Histopathology of the Retina from a Three Year-Old Suspected to Have Joubert Syndrome.

PURPOSE: To define the retinal pathology in a 3 year-old eye donor who died from complications of an undiagnosed genetic syndrome. METHODS: Eyes were fixed and analyzed using macroscopic fundus photography (MF), confocal scanning laser ophthalmoscopy (cSLO) and spectral-domain optical coherence tomography (SD-OCT). Small areas from the perifovea and periphery were processed for histology and indirect immunofluorescence, using antibodies specific to retinal proteins such as rhodopsin, cone arrestin, RPE65 and others. Available medical records were also reviewed. RESULTS: With all three imaging modalities, the affected donor's eyes lacked the distinct morphological detail typically observed with these techniques in postmortem control eyes. MF images showed a "photonegative effect" due to a hypopigmented macula relative to a hyperpigmented retinal background. cSLO imaging demonstrated a weak autofuorescence signal that was largely devoid of the usual retinal structures compared to the control. SD-OCT suggested disorganization of the affected retina, absence of a photoreceptor layer, and degeneration of the choroid in the macular area. Histologic findings indicated a highly disorganized photoreceptor layer in the macula and periphery. The RPE layer displayed thinning in some regions of the periphery and decreased pigmentation in most areas. Rods and cones were significantly reduced in the affected retina but a few cones were detected in the perifovea. Centrin-2 labeling was mostly absent from the connecting cilium of the photoreceptor cells. Medical record review pointed to a possible clinical diagnosis of Joubert syndrome. CONCLUSIONS: The retinal degenerative findings, and absence of centrin-2 labeling are compatible with the expected retinal phenotype in patients with Joubert syndrome.

SPACRCAN in the developing retina and pineal gland of the rat: spatial and temporal pattern of gene expression and protein synthesis.

SPACRCAN is a hyaluronan-binding proteoglycan that is present in the pineal gland and interphotoreceptor matrix of the retina. Here, we evaluate the pattern of SPACRCAN gene expression and protein appearance during retinal and pineal gland development in the rat. In situ hybridization histochemistry with SPACRCAN riboprobes indicates that hybridization signals are first evident in the retina over developing photoreceptor cells at embryonic day 16 (E16) and in the pineal gland at E21. Immunocytochemistry using a SPACRCAN antibody shows localization of SPACRCAN protein in the developing interphotoreceptor matrix by Postnatal day 5 (P5) and in the pineal gland by P6. These studies suggest that SPACRCAN mRNA expression may occur substantially earlier than the time when SPACRCAN protein is detectable in both the retina and the pineal gland. The period of retinal histogenesis when SPACRCAN is detected first is coincident with the time photoreceptors begin to extend from the outer retinal surface, suggesting that SPACRCAN may participate in the maturation and maintenance of the light-sensitive photoreceptor outer segment.

Interphotoreceptor matrix in the fovea and peripheral retina of the primate Macaca mulatta: distribution and glycoforms of SPACR and SPACRCAN.

SPACR and SPACRCAN localization in the interphotoreceptor matrix (IPM) of the fovea and peripheral retina of Macaca mulatta was established with antibodies to these core proteins and the chondroitin sulfate epitopes and lectin binding properties of these molecules were defined. The IPM of both rods and cones labeled with anti-SPACR, anti-SPACRCAN, anti-Delta Di6S antibodies and wheat germ agglutinin (WGA). Whereas anti-SPACR and anti-SPACRCAN antibodies labeled rod and cone matrix compartments with similar intensity, the Delta Di6S chondroitin antibody labeling was more intense around cones than rods. Peanut lectin (PNA) labeling was present only around cones. No IPM labeling was observed with Delta Di0S-chondroitin or Delta Di4S-chondroitin antibodies. Western blots of undigested IPM extracts showed anti-SPACR immunoreactivity at 150 kDa, colocalizing with the position of WGA and PNA binding. In Western blots of the chondroitinase ABC digested sample and samples double digested with chondroitinase ABC and AC II, anti-SPACR immunoreactivity, WGA and PNA labeling intensity were virtually identical to that in the undigested sample, with prominent staining of the 150 kDa SPACR band. In contrast, anti-SPACRCAN immunoreactivity was not present in the undigested sample, but was evident in both the chondroitinase ABC and double digested samples as a broad band at approximately 230 kDa. Delta Di6S, Delta Di4S, WGA and PNA labeling colocalized with the anti-SPACRCAN immunoreactivity in the chondroitinase ABC digested sample. These findings indicate that SPACR and SPACRCAN are present around cones in the fovea and both rods and cones in the peripheral retina, but that the specific glycoforms of these molecules are different depending on whether present in the cone or rod associated IPM.

Bestrophin, the product of the Best vitelliform macular dystrophy gene (VMD2), localizes to the basolateral plasma membrane of the retinal pigment epithelium.

Best vitelliform macular dystrophy is a dominantly inherited, early onset, macular degenerative disease that exhibits some histopathologic similarities to age-related macular degeneration. Although the vitelliform lesion is common in the fundus of individuals with Best disease, diagnosis is based on a reduced ratio of the light peak to dark trough in the electrooculogram. Recently, the VMD2 gene on chromosome 11q13, encoding the protein bestrophin, was identified. The function of bestrophin is unknown. To facilitate studies of bestrophin, we produced both rabbit polyclonal and mouse monoclonal antibodies that proved useful for Western blotting, immunoprecipitation, and immunocytochemistry. To characterize bestrophin, we initially probed the retinal pigment epithelium (RPE)-derived cell lines ARPE-19, D407, and RPE-J. All of the cell lines expressed bestrophin mRNA by reverse transcription-PCR, but not on Western blots. Bestrophin in human RPE partitioned in the detergent phase during Triton X-114 extraction and could be modified by biotin in intact cells, indicative of a plasma membrane localization. Immunocytochemical staining of macaque and porcine eyes indicated that bestrophin is localized at the basolateral plasma membrane of RPE cells. When expressed in RPE-J cells by adenovirus-mediated gene transfer, bestrophin again was determined by confocal microscopy and cell surface biotinylation to be a basolateral plasma membrane protein. The basolateral plasma membrane localization of bestrophin suggests the possibility that bestrophin plays a role in generating the altered electrooculogram of individuals with Best disease.

SPACR in the interphotoreceptor matrix of the mouse retina: molecular, biochemical and immunohistochemical characterization.

Mouse SPACR cDNA was cloned by screening a mouse retina cDNA library using a PCR probe derived from human SPACR cDNA. Mouse SPACR cDNA comprises 3675 bp containing an open reading frame coding for 742 amino acids. Multitissue Northern blot analysis and in situ hybridization studies indicate that SPACR expression is restricted to retinal photoreceptors. The SPACR core protein was identified with Western blotting following SDS-PAGE with a SPACR C-terminal peptide polyclonal antibody and a chondroitin-6-sulfate Deltadisaccharide monoclonal antibody. The 150 kD immunopositive band was isolated, digested with trypsin and the peptides analysed by MALDI mass spectroscopy. Peptide mass mapping confirmed the identity of the 150 kD immunopositive band to be mouse SPACR core protein. Alignment comparisons of the deduced amino acid sequence of mouse and human SPACR show 64% homology. Like SPACR in the human interphotoreceptor matrix, the mouse orthologue contains a large central mucin-like domain flanked by consensus sites for N-linked oligosaccharide attachment, one EGF-like domain and four hyaluronan-binding motifs. Unlike human SPACR, which contains no conventional consensus sites for glycosaminoglycan attachment, mouse SPACR contains three. Recent biochemical studies of human and mouse SPACR protein indicate that this novel interphotoreceptor matrix molecule is a glycoprotein in human and a proteoglycan in the mouse. The presence of consensus sites for glycosaminoglycan attachment in the deduced sequence of mouse SPACR and the absence of these sites in human SPACR provide molecular verification of our biochemical results, suggesting that differences in post-translational modifications of SPACR may be important in SPACR function in foveate and non-foveate retinas.

SPACRCAN, a novel human interphotoreceptor matrix hyaluronan-binding proteoglycan synthesized by photoreceptors and pinealocytes.

The interphotoreceptor matrix is a unique extracellular complex occupying the interface between photoreceptors and the retinal pigment epithelium in the fundus of the eye. Because of the putative supportive role in photoreceptor maintenance, it is likely that constituent molecules play key roles in photoreceptor function and may be targets for inherited retinal disease. In this study we identify and characterize SPACRCAN, a novel chondroitin proteoglycan in this matrix. SPACRCAN was cloned from a human retinal cDNA library and the gene localized to chromosome 3q11.2. Analysis of SPACRCAN mRNA and protein revealed that SPACRCAN is expressed exclusively by photoreceptors and pinealocytes. SPACRCAN synthesized by photoreceptors is localized to the interphotoreceptor matrix where it surrounds both rods and cones. The functional protein contains 1160 amino acids with a large central mucin domain, three consensus sites for glycosaminoglycan attachment, two epidermal growth factor-like repeats, a putative hyaluronan-binding motif, and a potential transmembrane domain near the C-terminal. Lectin and Western blotting indicate an M(r) around 400,000 before and 230,000 after chondroitinase ABC digestion. Removal of N- and O-linked oligosaccharides reduces the M(r) to approximately 160,000, suggesting that approximately 60% of the mass of SPACRCAN is carbohydrate. Finally, we demonstrate that SPACRCAN binds hyaluronan and propose that associations between SPACRCAN and hyaluronan may be involved in organization of the insoluble interphotoreceptor matrix, particularly as SPACRCAN is the major proteoglycan present in this matrix.

Chondroitin sulfate proteoglycan core proteins in the interphotoreceptor matrix: a comparative study using biochemical and immunohistochemical analysis.

This study characterizes the core proteins of chondroitin sulfate-type glycosaminoglycans located in the interphotoreceptor matrix and establishes the tissue distribution of chondroitin immunoreactivity in human, bovine, mouse and rat retinas. Monoclonal antibodies specific to unsulfated (DeltaDiOS), 4-sulfated (DeltaDi4S) and 6-sulfated (DeltaDi6S) chondroitin were employed. Retinal sections and IPM samples were either (a) digested with chondroitinase ABC to expose antibody specific epitopes, (b) double digested with chondroitinase ABC and chondroitinase AC II to remove specific epitopes, or (c) left undigested to evaluate mimotope labeling. In tissue sections from each species studied, positive immunoreactivity to the DeltaDi6S antibody was present in the IPM surrounding both rods and cones. In human and bovine, DeltaDi6S labeling of the cone matrix compartments was more intense than labeling of the matrix surrounding rods. Intense DeltaDi6S immunoreactivity was present surrounding the foveal cones. In mouse and rat, no differences in labeling intensity of IPM surrounding rod and cone photoreceptors were evident, although labeling of the IPM near the apical surface of the retinal pigment epithelium and around the photoreceptor inner segments was more pronounced than that surrounding the outer segments. All DeltaDi6S antibody labeling was eliminated with chondroitinase AC II digestion. No IPM immunoreactivity in tissue sections was observed when the DeltaDi0S or DeltaDi4S antibodies were used. In Western blots of IPM extracts treated with chondroitinase ABC, prominent DeltaDi6S immunoreactive bands were present at approximately 230 kD and 150 kD in each species studied, with the exception of the human, where the 150 kD component is not a chondroitin proteoglycan. Each of the prominent DeltaDi6S immunoreactive bands showed minor immunoreactivity to the DeltaDi4S antibody. No DeltaDi0S immunoreactivity was noted in Western blots of IPM samples from any species. All immunoreactivity was lost following chondroitinase AC II digestion. These observations document similarities in the electrophoretic mobility of IPM proteoglycan core proteins released following chondroitinase ABC digestion in the four species studied, but reveal pronounced differences in the tissue distribution. Bovine and human IPM show greater concentrations of DeltaDi6S immunoreactivity surrounding cones than rods, whereas rodent tissues show higher concentrations near the retinal pigment epithelium and around the photoreceptor inner segments than around the outer segments. The pattern of distribution of these proteoglycan molecules is highly conserved in these species, suggesting a common role in IPM structure and function.

Characterization of SPACR, a sialoprotein associated with cones and rods present in the interphotoreceptor matrix of the human retina: immunological and lectin binding analysis.

Rod and cone photoreceptors project from the outer retinal surface into a carbohydrate-rich interphotoreceptor matrix (IPM). Unique IPM glycoconjugates are distributed around rods and cones. Wheat germ agglutinin (WGA) strongly decorates the rod matrix domains and weakly decorates the cone matrix domains. This study characterizes the major WGA-binding glycoprotein in the human IPM, which we refer to as SPACR (sialoprotein associated with cones and rods). SPACR, which has a molecular weight of 147 kDa, was isolated and purified from the IPM by lectin affinity chromatography. A polyclonal antibody to SPACR was prepared that colocalizes in tissue preparations with WGA-binding domains in the IPM. Sequential digestion of SPACR with N- and O-glycosidases results in a systematic increase in electrophorectic mobility, indicating the presence of both N- and O-linked glycoconjugates. Complete deglycosylation results in a reduction in the relative molecular mass of SPACR by about 30%. Analysis of lectin binding allowed us to identify some of the structural characteristics of SPACR glycoconjugates. Treatment with neuraminidase exposes Galbeta1-3GalNAc disaccharide as indicated by positive peanut agglutinin (PNA) staining, accompanied by the loss of WGA staining. Maackia amurensis agglutinins (MAA-1 and MAA-2), specific for sialic acid in alpha2-3 linkage to Gal, bind SPACR, while Sambucus nigra agglutinin (SNA), specific for alpha2-6 linked sialic acid, does not, indicating that the dominant glycoconjugate determinant on SPACR is the O-linked carbohydrate, NeuAcalpha2-3Galbeta1-3GalNAc. The abundance of sialic acid in SPACR suggests that this glycoprotein may contribute substantially to the polyanionic nature of the IPM. The carbohydrate chains present on SPACR could also provide sites for extensive crosslinking and participate in the formation of the ordered IPM lattice that surrounds the elongate photoreceptors projecting from the outer retinal surface.

Hyaluronan in the interphotoreceptor matrix of the eye: species differences in content, distribution, ligand binding and degradation.

Hyaluronan (HA) distribution in the posterior eye wall from the vitreous through the sclera, with special consideration to localization in the retina and interphotoreceptor matrix (IPM), was evaluated in human, bovine, guinea pig, dog, rat and mouse tissues using a specific probe for HA (bHABC, biotinylated hyaluronan binding complex). The sclera, some regions of the choroid and vitreous body was positive for HA, as was the basal lamina of the retina (inner limiting membrane). bHABC binding was detected in the IPM of all species studied except the mouse. Predigestion with Streptomyces hyaluronidase for 3 hr before bHABC application eliminated binding in the vitreous, choroid, sclera and basal lamina of the retina, but did not eliminate bHABC binding in the IPM. In tissues from all species studied, incubation for 6 hr with hyaluronidase eliminated bHABC binding in the IPM, except for two human samples. In these two human samples, HA specific binding in the IPM persisted even after 24 hr enzyme treatment. bHABC failed to bind to any tissue layer when bHABC was preincubated with hyaluronan oligosaccharides before application. The resistance of the IPM HA to hyaluronidase digestion may reflect extensive coverage of HA binding sites by ligands present in this compartment which hinder enzyme access. The absence of bHABC binding to the IPM when the probe is preincubated with HA oligosaccharides indicates that the binding reflects specific interaction with HA. We conclude that, with the exception of the mouse, HA is a prominent constituent of the IPM, where it may serve to organize the matrix by functioning as a basic scaffold to which other macromolecules in the insoluble IPM are attached.

Hyaluronan localization in tissues of the mouse posterior eye wall: absence in the interphotoreceptor matrix.

The distribution of hyaluronan (HA) in the posterior eye wall from the vitreous through the sclera, with special consideration to localization in the retina and interphotoreceptor matrix (IPM), was evaluated in mouse tissues using an HA specific probe (bHABC, biotinylated hyaluronan binding complex). The vitreous body was positive for HA, as was Bruch's membrane, expansive areas within the choroid, sclera and perimysial connective tissue of extraocular muscle. No HA-staining was detected in the IPM or in any other retina layer except for the basal lamina (inner limiting membrane of the retina) which abuts the vitreous. Predigestion of sections with trypsin or chondroitinase ABC before bHABC application did not produce additional HA-staining in the retina or IPM.

Increased retinal synthesis of heparan sulfate proteoglycan and HNK-1 glycoproteins following photoreceptor degeneration.

In an earlier analysis of the retinal biosynthesis of proteoglycan, we noted that, following photoreceptor degeneration in the rd (retinal degeneration) mouse, the remaining inner retina exhibited a marked elevation in synthesis of heparan sulfate proteoglycan (HSPG), well above the level observed in the normal (nondegenerate) retina, as well as a pronounced increase in sulfation of protein substrates. Biochemical and autoradiographic results of 35S-amino acid utilization reported here confirm that the 35SO4(2-) differences seen previously are accompanied by increased protein synthesis in the rd retina. An intact photoreceptor cell layer is neither a barrier to nor a sink for the amino acid precursor. Further, we have examined sulfate utilization in four other rodent strains with photoreceptor degenerations. In each of the models examined, an increase in retinal synthesis of 35SO4(2-)-labeled HSPG and glycoproteins occurs following photoreceptor degeneration. We have metabolically labeled with Na2(35)SO4 isolated retinal cultures from the following: (a) mice with light-induced photoreceptor degeneration; (b) rd mice; (c) transgenic mice with photoreceptor degeneration; (d) RCS rats; and (e) rats with light-induced photoreceptor degeneration. Comparisons were made with concurrent cultures of control nondegenerate retinal tissues. Protein and proteoglycan-enriched fractions were prepared from the incubation media and guanidine HCl/detergent extracts of the retinas by ion-exchange chromatography. The 35SO4(2-)-proteoglycans were identified by chondroitinase ABC and nitrous acid treatments. Retinas lacking photoreceptors produced at least five times the amount of 35SO4(2-)-HSPG found in control incubations. The RCS and light-damaged rats also showed increased synthesis of 35SO4(2-)-chondroitin sulfate proteoglycan relative to the control, through the increase was of lesser magnitude than the HSPG effect. 35SO4(2-)-protein in degenerate and light-damaged retinas always contained at least twice the radioactivity found in comparable control preparations. The bulk of the increased radiolabeling was found in N-linked oligosaccharides, including several recognized by the HNK-1 antibody. These data suggest that a sustained increase in HSPG and HNK-1 glycoprotein synthesis is a consistent response of inner retinal cells following loss of photoreceptors and is independent of the cause of photoreceptor degeneration.

Photoreceptor outer segment development in Xenopus laevis: influence of the pigment epithelium.

Opsin gene expression, synthesis, and photoreceptor outer segment morphology were evaluated during retinal development in Xenopus laevis. Retinal rudiments were harvested during in vivo development from embryonic stages 31 through 46 or were allowed to develop in vitro after removal from stage 33/34 embryos for 1, 2, or 3 days either with or without an investing pigment epithelium. Opsin mRNA was detected at stage 33/34 and the transcript level increased until stage 40 and remained at this level through stage 46. Opsin was first detected at stage 37/38 and progressively increased through stage 46. Rudimentary photoreceptor outer segment membranes occasionally appeared as early as stage 33/34 and they gradually increased in length, forming well-defined stacks of collapsed membranous saccules (discs) during in vivo development. The maturation of eye rudiments in culture was followed to determine how closely in vivo and in vitro development compare and to examine the ability of photoreceptors to differentiate when maintained in the absence of an overlying pigment epithelium (PE) layer. With the PE present, opsin mRNA as well as opsin content steadily increased over the entire culture period. After 1 day of culture, short cilia with minimal amounts of outer segment membranous material were present. By Day 3, the degree of outer segment differentiation corresponded morphologically to approximately stage 43 of in vivo development. When cultured in the absence of an investing PE, the opsin mRNA level increased minimally during the 3 days in culture. Opsin content increased, yet the relative amount was approximately 50% less than that present in retinas developing in the presence of the PE. Membranous material was elaborated; however, the outer segments appeared to be highly disorganized and formed whorl-like structures rather than the normal stacked disc morphology. These results suggest that the PE may be involved in regulating opsin at the transcriptional and/or translational levels and also participates in the organization of rod outer segment membranes.

Interphotoreceptor retinoid-binding protein (IRBP), a major 124 kDa glycoprotein in the interphotoreceptor matrix of Xenopus laevis. Characterization, molecular cloning and biosynthesis.

We have demonstrated that the neural retina of Xenopus laevis secretes into the extracellular matrix surrounding the inner and outer segments of its photoreceptors a glycoprotein containing hydrophobic domains conserved in mammalian interphotoreceptor retinoid-binding proteins (IRBPs). The soluble extract of the interphotoreceptor matrix contains a 124 kDa protein that cross-reacts with anti-bovine IRBP immunoglobulins. In vitro [3H]fucose incorporation studies combined with in vivo light and electron microscopic autoradiographic analysis, showed that the IRBP-like glycoprotein is synthesized by the neural retina and secreted into the interphotoreceptor matrix. A 1.2 kb Xenopus IRBP cDNA was isolated by screening a stage 42 (swimming tadpole) lambda Zap II library with a human IRBP cDNA under low-stringency conditions. The cDNA hybridizes with a 4.2 kb mRNA in adult Xenopus neural retina, tadpole heads as well as a less-abundant mRNA of the same size in brain. During development, IRBP and opsin mRNA expression correlates with photoreceptor differentiation. The translated amino acid sequence of the Xenopus IRBP clone has an overall 70% identity with the fourth repeat of the human protein. Sequence alignment with the four repeats of human IRBP showed three highly conserved regions, rich in hydrophobic residues. This focal conservation predicts domains important to the protein's function, which presumably is to facilitate the exchange of 11-cis retinal and all-trans retinol between the pigment epithelium and photoreceptors, and to the transport of fatty acids through the hydrophilic interphotoreceptor matrix.

Characterization of the interphotoreceptor matrix surrounding rod photoreceptors in the human retina.

Previous studies have documented the presence of specific lectin-binding domains in the insoluble interphotoreceptor matrix (IPM) isolated from human retina. Peanut agglutinin (PNA) selectively binds to cone matrix domains whereas wheat germ agglutinin (WGA) binds to matrix domains surrounding rods. In the present study, the rod-associated WGA-binding domains are further characterized using lectin-based cytochemistry and polyacrylamide gel electrophoresis in combination with neuraminidase digestion. The lectin-binding patterns of non-neuraminidase-treated samples are similar to those described in previous reports. After neuraminidase treatment, both rod and cone matrix domains demonstrate PNA binding while the binding of WGA to rod matrix domains is reduced. However, the binding of WGA to photoreceptor outer segments is not affected by neuraminidase. Blots of IPM proteins probed with lectins indicate that the WGA-binding macromolecules are represented as a group of high molecular weight glycoproteins, whereas the PNA-binding components are represented as a group of lower molecular weight glycoproteins. The major WGA-binding glycoprotein (147 kDa) shows reduced binding affinity to WGA and increased binding affinity to PNA following neuraminidase treatment. Further, this 147-kDa glycoprotein, although similar in molecular weight to IRBP (interphotoreceptor retinol-binding protein) (141 kDa), is not recognized by the lectin, concanavalin A (Con A), or by an anti-IRBP antibody. Our data suggest that: (1) the major component of the WGA-binding domain demonstrated by polyacrylamide gel analysis is a glycoprotein with a molecular weight of 147 kDa containing galactose residues that are masked by terminal sialic acid residues; and (2) the binding of WGA to photoreceptor outer segments is resistant to neuraminidase, consistent with the earlier reports that WGA-binding domains of photoreceptor outer segments may not be sialyl-containing glycoconjugates.

Interphotoreceptor matrix in the human retina: cone-like domains surround a small population of rod photoreceptors.

The interphotoreceptor matrix (IPM) in mammalian retinas is subdivided into rod and cone specific compartments: peanut agglutinin (PNA) binding glycoconjugates are associated with cones, whereas wheat germ agglutinin (WGA) binding glycoconjugates are associated with rods. To establish the identity of a photoreceptor cell type in the human retina with rod dimensions but with a matrix domain which stains with PNA, double label studies, using PNA-ferritin to decorate the extracellular domains and immunocytochemical techniques using a rod specific anti-opsin antibody were conducted. The PNA-binding domains were observed in the cone-associated IPM as well as in the IPM surrounding a small population of rod-shaped photoreceptors. The outer segments of these rod-shaped photoreceptors showed intense labeling with a rod specific anti-opsin antibody as did all other rods which were free of PNA-labeling. A quantitative analysis of all retinal quadrants indicates that this novel rod represents approximately 0.3% of the total rod population in the human retina.

Proteoglycans in the mouse interphotoreceptor matrix. V. Distribution at the apical surface of the pigment epithelium before and after retinal separation.

The distribution of chondroitin sulfate proteoglycans in the mouse interphotoreceptor matrix (IPM) proximal to the retinal pigment epithelium (RPE) was evaluated with EM histochemical techniques using Cupromeronic Blue (CmB) before and after retinal separation. Densely packed, sheet-like processes surrounding vesicle-like compartments containing CmB staining were normal constituents of the IPM at the apical surface of the RPE. Most of the vesicle-like compartments, which appeared to be isolated from the IPM in single section profiles, were found to be continuous with the IPM when three-dimensional reconstructions of serial thin sections were evaluated. Analyses of stereo image pairs of semithin sections visualized by electron spectroscopic imaging (ESI) also demonstrated that the CmB stained proteoglycans in the lumen of these pseudovesicles were in continuity with the CmB stained components present in the IPM. Moreover, ESI demonstrated that the CmB stained profiles formed an elaborate interconnecting network extending from the apical border of the RPE to the level of the outer limiting membrane of the retina. After removal of the retina, rinsing of the RPE with Ringer's solution prior to fixation eliminated proteoglycan staining near the base of the apical processes, whereas staining near the tips of these processes remained. The CmB stained filaments remaining after rinsing were thicker and shorter, and made fewer interconnections than those in the non-separated preparations. These results suggest that two types of chondroitin sulfate proteoglycans are present in the IPM which differ in distribution and in the degree of aqueous solubility. Additionally, a closely associated retina appears to be required for maintenance of the normal structure of proteoglycans associated with the RPE surface. The elaborate topography at the RPE apical surface may play a role in the delivery and/or recovery of components of the IPM.

Regional variation within the interphotoreceptor matrix from fovea to the retinal periphery.

The interphotoreceptor matrix in the human retina comprises a highly organised extracellular compartment. Using fluorescent labelled peanut agglutinin and the cationic dyes Cupromeronic and Cuprolinic Blue, unique cylindrical domains surrounding cone photoreceptors can be demonstrated. These cone specific domains are relatively insoluble and are closely adherent to cone photoreceptors and to the pigment epithelium, suggesting that these structures may play a role in retinal attachment.

Serotoninergic neurons in the retina of Xenopus laevis: selective staining, identification, development, and content.

Uptake of 3H-serotonin followed by autoradiography, and uptake of the serotonin analog 5,7-dihydroxytryptamine (5,7-DHT), with subsequent staining, were each used to define a unique set of neurons in the retina of the African clawed frog, Xenopus laevis. Both techniques demonstrated the same population of neurons, on the basis of perikaryal size, shape, and position within the retina. Two classes of amacrine cells accumulated 5,7-DHT at the proximal (vitread) margin of the inner nuclear layer; the two classes were distinguished by the size of their perikarya. Two similar populations of cells, observed in the ganglion cell layer with lower frequency, may represent "displaced" counterparts of these two amacrine cell types. A class of bipolar cells whose perikarya were located in middle-to-distal regions of the inner nuclear layer also accumulated 5,7-DHT and 3H-serotonin. Processes of these cells contributed to a dense plexus of fine fibers that appeared evenly distributed throughout the inner plexiform layer. 3H-Serotonin-accumulating cells first appeared in the developing retina at stage 35/36, a time immediately after retinal stratification but before elaboration of either plexiform layer. Electron microscopic analysis permitted an identification of 3H-serotonin-accumulating terminals in the inner plexiform layer. Serotonin-labeled terminals containing conventional contacts, suggestive of amacrine cells, were presynaptic to unidentified processes and postsynaptic to bipolar cells. Labeled terminals containing ribbon contacts, indicative of bipolar cells, were postsynaptic to amacrine cells. The amount of serotonin contained in isolated retinas was 15 pmol/mg protein as measured by HPLC with electrochemical detection. We attempted to stimulate the release of accumulated 3H-serotonin from mature retinas by increasing the K+-concentration in the bathing medium. Although preloaded glycine is readily released from 14C-glycine-accumulating neurons, from the same retinas there was no calcium-dependent, K+-stimulated release of 3H-serotonin. This finding suggests that serotonin and glycine are processed differently by retinal neurons, the consequence of which results in differing responses to 40 mM K+.