Intracellular redox state modulates the T cell surface proteome – relevance to ageing

During ageing an altered redox balance has been observed in both intracellular and extracellular compartments, primarily due to glutathione depletion and metabolic stress. Maintaining redox homeostasis is important for controlling proliferation and apoptosis in response to specific stimuli for a variety of cells. For T cells, the ability to generate specific response to antigen is dependent on the oxidation state of cell surface and cytoplasmic protein-thiols. Here we describe the effects of depleting intracellular glutathione concentration for T cell exofacial expression of thioredoxin 1 and IL-2 production, and have determined the distribution of Trx1 with ageing. Using buthionine sulfoximine to deplete intracellular glutathione in Jurkat T cells we show using Western blotting that cell surface thioredoxin-1 is lowered and that the response to the lectin phytohaemagglutinin measured by ELISA as IL-2 production is also decreased. Using flow cytometry we show that the distribution of Trx1 on primary CD4+ T cells is age-dependent, with lower surface Trx1 expression and greater variability of surface expression observed with age. Together these data suggest that a relationship exists between the intracellular redox compartment and exofacial surface. Redox imbalance may be important for impaired T cell function during ageing.

The Influence of Age, Cardiorespiratory Fitness, Exercise and Sedentary Behaviours on Circulating Angiogenic Cells and Cell Surface Receptor Expression.

Cardiovascular disease (CVD) is the biggest killer of people in western civilisation. Age is a significant risk factor for the development for CVD, and treatments and therapies to address this increased risk are crucial to quality of life and longevity. Exercise is one such intervention which has been shown to reduce CVD risk. Age is also associated with endothelial dysfunction, reduced angiogenic capabilities, and reduced ability to repair the vessel wall. Circulating angiogenic cells (CACs) are a subset of circulating cells which assist in the repair and growth of the vasculature and in the maintenance of endothelial function. Reductions in these cells are observed in those with vascular disease compared to age-matched healthy controls. Exercise may reduce CVD risk by improvements in number and/or function of these CACs. Data was collected from human volunteers of various ages, cardiorespiratory fitness (CRF) levels and latent viral infection history status to investigate the effects of chronological age, CRF, viral serology and other lifestyle factors, such as sedentary behaviours and exercise on CACs. The levels of CACs in these volunteers were measured using four colour flow cytometry using various monoclonal antibodies specific to cell surface markers that are used to identify specific subsets of these CACs. In addition, the response to acute exercise of a specific subset of these CACs, termed ‘angiogenic T-cells’ (TANG) were investigated, in a group of well-trained males aged 20-40 years, using a strenuous submaximal exercise bout. Advancing age was associated with a decline in various subsets of CACs, including bone marrow-derived CD34+ progenitors, putative endothelial progenitor cells (EPCs) and also TANG cells. Individuals with a higher CRF were more likely to have higher circulating numbers of TANG cells, particularly in the CD4+ subset. CRF did not appear to modulate CD34+ progenitors or EPC subsets. Increasing sitting time was associated with reduction in TANG cells, but after correcting for the effects of fitness, sitting time no longer negatively affected the circulating number of these cells. Acute exercise was a powerful stimulus for increasing the number of TANG cells (140% increase), potentially through an SDF-1:CXCR4-dependent mechanism, but more studies are required to investigate this. Latent CMV infection was associated with higher number of TANG cells (CD8+), but only in 18-40 year old individuals, and not in an older age group (41-65 year old). The significance of this has yet to be understood. In conclusion, advancing age may contribute to increased CVD risk partly due to the observed reductions in angiogenic cells circulating in the peripheral compartment. Maintaining a high CRF may attenuate this CVD reduction by modulating TANG cell number, but potentially not CD34+ progenitor or EPC subsets. Acute exercise may offer a short window for vascular adaptation through the mobilisation of TANG cells into the circulation.

Cell surface Glut1 levels distinguish human CD4 and CD8 T lymphocyte subsets with distinct effector functions

International audience

Involvement of cell surface TG2 in the aggregation of K562 cells triggered by gluten

Gluten-induced aggregation of K562 cells represents an in vitro model reproducing the early steps occurring in the small bowel of celiac patients exposed to gliadin. Despite the clear involvement of TG2 in the activation of the antigen-presenting cells, it is not yet clear in which compartment it occurs. Herein we study the calcium-dependent aggregation of these cells, using either cell-permeable or cell-impermeable TG2 inhibitors. Gluten induces efficient aggregation when calcium is absent in the extracellular environment, while TG2 inhibitors do not restore the full aggregating potential of gluten in the presence of calcium. These findings suggest that TG2 activity is not essential in the cellular aggregation mechanism. We demonstrate that gluten contacts the cells and provokes their aggregation through a mechanism involving the A-gliadin peptide 31-43. This peptide also activates the cell surface associated extracellular TG2 in the absence of calcium. Using a bioinformatics approach, we identify the possible docking sites of this peptide on the open and closed TG2 structures. Peptide docks with the closed TG2 structure near to the GTP/GDP site, by establishing molecular interactions with the same amino acids involved in stabilization of GTP binding. We suggest that it may occur through the displacement of GTP, switching the TG2 structure from the closed to the active open conformation. Furthermore, docking analysis shows peptide binding with the β-sandwich domain of the closed TG2 structure, suggesting that this region could be responsible for the different aggregating effects of gluten shown in the presence or absence of calcium. We deduce from these data a possible mechanism of action by which gluten makes contact with the cell surface, which could have possible implications in the celiac disease onset.

Essential role for complex N-glycans in forming an organized layer of bronchial epithelium.

Mice lacking the complex subset of N-glycans due to inactivation of the Mgat1 gene die at mid-gestation, making it difficult to identify specific biological functions for this class of cell surface carbohydrates. To circumvent this embryonic lethality and to uncover tissue-specific functions for complex N-glycans, WW6 embryonic stem cells with inactivated Mgat1 alleles were tracked in chimeric embryos. The Mgat1 gene encodes N-acetylglucosaminyltransferase I (Glc-NAc-TI; EC 2.4.1.101), the transferase that initiates the synthesis of complex N-glycans. WW6 cells carry an inert beta-globin transgene that allows their identification in chimeras by DNA-DNA in situ hybridization. Independent Mgat1-/- and Mgat1+/- mutant WW6 isolates contributed like parent WW6 cells to the tissues of embryonic day (E) 10.5 to E16.5 chimeras. However, a cell type-specific difference was observed in lung. Homozygous null Mgat1-/- WW6 cells did not contribute to the epithelial layer in more than 99% bronchi. This deficiency was corrected by transfection of a Mgat1 transgene. Interestingly, heterozygous Mgat1+/- WW6 cells were also deficient in populating the layer of bronchial epithelium. Furthermore, examination of lung bud in E9.5 Mgat1-/- mutant embryos showed complete absence of an organized epithelial cell layer in the bronchus. Thus, complex N-glycans are required to form a morphologically recognizable bronchial epithelium, revealing an in vivo, cell type-specific function for this class of N-glycans.

A cell migration device that maintains a defined surface with no cellular damage during wound edge generation

Studying the rate of cell migration provides insight into fundamental cell biology as well as a tool to assess the functionality of synthetic surfaces and soluble environments used in tissue engineering. The traditional tools used to study cell migration include the fence and wound healing assays. In this paper we describe the development of a microchannel based device for the study of cell migration on defined surfaces. We demonstrate that this device provides a superior tool, relative to the previously mentioned assays, for assessing the propagation rate of cell wave fronts. The significant advantage provided by this technology is the ability to maintain a virgin surface prior to the commencement of the cell migration assay. Here, the device is used to assess rates of mouse fibroblasts (NIH 3T3) and human osteosarcoma (SaOS2) cell migration on surfaces functionalized with various extracellular matrix proteins as a demonstration that confining cell migration within a microchannel produces consistent and robust data. The device design enables rapid and simplistic assessment of multiple repeats on a single chip, where surfaces have not been previously exposed to cells or cellular secretions.

Surface Proteins of Lactobacillus crispatus: Adhesive Properties and Cell Wall Anchoring

Bacterial surface-associated proteins are important in communication with the environment and bacteria-host interactions. In this thesis work, surface molecules of Lactobacillus crispatus important in host interaction were studied. The L. crispatus strains of the study were known from previous studies to be efficient in adhesion to intestinal tract and ECM. L. crispatus JCM 5810 possess an adhesive surface layer (S-layer) protein, whose functions and domain structure was characterized. We cloned two S-layer protein genes (cbsA; collagen-binding S-layer protein A and silent cbsB) and identified the protein region in CbsA important for adhesion to host tissues, for polymerization into a periodic layer as well as for attachment to the bacterial cell surface. The analysis was done by extensive mutation analysis and by testing His6-tagged fusion proteins from recombinant Escherichia coli as well as by expressing truncated CbsA peptides on the surface of Lactobacillus casei. The N-terminal region (31-274) of CbsA showed efficient and specific binding to collagens, laminin and extracellular matrix on tissue sections of chicken intestine. The N-terminal region also contained the information for formation of periodic S-layer polymer. This region is bordered at both ends by a conserved short region rich in valines, whose substitution to leucines drastically affected the periodic polymer structure. The mutated CbsA proteins that failed to form a periodic polymer, did not bind collagens, which indicates that the polymerized structure of CbsA is needed for collagen-binding ability. The C-terminal region, which is highly identical in S-layer proteins of L. crispatus, Lactobacillus acidophilus and Lactobacillus helveticus, was shown to anchor the protein to the bacterial cell wall. The C-terminal CbsA peptide specifically bound to bacterial teichoic acid and lipoteichoic acids. In conclusion, the N-terminal domain of the S-layer protein of L. crispatus is important for polymerization and adhesion to host tissues, whereas the C-terminal domain anchors the protein to bacterial cell-wall teichoic acids. Lactobacilli are fermentative organisms that effectively lower the surrounding pH. While this study was in progress, plasminogen-binding proteins enolase and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were identified in the extracellular proteome of L. crispatus ST1. In this work, the cell-wall association of enolase and GAPDH were shown to rely on pH-reversible binding to the cell-wall lipoteichoic acids. Enolase from L. crispatus was functionally compared with enolase from L. johnsonii as well as from pathogenic streptococci (Streptococcus pneumoniae, Streptococcus pyogenes) and Staphylococcus aureus. His6-enolases from commensal lactobacilli bound human plasminogen and enhanced its activation by human plasminogen activators similarly to, or even better than, the enolases from pathogens. Similarly, the His6-enolases from lactobacilli exhibited adhesive characteristics previously assigned to pathogens. The results call for more detailed analyses of the role of the host plasminogen system in bacterial pathogenesis and commensalism as well of the biological role and potential health risk of the extracellular proteome in lactobacilli.

Isolation and partial characterization of host cell wall surface glycoproteins: their involvement in agglutination, attachment and appressorium formation by piptocephalis virginiana

Cell surface proteins obtained by alkaline extraction from isolated cell walls of Mortierella pusilla and M. candelabrum, host and nonhost, respectively, to the mycoparasite, Piptocephalis virginiana, were tested for their ability to agglutinate mycoparasite spores. The host cell wall protein extract had a high agglutinating activity (788 a.u. mg- t ) as compared with the nonhost extract (21 a.li. mg- t ). SDS-polyacrylamide gel electrophoresis of the cell wall proteins revealed four protein bands, a, b, c, and d (Mr 117, 100, 85 and 64 kd, respectively) at the host surface, but not at the nonhost surface, except for the faint band c. Deletion of proteins b or c from the host cell wall protein extract significantly reduced its agglutinating activity. Proteins band c, obtained as purified preparations by a series of procedures, were shown to be two glycoproteins. Carbohydrate analysis by gas chromatography demonstrated that glucose and Nacetylglucosamine were the major carbohydrate components of the glycoproteins. It was further shown that the agglutinating activity of the pure preparation containing both band c was 500-850 times that of the single glycoproteins, suggesting the involvement of both glycoproteins in agglutination. The results suggest that the glycoproteins band c are the two subunits of agglutinin present at the host cell surface. The two glycoproteins band c purified from the host cell wall protein extract were further examined after various treatments for their possible role in agglutination, attachment and appressorium formation by the mycoparasite. Results obtained by agglutination and attachment tests showed: (1) the two glycoprotein-s are not only an agglutinin responsible for the mycoparasite spore agglutination, but may also serve as a receptor for the specific recognition, attachment and appressorium formation by the mycoparasite; (2) treatment of the rnycoparasite spores with various sugars revealed that arabinose, glucose and N-acetylglucosamine inhibited the agglutination and attachment activity of the glycoproteins, however, the relative percentage of appressorium formation was not affected by the above sugars; (3) the two glycoproteins are relatively stable with respect to their agglutinin and receptor functions. The present results suggest that the agglutination and attachment may be mediated directly by certain sugars present at the host and mycoparasite cell surfaces while the appressorlum formation may be the response of complementary combinations of both sugar and protein, the two parts of the glycoproteins at the interacting surfaces of two fungi.

Probing cell surfaces of host and nonhost species of a mycoparasite, using FITC-labeled lectins

Cell surfaces of susceptible host species (Mortierella pusllla and Cboanepilora cucurbitarum ), resistant host (Pilascolomyces articulosus ), nonhost (Mortierella candelabrum ) and the mycoparasite (Piptocepilalis virginiana) were examined for sugar distribution patterns using light and fluorescent microscopy techniques. The susceptible host, resistant host and the mycoparasite species exhibited a similar sugar distribution profile; they all showed N-acetyl glucosamine and D-glucose on their cell wall surfaces. The nonhost cell wall surface showed a positive binding reaction to FITClectins specific for N-acetyl glucosamine and also for OI.-fucose, N-acetyl galactosamine and galactose. Treatment of these fungi with mild concentrations of proteinases (both commercial as well as the mycoparasiteproteinase) resulted in the revelation of additional sugars on the fungal cell walls. The susceptible host treated with proteinase expressed higher levels of N-acetyl glucosamine and D-glucose. The susceptible host also showed the presence of OI.-fucose, N-acetyl galactosamine and galactose. The proteinasetreated susceptible host cell walls also showed an increase in the levels of attachment with the mycoparasite. Treatment of the resistant host with proteinases revealed OI.-fucose in addition to N-acetyl glucosamine and D-glucose. Treatment of the nonhost cell wall with proteinase resulted in the exposure of low levels of D-glucose, in addition to sugars found on the untreated nonhost cell wall surface. The mycoparasite treated with proteinase revealed OI.-fucose, N-acetyl galactosamine and galactose on its cell surface in addition to the sugars N-acetyl glucosamine and D-glucose. Protoplasts were isolated from hosts and nonhost fungi and their surfaces were examined for sugar distribution patterns. The susceptible host and nonhost protoplast membranes showed all the sugars (N-acetyl glucosamine, D-glucose, (It.-fucose, N-acetyl galactosamine and galactose) tested for. The resistant host protoplast membrane however, had only N-acetyl glucosamine and D-glucose exposed. This sugar distribution profile resembles that exhibited by the untreated resistant host cell wall, as well as that shown by the untreated mycoparasite cell surface. Only susceptible host protoplasts were successful in attaching to the mycoparasite surface. Resistant host protoplasts did not show any interaction with the i mycoparasite cell surface. Both susceptible as well as resistant host protoplasts were incapable of attaching to agarose beads surface-coated with specific carbohydrates. The mycoparasite however, did attach to agarose beads surface-coated with either N-acetyl glucosamine, D-glucose/Dmannose or o:,- methyl-D-mannose. The relevance of the cell wall and the protoplast membrane in the light of the present results, in reacting appropriately to bring about either a susceptible, a resistant or a nonhost response has been discussed.

Surface engineering of silica nanoparticles for oral insulin delivery: characterization and cell toxicity studies

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Enhanced T cell transmigration across the murine liver sinusoidal endothelium is mediated by transcytosis and surface presentation of chemokines

Transmigration through the liver endothelium is a prerequisite for the homeostatic balance of intrahepatic T cells and a key regulator of inflammatory processes within the liver. Extravasation into the liver parenchyma is regulated by the distinct expression patterns of adhesion molecules and chemokines and their receptors on the lymphocyte and endothelial cell surface. In the present study, we investigated whether liver sinusoidal endothelial cells (LSEC) inhibit or support the chemokine-driven transmigration and differentially influence the transmigration of pro-inflammatory or anti-inflammatory CD4(+) T cells, indicating a mechanism of hepatic immunoregulation. Finally, the results shed light on the molecular mechanisms by which LSEC modulate chemokine-dependent transmigration. LSEC significantly enhanced the chemotactic effect of CXC-motif chemokine ligand 12 (CXCL12) and CXCL9, but not of CXCL16 or CCL20, on naive and memory CD4(+) T cells of a T helper 1, T helper 2, or interleukin-10-producing phenotype. In contrast, brain and lymphatic endothelioma cells and ex vivo isolated lung endothelia inhibited chemokine-driven transmigration. As for the molecular mechanisms, chemokine-induced activation of LSEC was excluded by blockage of G(i)-protein-coupled signaling and the use of knockout mice. After preincubation of CXCL12 to the basal side, LSEC took up CXCL12 and enhanced transmigration as efficiently as in the presence of the soluble chemokine. Blockage of transcytosis in LSEC significantly inhibited this effect, and this suggested that chemokines taken up from the basolateral side and presented on the luminal side of endothelial cells trigger T cell transmigration. CONCLUSION: Our findings demonstrate a unique capacity of LSEC to present chemokines to circulating lymphocytes and highlight the importance of endothelial cells for the in vivo effects of chemokines. Chemokine presentation by LSEC could provide a future therapeutic target for inhibiting lymphocyte immigration and suppressing hepatic inflammation.

Polyvalent binding to carbohydrates immobilized on an insoluble resin

Numerous studies have established that polyvalency is a critical feature of cell surface carbohydrate recognition. Nevertheless, carbohydrate–protein interactions are typically evaluated by using assays that focus on the behavior of monovalent carbohydrate ligands in solution. It has generally been assumed that the relative affinities of monovalent carbohydrate ligands in solution correlate with their polyvalent avidities. In this paper we show that carbohydrate ligands synthesized directly on TentaGel beads interact with carbohydrate-binding proteins in a polyvalent manner. The carbohydrate-derivatized beads can, therefore, be used as model systems for cell surfaces to evaluate polyvalent carbohydrate–protein interactions. By using a combinatorial approach to synthesize solid-phase libraries of polyvalent carbohydrates, one can rapidly address key issues in the area of cell surface carbohydrate recognition. For example, studies reported herein demonstrate that there is an unanticipated degree of specificity in recognition processes involving polyvalent carbohydrates. However, the correlation between polyvalent avidities and solution affinities is poor. Apparently, the presentation of carbohydrates on the polymer surface has a profound influence on the interaction of the ligand with the protein receptor. These findings have implications for how carbohydrates function as recognition signals in nature, as well as for how polyvalent carbohydrate–protein interactions should be studied.

Expression and putative role of lactoseries carbohydrates present on NCAM in the rat primary olfactory pathway

Primary olfactory neurons project axons from the olfactory neuroepithelium lining the nasal cavity to,the olfactory bulb in the brain. These axons grow within large mixed bundles in the olfactory nerve and then sort out into homotypic fascicles in the nerve fiber layer of the olfactory bulb before terminating in topographically fixed glomeruli. Carbohydrates expressed on the cell surface have been implicated in axon sorting within the nerve fiber layer. We have identified two novel subpopulations of primary olfactory neurons that express distinct alpha-extended lactoseries carbohydrates recognised by monoclonal antibodies LA4 and KH10. Both carbohydrate epitopes are present on novel glycoforms of the neural cell adhesion molecule, which we have named NOC-7 and NOC-8. Primary axon fasciculation is disrupted in vitro when interactions between these cell surface lactoseries carbohydrates and their endogenous binding molecules are inhibited by the LA4 and KH10 antibodies or lactosamine sugars. We report the expression of multiple members of the lactoseries binding galectin family in the primary olfactory system. In particular, galectin-3 is expressed by ensheathing cells surrounding nerve fascicles in the submucosa and nerve fiber layer, where it may mediate cross-linking of axons. Galectin-4, -7, and -8 are expressed by the primary olfactory axons as they grow from the nasal cavity to the olfactory bulb. A putative role for NOC-7 and NOC-8 in axon fasciculation and the expression of multiple galectins in the developing olfactory nerve suggest that these molecules may be involved in the formation of this pathway, particularly in the sorting of axons as they converge towards their target. (C) 2004Wiley-Liss, Inc.

A collagen IV matrix is required for guinea pig gastric epithelial cell monolayers to provide an optimal model of the stomach surface for biopharmaceutical screening

The surface epithelial cells of the stomach represent a major component of the gastric barrier. A cell culture model of the gastric epithelial cell surface would prove useful for biopharmaceutical screening of new chemical entities and dosage forms. Primary cultures of guinea pig gastric mucous epithelial cells were grown on filter inserts (Transwells®) for 3 days. Tight-junction formation, assessed by transepithelial electrical resistance (TEER) and permeability of mannitol and fluorescein, was enhanced when collagen IV rather than collagen I was used to coat the polycarbonate filter. TEER for cells grown on collagen IV was close to that obtained with intact guinea pig gastric epithelium in vitro. Differentiation was assessed by incorporation of [ 3H]glucosamine into glycoprotein and by activity of NADPH oxidase, which produces superoxide. Both of these measures were greater for cells grown on filters coated with collagen I than for cells grown on plastic culture plates, but no major difference was found between cells grown on collagens I and IV. The proportion of cells, which stained positively for mucin with periodic acid Schiff reagent, was greater than 95% for all culture conditions. Monolayers grown on membranes coated with collagen IV exhibited apically polarized secretion of mucin and superoxide, and were resistant to acidification of the apical medium to pH 3.0 for 30 min. A screen of nonsteroidal anti-inflammatory drugs revealed a novel effect of diclofenac and niflumic acid in reversibly reducing permeability by the paracellular route. In conclusion, the mucous cell preparation grown on collagen IV represents a good model of the gastric surface epithelium suitable for screening procedures. © 2005 The Society for Biomolecular Screening.

Discovery and characterization of IGFBP-mediated endocytosis in the human retinal pigment epithelial cell line ARPE-19

Insulin-like growth factor binding proteins (IGFBPs) are prime regulators of IGF-action in numerous cell types including the retinal pigment epithelium (RPE). The RPE performs several functions essential for vision, including growth factor secretion and waste removal via a phagocytic process mediated in part by vitronectin (Vn). In the course of studying the effects of IGFBPs on IGF-mediated VEGF secretion and Vn-mediated phagocytosis in the RPE cell line ARPE-19, we have discovered that these cells avidly ingest synthetic microspheres (2.0 μm diameter) coated with IGFBPs. Given the novelty of this finding and the established role for endocytosis in mediating IGFBP actions in other cell types, we have explored the potential role of candidate cell surface receptors. Moreover, we have examined the role of key IGFBP structural motifs, by comparing responses to three members of the IGFBP family (IGFBP-3, IGFBP-4 and IGFBP-5) which display overlapping variations in primary structure and glycosylation status. Coating of microspheres (FluoSpheres®, sulfate modified polystyrene filled with a fluorophore) was conducted at 37 °C for 1 h using 20 μg/mL of test protein, followed by extensive washing. Binding of proteins was confirmed using a microBCA assay. The negative control consisted of microspheres treated with 0.1% bovine serum albumin (BSA), and all test samples were post-treated with BSA in an effort to coat any remaining free protein binding sites, which might otherwise encourage non-specific interactions with the cell surface. Serum-starved cultures of ARPE-19 cells were incubated with microspheres for 24 h, using a ratio of approximately 100 microspheres per cell. Uptake of microspheres was quantified using a fluorometer and was confirmed visually by confocal fluorescence microscopy. The ARPE-19 cells displayed little affinity for BSA-treated microspheres, but avidly ingested large quantities of those pre-treated with Vn (ANOVA; p < 0.001). Strong responses were also observed towards recombinant formulations of non-glycosylated IGFBP-3, glycosylated IGFBP-3 and glycosylated IGFBP-5 (all p < 0.001), while glycosylated IGFBP-4 induced a relatively minor response (p < 0.05). The response to IGFBP-3 was unaffected in the presence of excess soluble IGFBP-3, IGF-I or Vn. Likewise, soluble IGFBP-3 did not induce uptake of BSA-treated microspheres. Antibodies to either the transferrin receptor or type 1 IGF-receptor displayed slight inhibitory effects on responses to IGFBPs and Vn. Heparin abolished responses to Vn, IGFBP-5 and non-glycosylated IGFBP-3, but only partially inhibited the response to glycosylated IGFBP-3. Our results demonstrate for the first time IGFBP-mediated endocytosis in ARPE-19 cells and suggest roles for the IGFBP-heparin-binding domain and glycosylation status. These findings have important implications for understanding the mechanisms of IGFBP actions on the RPE, and in particular suggest a role for IGFBP-endocytosis.