina by using immunocytochemical labeling for antisera known to be specific for primate cone and rod proteins. In all three species, antisera to long/medium (L/M) -wavelength specific cone opsin and cone-specific alpha-transducin detected a single row of cones. Only Macaca and tarsier retina contained cones labeled by antiserum to short (S) -wavelength specific cone opsin. Tarsier rod cell bodies were 6-12 deep, depending on retinal eccentricity. Tarsier central cones had 2-microm-wide outer (OS) and inner segments, which came straight off the cell body. Cone morphology differed little from rods except OS were shorter. Macaca cones labeled for 7G6 and calbindin, Aotus cones did not label for calbindin, and Tarsius cones did not label for 7G6 or calbindin. In tarsier retinal whole-mounts, peak cone density ranged from 11,600-14,200/cones mm(2). The 11- to 12-mm-wide peak region centered roughly on the optic disc, although foveal counts remain to be completed. Density decreased symmetrically to a far peripheral band of 4,200-7, 000/cones mm(2). In contrast, S cone density was very low in central retina (0-300/mm(2)), rose symmetrically with eccentricity, and peaked at 1,100-1,600/mm(2) in a 2- to 3-mm-wide zone in the far periphery. In this zone, S cones were 9-14% of all cones. L/M cones were regularly spaced, whereas S cones showed no regular distribution pattern. Although the functional characteristics of the tarsier S and L/M cone systems are yet to be determined, tarsier cone proteins and distribution have some similarities to both New and Old World monkey retinas.

Cell fate decisions in a human retinal precursor cell line: basic fibroblast growth factor- and transforming growth factor-alpha-mediated differentiation

The purpose of this study was to determine if immortalized human retinal precursor cells could serve as a model to investigate cues that modulate cell fate and differentiation. We investigated the effects of a variety of growth factors broadly but specifically tested the effects of basic fibroblast growth factor (bFGF) and transforming growth factor (TGF)a in retinal cell differentiation and commitment. To determine the role of exogenously added growth factors in a human retinal precursor cell line (KGLDMSM), established from a first-trimester retina, cells were adapted to grow in a defined medium and exposed to a variety of trophic factors (epidermal growth factor [EGF], neuron growth factor [NGF], TGFalpha, TGFbeta, acidic FGF, and bFGF). Dose-response curves were developed to arrive at optimal concentrations. The neurotrophic potential of growth factors was determined by 3H-thymidine incorporation and bromodeoxyuridine (BrdU) labeling. The identity of the emerging neuronal phenotypes were determined by phase-contrast microscopy, immunolabeling for the neuron-specific antigens neurofilament protein (NF) and neuron-specific enolases (NSE), and photoreceptor-specific antigens (Rho1D4, 7G6) using immunocytochemistry and Western blot analysis. To identify some of the early response genes (c-fos, c-myc) expressed in response to growth factors, Northern blot analysis was performed. Almost all of the factors tested increased the total number of cells with a neuronal phenotype. Potency of growth factors to generate neurons was TGFalpha > bFGF > EGF > NGF. Both TGFalpha and bFGF, alone or in combination, increased the total number of neurons. Most of the neurons generated were photoreceptors, as depicted by the polarized phenotype, expression of photoreceptor-specific antigens, and processes resembling rudimentary outer segments. The increase in photoreceptor-like neurons is possibly attributable to an increase in numbers rather than greater survival. Additionally, the majority of the photoreceptors generated labeled with BrdU and for photoreceptor-specific antigens, suggesting that an inductive effect of bFGF and TGFalpha could occur in the cell cycle or shortly thereafter. Both bFGF and TGFalpha induced the expression of the early response gene c-fos while not altering the expression of c-actin or c-myc. The emergence of a photoreceptor phenotype was confirmed by both immunocytochemistry and Western blot analysis. The immortalized retinal precursor cell line could prove valuable in determining the role of exogenously added growth factors in retinal development and differentiation. Both bFGF and TGFalpha enhance the photoreceptor phenotype in medium-density cultures under conditions of defined medium. The same was confirmed by phase-contrast microscopy, immunocytochemistry, and Western blot analysis. Furthermore, cell fate determination in cultured precursor cells could occur during the late part of the cell cycle or shortly after completion of cell division. The effects of TGFalpha and bFGF seem to be slightly additive. The cell line will be extremely valuable in studying mechanisms of cell commitment and generation of retinal cell types, which could be tested for their potential for transplantation.

Local, possibly contact-mediated signalling restricted to homotypic neurons controls the regular spacing of cells within the cholinergic arrays in the developing rodent retina

In the vertebrate retina neurons of the same type commonly form non-random arrays, assembled by unknown positional mechanisms during development. Computational models in which no two cells are closer than a minimal distance, simulate many retinal arrays. These findings have important biological implications, since they suggest that cells are determined as neurons of specific types before entering their arrays, and that local, possibly contact-mediated interactions acting exclusively among the elements of an array account for its assembly. This is here verified by combining experimental manipulations in normal and transgenic models with computational analysis for the cholinergic mosaics, the only arrays so far for which the development of spatial ordering is known quantitatively. When generalised, these findings suggest a plan for vertebrate retinal patterning, where homotypic interactions organise retinal arrays first, then local interactions between synaptic partners suffice to establish the topographical connections that support retinal processing.

Photoreceptor distribution in the retinas of subprimate mammals

Relevant data on the distribution of color cones are summarized, with special emphasis on the marked dorsoventral asymmetries observed in a number of mammalian species. In addition, an overview is given of studies that demonstrate the coexistence of two visual pigments within the same cone cell. The biological significance of these phenomena is discussed in conjunction with comparative immunocytochemical analyses of subprimate retinas. Based on various cone distribution patterns and temporal and spatial visual pigment coexpression, two models of cone photoreceptor differentiation are suggested.

Retinal mosaics: new insights into an old concept

It has been known since the middle of the 19th century that different neuronal types are distributed across the retinal surface in non-random arrays: indeed, these arrays, called 'mosaics', have long been considered to be a fundamental feature of retinal organization. However, until recently, little was known about how such mosaics are established during development. In the hope of stimulating further research, this article reviews the current status of three very different approaches to this intriguing general problem. The first postulates arrays of molecular markers, which are produced by specific cell types shortly after their final mitotic divisions and could be influential in the differentiation of other cell types. The second invokes a tangential dispersion of differentiating cells to generate spatial order, either while these cells are still migrating or soon after they reach their laminar destinations. The third involves the elimination of wrongly positioned cells through the process of naturally occurring cell death.

The development of the pattern of retinal ganglion cells in the chick retina: mechanisms that control differentiation

Neurons in both vertebrate and invertebrate eyes are organized in regular arrays. Although much is known about the mechanisms involved in the formation of the regular arrays of neurons found in invertebrate eyes, much less is known about the mechanisms of formation of neuronal mosaics in the vertebrate eye. The purpose of these studies was to determine the cellular mechanisms that pattern the first neurons in vertebrate retina, the retinal ganglion cells. We have found that the ganglion cells in the chick retina develop as a patterned array that spreads from the central to peripheral retina as a wave front of differentiation. The onset of ganglion cell differentiation keeps pace with overall retinal growth; however, there is no clear cell cycle synchronization at the front of differentiation of the first ganglion cells. The differentiation of ganglion cells is not dependent on signals from previously formed ganglion cells, since isolation of the peripheral retina by as much as 400 μm from the front of ganglion cell differentiation does not prevent new ganglion cells from developing. Consistent with previous studies, blocking FGF receptor activation with a specific inhibitor to the FGFRs retards the movement of the front of ganglion cell differentiation, while application of exogenous FGF1 causes the precocious development of ganglion cells in peripheral retina. Our observations, taken together with those of previous studies, support a role for FGFs and FGF receptor activation in the initial development of retinal ganglion cells from the undifferentiated neuroepithelium peripheral to the expanding wave front of differentiation.

Ontogenetic changes in the retinal photoreceptor mosaic in a fish, the black bream, Acanthopagrus butcheri

The morphological development of the photoreceptor mosaic was followed by light and electron microscopy in a specific region of dorsal retina of the black bream, Acanthopagrus butcheri (Sparidae, Teleostei), from hatching to eight weeks of age. The retina was differentiated when the larvae reached a total length of 3 mm (3-5 days posthatch). Single cones, arranged in tightly packed rows, were the only morphologically distinct type of photoreceptor present until the larvae were 6 mm (day 15) in standard length (SL). At this time, the rod nuclei had become differentiated and the ellipsoids of selected cones began to form subsurface cisternae along neighbouring cone membranes. In this way, double, triple, quadruple, and occasionally photoreceptor chains of up to 10 cones were formed. At 8 mm SL, there was little apparent order in the photoreceptor mosaic. However, concomitant with subsequent growth, quadruple and other multiple cone receptors disappeared, with the exception of the triple cones, which gradually reduced in both number and retinal coverage to be restricted to central retina by 15 mm SL (days 40-55). Following this stage, the arrangement of double and single cones peripheral to the region of triple cones in dorsal retina was transformed into the adult pattern of a regular mosaic of four double cones surrounding a single cone. These results demonstrate that an established photoreceptor mosaic of rows of single cones can be reorganised to form a regular square mosaic composed of single and double cones.

Modelling the mosaic organization of rod and cone photoreceptors with a minimal-spacing rule

The mosaic of photoreceptors is regarded as a prime example of the precise control of cellular positioning in the vertebrate nervous system. This study was undertaken with the idea that understanding the intrinsic geometrical features of photoreceptor mosaics is a necessary step to unveil the biological mechanisms governing their formation. We show in the retina of the ground squirrel that the arrays of both the rods and S cones are non-random, but that nothing more than a simple minimal-spacing rule constraining receptor positioning is sufficient to account for the spatial organization of both mosaics. The size of this 'exclusion zone' is an intrinsic characteristic of each cell type, and it is simply the difference in the size of this domain that accounts for the regularity of the S cone array and the irregularity of the rod array at identical density. Consequently, regularity in receptor mosaics is produced by two independent biological events, one embodying the exclusion zone, and another specifying the local density of a given receptor type.

Analysis of the short wavelength-sensitive ("blue") cone mosaic in the primate retina: comparison of New World and Old World monkeys

The distribution of short wavelength-sensitive (SWS or "blue") cone photoreceptors was compared in primates with dichromatic ("red-green colour blind") and trichromatic colour vision. We compared a New World species, the marmoset (Callithrix jacchus), with an Old World species, the macaque monkey (Macaca nemestrina). The SWS cones were identified by their immunoreactivity to an antiserum against the human SWS cone opsin. A single retina from a male capuchin monkey (Cebus apella) also was studied. The SWS cones make up less than 10% of all cone photoreceptors throughout the retina of all animals studied. In marmoset, the peak spatial density of SWS cones is close to 10,000/mm2 at the foveola. In macaque, the peak spatial density of SWS cones, close to 6,000/mm2, is at the fovea, but SWS cones are absent within 50 microm of the centre of the foveola. In both species, the density of SWS cones is higher on the nasal retinal axis than at corresponding eccentricities on the other retinal axes. The SWS cones in macaque are arranged in a semiregular array, but they are distributed randomly in marmoset. There is no difference in the spatial density or local arrangement of SWS cones between dichromatic and trichromatic marmosets. The results suggest that the SWS cone photoreceptor system is subject to different developmental and evolutionary constraints than those that have led to the formation of the red-green photoreceptor systems in primate vision.

Cellular diversification in the vertebrate retina

The discovery of heterogeneous populations of retinal precursors with sequentially modified fates may help solve the conundrum of conserved histogenesis in the absence of determination either by birthdate or lineage. Combined with a wealth of new data on the exogenous and endogenous factors that influence cellular fate in the retina, models of how complexity is generated are beginning to emerge.