Biosynthesis and intracellular sorting of growth hormone-viral envelope glycoprotein hybrids.

Various aspects of the biogenetic mechanisms that are involved in the insertion of nascent plasma membrane proteins into the endoplasmic reticulum (ER) membrane and their subsequent distribution through the cell have been investigated. For these studies chimeric genes that encode hybrid proteins containing carboxy-terminal portions of the influenza virus hemagglutinin (154 amino acids) or the vesicular stomatitis virus envelope glycoprotein (G) (60 amino acids) linked to the carboxy terminus of a nearly complete secretory polypeptide, growth hormone (GH), were used. In in vitro transcription-translation experiments, it was found that the insertion signal in the GH portion of the chimeras led to incorporation of the membrane protein segments into the ER membrane. Effectively, GH became part of the luminal segment of membrane proteins of which only very small segments, corresponding to the cytoplasmic portions of the G or HA proteins, remained exposed on the surface of the microsomes. When the chimeric genes were expressed in transfected cells, the products, as expected, failed to be secreted and remained cell-associated. These results support the assignment of a halt transfer role to segments of the membrane polypeptides that include their transmembrane portions. The hybrid polypeptide containing the carboxy-terminal portion of HA linked to GH accumulated in a juxtanuclear region of the cytoplasm within modified ER cisternae, closely apposed to the Golgi apparatus. The location and appearance of these cisternae suggested that they represent overdeveloped transitional ER elements and thus may correspond to a natural way station between the ER and the Golgi apparatus, in which further transfer of the artificial molecules is halted. The GH-G hybrid could only be detected in transfected cells treated with chloroquine, a drug that led to its accumulation in the membranes of endosome or lysosome-like cytoplasmic vesicles. Although the possibility that the chimeric protein entered such vesicles directly from the Golgi apparatus cannot be ruled out, it appears more likely that it was first transferred to the cell surface and was then internalized by endocytosis.

Polarization of the Na+, K(+)-ATPase in epithelia derived from the neuroepithelium.

The neuroepithelium generates a fascinating group of epithelia. One of their intriguing properties is how they polarize the distribution of the Na+, K(+)-ATPase. Typically, this ion pump is concentrated in the basolateral membrane, but it is concentrated in the apical membranes of the retinal pigment epithelium and the epithelium of the choroid plexus. A comparison of their development with that of systemic epithelia yields insights into how cells polarize the distribution of this and other membrane proteins. The polarization of the Na+, K(+)-ATPase depends upon the interplay between different sorting signals and different types of polarity mechanisms. These include intracellular targeting signals that direct the delivery of newly synthesized proteins, and maintenance signals that stabilize proteins in the proper membrane domain. Conflicting signals appear to be arranged in a hierarchy that can be rearranged as cells respond to certain environmental stimuli. Part of this response is mediated by changes in the distribution and composition of the cortical cytoskeleton.

Differentiation and transdifferentiation of the retinal pigment epithelium.

The retinal pigment epithelium (RPE) lies between the retina and the choroid of the eye and plays a vital role in ocular metabolism. The RPE develops from the same sheet of neuroepithelium as the retina and the two derivatives become distinguished by different expression patterns of a number of transcription factors during embryonic development. As the RPE layer differentiates it expresses a set of unique molecules, many of which are restricted to certain regions of the cell. PRE cells undergo both a loss of polarity and a loss of expression of many of these cell type-specific molecules when placed in monolayer culture. The RPE of many species, including mammals, can be induced to transdifferentiate by growth factors such as basic fibroblast growth factor. Under the influence of such factors the RPE is triggered to alter expression of a wide array of molecules and to take on a retinal epithelium fate, from which differentiated retinal cell types including rod photoreceptors can be produced.

A growth hormone-vesicular stomatitis virus G hybrid protein is rapidly degraded in lysosomes following transport to the cell surface.

We have expressed the hybrid protein, GHG3, in baby hamster kidney cells to study protein turnover. GHG3 contains the cytoplasmic and transmembrane domains of vesicular stomatitis virus G protein linked to the C-terminus rat growth hormone. Turnover of GHG3 was prevented by lysosomal inhibitors (leupeptin, chloroquine, primaquine or monensin), while the accumulated GHG3 was localized to intracellular vesicles, results indicating that degradation occurred in lysosomes. The kinetics of degradation at 34 degrees C were determined in pulse-chase studies of metabolically labeled cells. After a lag period of 1 h, degradation was rapid (t1/2 = 1.25 h). The fate of GHG3 during the lag period was determined by immunofluorescence. We detected GHG3 on the cell surface when growth hormone antiserum was added to the growth medium 90 min prior to fixation and staining. No staining was observed if protein synthesis was inhibited with cycloheximide 90 min prior to the addition of growth hormone antiserum, a result indicating that GHG3 was rapidly removed from the cell surface. Unless the cells were pretreated with cycloheximide, antiserum was also detected in intracellular vesicles, which showed that GHG3 was endocytosed. These data indicate that a pool of GHG3 is transported rapidly to the cell surface, endocytosed and with little or no recycling directed to lysosomes for degradation.

The neural retina maintains integrins in the apical membrane of the RPE early in development.

PURPOSE: The retinal pigment epithelium (RPE) lines the interface between the neural retina and the choroid. Early in chicken development, the beta 1 family of integrins resides in the apical (facing the neural retina) and basolateral (facing the choroid) membranes of RPE. Later in development, integrins reside only in the basolateral membranes, which is more typical of simple transporting epithelia. The authors examined whether the distribution of integrins is regulated by the neural retina. METHODS: Individual integrins were examined by studying the individual alpha-subunits that form heterodimers with the beta 1 subunit. The expression and distribution of these subunits were determined by immunoblotting and immunocytochemistry. RESULTS: Subunits alpha 3 and alpha 6 exemplified contrasting behaviors. Early and late in development, alpha 3 was found only in the basal membranes. As was beta 1, the distribution of alpha 6 was nonpolarized early in development but was basal later in development. The effect of the immature neural retina was determined by reconstituting the RPE:neural retinal interface in explant culture. Absent the neural retina, alpha 6 and beta 1 were removed from the apical membrane. When present, the immature neural retina maintained both subunits in the apical membrane. The neural retina was effective only if the outer (primordial photoreceptor) surface of the retina apposed the RPE. CONCLUSIONS: These data suggest that matrix or intercellular interactions determine the distribution of individual integrins. Further, the changes in integrin distribution during development reflect the maturation of the primordial interphotoreceptor matrix or photoreceptor cell layer.

A culture model of development reveals multiple properties of RPE tight junctions.

PURPOSE: A culture model was used to examine the development of tight junctions in the retinal pigment epithelium (RPE). METHODS: Chick RPE was isolated on embryonic day 7 (E7), E10 or E14 and cultured on laminin-coated filters. Barrier properties were stimulated with E14 retinal conditioned medium. Morphology was characterized by confocal microscopy. Permeability was determined by measuring the flux of horseradish peroxidase (HRP), radiolabeled inulin and mannitol, and the transepithelial electrical resistance (TER). Changes in the expression of ZO-1 and a related protein, ZO-1LP, were determined by immunoblotting. RESULTS: RPE from each age formed epithelial monolayers of similar height, but the density of the cultures varied in parallel with density changes in vivo. The cultures appeared to regulate the permeability to ions and nonionic solutes independently. With embryonic age, there was a progressive decrease in permeability that first affected larger and then smaller tracers. Despite a small decrease in the permeability to mannitol, there was a large decrease in the permeability to ions. This suggests that in E14 cultures tight junctions discriminated by charge, as well as size. Although E14 retinal conditioned medium reduced the permeability to all solutes, it appeared to regulate size discrimination more than charge discrimination. Despite large effects on permeability, conditioned medium had no effect on the expression of ZO-1 or ZO-1LP. CONCLUSIONS: The ability of tight junctions to discriminate on the basis of charge and size is regulated independently during development. The permeability of tight junctions cannot be predicted by the level of ZO-1 expression.

Polarity and the development of the outer blood-retinal barrier.

The retinal pigment epithelium (RPE) is a monolayer that separates the outer surface of the neural retina from the choriocapillaris. Because the choriocapillaris is fenestrated, it is the RPE that forms the outer blood-retinal barrier and regulates the environment of the outer retina. Like all epithelia and endothelia, the ability of RPE to regulate transepithelial transport depends upon two properties: apical tight junctions to retard diffusion through the paracellular spaces of the monolayer, and an asymmetric distribution of proteins to regulate vectorial transport across the monolayer. During development, these properties form gradually. Initially, the tight junctions are leaky, and the RPE exhibits only partial polarity. As the neural retina and choriocapillaris develop, there are progressive changes in the composition of the apical junctional complexes, the expression of cell adhesion proteins, and the distribution of membrane and cytoskeletal proteins. Development can be used to dissect the multiple mechanisms that establish and maintain polarity and barrier function. These mechanisms are regulated by the interactions that develop between the RPE and its neighboring tissues. This review discusses the remodeling of the apical, lateral and basal plasma membranes of RPE that occurs during normal development, and establishes a framework to integrate the data obtained from multiple species. It examines the progress in understanding how environmental interactions regulate this development.

Regulation of glucose transporters during development of the retinal pigment epithelium.

The retinal pigment epithelium (RPE) separates the outer retina from its blood supply. To satisfy the retina's large requirement for glucose, the RPE expresses high levels of glucose transporters. In most rat cells, the transporter GLUT3 provides a basal level of transport, but the expression of GLUT1 can be regulated. The opposite is true in chicken (P. Wagstaff, H.Y. Kang, D. Mylott, P.J. Robbins, M.K. White, Characterization of the avian GLUT1 glucose transporter: differential regulation of GLUT1 and GLUT3 in chicken embryo fibroblasts, Mol. Biol. Cell 6 (1995) 1575-1589). We examined chick RPE to determine which isoform is regulated during development, and if the neural retina regulates GLUT expression. By RT-PCR, RPE expressed GLUT1 and GLUT3, but not GLUT2. Only the level of GLUT1 increased between E5 and E18. A corresponding increase in GLUT1 protein was observed by immunoblotting. Most of the increase occurred between E14 and E18, which corresponds to the late stage of tight junction development. A culture model of development was used to examine the intermediate phase, which extends from E7 to E14. While medium conditioned by the neural retina decreased paracellular diffusion across the tight junctions, it increased diffusion through the glucose transporters. Unlike mammals, chick upregulates different isoforms in quiescent RPE and proliferating fibroblasts. Further, the upregulation of glucose transport is coordinated with the development of tight junctions in the blood-retinal barrier.

Two secreted retinal factors regulate different stages of development of the outer blood-retinal barrier.

The retinal pigment epithelium (RPE) lies at the interface between the neural retina and the choriocapillaries where it forms a blood-retinal barrier. Like endothelial regions of the blood-brain barrier, the development of the RPE barrier is a gradual, multistep process. A culture model of chick RPE was used to study this development. The permeability of the tight junctions that limit diffusion between neighboring RPE cells was measured as the transepithelial electrical resistance (TER). Embryonic day 14 (E14) retinas were used to make a conditioned medium that lowered the permeability of cultured RPE. The TER of cultures prepared from E14 RPE was twice that of E7 RPE. In each culture, retinal conditioned medium increases the TER 2-2.5 fold. The active factors of conditioned medium that affected each culture had different physical properties. The factor that affected E7 was protease-resistant with a Mr<10 kDa, but the factor that affected E14 appeared to be a protein of approximately 49 kDa. Unlike the effect of astrocyte conditioned medium on endothelia, retinal conditioned medium did not act synergistically with cAMP. These data indicate that the chick retina, which lacks astrocytes, uses different diffusible factors to regulate different stages of tight junction development.

Kinetics and protein subunit interactions of Escherichia coli phosphatidylserine decarboxylase in detergent solution.

Phosphatidylserine decarboxylase from Escherichia coli, an intrinsic membrane protein, catalyzes the conversion of phosphatidylserine to phosphatidylethanolamine. The physical and kinetic properties of the purified enzyme were studied in several detergents under assay conditions. The active form of the enzyme is an oligomer, probably a trimer, and the enzyme activity was unaffected by the concentration of the nonionic poly(oxyethylene) ether detergent present in the assay medium, so long as the detergent micelle/substrate mole ratio was less than one. When this ratio was greater than one, the detergent acted as an inhibitor by competing with enzyme-containing micelles for substrate. The zwitterionic and bile salt detergents that were tested inactivated the enzyme by dissociating the oligomer. The native, Triton X-100 solubilized, enzyme was modified with a cross-linking reagent. Activity of the cross-linked enzyme was retained after the Triton X-100 was replaced by a zwitterionic sulfobetaine detergent and conformed to the same kinetic model as with the poly(oxyethylene) ether detergents. The cross-linked enzyme was also active when solubilized by the bile salt detergents although the activity did not conform to any simple kinetic model. These data indicate that the oligomer is the active form of the enzyme under assay conditions and that certain nondenaturing detergents can inactivate this enzyme by dissociating the enzyme complex.

The distribution of Na+,K(+)-ATPase in the retinal pigmented epithelium from chicken embryo is polarized in vivo but not in primary cell culture.

The polarity of retinal pigmented epithelia (RPE) from chicken embryos was studied in primary cell culture. Since cultured RPE approximates the morphological polarity of RPE in vivo, we investigated whether this polarity extends to the distribution of plasma membrane proteins that are peculiar to RPE. In contrast to other epithelia, the Na+,K(+)-ATPase of RPE is located in the apical rather than basolateral plasma membrane. To examine this property, we cultured RPE on extracellular matrix-coated filters. Primary cultures were compared to embryonic RPE in situ using electron microscopy and indirect immunofluorescence of frozen sections. The viability and morphology of RPE was improved by using a serum-free medium containing a bovine pituitary extract in conjunction with an extracellular matrix coating derived from Engelbreth-Holm-Swarm tumors. Cultured RPE mimicked the morphology of RPE in vivo with microvilli and junctional complexes on the apical pole and infoldings along the basolateral plasma membrane. Functional tight junctions formed as demonstrated by an EDTA-sensitive, transepithelial electrical resistance, and by the retention of [3H]inulin added to the apical chamber. In 2 hr, only 4-6% of the [3H]inulin crossed the monolayer, compared to 24% in control filters. Despite these features of polarity, the Na+,K(+)-ATPase was detected in both apical and basolateral membranes by immunofluorescence. In embryonic eyes in which the neural retina was removed, the Na+,K(+)-ATPase was confined to the apical membrane. In addition, the polarity of cultured RPE was probed with vesicular stomatitis virus. In contrast to other epithelia, budding virus particles were observed emerging from the apical, as well as basolateral, domain further suggesting the cultured cells were only partially polarized. These data indicate that structural criteria are inadequate to determine if cultured RPE have become polarized in the same manner as the epithelium in vivo.

Basement membrane stimulates the polarized distribution of integrins but not the Na,K-ATPase in the retinal pigment epithelium.

The basement membrane stimulates the differentiation and polarity of simple transporting epithelia. We demonstrated for the retinal pigment epithelium (RPE) of chicken embryos that polarity develops gradually. Although the RPE and an immature basement membrane are established on embryonic day 4 (E4), the distribution of the Na,K-ATPase and a family of basement membrane receptors containing the beta 1 subunit of integrin is nonpolarized. The percentage of polarized cells increases gradually until cells in all regions of the epithelium are polarized on E11. During this time, the basement membrane increases in size and complexity to form Bruch's membrane. To study the ability of the basement membrane to stimulate the polarized distribution of the beta 1 integrins or the Na,K-ATPase, RPE was harvested from E7, E9, or E14 embryos and cultured on Bruch's membrane isolated (in association with the choroid) from E14 embryos. As a control, the RPE was plated on the side of the choroid lacking a Bruch's membrane. The distribution of the beta 1 integrins and the Na,K-ATPase was determined by indirect immunofluorescence. Bruch's membrane stimulated the polarized distribution of the beta 1 integrins regardless of the developmental age of the RPE even though E7 RPE is nonpolarized in vivo. To examine the role of individual matrix components, RPE was plated on matrix-coated filters. The polarized distribution of the beta 1 integrins was stimulated by laminin, collagen IV, and Matrigel but not by fibronectin. Interestingly, laminin and collagen IV are present in the basement membrane on E4 when RPE is not polarized in vivo. Under no circumstances was the distribution of the Na,K-ATPase polarized. These data indicate that the basement membrane influences the distribution of a subset of plasma membrane proteins but that other factors are required for full polarity.

Diffusible, retinal factors stimulate the barrier properties of junctional complexes in the retinal pigment epithelium.

The retinal pigment epithelium lies at the interface between the neural retina and the choriocapillaris where it forms a blood-retinal barrier. Barrier function requires a polarized distribution of plasma membrane proteins and 'tight' tight junctions. During chicken embryogenesis, these features develop gradually. Although terminal junctional complexes are established by embryonic day 4, the distribution of the Na+/K(+)-APTase is not polarized in all cells of the epithelium until embryonic day 11. Similarly, the tight junctions of early embryos are leaky, but become tight by hatching (embryonic day 21). We used primary cell culture to examine the molecular basis of this gradual induction of polarized function. Pigment epithelium harvested from embryonic day 7, and cultured on filters, formed monolayers coupled by junctional complexes. The distribution of the Na+/K(+)-ATPase was non-polarized and the tight junctions were leaky with a transepithelial electrical resistance of 20-30 omega cm2. To isolate diffusible factors that stimulate the transepithelial electrical resistance, neural retinas from embryonic day 7, 14 or 16 embryos were incubated at 37 degrees C in base medium for 6 hours. The conditioned medium was added to the apical chamber of freshly cultured pigment epithelium. The distribution of the Na+/K(+)-ATPase became basolateral, and the electrical resistance gradually increased two to three times over 6 days. The increase in electrical resistance corresponded to a decrease in the rate of [3H]inulin diffusion across the monolayer. The effectiveness of the conditioned medium increased steadily with increasing age of the neural retina. Rather than increased production of an active factor, apparently different active factors were produced at different ages.(ABSTRACT TRUNCATED AT 250 WORDS)

Post-translational protein modification in the endoplasmic reticulum. Demonstration of fatty acylase and deoxymannojirimycin-sensitive alpha-mannosidase activities.

We have previously described a hybrid protein, GHHA, that contains a fragment of the influenza hemagglutinin joined to the C terminus of a nearly complete rat growth hormone (Rizzolo, L.J., Finidori, J., Gonzalez, A., Arpin, M., Ivanov, I.E., Adesnik, M., and Sabatini, D.D. (1985) J. Cell Biol. 101, 1351-1362). GHHA was transported from the rough endoplasmic reticulum (ER) to a smooth cisterna, continuous with the rough ER, but proximal to the Golgi apparatus. We have now labeled GHHA with [3H]palmitate, demonstrating that fatty acylation can occur in the ER. As expected for a thioester linkage, the label was released from GHHA by hydroxylamine and identified as palmitic acid by thin-layer chromatography. In a second study, we analyzed the structure of the N-linked carbohydrate chain of GHHA. The N-linked oligosaccharides, all high-mannose type, were released by endoglycosidase H and size-fractionated by high pressure liquid chromatography. The predominant structures were Glc1Man8GlcNAc and Man8GlcNAc, indicating that only 2 or 3 glucose and 1 mannose residues were removed from the original Glc3Man9GlcNAc2. Determination of the structure by acetolysis fragmentation indicated that a single Man8GlcNAc isomer was formed by a deoxymannojirimycin-sensitive alpha-mannosidase. This contrasts with a previously characterized ER alpha-mannosidase (Bischoff, J., Liscum, L., and Kornfeld, R. (1986) J. Biol. Chem. 261, 4766-4774) that generates the same isomer, but is deoxymannojirimycin-resistant. These data suggest the possibility that different enzymes are partitioned within the ER.

Denaturation of the tryptic fragments of the calcium (II) adenosine triphosphatase from sarcoplasmic reticulum by guanidinium hydrochloride.

Primary and secondary fragments of the Ca2+-adenosine triphosphatase from sarcoplasmic reticulum are resistant to complete denaturation by guanidinium hydrochloride, a property characteristic of many intrinsic membrane proteins. None of the fragments display a single cooperative transition from ordered structure to random coil suggesting each fragment contains several domains of differing resistance to guanidinium hydrochloride denaturation. The data suggest that the native enzyme has at least three membrane-embedded domains, with an externally accessible link between each.

Behavior of fragmented calcium (II) adenosine triphosphatase from sarcoplasmic reticulum in detergent solution.

The behavior of Ca2+-ATPase from sarcoplasmic reticulum in detergent solution was compared with that of Ca2+-ATPase which had been cleaved in half by limited trypsin digestion. Attempts to dissociate the fragments (I and II) with an excess of detergent micelles demonstrated that fragments I and II are structurally dependent upon each other, and that they must be denatured in order to be dissociated. Partial dissociation of the fragmented ATPase was found to occur in the bile salt detergents, deoxycholate and cholate, and optical data showed that there was an accompanying change in conformation. No dissociation of the fragmented ATPase was observed in nonionic detergents. The fragmented ATPase retained the same specific activity and stability as the intact ATPase under a variety of conditions when solubilized in Tween 80 or dodecyl octaoxyethylene glycol monoether. The data demonstrate that the noncovalent interactions that maintain the native conformation of the ATPase are not affected by either trypsin cleavage or solubilization in nonionic detergent solution.

Differential regulation of tight junction permeability during development of the retinal pigment epithelium.

The retinal pigment epithelium (RPE) is an epithelial region of the blood-brain barrier. During embryogenesis, permeability of the barrier gradually decreases. A culture model of RPE development revealed differences in how tight junctions regulate the paracellular diffusion of ionic and nonionic solutes (Ban Y and Rizzolo LJ. Mol Vis 3: 18, 1997). To examine these differences, the permeation of ionic and nonionic monosaccharides was compared with mannitol, and the permeation of the alkali metals was compared with sodium. The order of permeation was 3-O-methlyglucose = glucosamine = mannitol > N-acetylneuraminic acid. The ratio of N-acetylneuraminic acid to mannitol permeability decreased with embryonic age of the RPE or exposure to retinal-conditioned medium. Neither the ratio nor the permeability was affected by inhibiting transcytosis. The ratio increased if tight junctions were disrupted in low-calcium medium. The permeation of cations followed the sequence cesium > rubidium > potassium = sodium > lithium and was unaffected by embryonic age or retinal-conditioned medium. These results are considered in terms of a model in which the size distribution, charge, or number of open junctional pores could be modulated. It suggests that different subpopulations of pores can be regulated independently during development.

Apical orientation of the microtubule organizing center and associated gamma-tubulin during the polarization of the retinal pigment epithelium in vivo.

Simple epithelial cells express a morphological and functional polarity along their apical-to-basal axis. During the development of epithelia, a unique reorganization of microtubule arrays is thought to play a fundamental role in the establishment of cell polarity. To begin to understand this process in vivo, we have determined the distribution of gamma-tubulin within developing chicken retinal pigment epithelium (RPE). gamma-Tubulin is a recently discovered centrosomal protein that plays a role in nucleating microtubule growth from the centrosome. Although the RPE monolayer becomes established during embryonic Day 3, cell polarity gradually develops and matures over the next 10-13 days. Our studies reveal that gamma-tubulin is located in a distinct focus subjacent to the apical membrane by embryonic Day 3, the beginning of the polarization process. Using primary cell cultures, we examined the relationship between the establishment of junctional complexes and the reorganization of microtubule arrays. Despite the recovery of junctional complexes and a transepithelial electrical resistance, cultured cells failed to relocate gamma-tubulin foci to a position subjacent to the apical membrane. Rather, these foci remained in the juxtanuclear region. These data indicate that the rearrangement of unique, epithelial microtubule arrays requires more than cell-cell and cell-basement membrane interactions.

Protein-binding domains of the tight junction protein, ZO-2, are highly conserved between avian and mammalian species.

The tight junction is composed of many proteins and includes three members of the MAGUK (membrane-associated, guanylate kinase-like) protein family: ZO-1, ZO-2, and ZO-3. ZO-2 was cloned and sequenced from embryonic chicken retina. Antibodies against a short ZO-2 peptide immunolabeled the outer limiting membrane (an adherens junction of the neural retina) and the apical junctional complexes of the retinal pigment epithelium. Each ZO family member contains a homologous series of protein-binding domains: three distinct PDZ domains and src homology 3 (SH3), guanylate kinase-like (GuK), and acidic domains. Compared with human and canine ZO-2s, the PDZ and SH3 domains are the most conserved (90-95% amino acid sequence identity). These domains are only 50-71% identical with the homologous domains of ZO-1 and ZO-3. Although the sequence is less conserved for regions that link the protein-binding domains, the length of those regions is conserved in ZO-2s. The postacidic (C-terminal) region is the least conserved. The evolutionary pressure to maintain the sequence of the protein-binding domains suggests that homologous domains have different functions in ZO-1, ZO-2, and ZO-3.

Remodeling of junctional complexes during the development of the outer blood-retinal barrier.

BACKGROUND: The retinal pigment epithelium (RPE) forms the outer blood-retinal barrier by separating the neural retina from fenestrated capillaries in the choroid. The barrier depends upon tight junctions within the apical junctional complexes that bind neighboring cells. During development, permeability decreases as the apical junctional complex gradually matures. To investigate this process, the composition of the apical junctional complex was monitored during RPE development in chicken embryos. METHODS: Permeability was monitored by incubating freshly isolated RPE/choroid in medium containing horseradish peroxidase followed by histochemical staining and electron microscopy. The expression of the tight junction proteins, ZO-1 and occludin, was determined by immunofluorescence and immunoblotting. Development of the RPE apical junctional complex was to compared to the homologous complex that forms the outer limiting membrane of the neural retina. RESULTS: The apical junctional complex of the RPE was permeable to horseradish peroxidase until embryonic day 10-12. Two putative forms of ZO-1 had approximately the same molecular mass as mammalian ZO-1 and were present in the apical junctional complexes at different stages of development. We identified one form as ZO-1, because it was present in mature RPE and shared an epitope with the rodent isoforms, ZO-1 alpha+ and ZO-1 alpha-. The second form lacked this epitope but was identified by a polyclonal antibody to ZO-1. It was designated the ZO-1-like protein (ZO-1LP). On embryonic day 3, occludin and ZO-1LP were observed along the apical surface of the neuroepithelium that gave rise to the RPE and the neural retina. In the neural retina, occludin expression decreased just before inner segments were formed, but ZO-1LP expression continued in the outer limiting membrane throughout development. During RPE development, occludin expression was constant or increased slightly. By contrast, ZO-1LP was gradually replaced by ZO-1 and total ZO-1 immunoreactive proteins decreased more than 10x. CONCLUSIONS: A gradual change in the composition of the apical junctional complexes accompanied the period of barrier formation. In RPE, ZO-1 gradually replaced ZO-1LP, but the decrease in ZO-1 expression suggests its functions during junction formation are not directly related to junction permeability. By contrast, occludin was lost and ZO-1LP retained where an adherens junction forms the permeable, outer limiting membrane.