An immunologically privileged retinal antigen elicits tolerance: major role for central selection mechanisms.

Immunologically privileged retinal antigens can serve as targets of experimental autoimmune uveitis (EAU), a model for human uveitis. The tolerance status of susceptible strains, whose target antigen is not expressed in the thymus at detectable levels, is unclear. Here, we address this issue directly by analyzing the consequences of genetic deficiency versus sufficiency of a uveitogenic retinal antigen, interphotoreceptor retinoid-binding protein (IRBP). IRBP-knockout (KO) and wild-type (WT) mice on a highly EAU-susceptible background were challenged with IRBP. The KO mice had greatly elevated responses to IRBP, an altered recognition of IRBP epitopes, and their primed T cells induced exacerbated disease in WT recipients. Ultrasensitive immunohistochemical staining visualized sparse IRBP-positive cells, undetectable by conventional assays, in thymi of WT (but not of KO) mice. IRBP message was PCR amplified from these cells after microdissection. Thymus transplantation between KO and WT hosts demonstrated that this level of expression is functionally relevant and sets the threshold of immune (and autoimmune) reactivity. Namely, KO recipients of WT thymi generated reduced IRBP-specific responses, and WT recipients of KO thymi developed enhanced responses and a highly exacerbated disease. Repertoire culling and thymus-dependent CD25+ T cells were implicated in this effect. Thus, uveitis-susceptible individuals display a detectable and functionally significant tolerance to their target antigen, in which central mechanisms play a prominent role.

MicroRNA expression in human retinal pigment epithelial (ARPE-19) cells: increased expression of microRNA-9 by N-(4-hydroxyphenyl)retinamide.

PURPOSE: MicroRNAs (miRNAs) are important regulators of many cellular functions due to their ability to target mRNAs for degradation or translational inhibition. Previous studies have reported that the expression of microRNA-9 (miR-9) is regulated by retinoic acid and reactive oxygen species (ROS). We have previously shown that N-(4-hydroxyphenyl)-retinamide (4HPR), a retinoic acid derivative, induces ROS generation and apoptosis in cultured human retinal pigment epithelial (RPE) cells, known as ARPE-19 cells. The aim of the present study was to investigate the expression of miR-9 in ARPE-19 cells in response to 4HPR treatment, and to identify other miRNAs normally expressed in these cells. METHODS: ARPE-19 cells in culture were treated with 4HPR, the total RNA fractions were isolated, and the expression of various miRNAs and mRNAs was analyzed using real-time PCR. The miRNA expression profile of ARPE-19 cells was analyzed using microarray hybridization. RESULTS: Treatment of ARPE-19 cells with 4HPR resulted in apoptosis characterized by the increased expression of HMOX1 and GADD153 genes. A twofold increase in the expression of miR-9 was also observed during this response. Potential binding sites for the transcription factors encoded by CEBPA and CEBPB genes were found to be present in the putative promoter regions of all three genes encoding miR-9. 4HPR-induced miR-9 expression was associated with parallel increases in the expression of these transcription factor genes. 5-Aza-2'-deoxycytidine, a methyl transferase inhibitor, also increased the expression of miR-9 in ARPE-19 cells. Microarray hybridization analysis identified let-7b, let-7a, miR-125b, miR-24, miR-320, miR-23b, let-7e, and let-7d as the most abundant miRNAs normally expressed in ARPE-19 cells. These miRNAs are known to regulate cell growth, differentiation or development. The 4HPR treatment increased the expression of miR-16, miR-26b, miR-23a, and miR-15b in ARPE-19 cells, although these increases were modest when compared to the increase in the expression of miR-9. CONCLUSIONS: Our studies demonstrate that miR-9 is expressed in the RPE cell line ARPE-19, and its expression is increased by a retinoic acid derivative and by an inhibitor of promoter hypermethylation. Several miRNAs with inherent ability to regulate cell growth, differentiation and development are also normally expressed in ARPE-19 cells. Thus, miR-9 and other miRNAs could be important in maintaining RPE cell function.

Mitogen-activated protein kinase pathway mediates N-(4-hydroxyphenyl)retinamide-induced neuronal differentiation in the ARPE-19 human retinal pigment epithelial cell line.

We have shown previously that N-(4-hydroxyphenyl)retinamide (4HPR, fenretinide), a retinoic acid derivative, induces neuronal differentiation in cultured human retinal pigment epithelial (RPE) cells [Chen et al., J. Neurochem., 84 (2003), 972]. We asked the question whether the mitogen-activated protein kinase (MAPK) pathway is involved in the regulation of the 4HPR-induced neuronal differentiation of RPE (ARPE-19) cells. When we treated ARPE-19 cells with 4HPR, c-Raf and MEK1/2 kinase were activated resulting in activation of the downstream effector ERK1/2 and of SAPK/JNK. By blocking the upstream kinase MEK1/2 with specific inhibitor U0126 we abrogated the 4HPR-induced phosphorylation of ERK1/2 and SAPK/JNK, indicating that the neuronal differentiation occurs through a positive cross-talk between the ERK and the SAPK/JNK pathways. Both U0126 and the suppression of ERK1/2 expression with small interfering RNA effectively blocked the 4HPR-induced neuronal differentiation of RPE cells and the expression of calretinin. The activated ERK1/2 then induced a sequential activation of p90RSK, and increase in phosphorylation of transcription factors c-fos and c-jun leading to transcriptional activation of AP-1. Taken together, our results clearly demonstrate that c-Raf/MEK1/2 signaling cascade involving ERK1/2 plays a central role in mediating the 4HPR-induced neuronal differentiation and calretinin expression in the human ARPE-19 retinal pigment epithelial cell line.

A humanized model of experimental autoimmune uveitis in HLA class II transgenic mice.

Experimental autoimmune uveitis (EAU) is a disease of the neural retina induced by immunization with retinal antigens, such as interphotoreceptor retinoid-binding protein (IRBP) and arrestin (retinal soluble antigen, S-Ag). EAU serves as a model for human autoimmune uveitic diseases associated with major histocompatibility complex (HLA) genes, in which patients exhibit immunological responses to retinal antigens. Here we report the development of a humanized EAU model in HLA transgenic (TG) mice. HLA-DR3, -DR4, -DQ6, and -DQ8 TG mice were susceptible to IRBP-induced EAU. Importantly, HLA-DR3 TG mice developed severe EAU with S-Ag, to which wild-type mice are highly resistant. Lymphocyte proliferation was blocked by anti-HLA antibodies, confirming that antigen is functionally presented by the human MHC molecules. Disease could be transferred by immune cells with a Th1-like cytokine profile. Antigen-specific T cell repertoire, as manifested by responses to overlapping peptides derived from S-Ag or IRBP, differed from that of wild-type mice. Interestingly, DR3 TG mice, but not wild-type mice, recognized an immunodominant S-Ag epitope between residues 291 and 310 that overlaps with a region of S-Ag recognized by uveitis patients. Thus, EAU in HLA TG mice offers a new model of uveitis that should represent human disease more faithfully than currently existing models.

Visual cycle: Dependence of retinol production and removal on photoproduct decay and cell morphology.

The visual cycle is a chain of biochemical reactions that regenerate visual pigment following exposure to light. Initial steps, the liberation of all-trans retinal and its reduction to all-trans retinol by retinol dehydrogenase (RDH), take place in photoreceptors. We performed comparative microspectrophotometric and microfluorometric measurements on a variety of rod and cone photoreceptors isolated from salamander retinae to correlate the rates of photoproduct decay and retinol production. Metapigment decay rate was spatially uniform within outer segments and 50-70 times faster in the cells that contained cone-type pigment (SWS2 and M/LWS) compared to cells with rod-type pigment (RH1). Retinol production rate was strongly position dependent, fastest at the base of outer segments. Retinol production rate was 10-40 times faster in cones with cone pigments (SWS2 and M/LWS) than in the basal OS of rods containing rod pigment (RH1). Production rate was approximately five times faster in rods containing cone pigment (SWS2) than the rate in basal OS of rods containing the rod pigment (RH1). We show that retinol production is defined either by metapigment decay rate or RDH reaction rate, depending on cell type or outer segment region, whereas retinol removal is defined by the surface-to-volume ratio of the outer segment and the availability of retinoid binding protein (IRBP). The more rapid rates of retinol production in cones compared to rods are consistent with the more rapid operation of the visual cycle in these cells.

Confocal immunolocalization of bovine serum albumin, serum retinol-binding protein, and interphotoreceptor retinoid-binding protein in bovine retina.

PURPOSE: Recently it has been shown that the transport as well as clearance of retinol from isolated rod photoreceptors requires an extracellular factor. Interphotoreceptor retinoid-binding protein (IRBP) is a component of the interphotoreceptor matrix (IPM) and is known to bind visual cycle retinoids. Serum albumin and serum retinol-binding protein (sRBP), proteins capable of binding retinoids, have also been reported to be components of the IPM. It is of interest to know the components present in the IPM that are capable of binding visual cycle retinoids and that also facilitate rhodopsin regeneration. The purpose of this study was to determine the localization of serum albumin, sRBP, and IRBP in bovine retina using immunofluorescence analysis. METHODS: Fresh bovine eyes, obtained from a local abattoir, were fixed immediately after enucleation. Tissue sections (100 microm) were incubated with primary antibodies to bovine serum albumin (BSA), sRBP, and IRBP. Sections were washed then incubated 4 h with 4'-6-Diamidino-2-phenylindole (DAPI), Alexa Fluor(R) 488 goat antimouse, and Alexa Fluor 568 goat antirabbit secondary antibodies. Sections were analyzed using a laser scanning confocal microscope equipped with Nomarski optics. Western immunoblot analysis of bovine retinal tissues and protein standards was performed using the primary antibodies to BSA, sRBP, and IRBP to show specificity to their respective antigens. RESULTS: Immunoblot analysis showed that monoclonal anti-BSA was highly specific for BSA detecting only a single band at about 67 kDa. Antihuman sRBP and antibovine IRBP were also highly specific, recognizing a single band at about 25 and about 133 kDa, respectively. No immunopositive bands were observed in bovine neural retinal when probed with the anti-sRBP antibody; however, a single immunoreactive band at about 67 and about 133 kDa was detected in bovine neural retina by the anti-BSA and IRBP antibodies, respectively. Immunofluorescence analysis showed labeling for IRBP throughout the IPM. IRBP labeling was especially associated with the outer segments of photoreceptors and also with the apical surface of the retinal pigment epithelium. Immunofluorescence labeling for serum albumin was associated only with the lumen of retinal and choroidal blood vessels. Staining for both serum albumin and sRBP in the IPM was negative. CONCLUSIONS: Immunofluorescence analysis of fresh bovine eyes using antibodies to BSA and sRBP clearly shows that serum albumin and sRBP are not components of bovine IPM. IRBP, on the other hand, is localized to the IPM where it is available for the binding and transport of visual cycle retinoids. From these data we conclude that serum albumin and sRBP are not factors that could participate in the binding as well as transport of visual cycle retinoids in the IPM of bovine retina.

Fenretinide-induced neuronal differentiation of ARPE-19 human retinal pigment epithelial cells is associated with the differential expression of Hsp70, 14-3-3, pax-6, tubulin beta-III, NSE, and bag-1 proteins.

PURPOSE: We reported earlier that fenretinide can induce neuronal differentiation of ARPE-19 human retinal pigment epithelial cells in culture. The purpose of this study was to investigate the potential involvement of key proteins involved in gene transcription, signal transduction, cell cycle check point, differentiation, neuronal cell survival, and stress response in the neuronal differentiation of ARPE-19 cells by fenretinide. METHODS: Cells in culture were treated with 1.0 microM fenretinide. Cells were analyzed using antibodies against pax-6, neuronal specific enolase (NSE), tubulin beta-III, 14-3-3, bag-1, and Hsp-70 proteins using immunocytochemistry, western blot and ELISA methodologies. RESULTS: We found that pax-6 and NSE were both expressed in the control ARPE-19 cells. Fenretinide induced neuronal differentiation of ARPE-19 cells led to a decrease in pax-6 protein and an increase in tubulin beta-III protein expression after 5 days fenretinide treatment. There was a translocation of 14-3-3 from the cytoplasm to the nucleus, and an increase in nuclear expression of bag-1 after treatment. We also found a time-dependent increase in Hsp70 protein expression in ARPE-19 cells treated with fenretinide. D-407, another human retinal pigment epithelial cell line, but not either Y-79 or PC-12 cells, was also able to be induced into neuronal morphologies by fenretinide. CONCLUSIONS: The fenretinide-induced neuronal differentiation of ARPE-19 cells is associated with an increase in expression of the neuronal specific protein tubulin beta-III, and a decrease in expression of the progenitor cell marker pax-6. Neuronal differentiation of ARPE-19 cells is also associated with nuclear translocation of 14-3-3, a protein involved in signal transduction, cell cycle check point and cell growth, and an increase in expression of bag-1, a protein involved in neuronal cell survival and axon elongation. These results suggest that ARPE-19 cells could be a progenitor cell line that can be differentiated into neuronal cells when treated with factors such as fenretinide.

Immunofluorescence analysis of the expression of Norpeg (Rai14) in retinal Müller and ganglion cells.

Novel retinal pigment epithelial cell gene (Norpeg, Rai14), a developmentally regulated mouse gene, encodes a protein containing six ankyrin repeats and a coiled-coil domain. The objective of the present study was to characterize the Norpeg protein and to analyze its expression in mouse retina using an antibody preparation that we developed. The approximately 110kDa Norpeg was immunoprecipitated and characterized by mass spectrometry. Primary cultures of Müller and ganglion cells isolated from the mouse retina were found to express Norpeg when analyzed by immunoblotting and immunofluorescence microscopy. Immunofluorescence analysis of normal mouse retina sections demonstrated that Norpeg is expressed in cells of the ganglion cell layer, inner nuclear layer as well as in the retinal pigment epithelium. Immunoreactivity was also evident in the radial glial (Müller) cell fibers.

Effect of visible light on normal and P23H-3 transgenic rat retinas: characterization of a novel retinoic acid derivative present in the P23H-3 retina.

Transgenic rats with the P23H mutation in rhodopsin exhibit increased susceptibility to light damage, compared with normal animals. It is known that light-induced retinal damage requires repetitive bleaching of rhodopsin and that photoreceptor cell loss is by apoptosis; however, the underlying molecular mechanism(s) leading to photoreceptor cell death are still unknown. Photoproducts, such as all-trans retinal or other retinoid metabolites, released by the extensive bleaching of rhodopsin could lead to activation of degenerative processes, especially in animals genetically predisposed to retinal degenerations. Using wild-type and transgenic rats carrying the P23H opsin mutation, we evaluated the effects of acute intense visible light on retinoid content, type and distribution in ocular tissues. Rats were exposed to green light (480-590 nm) for 0, 5, 10, 30 and 120 min. Following light treatment, rats were sacrificed and neural retinas were dissected free of the retinal pigment epithelium. Retinoids were extracted from retinal tissues and then subjected to HPLC and mass spectral analysis. We found that the light exposure affected relative levels of retinoids in the neural retina and retinal pigment epithelium of wild-type and P23H rat eyes similarly. In the P23H rat retina but not the wild-type rat retina, we found a retinoic acid-like compound with an absorbance maximum of 357 nm and a mass of 304 daltons. Production of this retinoic acid-like compound in transgenic rats is influenced by the age of the animals and the duration of light exposure. It is possible that this unique retinoid may be involved in the process of light-induced retinal degeneration.

Physiological and microfluorometric studies of reduction and clearance of retinal in bleached rod photoreceptors.

The visual cycle comprises a sequence of reactions that regenerate the visual pigment in photoreceptors during dark adaptation, starting with the reduction of all-trans retinal to all-trans retinol and its clearance from photoreceptors. We have followed the reduction of retinal and clearance of retinol within bleached outer segments of red rods isolated from salamander retina by measuring its intrinsic fluorescence. Following exposure to a bright light (bleach), increasing fluorescence intensity was observed to propagate along the outer segments in a direction from the proximal region adjacent to the inner segment toward the distal tip. Peak retinol fluorescence was achieved after approximately 30 min, after which it declined very slowly. Clearance of retinol fluorescence is considerably accelerated by the presence of the exogenous lipophilic substances IRBP (interphotoreceptor retinoid binding protein) and serum albumin. We have used simultaneous fluorometric and electrophysiological measurements to compare the rate of reduction of all-trans retinal to all-trans retinol to the rate of recovery of flash response amplitude in these cells in the presence and absence of IRBP. We find that flash response recovery in rods is modestly accelerated in the presence of extracellular IRBP. These results suggest such substances may participate in the clearance of retinoids from rod photoreceptors, and that this clearance, at least in rods, may facilitate dark adaptation by accelerating the clearance of photoproducts of bleaching.

Methimazole protects from experimental autoimmune uveitis (EAU) by inhibiting antigen presenting cell function and reducing antigen priming.

Methimazole (methyl-mercapto-imidazole, MMI), a compound used clinically in therapy of Graves' thyroiditis, was found to inhibit development of several autoimmune diseases in animal models. It was suggested on the basis of in vitro data that inhibition is through down-regulation of interferon-gamma (IFN-gamma)-induced expression of major histocompatibility complex class I and class II molecules. Here, we investigate the effect of MMI on experimental autoimmune uveoretinitis (EAU) and study its mechanism(s). Treatment of EAU with MMI administered in drinking water inhibited induction of the disease and associated antigen (Ag)-specific proliferation and cytokine production by draining lymph node cells (LNCs). The treatment was protective only if administered during the first but not during the second week after immunization, suggesting an effect on the induction phase of EAU. It is interesting that MMI inhibited disease in IFN-gamma knockout mice, indicating that the in vivo protective effect is IFN-gamma-independent. Flow cytometric analysis of draining LNCs extracted 5 days after immunization showed that MMI partly to completely reversed the increase in Mac-1(+)/class I(+)/class II(+) cells induced by immunization and reduced the proportion of B7-1 and CD40-positive cells, suggesting a deficit in the Ag-presenting cell (APC) population. APC from untreated mice largely restored antigen-specific proliferation of MMI-treated LNCs. We suggest that MMI inhibits EAU at least in part by preventing the recruitment and/or maturation of APC, resulting in reduced generation of Ag-specific T cells.

The immunomodulator vasoactive intestinal peptide (VIP) does not affect experimental autoimmune uveitis (EAU) in B10.RIII mice.

PURPOSE: Vasoactive intestinal peptide (VIP) exhibits immunomodulatory activities both in vivo and in vitro, including efficient inhibition of murine experimental arthritis. In this study, we investigated the effects of VIP treatment on the induction of experimental autoimmune uveoretinitis (EAU). METHODS: EAU was induced in B10.RIII mice by immunization with interphotoreceptor retinoid-binding protein (IRBP) using routine methods, but without treatment with pertussis toxin (PTX). VIP was injected i.p. at different doses into mice on alternate days. Mice were tested by conventional methods for ocular inflammation, antibody levels, lymphocyte proliferation, and cytokine release by cultured lymphocytes. RESULTS: Treatment with VIP, at different doses, had essentially no effect on the development of EAU or antibody production in the B10.RIII mice. The treatment did have variable effects on the low interferon-gamma production by lymphocytes of these mice. CONCLUSION: Unlike its inhibitory effect in the experimental arthritis system, VIP did not modulate the development of EAU in B10.RIII mice.

Susceptibility to retinal light damage in transgenic rats with rhodopsin mutations.

PURPOSE: To determine relative light-induced retinal damage susceptibility in transgenic rats expressing mutations in the N- or C-terminal region of rhodopsin. METHODS: Heterozygous transgenic rats, including P23H sublines 2 and 3 and S334ter sublines 4 and 9, were reared in dim cyclic light or in darkness before visible light exposure starting at various times of the day or night. Before exposure to light, some rats were given the synthetic antioxidant dimethylthiourea (DMTU). At various times after intense light treatment, rats were killed for determinations of rhodopsin and retinal DNA recovery, DNA fragmentation patterns, and Northern blot analysis of retinal heme oxygenase (HO)-1 and interphotoreceptor retinol binding protein (IRBP). Rod outer segments (ROSs) were isolated for Western blot analysis of rhodopsin using N- and C- terminal-specific monoclonal antibodies. RESULTS: All rats incurred greater photoreceptor cell damage from exposure to light starting at 1 AM than from exposure at 5 PM. Among cyclic-light-reared rats, P23H line 3 animals were more susceptible to light-induced damage than P23H line 2 animals. S334ter rats exhibited retinal light damage profiles similar to those in normal rats. Dark-rearing potentiated retinal damage by light. However, dark-rearing alone prolonged photoreceptor cell life in P23H rats, but had no such effect in S334ter animals. DMTU pretreatment was effective in preventing or reducing light-induced retinal damage in all transgenic rats. S334ter rat ROSs contained the truncated form of rhodopsin. Intense light exposure resulted in DNA ladders typical of apoptotic cell death and the simultaneous induction of retinal HO-1 mRNA and reduced expression of IRBP. CONCLUSIONS: Light-induced retinal damage in transgenic rats depends on the time of day of exposure to light, prior light-or dark-rearing environment, and the relative level of transgene expression. Retinal light damage leads to apoptotic visual cell loss and appears to result from oxidative stress. These results suggest that reduced environmental lighting and/or antioxidant treatment may delay retinal degenerations arising from rhodopsin mutations.

Chemokine gene expression in iris-ciliary body during experimental autoimmune uveoretinitis.

PURPOSE: To evaluate, through differential gene expression of chemokines in iris-ciliary body (I/CB), the extent to which leukocyte recruitment to the ocular anterior segment participates in the pathogenesis of experimental autoimmune uveoretinitis (EAU) in mice. METHODS: B10.A mice were immunized with 50 microg of interphotoreceptor retinoid binding protein (IRBP) mixed with complete Freund's adjuvant. Aqueous humor (AqH) was collected at 0, 11, 17, and 28 days and assayed for leukocyte content and protein levels. Enucleated eyes were subjected to histologic analysis. Chemokine gene expression in I/CB was determined at these same time points by a multiprobe ribonuclease protection assay (RPA) system. RESULTS: Inflammation was detected in the anterior chamber (AC) at 11 days and leukocyte recruitment continued thereafter. Polymorphonuclear neutrophils (PMN) were the predominant cell type in the AC, whereas macrophages/monocytes and lymphocytes were predominant in the retina/subretinal space at 17 days. Peak gene expression of macrophage inflammatory protein (MIP)-1alpha, MIP-1beta, and MIP-2 and interferon-gamma-inducible protein of 10 kd (IP-10) was detected in I/CB on day 11, whereas peak expression of RANTES and eotaxin was observed at 17 days. CONCLUSIONS: Early I/CB peak expression of mRNA for MIP-2 followed by PMN recruitment into the AC, suggests that PMN may play an important role in EAU pathogenesis.

Interphotoreceptor retinoid-binding protein is the physiologically relevant carrier that removes retinol from rod photoreceptor outer segments.

Light detection by vertebrate rod photoreceptor outer segments results in the destruction of the visual pigment, rhodopsin, as its retinyl moiety is photoisomerized from 11-cis to all-trans. The regeneration of rhodopsin is necessary for vision and begins with the release of the all-trans retinal and its reduction to all-trans retinol. Retinol is then transported out of the rod outer segment for further processing. We used fluorescence imaging to monitor retinol fluorescence and quantify the kinetics of its formation and clearance after rhodopsin bleaching in the outer segments of living isolated frog (Rana pipiens) rod photoreceptors. We independently measured the release of all-trans retinal from bleached rhodopsin in frog rod outer segment membranes and the rate of all-trans retinol removal by the lipophilic carriers interphotoreceptor retinoid binding protein (IRBP) and serum albumin. We find that the kinetics of all-trans retinol formation in frog rod outer segments after rhodopsin bleaching are to a good first approximation determined by the kinetics of all-trans retinal release from the bleached pigment. For the physiological concentrations of carriers, the rate of retinol removal from the outer segment is determined by IRBP concentration, whereas the effect of serum albumin is negligible. The results indicate the presence of a specific interaction between IRBP and the rod outer segment, probably mediated by a receptor. The effect of different concentrations of IRBP on the rate of retinol removal shows no cooperativity and has an EC50 of 40 micromol/L.

N-(4-hydroxyphenyl)retinamide induces apoptosis in human retinal pigment epithelial cells: retinoic acid receptors regulate apoptosis, reactive oxygen species generation, and the expression of heme oxygenase-1 and Gadd153.

N-(4-hydroxyphenyl)retinamide (4HPR, fenretinide), a retinoic acid (RA) derivative and a potential cancer preventive agent, is known to exert its chemotherapeutic effects in cancer cells through induction of apoptosis. Earlier work from our laboratory has shown that relatively low concentrations of 4HPR induce neuronal differentiation of cultured human retinal pigment epithelial (ARPE-19) cells (Chen et al., 2003, J Neurochem 84:972-981). However, at higher concentrations of 4HPR, these cells showed morphological changes including cell shrinkage and cell death. Here we demonstrate that ARPE-19 cells treated with 4HPR exhibit a dose- and time-dependent induction of apoptosis as evidenced by morphological changes, mono- and oligonucleosome generation, and increased activity of caspases 2 and 3. The 4HPR-induced apoptosis as well as the activation of caspases 2 and 3 were blocked by both retinoic acid receptors (RAR) pan-antagonists, AGN193109 and AGN194310, and by an RARalpha-specific antagonist AGN194301. 4HPR treatment also increased reactive oxygen species (ROS) generation in ARPE-19 cells in a time-dependent manner as determined from the oxidation of 2',7'-dichlorofluorescin. In addition, the increase in the expression of heme oxygenase-1 (HO-1), a stress response protein, and the growth arrest and DNA damage-inducible transcription factor 153 (Gadd153) in response to the ROS generation were also blocked by these receptor antagonists. Pyrrolidine dithiocarbamate (PDTC), a free-radical scavenger, inhibited 4HPR-induced ROS generation, the expression of its downstream mediator, Gadd153, and apoptosis in the pretreated cells. Therefore, our results, clearly demonstrate that 4HPR induces apoptosis in ARPE-19 cells and that RARs mediate this process by regulating ROS generation as well as the expression of Gadd153 and HO-1.

Cell density-dependent nuclear/cytoplasmic localization of NORPEG (RAI14) protein.

NORPEG (RAI14), a developmentally regulated gene induced by retinoic acid, encodes a 980 amino acid (aa) residue protein containing six ankyrin repeats and a long coiled-coil domain [Kutty et al., J. Biol. Chem. 276 (2001), pp. 2831-2840]. We have expressed aa residues 1-287 of NORPEG and used the recombinant protein to produce an anti-NORPEG polyclonal antibody. Confocal immunofluorescence analysis showed that the subcellular localization of NORPEG in retinal pigment epithelial (ARPE-19) cells varies with cell density, with predominantly nuclear localization in nonconfluent cells, but a cytoplasmic localization, reminiscent of cytoskeleton, in confluent cultures. Interestingly, an evolutionarily conserved putative monopartite nuclear localization signal (P(270)KKRKAP(276)) was identified by analyzing the sequences of NORPEG and its orthologs. GFP-NORPEG (2-287 aa), a fusion protein containing this signal, was indeed localized to nuclei when expressed in ARPE-19 or COS-7 cells. Deletion and mutation analysis indicated that the identified nuclear localization sequence is indispensable for nuclear targeting.

Interphotoreceptor retinoid-binding protein (IRBP) promotes the release of all-trans retinol from the isolated retina following rhodopsin bleaching illumination.

All-trans retinol generated in rod photoreceptors upon the bleaching of rhodopsin is known to move from the rods to the retinal pigment epithelium (RPE), where it is enzymatically converted to 11-cis retinal in the retinoid visual cycle. Interphotoreceptor retinoid-binding protein (IRBP) contained in the extracellular compartment (interphotoreceptor matrix) that separates the retina and RPE has been hypothesized to facilitate this movement of all-trans retinol, but the precise role of IRBP in this process remains unclear. To examine the activity of IRBP in the release of all-trans retinol from the rods, initially dark-adapted isolated retinas obtained from toad (Bufo marinus) eyes were bleached and then incubated in darkness for defined periods (5-180 min) in physiological saline (Ringer solution) supplemented with IRBP (here termed 'IRBP I') at defined concentrations (2-90 microm). Retinoids present in the retina and extracellular medium were then determined by extraction and HPLC analysis. Preparations incubated with > or =10 microm IRBP I showed a pronounced release of all-trans retinol with increasing period of incubation. As determined with 25 microm IRBP I, the increase of all-trans retinol in the extracellular medium was accompanied by a significant decrease in the combined amount of all-trans retinal and all-trans retinol contained in the retina. This effect was not mimicked by unsupplemented Ringer solution or by Ringer solution containing 25 or 90 microm bovine serum albumin. However, incubation with 'IRBP II', a previously described variant of IRBP with altered lectin-binding properties, led to the appearance of substantial all-trans retinol in the extracellular medium. The results suggest that in vivo, IRBP plays a direct role in the release of all-trans retinol from the rods during operation of the visual cycle.

Differentiation of human retinal pigment epithelial cells into neuronal phenotype by N-(4-hydroxyphenyl)retinamide.

ARPE-19, a human retinal pigment epithelial (RPE) cell line, has been widely used in studies of RPE function as well as gene expression. Here, we report the novel finding that N-(4-hydroxyphenyl)retinamide (fenretinide), a synthetic retinoic acid derivative and a potential chemopreventive agent against cancer, induced the differentiation of ARPE-19 cells into a neuronal phenotype. The treated cells lost their epithelial phenotype and exhibited a typical neuronal shape with long processes (four to five times longer than the cell body). The onset of fenretinide-induced neuronal differentiation was dose and time dependent, started within 1-2 days, and lasted at least 4 weeks. Immunohistochemical studies indicated that the expression of neurofilament proteins (NF160 and NF200), calretinin and neural cell adhesion molecule was increased in these differentiated cells. Western blot analysis indicated that cellular retinaldehyde-binding protein, which is normally expressed in RPE cells, was decreased in treated cells. Protein analysis on a two-dimensional gel followed by matrix-assisted laser desorption ionization-time of flight mass spectrometric analysis demonstrated that heat-shock protein 70 was increased after fenretinide treatment. Thus, fenretinide, a synthetic retinoid, is able to induce neuronal differentiation of human RPE cells in culture.