The retinal pigment epithelium: a versatile partner in vision

Abstract

SUMMARY The retinal pigment epithelium (RPE) is a monolayer of cuboidal cells that lies in close association with the rod and cone photoreceptors. This epithelium has diverse features, three of which are discussed in some detail in this review, namely the daily phagocytosis of rod and cone outer segment fragments that are shed from their distal ends; the uptake, processing, transport and release of vitamin A (retinol) and some of its visual cycle intermediates (retinoids); and some of the aspects of its apical and basolateral membrane polarity that are the reverse of most other epithelia. Phagocytosis takes place at the apical surface via membrane receptor-mediated processes that are not yet well defined. Retinol uptake occurs at both the basolateral and apical surfaces by what appear to be separate receptor-mediated processes. The release of a crucial retinoid, 11-cis retinaldehyde (11-cis retinal), occurs solely across the apical membrane. Delivery of retinol across the basolateral membrane is mediated by a retinol binding protein (RBP) that is secreted by the liver as a complex with retinol (vitamin A). Within the cell, retinol and its derivatives are solubilized by intracellular retinoid binding proteins that are selective for retinol (cellular retinol binding protein, CRBP) and 11-cis retinoids (cellular retinal binding protein, CRALBP). Release of 11-cis retinal across the apical membrane and re-uptake of retinol from the photoreceptors during the visual cycle is promoted by an intercellular retinoid binding protein (IRBP). Na,K-ATPase, the membrane-integrated enzyme required to set up the ion gradients that drive other ion transporters, is largely localized to the apical membrane. This is the reverse of most epithelia. The RPE expresses the enveloped viral G protein and hemagglutinin on its basolateral and apical surface, respectively and does not appear to possess a general scheme for reversal of memrane protein polarity. Therefore possible alternative mechanisms for this reversal in Na,K-ATPase polarity are discussed. They include unique domains in the primary amino acid sequence of Na,K-ATPase subunits, cytoskeletal elements and components of the extracellular matrix. The precise mechanism remains unresolved.