Photoreceptor outer segment phagocytosis attenuates oxidative stress-induced apoptosis with concomitant neuroprotectin D1 synthesis

Abstract

Photoreceptor cell (rods and cones) renewal is accompanied by intermittent shedding of the distal tips of the outer segment followed by their phagocytosis in the retinal pigment epithelial (RPE) cells. This renewal is essential for vision, and it is thought that it fosters survival of photoreceptors and of RPE cells. However, no specific survival messenger/mediators have as yet been identified. We show here that photoreceptor outer segment (POS) phagocytosis markedly attenuates oxidative stress-induced apoptosis in ARPE-19 cells in culture. This phenomenon does not seem to be a generalized outcome of phagocytosis because nonbiological (polystyrene microsphere) phagocytosis did not elicit protection. The free docosahexaenoic acid (DHA) pool size and neuroprotectin D1 (NPD1) content increased during POS phagocytosis but not during microspheres phagocytosis. We have also explored other lipid mediators [lipoxin A4 and 15(S)- and 12(S)-hydroxyeicosatetraenoic acids] under these conditions and found them unchanged upon POS phagocytosis. Moreover, oxidative stress challenge to RPE cells undergoing POS phagocytosis further increased DHA and NPD1 content. Under these conditions, NPD1 was found within the RPE cells as well as in the culture medium, suggesting autocrine and paracrine bioactivity. Furthermore, using deuterium-labeled DHA, we show that as the availability of free DHA increases during oxidative stress, NPD1 synthesis is augmented in ARPE-19 cells. Our data suggest a distinct signaling that promotes survival of photoreceptor and RPE cells by enhancing the synthesis of NPD1 during phagocytosis. Taken together, NPD1 may be a mediator that promotes homeostatic regulation of cell integrity during photoreceptor cell renewal.

Photoreceptor cells (rods and cones) are highly specialized and differentiated neurons with stacks of photosensitive disks that contain rhodopsin and other quantitatively minor proteins in their outer segments. There is a constant rebuilding of photoreceptor cells, mainly through the shedding of the outer segment tips, followed by their phagocytosis in the retinal pigment epithelial (RPE) cells. This renewal is daily and circadian in mammalians, and as a result of disk membrane biogenesis at the base of the outer segment, the photoreceptor outer segment length remains constant. During this process, proteins turn over and are continuously replaced (1, 2). In contrast, docosahexaenoyl chains of disk membrane phospholipids are tenaciously retained during the outer segment renewal. Docosahexaenoic acid (DHA), after phagolysosomal digestion in the RPE cell, is recycled back through the interphotoreceptor matrix to the inner segment of photoreceptors for its reutilization in membrane biogenesis (3–5).

DHA belongs to the essential omega-3 fatty acid family and therefore cannot be made de novo in the body. The photoreceptor cells, unlike most other cells of the body except brain and sperm cells, are highly enriched in DHA and tenaciously retain DHA even during very prolonged periods of omega-3 fatty acid deprivation. Studies in mice correlating photoreceptor and synapse biogenesis during the postnatal development demonstrate that dietary linolenic acid is actively elongated and desaturated in the liver before its distribution through the bloodstream to the retina and brain (6). Several studies have shown that DHA is required for photoreceptor function and vision both in animals (7, 8) as well as in humans (9, 10). Moreover, the essentiality of DHA has been documented for vision and brain maturation in premature babies and in newborns (8, 11).

It is not clear how the supply of DHA from the liver is regulated. It has been hypothesized that a “signal” from the retina and/or brain to the liver may evoke DHA supply (6). Under normal conditions, DHA is retained and protected from peroxidation. However, in experimental models of degenerations (12), when lipid peroxidation occurs, perturbations of photoreceptor function, damage, and cell death take place. In several forms of retinitis pigmentosa (13–17) and in Usher syndrome (18), a decrease of DHA content in blood has been reported. A possible implication of these studies is that decreased DHA supply to the retina may impair photoreceptor function by decreasing the availability of DHA to photoreceptors. However, the relationship between decreased DHA blood supply and disease initiation and progression remains to be ascertained. Moreover, specific structural roles for DHA in photoreceptor outer segments (POS) have been suggested, such as an elongation product of DHA that is found tightly bound to rhodopsin that may favor a membrane organization unlike the classical bilayer (19). The functional significance and the pathological consequences (when peroxidation or shortage in DHA supply occurs) of this DHA elongation product have not been defined. Also, phospholipids containing DHA have been suggested to provide an appropriate environment for G protein-coupled events to take place in the POS (20). In addition, several studies have suggested a neuroprotective role for DHA (20–25).

Although it has been known that POS phagocytosis in RPE cells is essential for photoreceptor cell function and survival, thus far no specific messengers/mediators of this process have been identified. We hypothesized that POS phagocytosis makes RPE cells less susceptible to oxidative stress-induced apoptosis by providing DHA for neuroprotectin D1 (NPD1) synthesis. NPD1 is derived from DHA and is a potent, stereospecific inhibitor of cytokine-induced proinflammatory gene induction and apoptosis (26). Here, we show that feeding POS to ARPE-19 cells results in enhanced refractoriness to oxidative stress-induced apoptosis. This action is specific for POS because it was observed that the phagocytosis of polystyrene microspheres by ARPE-19 cells did not lead to a protective response against oxidative stress. Moreover, we show that POS, but not polystyrene microspheres, induces DHA release and NPD1 synthesis. Our findings reveal that POS phagocytosis through DHA supply and NPD1 synthesis represents a homeostatic regulatory process of RPE cell integrity.

Results and Discussion

POS and Polystyrene Microsphere Phagocytosis in ARPE-19 Cells.

ARPE-19 cells fed with either POS or polystyrene microspheres actively ingest these particles under the present experimental conditions (Fig. 1 A–F). Rhodopsin immunoreactivity from RPE cells incubated with POS shows internalization of these membranes in the proximity of the nuclei or below the nuclei when stacks of 50 confocal optical images were visualized (Fig. 1 B and C; bottom of the cell culture monolayer is indicated by the white arrows). Similar images, demonstrating internalization, were obtained by analyzing cells incubated with FITC-labeled polystyrene microspheres (Fig. 1 D–F). The relative fluorescence abundance was much higher in ARPE-19 cells fed microspheres than POS even though 10 million particles of each were added to wells containing similar cell density (Fig. 1). The differences may be because incubations were for 16 h, a time at which rhodopsin degradation was ongoing, whereas microspheres are not degradable.

Fig. 1.

Phagocytosis of POS and of polystyrene microspheres by ARPE-19 cells. (A–F) Confocal microscopy demonstrating phagocytosis of POS and microspheres by ARPE-19 cells. (A) Confocal image stack showing rhodopsin-labeled POS (green) and ARPE-19 cell Hoechst-labeled nuclei (blue). (B) View through the depth of the image stack at the position represented by the green line in A. (C) View through the same stack at the location of the red line in A. The depth of image A within the stack is indicated by blue lines in B and C. (D) Confocal image stack showing FITC-labeled latex microspheres (green) and ARPE-19 cell Hoechst-labeled nuclei (blue). (E) View through the depth of this image stack at the location of the green line in D. (F) View through the same stack at the place denoted by the red line in D. As above, the image plane in D is denoted by blue lines in E and F. Both POS and latex microspheres occur at a level corresponding to the lateral sides of the cell nuclei or below, as viewed in B, C, E, and F, indicating that they are contained within the cytoplasm of these cell cultures. White arrowheads point to the bottom of the culture well. Both images are presented at the same magnification. (Original magnification: ×400.) Polystyrene beads are 2 μm in diameter. (G) ARPE-19 cells that were grown in culture for 72 h, phagocytized isolated FITC-labeled bovine rod outer segments. Phase-contrast (Left) and green fluorescence (Right) images of the same cells show internalized POS in ring-like configurations, tightly surrounding the nuclei. (H) Some cultures were counterstained red for better contrast with the phagocytized FITC-labeled outer segments (yellow).

POS Phagocytosis Selectively Attenuates Oxidative Stress-Induced Apoptosis.

We found that ARPE-19 cells fed with bovine POS were more resistant to oxidative stress than cells that did not phagocytize rod outer segments (Fig. 2 A and B). Fig. 2 shows that neither POS nor microspheres alone trigger Hoechst-positive cells, and POS combined with oxidative stress markedly decreased oxidative stress-induced apoptosis, unlike microspheres.

Fig. 2.

POS phagocytosis attenuates oxidative stress-induced apoptosis. Cells were plated and maintained in culture for 72 h followed by 8 h of serum starvation before the addition of POS or polystyrene microspheres. (A) Hoechst staining shows an increase in cell death after 16 h of H₂O₂/TNF-α (oxidative stress) exposure, which is ameliorated by POS phagocytosis but not by polystyrene microspheres. (B) Quantitative analysis of Hoechst-stained ARPE-19 cells indicates that POS phagocytosis significantly decreases the amount of apoptosis observed during oxidative stress. Phagocytosis of polystyrene microspheres during oxidative stress does not alter the amount of apoptosis observed during oxidative stress alone. Results represent averages ± SEM of repeats of two independent experiments.

Free DHA and NPD1 Synthesis During Phagocytosis in RPE Cells.

The RPE cell recycles DHA from phagocytized disk membranes back to the inner segment of the photoreceptor cell through the interphotoreceptor matrix (2–5). A salient property of the retina and brain is their unusual ability to retain DHA tenaciously even during prolonged periods of omega-3 fatty acid deprivation (2, 22, 27). The RPE cell, in addition, contributes to the DHA-enriching ability of the photoreceptor cells by taking up DHA from the bloodstream through the choriocapillaris (2). However, the bulk of DHA in the RPE cell is a component of photoreceptor disk membrane phospholipids that, after phagocytosis, is recycled as part of outer segment renewal. Because DHA is the initial precursor of NPD1 synthesis, we next measured the pool size of DHA (Fig. 3) and of NPD1 (Fig. 4) during phagocytosis and oxidative stress. Data shown in Fig. 3 indicate that significant levels of DHA accumulate in cells and medium 6 h after the onset of phagocytosis. Oxidative stress triggered further increases in free DHA at 6 and 16 h.

Fig. 3.

Free (unesterified) DHA pool size changes as a function of time after POS phagocytosis or microspheres: effect of oxidative stress. DHA has been quantified in cells as well as in incubation media. Data represent the average ± SEM of two independent studies; statistical analysis was done by Student's t test. NS, not statistically significant.

Fig. 4.

NPD1 changes as a function of time after POS phagocytosis or microspheres: effect of oxidative stress. NPD1 has been quantified in cells as well as in incubation media. Data analysis was as in Fig. 3. NS, not statistically significant.

Fig. 4 depicts a remarkable phagocytosis-dependent NPD1 synthesis, particularly in the presence of oxidative stress. Rod outer segment tips are the biologically relevant ligand for the RPE. When RPE cells undergoing POS phagocytosis were subjected to oxidative stress, in as early as 3 h, a 3-fold accumulation of NPD1 was observed compared with cells that were not fed POS. This increase in NPD1 was almost 6.8 times higher than RPE cells that phagocytized microspheres. This accumulation peaked after 6 h. In the incubation medium, accumulation of NPD1 also occurred at 6 h and increased at 16 h, which represented a 36.6-fold increase for the POS-treated cells after oxidative stress. Control cells showed only a 15-fold increase after oxidative stress. This enhanced synthesis of NPD1 after POS phagocytosis is concomitant with POS-induced attenuation of oxidative stress-mediated apoptosis (Fig. 2). Although ARPE-19 cells also phagocytized the biologically inert polystyrene microspheres, NPD1 content was not affected in RPE cells or in the incubation medium (Fig. 4). Moreover, although oxidative stress did stimulate NPD1 accumulation, it was also not affected by microsphere phagocytosis. These results correlate with the observed lack of cytoprotection by microspheres phagocytosis (Fig. 2). In addition, POS-mediated RPE protection against oxidative stress, with concurrent NPD1 synthesis, takes place in primary human RPE cells at passage 3, prepared from NDRI-supplied human eyes (P.K.M., V.L.M., J.C.d.R.V., W.C.G., and N.G.B., unpublished work).

Unlike nonspecific, nonbiological ligand microsphere phagocytosis, POS also has been reported to trigger early-response gene induction in the RPE (28), including cyclooxygenase 2 (29) and peroxisome proliferator-activated receptor γ expression (30). Whether any of these events is related to the NPD1 survival signaling described here remains to be ascertained.

We simultaneously measured the free DHA pool size by tandem MS and found that it increases as a function of time of exposure to oxidative stress in ARPE-19 cells (Fig. 3). Free DHA in cells showed a moderate increase after 6 h when cells were subjected only to POS phagocytosis (10.5-fold increase). Oxidative stress, however, strongly enhanced free DHA accumulation in a time-dependent fashion, peaking at 16 h. Interestingly, although the overall increase reached 10-fold, POS phagocytosis kept the DHA pool size at a constant 2.4-fold increased level, which implies that NPD1 synthesis reflects an event other than simply enhanced overall availability of free DHA upon phagocytosis. There is a correlation between increases in free DHA pool size and in NPD1 synthesis. POS phagocytosis stimulates NPD1 synthesis at 3–6 h in cells and accumulation in medium after 16 h (Fig. 4), whereas free DHA increases earlier and keeps accumulating up to 16 h (Fig. 3). These enhancements in DHA and NPD1 pool size are much larger when POS phagocytosis takes place on RPE cells exposed to oxidative stress. Interestingly, microsphere phagocytosis does not cause enhanced changes in DHA and NPD1. Thus, a very specific free DHA pool may be the precursor for NPD1. Arachidonic acid is also an active precursor of several bioactive lipids, including prostaglandins and lipoxygenase products, and they have been correlated with photoreceptor phagocytosis (31, 32). Because arachidonic acid is released under the present experimental conditions (data not shown), it was therefore of interest to explore some of the arachidonic acid cascade members. We found that lipoxin A4 [supporting information (SI) Fig. 6], 12(S)-hydroxyeicosatetraenoic acid (HETE) (SI Fig. 7), and 15(S)-HETE (SI Fig. 8) were unchanged during POS phagocytosis. These findings suggest that an unrecognized function of photoreceptor phagocytosis is to induce NPD1 synthesis and thereby elicit cytoprotection. The RPE is highly susceptible to oxidative stress because of the high oxygen consumption of the retina, active flux of polyunsaturated fatty acid (DHA)-containing phospholipids, and exposure to light (2, 33). Thus, NPD1 may be a major endogenous promoter of RPE cell survival during POS renewal.

Synthesis of Neuroprotectin D1 in ARPE-19 Cells Undergoing Oxidative Stress.

The studies depicted in Figs. 3 and 4 demonstrate a positive correlation between enhanced free DHA and increased NPD1 content in ARPE-19 cells. To determine whether enhanced availability of free DHA leads to the synthesis of NPD1 in these cells undergoing oxidative stress, we have used deuterium-labeled DHA ([²H₅]DHA). ARPE-19 cells were incubated with [²H₅]DHA and sampled as a function of time, and [²H₅]NPD1 was identified and quantified. Fig. 5 illustrates the characterization of [²H₅]DHA, in cells and incubation media, by tandem liquid chromatography (LC)–photodiode array–electrospray ionization–tandem MS-based lipidomic analysis. This approach allows us to follow DHA conversion specifically because the deuterium is on the methylene carbons 21 and 22, which are not metabolically altered. Also, the products are heavier (by a mass unit of 1) than the same nondeuterated molecule and can be detected by tandem MS. Fig. 5 and SI Fig. 9 illustrate aspects of the characterization of [²H₅]NPD1 (negative molecular ion m/z 364.2) and of endogenous nondeuterated NPD1 (negative molecular ion m/z 359.2). These observations support the notion that as free DHA accumulates in the ARPE-19 cells during POS phagocytosis, it is a substrate for NPD1 synthesis. At the time of the initial identification of lipoxygenase products from DHA, mono-, di-, and trihydroxylated derivatives were identified (33). Further studies are needed to determine whether other docosanoids may also be formed under these conditions.

Fig. 5.

Deuterated DHA, added to the ARPE-19 cell medium, is incorporated and converted to deuterated NPD1. (A) Time course of synthesis of d5-NPD1 in ARPE-19 cells and medium, after 100 nM [²H₅]DHA was added to cells at the onset of oxidative stress. Results are the average ± SEM (n = 3). (B) Molecular structure of [²H₅]NPD1, showing molecular ion m/z 364.15; the localization of deuterium at carbon 21 and 22 is also indicated. Dashed lines show typical product ion breakpoints at MS collision cell. (C) Molecular structure of endogenous NPD1, with molecular and product ions as indicated in B. (D and E) Typical TIC curve of detection for [²H₅]NPD1 and endogenous NPD1 (E) in ARPE-19 cells treated as described in A.

Conclusions

We have demonstrated that POS phagocytosis selectively enhances the ability of ARPE-19 cells to withstand oxidative stress. Under these conditions, the free DHA pool size is up-regulated in concert with NPD1 increases. In addition, using deuterium-labeled DHA we have ascertained that ARPE-19 cells use free DHA for NPD1 synthesis when exposed to oxidative stress. Remaining to be defined are the regulatory events triggered by POS phagocytosis, including signals that activate phospholipase A₂, to cleave docosahexaenoic chains from phospholipids of the POS disk membranes. Several phospholipases A₂ are present in the RPE cells, and recently a low molecular weight phospholipase A₂ was suggested to be engaged in POS phagocytosis (34). Another question is how the release of DHA is coupled to NPD1 synthesis through a 15-lipoxygenase enzyme (26).

It has been suggested that shedding the tips of POS may reflect removal of older disks and/or the presence of oxidized phospholipids. In fact, oxidized phospholipids have been proposed as a signal for the initiation of phagocytosis by the RPE cell (35). Studies indicate that oxidized POS promotes down-regulation of factor H in the RPE (36). Factor H is an inhibitor of the alternative pathway of complement system activation and, as a result, has the property of limiting cell injury and inflammation (36, 37). It has also been implicated as a major risk factor for age-related macular degeneration. Here, we demonstrate that outer segment phagocytosis elicits a specific cytoprotective response in the RPE involving the synthesis of the antiinflammatory and antiapoptotic survival mediator NPD1. This response may represent a homeostatic regulatory mechanism that, during outer segment renewal, promotes maintenance of cell integrity and function despite an environment prone to oxidative damage. Also, neurotrophins, particularly pigment epithelium-derived factor, are potent agonists of NPD1 formation and of its release through the apical RPE side (38). Taken together, these findings suggest that the DHA–NPD1 signaling may participate in promoting cellular integrity, maintaining the function of photoreceptors and of the RPE, as well as participating in regulating differentiation and organizational patterning of photoreceptor cells and the RPE. The active secretion of NPD1 from RPE cells, reported here as well as upon the exposure to neurotrophins (38), suggests autocrine and paracrine bioactivity of this lipid mediator. The breakdown of this homeostatic DHA–NPD1 regulation may be involved in the initiation and progression of retinal degenerative diseases.

Materials and Methods

ARPE-19 Cells.

Cells at 80–85% confluence (72-h growth in DMEM/F12 + 10% FBS at 37°C) were serum-starved for 8 h before exposure to POS or microspheres. The serum-starved cells were treated with TNF-α (Sigma–Aldrich, St. Louis, MO) (10 ng/ml) and H₂O₂ to induce oxidative stress and, as a function of time, simultaneously with oxidative stress for 0, 1, 3, 6, and 16 h before harvesting for lipidomic or cell death analysis. Cells were analyzed after 16 h of oxidative stress to detect Hoechst-positive apoptotic cells (26).

ARPE-19 cells were incubated with bovine POS (10 million per well), or 2-μm diameter FITC-labeled polystyrene microspheres (Molecular Probes, Eugene OR) (10 million per well), for 16 h, after which the cultures were washed and methanol-fixed. POS were then labeled with rabbit anti-rhodopsin (1:50; Santa Cruz Biotechnology, Santa Cruz, CA) and visualized with Alexa Fluor 488 donkey anti-rabbit IgG (1:200; Molecular Probes). ARPE-19 cell cultures were viewed by confocal microscopy (LSM 510 Meta system; Zeiss, Thornwood, NY) with an LD-Acroplan ×40/0.6 corr Ph2 objective. A stack of 50 optical sections was collected from the base of the culture well toward the top of the plate to include the entire cell layer. The Zeiss LSM 5 (version 3.2) microscope controller and the LSM Image Browser (version 4.0.0.157; Zeiss) software were used, and final images were assembled with Photoshop 7.0 (Adobe, Mountain View, CA) on a Windows (Microsoft, Redmond, WA) platform.

Hoechst Staining.

ARPE-19 cells were fixed with methanol for 15 min at room temperature, washed with PBS at room temperature, and incubated with 2 μM Hoechst reagent dissolved in Lock's solution (Promega, Madison, WI) at room temperature for 15 min before imaging. Cells were washed once with PBS and photographed with a DIAPHOT 200 microscope (Nikon, Melville, NY) with fluorescence optics. Images were recorded by a color-chilled 3CCD camera (Hamamatsu, Bridgewater, NJ) and Photoshop software (Adobe).

Lipidomic Analysis.

Media and ARPE-19 cells were collected separately, and 5 μl of internal standard (PGD2-d4, 0.01 μg/μl) was added to each sample immediately. The cells were extracted with 1 ml of methanol. Lipid extracts removed from plates and collected in test tubes were added to chloroform to yield a ratio of 2:1 chloroform/methanol. Medium was spun-down to separate cell debris, then 1 ml was collected in 9 ml of cold chloroform/methanol (1:1). Protein precipitates were then separated by centrifugation at 1,000 × g (30 min, 4°C). Lipids were extracted from cells and medium (26). Eluates were concentrated on an N₂ stream evaporator and resuspended in 100 μl of methanol before MS analysis. Samples were loaded to a liquid chromatograph–tandem mass spectrometer (LC-TSQ Quantum; Thermo Scientific Co., Waltham, MA) installed with a Pursuit 5-μm C18 column (100 mm × 2.1 mm; Thermo Scientific Co.), and eluted in a linear gradient [100% solution A (40:60:0.01 methanol/water/acetic acid, pH 3.5) to 100% solution B (99.99:0.01 methanol/acetic acid)] at a flow rate of 300 μl/min for 45 min. LC effluents were diverted to an electrospray-ionization probe on a TSQ Quantum triple quadrupole mass spectrometer. Lipid standards (Cayman Chemical Company, Ann Arbor, MI) were used for tuning and optimization and to create calibration curves. The instrument was set on full-scan mode to detect parent ions and selected-reaction mode for quantitative analysis, to detect product ions, simultaneously. The selected parent/product ions (m/z) and collision energy (v) obtained by running on negative ion detection mode were: 359/153/22 for NPD1, 364/153/22 for d5-NPD1, 327/283/16 for DHA, 332/288/16 for d5-DHA, 375/193/25 for RvD3, 319/219/18 for 15(S)-HETE, 319/179/20 for 12(S)-HETE, 351/217/26 for lipoxin A4, 355/275/22 for PGD2-d4 (used as internal standard).

Data Analysis.

All data are expressed as means ± SEM of at least two independent experiments (n = indicates the no. of individual samples). Statistical comparisons were performed with Student's t test. Asterisks indicate P < 0.05, which was considered significant for all comparison; when it was not indicated, nonstatistical returns were obtained.

Abbreviations

DHA

docosahexaenoic acid

HETE

hydroxyeicosatetraenoic acid

LC

liquid chromatography

NPD1

neuroprotectin D1

POS

photoreceptor outer segments

RPE

retinal pigment epithelial.