Outer segment phagocytosis by cultured retinal pigment epithelial cells requires Gas6.

The function and viability of vertebrate photoreceptors requires the daily phagocytosis of photoreceptor outer segments (OS) by the adjacent retinal pigment epithelium (RPE). We demonstrate here a critical role in this process for Gas6 and by implication one of its receptor protein tyrosine kinases (RTKs), Mertk (Mer). Gas6 specifically and selectively stimulates the phagocytosis of OS by normal cultured rat RPE cells. The magnitude of the response is dose-dependent and shows an absolute requirement for calcium. By contrast the Royal College of Surgeons (RCS) rat RPE cells, in which a mutation in the gene Mertk results in the expression of a truncated, non-functional receptor, does not respond to Gas6. These data strongly suggest that activation of Mertk by its ligand, Gas6, is the specific signaling pathway responsible for initiating the ingestion of shed OS. Moreover, photoreceptor degeneration in the RCS rat retina, which lacks Mertk, and in humans with a mutation in Mertk, strongly suggests that the Gas6/Mertk signaling pathway is essential for photoreceptor viability. We believe that this is the first demonstration of a specific function for Gas6 in the eye.

Nitric oxide as a second messenger in phagocytosis by cultured retinal pigment epithelial cells.

PURPOSE: To investigate a possible role of the nitric oxide (NO)-cGMP signal transduction system in phagocytosis of rod outer segments (ROS) by cultured retinal pigment epithelial (RPE) cells. METHODS: Primary cultures of RPE cells from 10-day-old Brown Norway rats were used to study the phagocytosis of ROS by these cells. Phagocytosis of ROS was evaluated with or without an inhibitor of nitric oxide synthase (NOS), N(G)-nitro-L-arginine (L-NNA), and the reverse effects of L-NNA by L-arginine and 8-bromo-cGMP on phagocytosis were also studied. NO-associated cGMP production by RPE cells was monitored during phagocytosis using L-NNA. NOS activity was assayed in RPE cells and ROS to locate the source of NO. RESULTS: Phagocytosis of ROS was inhibited by L-NNA but not by D-NNA. L-NNA inhibited the ingestion in a dose-dependent manner, but not the binding of ROS. The inhibition was reversed by L-arginine and also by an NO donor, SIN-1. RPE cells challenged with ROS showed increased cGMP activity, which was significantly reduced by L-NNA and again restored by an overdose of L-arginine. NOS activity was found in RPE cells but not in ROS. CONCLUSIONS: Our data show that cGMP plays a role in the ingestion phase of ROS phagocytosis by RPE cells via a cGMP second-messenger system.

Cell polarity, phagocytosis and viral gene transfer in cultured human retinal pigment epithelial cells.

PURPOSE: To investigate whether there is a difference in the expression of adenovirus transgenes in human retinal pigment epithelial cells when the vector was exposed to the apical or basal surface, the effect of transgene expression on rod outer segment (ROS) phagocytosis and finally, the role of phagocytosis in gene transfer to RPE cells, using the Royal College of Surgeons (RCS) rat. METHODS: Monolayers of human retinal pigment epithelium (HRPE) or an RPE cell line (A407) had the apical or basal surfaces exposed to 10(7) pfu/ml of replication deficient adenovirus (Ad.RSV.betagal) carrying the beta-galactosidase marker gene, and the numbers of expressing cells were compared. Parallel cultures were infected and challenged with fluorescein-labelled bovine rod outer segments (FBROS). The fluorescence of infected versus uninfected cells was recorded for both challenged and unchallenged states, using fluorophotometric flow cytometry. Primary cultures of RCS rat RPE were established and the transgene uptake dynamics compared to control Long Evans rat RPE cells. RESULTS: The expression of transgene in HRPE and A407 cell cultures was an order of magnitude greater when the vector was exposed apically (analysis of variance p < 0.05). There was no difference in the phagocytic capacity of Ad.RSV.betagal-infected and -noninfected cells when challenged with FBROS. There was also no difference in the number of cells expressing transgene, when compared to the RCS or Long Evans control rat RPE. CONCLUSIONS: The surface of exposure in polarized retinal pigment epithelial cells affects the rate of uptake and expression of adenovirus. The defective ROS phagocytosis in RCS rat RPE cells did not lead to a decrease in transgene expression relative to the Long Evans control cells. Finally we have found that phagocytosis is not significantly altered with adenoviral transgene expression in this in vitro model.

Further studies on the identification of the phagocytosis receptor of rat retinal pigment epithelial cells.

We have previously produced a polyclonal antiserum (R1S5) against a plasma membrane-enriched fraction of rat retinal pigment epithelial (RPE) cells which inhibits the phagocytosis of photoreceptor outer segments (OS) by these cells. This antiserum has now been used to purify a subset of RPE membrane glycoproteins. Using a combination of lectin affinity chromatography, and chromatography on an affinity column made with R1S5-IgG, we have enriched an RPE membrane extract about 100-fold. This enriched extract contains only 12 components, all of which are glycoproteins, and retains the ability to adsorb out the inhibitory activity of antiserum R1S5. This shows that one or more of these glycoproteins recognizes an inhibitory IgG in R1S5 and suggests that one or more of these glycoproteins may participate in the phagocytosis of OS by RPE cells, possibly as the phagocytosis receptor. We have performed N-terminal microsequencing of seven of these glycoproteins: four of the seven, with Mrs of 34, 36, 51 and 55 kDa, show no sequence homology to any known proteins.

Carbachol does not correct the defect in the phagocytosis of outer segments by Royal College of Surgeons rat retinal pigment epithelial cells.

PURPOSE: To investigate the effect of carbachol on the phagocytosis of photoreceptor outer segments (OS) in cultures of normal Long-Evans and dystrophic Royal College of Surgeons (RCS) rat retinal pigment epithelial (RPE) cells. METHODS: Retinal pigment epithelial cells from normal and RCS rats were grown in tissue culture. On reaching confluence, they were presented with OS suspended in Krebs-Henseleit buffer in the presence or absence of carbachol and LiCl. The number of bound and ingested OS was quantitated using double immunofluorescence staining. RESULTS: LiCl inhibited the ingestion of OS by more than 90% but had no effect on the binding of OS by Long-Evans RPE cells. The addition of carbachol further reduced OS ingestion. Carbachol alone decreased OS ingestion by normal RPE cells by 30% but had no effect on OS binding. The effect of LiCl and carbachol on RCS RPE cells was similar to their effect on normal RPE cells. CONCLUSIONS: Carbachol does not increase OS phagocytosis in normal or RCS rat RPE cells. The phagocytic defect in RCS rat RPE cannot be reversed or overcome by stimulation of the IP3 pathway by carbachol. LiCl strongly inhibits the ingestion of OS by normal and by RCS RPE cells, and this effect is enhanced by carbachol.

Single-dose compared with fractionated-dose radiation of the OM431 choroidal melanoma cell line.

PURPOSE: To compare single-dose and fractionated-dose radiotherapeutic effects on choroidal melanoma cells. METHODS: We determined the effects of gamma radiation on OM431 cell survival by exposing cells to either a single 9-Gy dose or two 4.5-Gy fractionated doses at intervals of 20 minutes to eight hours. The effects of single dosing and fractionated dosing at six hours were compared at doses of 2 to 12 Gy. RESULTS: Tumor cell repair was most rapid during the first two hours. Maximum repair had occurred by six hours after radiation. Cell survival curves showed doses greater than 3 Gy of single-dose gamma radiation resulted in a greater number of cells killed than did equivalent fractionated doses. CONCLUSIONS: Ocular melanoma in vitro is relatively radioresistant to low-dose fractionated radiotherapy. High single-dose radiotherapy would be more effective but would also result in more damage to normal tissue unless more focused modalities of radiotherapy are used.

Effects of gamma radiation on the OM431 human ocular melanoma cell line.

In order to determine the dose responsiveness to radiation of ocular melanoma, we conducted an in vitro dose-response study on a monolayer cell culture using a clonogenic assay. The effects on cell survival were determined relative to unirradiated controls. A human epithelioid ocular melanoma cell line, OM431, was maintained in tissue culture and serial dilutions of viable cells were plated in flasks, allowed to settle and attach for 48 h, and subsequently irradiated with 1-10 Gy in single fractions. After 2 weeks, the number of reproducing clones (forming colonies with greater than 32 cells or five generations) were counted. The surviving fractions of cells were plotted on a cell survival curve using the linear quadratic model. The survival curve showed a large initial shoulder followed by an exponential decline in growth. Our data suggest that the OM431 ocular melanoma cell line responds to irradiation in a manner similar to other melanoma cell lines and is relatively radioresistant especially at lower doses.

Isoproterenol inhibits rod outer segment phagocytosis by both cAMP-dependent and independent pathways.

PURPOSE: The authors studied the involvement of cAMP-dependent second messenger systems in the inhibition of rod outer segment (ROS) phagocytosis by isoproterenol (ISO) and forskolin (FSK) using two membrane-permeant analogs of cyclic adenosine monophosphate (cAMP), the Rp and Sp diastereoisomers of cyclic adenosine 3',5' monophosphothioate (cAMPS). Rp-cAMPS is a potent competitive inhibitor of cAMP-dependent protein kinase I and II (PKA I and II), whereas Sp-cAMPS is a potent activator of these enzymes. METHODS: ROS phagocytosis was quantitated in cultured rat RPE cells using a previously described double immunofluorescence assay. RESULTS: Sp-cAMPS showed a dose-dependent inhibition of ROS phagocytosis, whereas 100 microM Rp-cAMPS had no effect on this process. Rp-cAMPS fully prevented the inhibitory effect of Sp-cAMPS and FSK but was able to prevent only partially the inhibition of ROS phagocytosis induced by ISO. Isoproterenol plus FSK showed an additive effect on the inhibition of phagocytosis, suggesting that they act at two independent sites. However, ISO plus Sp-cAMPS or FSK plus Sp-cAMPS showed no additivity. CONCLUSIONS: Results suggest that FSK inhibits ROS phagocytosis by RPE cells through a cAMP-dependent pathway, whereas ISO inhibits ROS phagocytosis by RPE cells through cAMP-dependent and cAMP-independent pathways.

Stimulation of A2 adenosine receptors inhibits the ingestion of photoreceptor outer segments by retinal pigment epithelium.

PURPOSE: Recent studies have shown that A2 adenosine receptors are present in retinal pigment epithelium (RPE). In this study, the effect of adenosine and adenosine analogues on photoreceptor outer segment (ROS) phagocytosis by RPE was investigated. METHODS: Primary cultures of RPE cells were incubated with isolated outer segments in the presence of various adenosine derivatives. Changes in adenylyl cyclase activity was measured by cyclic adenosine monophosphate (cAMP) production using a radioimmunoassay detection system. RESULTS: Adenosine inhibited the ingestion phase of phagocytosis (IC50 = 50 microM), and this effect was potentiated 80-fold in the presence of dipyridamole (IC50 = 0.6 microM). In the presence of 10 microM 8-phenyltheophylline, the inhibitory effect of 100 microM adenosine was reduced from 80% inhibition of ROS ingestion to 33% inhibition. The rank order of potency of adenosine analogues to inhibit ROS ingestion by RPE was N6-cyclohexyladenosine/5'-[N-ethylcarboxamido]-adenosine (NECA) = NECA > adenosine >> [R]-N6-[2-phenylisopropyl]-adenosine. The greatest stimulation of cAMP production was observed with 33.3 microM NECA: The production of cAMP reached its maximum level after 2 minutes of incubation, and after 10 minutes the levels of cAMP were back to basal. CONCLUSIONS: These results suggest that adenosine and adenosine analogues modulate ROS ingestion by RPE via activation of adenosine A2b receptors, possibly through the cAMP intracellular signaling pathway.

The phagocytosis of rod outer segments is inhibited by drugs linked to cyclic adenosine monophosphate production.

PURPOSE: To study the effect of drugs that increase intracellular cyclic adenosine monophosphate on the ability of rat retinal pigment epithelial cells to phagocytize rod outer segments (ROS). METHODS: Cultured rat retinal pigment epithelial cells were treated with cholera toxin, forskolin, isoproterenol, or isobutylmethylxanthine and the phagocytosis of ROS by such treated cells was compared to that of control specimens. RESULTS: All of the drugs examined inhibited the ingestion, but not the binding of ROS by cultured retinal pigment epithelial cells. Cell viability was not compromised by the drug treatment because they rapidly recovered their ability to ingest ROS when the drug was removed. Dose-response curves for the inhibition of ROS phagocytosis by forskolin and isoproterenol demonstrated that this process is exquisitely sensitive to these agonists, with an IC50 for these drugs of 33 nmol/l. The results showed no measurable quantitative correlation between cyclic adenosine monophosphate levels and the inhibition of ROS phagocytosis. CONCLUSIONS: Results showed that the ingestion of ROS by retinal pigment epithelial cells was inhibited by agents that increase intracellular cyclic adenosine monophosphate, but seems to be independent of the level of this second messenger. Alternatively, ROS phagocytosis may be exquisitely sensitive to changes in the intracellular concentration of cyclic adenosine monophosphate, which are too small to measure by available methods.

cAMP production via the adenylyl cyclase pathway is reduced in RCS rat RPE.

cAMP production was investigated in retinal pigment epithelium (RPE) cells isolated from normal rats and from rats with an inherited retinal dystrophy (Rdy/p+). In normal RPE cells, 5'-[N-Ethylcarboxamido]-adenosine (A2 receptors) produced a fivefold increase in the level of cyclic adenosine monophosphate (cAMP) over basal levels. However, only a onefold increase in cAMP was observed in dystrophic cells. cAMP production by prostaglandins E1 and E2 (prostaglandin receptors) in dystrophic RPE cells was only 29-38% of the level observed in normal cells. Direct stimulation of adenylyl cyclase by 10 mumol/l forskolin increased cAMP levels in normal RPE cells by 90 fold over basal, but only by sixfold in the dystrophic cells. These data suggest there may be a defect in the adenylyl cyclase signaling pathway in dystrophic RPE cells.

The phagocytosis of ROS by RPE cells is inhibited by an antiserum to rat RPE cell plasma membranes.

A polyclonal antiserum to a rat retinal pigment epithelium (RPE) plasma membrane-enriched fraction has been utilized to identify candidate receptor proteins which may be involved in the phagocytosis of rod outer segments (ROS) by the RPE. Immunoblots of RPE cell extracts show that the the antiserum recognizes a number of glycoproteins, including two with M(r)s of 174 and 75 kDa. The antiserum also recognizes their non-glycosylated counterparts, with M(r)s of 169 and 65 kDa, respectively, which are synthesized after treatment of the cells with tunicamycin B2. Immuno-precipitation of [35S]-methionine-labeled RPE cell extracts also demonstrates the presence of antibodies to these same glycoproteins as well as to other proteins. The antiserum inhibits the binding of ROS to the RPE, which subsequently results in a decrease in the ingestion of ROS. ROS phagocytosis by the RPE is inhibited by 97% in the presence of a 1:10 dilution of the IgG fraction of the antiserum. Phagocytosis recovers to normal levels after 4-6 hr of chase in the absence of antibodies. After sequential adsorption of the IgG fraction to monolayers of fixed RPE cells, which removes RPE surface-specific IgGs, the extent of inhibition of ROS phagocytosis produced by the IgG fraction is reduced. Using immunoblotting we have identified a number of surface-specific immunoreactive bands which are adsorbed out of the antiserum, including the 174 and 75 kDa bands. These data give further support to the hypothesis that ROS phagocytosis is a receptor-mediated process, which occurs via specific cell surface glycoprotein receptors.

The phagocytosis of ROS by RPE cells is not inhibited by mannose-containing ligands.

We have examined the ability of mannose and the mannose-rich ligands, mannan and mannosylated BSA, to inhibit the phagocytosis of rod outer segments (ROS) by cultured rat retinal pigment epithelial (RPE) cells. Mannose, at concentrations up to 0.25 M, had no effect on either the binding or the ingestion of ROS. At concentrations above 0.25 M, the cells were rounded and showed detachment from the substrate, and phagocytosis was markedly inhibited. Neither mannan (2 mg ml-1), nor mannosylated BSA(0.8 mg ml-1), affected the phagocytosis of ROS. These results suggest that the phagocytosis of ROS is probably not mediated by a mannose receptor on the surface of the RPE cells.

ROS ingestion by RPE cells is turned off by increased protein kinase C activity and by increased calcium.

The activation of protein kinase C (PKC) by phorbol myristate acetate (PMA) rapidly inhibits the phagocytosis of rod outer segments (ROS) by cultured rat retinal pigment epithelial (RPE) cells. PMA, at a concentration between 3.3 and 10 nM, blocks ROS ingestion by 50%, but does not inhibit the binding of ROS. The Ca2+ ionophore, A23187, also inhibits ROS phagocytosis, with an IC50 of about 0.5-1.0 microM and interferes with the ability of RPE cells to bind ROS. The effects of both of these drugs are reversible after drug washout. When PMA and A23187 are applied to cells consecutively, the effects are additive. These results suggest either that PMA and A23187, act upon the same proteins in the pathway which controls ROS ingestion, or that A23187 affects phagocytosis at the ROS binding level, while PKC affects steps further along the ingestion path. The effect of this process is to shut down the ingestion of ROS, as is seen during the prolonged feeding of ROS to RPE cells in culture.

RPE cells from normal rats do not secrete a factor which enhances the phagocytosis of ROS by dystrophic rat RPE cells.

Retinal pigment epithelial (RPE) cells from normal and dystrophic rats were grown separately and in mixed culture for 7 days, without a change of growth medium. Isolated rod outer segments (ROS) were suspended in the conditioned medium from these cells, and were fed to the mixed or pure RPE cell cultures. No increase or decrease in the phagocytosis of ROS by dystrophic or normal RPE cells, respectively, was observed. These results suggest that normal RPE cells do not secrete a diffusible factor(s) which enhances the phagocytosis of ROS by dystrophic RPE cells.

The effect of inhibitors of glycoprotein synthesis and processing on the phagocytosis of rod outer segments by cultured retinal pigment epithelial cells.

Retinal pigment epithelial cells selectively phagocytize rod outer segments by a process that may be mediated by specific cell surface receptors. Since many receptors are glycoproteins, we have studied the effect of tunicamycin, an inhibitor of N-linked oligosaccharide synthesis, and of castanospermine and swainsonine, which are inhibitors of oligosaccharide processing, on the ability of cultured retinal pigment epithelial cells to phagocytize rod outer segment. Tunicamycin inhibits the glycosylation of newly synthesized glycoproteins by 85-90%; concomitantly, the phagocytosis of rod outer segments is inhibited by 70-80%. The effect of tunicamycin is to initially reduce rod outer segments binding, and therefore the subsequent ingestion of rod outer segments. SDS-PAGE analysis and autoradiography of [35S]methionine labelled extracts of tunicamycin-treated cells, demonstrates the disappearance of a number of glycoprotein bands, and the appearance of a number of protein bands of lower Mr. Kinetic analysis of the disappearance and reappearance of specific glycoproteins suggests that the lower Mr bands are the non-glycosylated forms of the higher Mr bands. By contrast, castanospermine and swainsonine have no effect on the ability of retinal pigment epithelial cells to phagocytize rod outer segments, or on the SDS-PAGE pattern of treated cells, although they were shown to inhibit oligosaccharide processing as expected. These results support the hypothesis that rod outer segment phagocytosis by retinal pigment epithelial cells is mediated by specific glycoprotein receptors. N-Glycosylation of these receptors is required for their function, or for their insertion into the plasma membrane, whereas processing of the N-linked oligosaccharide chains of these receptors is not crucial for rod outer segment phagocytosis by retinal pigment epithelial cells.

Kinetic studies of rod outer segment binding and ingestion by cultured rat RPE cells.

Retinal pigment epithelial (RPE) cells selectively phagocytize rod outer segments (ROS) by a process which may be mediated by specific cell surface receptors. We have studied the kinetics of this process using rat RPE cells grown in tissue culture. By cooling RPE cells to 17 degrees C, the binding and ingestion phases of phagocytosis can be separated. Maximum ROS binding with minimum ingestion occurs at 17 degrees C; above 17 degrees C the rate of ingestion increases markedly. Thus it is possible to measure the kinetics of ROS binding to RPE cells at 17 degrees C and of ROS ingestion at 37 degrees C. At 17 degrees C, ROS binding is saturable, both with respect to time and to ROS concentration. ROS ingestion saturates after 4 hr of incubation at 37 degrees C, after which the cells are refractory to further ROS ingestion for 1-2 hr. During this recovery period, rapid digestion of the internalized ROS takes place. Cycloheximide, when present at a concentration (2 x 10(-5) M) which inhibits protein synthesis by 92%, has no effect on ROS phagocytosis or on the recovery of ROS ingestion at 37 degrees C. This suggests that if receptors mediate the ingestion of ROS by RPE cells, they are not degraded after the ROS are internalized. Dystrophic rat (RCS-p+) RPE cells exhibit normal binding, but very limited ingestion of ROS at 37 degrees C. The rate and amount of ROS binding to these cells at 37 degrees C is comparable with that occurring to normal cells at 17 degrees C. These observations support the hypothesis that there are a limited number of receptors which are specific for ROS binding on the surface of normal and dystrophic rat RPE cells.

Surface modification of retinal pigment epithelial cells: effects on phagocytosis and glycoprotein composition.

Proteases have been used as a tool to investigate the role of cell-surface molecules of cultured retinal pigment epithelial cells (RPE) in the phagocytosis of rod outer segments (ROS). Proteolytic digestion of RPE cells by pronase, thermolysin and Staphylococcus aureus V8 protease (V8 protease) inhibited the phagocytosis of ROS without affecting the viability of the RPE cells. A particular feature of RPE cell proteolysis was that those macromolecules responsible for ROS ingestion were susceptible, while those macromolecules that mediated ROS binding were resistant to cleavage by all three proteases. By taking advantage of this phenomenon, ROS were used as affinity particles to obtain a plasma membrane-enriched fraction of RPE cells before and after proteolytic digestion. All three proteases partially or completely removed several glycoproteins from the cell surfaces. Removal of these glycoproteins was correlated with a loss in phagocytic ability by RPE cells. Two high-molecular-weight (MW) glycoproteins of MWs 160,000 and 214,000 were consistently removed by all proteases tested. Protease-treated RPE cells restored their phagocytic capabilities and normal glycoprotein composition within 24 hr after proteolytic treatment. These data suggest that glycoproteins located on the surfaces of RPE cells may be involved in mediating the phagocytosis of ROS by these cells.

Rat retinal pigment epithelial cells show specificity of phagocytosis in vitro.

The retinal pigment epithelial (RPE) cell of the eye normally phagocytozes only retinal rod outer segments (ROS). The specificity of this phagocytic process was examined by incubating RPE cells with a variety of particle types. Confluent RPE cell cultures were incubated for 3 h at 37 degrees C in the presence of rat ROS, rat red blood cells (RBC), algae, bacteria, or yeast. Other cell cultures were incubated with equal numbers of ROS and one other particle type. Quantitative scanning electron microscopy was used to determine the numbers and morphology of particles bound to RPE cells, while double immunofluorescence labeling (Chaitin, M. H., and M. O. Hall, 1983, Invest. Ophthalmol. Vis. Sci., 24:812-820) was used to quantitate particle binding and ingestion. Both assays demonstrated phagocytosis to be a highly specific process. RPE cells bound 40-250 X more ROS than RBC, 30 X more ROS than algae, and 5 X more ROS than bacteria or yeast. Ingestion was more specific than binding; RPE cells ingested 970 X more ROS than RBC, 140 X more ROS than bacteria, and 35 X more ROS than yeast. The phagocytic preference for ROS was maintained in competition experiments with other particle types. Serum was found to be essential for phagocytosis. This study demonstrates that both the binding and ingestion phases of phagocytosis are highly specific processes.

Phagocytosis of light- and dark-adapted rod outer segments by cultured RPE cells: a reassessment.

Rod outer segments (ROS) isolated from adult rat retinas are phagocytized by cultured rat retinal pigment epithelial (RPE) cells. Using a double immunofluorescent labeling procedure, we have compared the binding and ingestion of ROS isolated at different times of the day. After 2 hr of incubation, approximately 98% of the ROS are ingested, while 2% are still attached to the RPE cell surface, irrespective of the time of day or lighting conditions under which the ROS are isolated. These findings differ from those reported earlier, using a radioactive method for quantitating ROS phagocytosis (Hall, 1978).