S.-L. Fong

Vitamin a transport between retina and pigment epithelium—an interstitial protein carrying endogenous retinol (interstitial retinol-binding protein)

We have demonstrated and partially characterized an interstitial retinol-binding protein (IRBP) confined to bovine interphotoreceptor matrix (IPM). The native protein is a concanavalin A-binding glycoprotein with a mol. wt of 260 k as measured by gel-filtration and size-exclusion high-performance liquid chromatography. On SDS-gels, its mol. wt is 140-145 k. Since the protein is glycosylated, this value is probably too high. Hence, the native protein may be a dimer consisting of two identical subunits. The endogenous ligand has been analyzed by high-performance liquid chromatography--it consists mainly of all-trans retinol. Occasionally, retinal and 11-cis retinol are also associated with it. The amount of retinol bound to IRBP increases when the eyes are illuminated. The total binding capacity was estimated to represent 4-5% of the retinol released from a total rhodopsin bleach. We have established that, like serum retinol-binding protein, IRBP can be also bind retinoic acid, although it has not been established that retinoic acid is an endogenous ligand. The fluorescence emission lambda max for IRBP with its native ligand is at 470 nm and the excitation lambda max for this fluorescence is at 333 nm. Other retinoid carriers in the interphotoreceptor matrix have molecular weights of about 15 and 33 k. These probably correspond to cellular retinol- and retinal-binding proteins, respectively. Since both proteins have been identified in the pigment epithelium and retina cytosols, their presence in the IPM could be a result of cell damage. We conclude that interstitial retinol-binding protein is the best candidate for a transport protein carrying retinol between the rod outer segments and the pigment epithelium.

Rhodopsin, 11-cis vitamin A, and interstitial retinol-binding protein (IRBP) during retinal development in normal and rd mutant mice

Biochemical and immunological techniques were used to determine the emergence of interstitial retinol binding protein (IRBP), rhodopsin, and stored retinyl esters (all-trans and 11-cis) during retinal development in normal and rd mice. IRBP could be demonstrated at embryonic Day 17 (E17), corresponding to an early stage of inner segment development. Although all-trans retinyl esters were present earlier, 11-cis retinyl esters did not appear until postnatal Days 6-7 (P6-P7), corresponding to rod outer segment (ROS) disc formation. Rhodopsin was detected at the same developmental stage. The proportion of 11-cis retinyl esters reached a maximum of 40-50% at P15-P20. Thereafter, the proportion dropped, due to more rapid accumulation of the all-trans isomer. Rhodopsin and IRBP increased in parallel with ROS elongation up to P25, when the ROS had reached their mature lengths. The increases then continued up to P40-P50. In rd (retinal degeneration) mice, IRBP and rhodopsin were identical with the controls until P12, but then dropped as the photoreceptors degenerated. Synthesis and secretion of IRBP in vitro was less than 10% of the controls in rd retinas at P26, when only 4-5% of the photoreceptors survived. The quantities of retinyl esters (mainly stearate and palmitate in the ratio of 6:1, respectively) stored in dark-adapted mouse eyes progressively increased as the animals aged, representing 0.5 mole eq. of the rhodopsin at 8 months. Although retinyl esters (11-cis and all-trans) also accumulated in rd mouse eyes up to P12, little further increase occurred. At P93, the retinyl esters (0.01 nmole X eye-1) were only 4% of the controls at P91. A peak in the proportion of 11-cis isomer occurred at P10-P20, but it averaged only 15% of the total ester and declined to 5% at P93. These findings support the hypothesis that IRBP is synthesized by the rods and cones, and suggest that its synthesis and secretion are initiated when the photoreceptor inner segments start to differentiate. 11-cis Retinoids and rhodopsin do not appear until the outer segments start to form. It is suggested that in the rd mouse the absence of photoreceptors, perhaps coupled with lack of normal interphotoreceptor matrix, leads to a loss in the ability of the pigment epithelium to store retinyl esters.

[20] Detergents for extraction of visual pigments: Types, solubilization, and stability

Publisher Summary This chapter discusses the detergents for extraction of types, solubilization, and stability of visual pigments. Visual pigments are intrinsic membrane proteins. They are insoluble in water; hence their characterization can only be accomplished with the aid of appropriate solubilizing agents, usually detergents. The choice of detergent is critically dependent on the needs of each individual investigator. Factors to be considered include detergent solubility, efficiency of membrane solubilization, visual pigment stability, regenerability, acceptable UV transmission, ease of removal, and defined chemical structure and purity. The majority of comparative studies on visual pigments from various species, however, have been carried out with digitonin. Commercial-quality digitonin usually precipitates in 1–2% aqueous solutions. The problem has been substantially overcome by isolating the soluble component, but this material has not been defined chemically. Although visual pigments are very stable in digitonin extracts, the agent is a very inefficient extractant.

Visual cycle in the mammalian eye. Retinoid-binding proteins and the distribution of 11-cis retinoids

This work was designed to provide an insight into the mammalian visual cycle by investigating the possible function of retinoid-binding proteins in this system, and the distribution and type of 11-cis retinoids present in the interphotoreceptor matrix and the cytosols of the retinal pigment epithelium and retina. The total retinol and retinal in the soluble fractions from these three compartments was 8% (3.31 nmol/eye) of the retinyl palmitate and stearate stored in the pigment epithelium membrane fractions (39 nmol/eye). Only small amounts of retinoids were detected in the rod outer segment cytosol. The insoluble fractions also contained retinol, nearly all of which was found in the retina. The retinoids in the soluble fractions appeared to be bound to cellular retinol-binding protein (CRBP), cellular retinal-binding protein (CRA1BP) and interstitial retinol-binding protein (IRBP, a high-Mr glycoprotein). Using immunospecific precipitation, immunoblot and immunocytochemical techniques it was demonstrated that IRBP was localized in the interphotoreceptor matrix and was synthesized and secreted by the retina, a process that did not require the protein to be glycosylated. The amount of retinol bound to IRBP increased if the eyes were exposed to light, when it was estimated that the protein carried up to 30% of its full capacity for all-trans retinol. In addition to all-trans retinol, IRBP carried smaller amounts of 11-cis retinol. The proportion of 11-cis retinol was frequently higher in eyes that had been protected from illumination, suggesting that IRBP plays a role in rhodopsin regeneration during dark-adaptation. Additionally, endogenous 11-cis retinoids in the retina and RPE cytosols were bound to an Mr 33,000 protein tentatively identified as CRA1BP. The 11-cis retinoid in the retina cytosol was mainly in the form of retinol, while in the RPE cytosol it was mainly in the form of retinal. Substantial amounts of 11-cis retinol were also found in the insoluble (membrane) fraction from the retina. It is suggested that in the mammalian retina 11-cis retinol is generated from all-trans retinol (possibly in the Muller cells). Lack of an 11-cis retinol oxidoreductase in the retina prevents it from being utilized for rhodopsin regeneration until it has been transported to the pigment epithelium, where it is converted to 11-cis retinal and returned to the rod outer segments. It is also suggested that IRBP may be implicated in the transport of retinoids between the rod outer segments, the Muller cells and the pigment epithelium.(ABSTRACT TRUNCATED AT 400 WORDS)

Human interstitial retinol-binding protein (IRBP): cloning, partial sequence, and chromosomal localization

A cloned 2184-bp cDNA coding for human interstitial retinol-binding protein (IRBP) has been isolated and sequenced. The probe hybridized to a 5.2-kb poly(A) RNA from human retinas. Nineteen tryptic peptides (363 amino acids) sequenced and purified from bovine IRBP could be aligned with 86–88% homology to the translated sequence. Two segments approximately 200 amino acids long were found to have a 41% residue identity,suggesting an internal duplication event. This cloned cDNA was used to probe DNA samples from a panel of 29 rodent-human somatic cell hybrids, mapping the structural gene for IRBP to chromosome 10. In situ hybridization suggested a regional localization near the centromere (p11.2→q11.2), although a secondary site of hybridization at q24→25 was also observed.

Interstitial retinol-binding protein and cellular retinal-binding protein in the mammalian pineal

Antibodies against bovine interstitial retinol-binding protein (IRBP) and cellular retinal-binding protein (CRA1BP) were used in immunochemical and immunocytochemical studies of the pineal glands of cattle, hamsters and rats (RCS and RCS-rdy+). On immunoblots, IRBP (Mr 144,000) was identified in cattle, hamster and rat pineal extracts. The abundance of IRBP in bovine pineals was 33 +/- 6 ng.mg-1 (mean +/- SD, n = 12) soluble protein. RCS (Royal College of Surgeons) rat pineals gave a strong IRBP reaction on immunoblots, even when virtually no IRBP could be found in the eye due to photoreceptor degeneration. In the hamster retina IRBP immunostaining was distributed throughout the entire interphotoreceptor matrix and the outer segment layer. The pineal also showed strong IRBP-like immunostaining scattered uniformly throughout the gland. Other hamster brain regions showed no specific immunostaining; however, an immunoreactive protein with the same Mr as IRBP was detected on Western blots of bovine cerebral cortex, spinal cord and brainstem soluble proteins. Immunoreactive proteins at lower Mr were also detected in these tissues. CRA1BP immunoreactivity (Mr about 32,000) was observed in immunoblots of bovine, hamster and rat pineal proteins. These findings suggest that some mammalian pinealocytes are related to the retinal cells that contain CRA1BP (i.e. pigment epithelium, Muller cells) while others are related to the photoreceptors, which synthesize IRBP.

Utilization of exogenous retinol by frog pigment epithelium.

High-performance liquid chromatography was used to investigate the utilization of exogenous 11,12-3H2-retinol by frog pigment epithelium (RPE) in vitro or after intraocular injection into the dark-adapted, whole animal. Isolated frog RPE contains an adequate supply of acyl donors and can esterify all-trans, 11-cis and 13-cis isomers of retinol. The esterifying activity is restricted to the particulate fraction. Homogenates of choroid cannot esterify retinol. The ester formed by the RPE is primarily palmitate, and is therefore identical with the endogenous retinyl ester. Frog RPE also formed 13-cis retinyl palmitate from all-trans retinol, probably by esterification of 13-cis retinol formed non-enzymatically from the all-trans isomer. None of the in vitro experiments provided any evidence for the formation of 11-cis retinoid. There was slow appearance of label in 11-cis retinyl palmitate when 3H-all-trans retinol was injected intraocularly into the intact frog. After 15 hr its specific activity was only 20% of that of the all-trans retinyl palmitate. This rate of formation of 11-cis retinoid is inadequate for rhodopsin regeneration. However, it is more than an order of magnitude too fast to be accounted for by phagocytosis of rhodopsin. It is suggested that 11-cis retinoid is generated in the retina and is slowly transferred to the site of esterification in the RPE.

An extracellular retinol-binding glycoprotein in the rat eye-characterization, localization and biosynthesis

We have identified and partially purified interstitial retinol-binding protein (IRBP) from the subretinal space of the rat. It appeared to be glycosylated. Its apparent mol. wt was 270,000 by gel-filtration and 144,000 by sodium dodecyl sulfate polyacrylamide gel electrophoresis. Rat IRBP cross-reacted with anti-bovine IRBP sheep and rabbit sera, bound all-trans-[15-(3)H] retinol and was bound by concanavalin A. IRBP was not detected in the cytosols of the neural retina or retinal pigment epithelium and choroid. This distribution was confirmed by immunocytochemistry using a fluorescence-labeled second antibody. Immunospecific fluorescence was most intense in the interphotoreceptor matrix in a 6.5 ?m band adjacent to the retinal pigment epithelium. It was less intense over the remainder of the rod outer segment layer and was comparatively faint over the inner segment region. Its occurrence in the interstitial space between the photoreceptors and retinal pigment epithelium coupled with the fact it bound all-trans-[15-(3)H] retinol supports the concept that it may be implicated in the transport of retinoids between the retina and the retinal pigment epithelium during the visual cycle. When incubated with [(3)H]leucine or [(3)H]glucosamine, isolated retinas (but not retinal pigment epithelium and choroid) secreted labeled IRBP into the medium. This suggests that the retina plays a role in regulating the amount of IRBP in the subretinal space.

Bovine interstitial retinol-binding protein (IRBP)—isolation and sequence analysis of cDNA clones, characterization and in vitro translation of mRNA

Three clones for b-IRBP were isolated by anti b-IRBP screening of two bovine retina libraries in the expression vector lambda gt11. The cDNA inserts were then used as hybridization probes to screen and isolate three more clones in a bovine retina library in the non-expression vector lambda gt10. The six overlapping clones generated a b-IRBP cDNA sequence of 3400 nucleotides. An open reading frame encoded the complete amino acid sequences of 8 of the 35 b-IRBP tryptic peptides purified in the present study. One tentative glycosylation site was identified. The coding region was followed by TAG translation terminating codon and an untranslated stretch of about 1700 nucleotides that ended in a sequence containing a presumptive AATAAA polyadenylation signal that was 18 nucleotides upstream from a 10 nucleotide oligo(A) tract. The coding region for b-IRBP would be expected to be 3300 bp long, but Northern blot hybridization experiments performed with bovine retina polyadenylated RNA and probes containing part of the coding region established that the mRNA for b-IRBP consisted of a major species of about 6300 bp, and a minor species of 5200 bp. In vitro translation of bovine retina polyadenylated RNA in a rabbit reticulocyte lysate system yielded an immunoreactive protein that was comparable in size with nonglycosylated, mature IRBP, showing that it is not synthesized from a large precursor, and supporting our finding that the mRNA contains an extensive non-coding region.

Interstitial retinol-binding protein (IRBP) in retinoblastoma

Interstitial retinol-binding protein (IRBP) is an extracellular glycoprotein that appears to be synthesized by the photoreceptors in the normal retina and may be involved in shuttling retinoids through the interphotoreceptor matrix between the retina and pigment epithelium. The present work demonstrated immunochemically that IRBP of the same molecular weight as normal human IRBP (135,000) was present in three retinoblastomas, irrespective of their degree of differentiation. Tumor 1 was classified as differentiated; Tumors 2 and 3 were classified as poorly differentiated. The level of IRBP in the soluble proteins of Tumor 2 was about 8 times that in Tumor 1 and about one-half that in the soluble proteins (including adhering interphotoreceptor matrix) from a pair of normal retinas from a 31-year-old donor. IRBP occurred in the interstitial space of Tumor 3. Most of the IRBP in this tumor was recovered from the medium in which the undifferentiated cells were dispersed. Incubation of the isolated cells from Tumor 3 in medium containing [(3)H]leucine demonstrated that [(3)H]IRBP was secreted into the medium. The [(3)H]IRBP was immunoreactive with rabbit antibovine IRBP antibodies. It was inferred that the [(3)H]IRBP was glycosylated because it was bound by immobilized concanavalin A and could be displaced with ?-methylmannopyranoside. Since IRBP is not normally found in retinas until the time of photoreceptor differentiation, regulation of its gene may be defective in this malignant neuroectodermal neoplasm. The present findings are relevant to the possible role of retinoids and their binding proteins in neoplastic cells, because they demonstrate for the first time the presence of an extracellular retinoid-binding protein in tumor tissue.