INTRODUCTION

The retina is a part of the central nervous system directly exposed to external environmental stimuli. The outer segments of photoreceptors contain rhodopsin and other molecules that capture light signals and transmit this information through the retinal synaptic circuitry to the brain. The outer segments are damaged by oxidative stress and need to undergo a renewal process during which the tip is shed daily and then phagocytized by the retinal pigment epithelial (RPE)¹ cells. Retinol and docosahexaenoic acid, essential for visual cell function, are recycled to the inner segment by the RPE cells (1, 2, 3, 4, 5, 6, 7, 8), which also regulate the intake of nutrients, removal of waste, and maintenance of photoreceptor cells in a daily, rhythmic manner, which is light-regulated in amphibians and circadian in mammals.

The ability of RPE to phagocytize the shed discs of outer segments is essential for photoreceptor integrity and function. If this ability is compromised, as in mutant RCS (Royal College of Surgeons) rats, irreversible degeneration of the photoreceptor cells takes place (9).

RPE cells are the most active phagocytes in the body. In the rat, an RPE cell ingests and degrades 25,000-30,000 membrane outer segment discs each day. Virtually all are degraded efficiently within a few hours after ingestion. In addition, RPE cells sort, process, and retrieve compounds essential for visual cell function, such as vitamin A (1, 2, 3, 4, 5) and docosahexaenoic acid (6, 7, 8). The RPE cells then prepare for the next day's cycle of phagocytosis. To perform this task, RPE cells must have tightly regulated intracellular signaling pathways that couple phagocytosis with expression of genes encoding the rod outer segment (ROS)-degradative enzymes, transport proteins, etc. There is no information available on gene expression in RPE cells early during ROS phagocytosis.

Even though RPE cells normally phagocytize only the outer segments of photoreceptors, they can also nonspecifically phagocytize particles such as latex beads (10). Phagocytosis of ROS is thought to be mediated by a specific receptor in RPE cells, although identification of this receptor remains elusive, as do the associated signal transduction pathways (11, 12).

The products of early response genes (ERGs) or immediate-early genes are engaged in receptor-mediated transduction events by which extracellular stimuli modulate gene expression. ERGs are a class of genes whose expression is low or undetectable in quiescent cells but whose transcription is activated rapidly after extracellular stimulation, independent of de novo protein synthesis. Several ERGs encode transcription factor proteins that modulate the subsequent opening of gene cascades (13). c-fos proto-oncogene encodes a component of the AP-1 transcription factor (14). Among the external stimuli that turn the expression of c-fos on in mouse peritoneal macrophages is Fc- and C3b-mediated phagocytosis (15). Other ERGs include zif-268, which is a zinc finger-containing transcription factor (also known as tis-8, egr-1, krox24, d2, NGF1-A) (16, 17); tis-1, a transcription factor from orphan steroid/thyroid hormone superfamily (NGF1-B, nur77, N10) (18, 19, 20); AP-2 transcription factor (21, 22); and cox-2 (tis-10, pgs-2, ptgs-2, pghs-B), an ERG that codes for the inducible form of prostaglandin synthase (23, 24).

We now report that phagocytosis of ROS by RPE cells selectively activates certain early response gene transcription factors. These genes are not activated by nonspecific phagocytosis or in dystrophic RPE cells. Part of these results have appeared in abstract form (25).

EXPERIMENTAL PROCEDURES

Long Evans female rats with timed litters (from Harlan Sprague-Dawley) were kept in plastic cages at the LSU Medical Center animal care facility until the pups reached the age of 10 days. The retinal dystrophic RCS-p+ strain and congenic control RCS-rdy+p+ strain of rats (26) were kindly provided by Dr. Matthew M. LaVail (University of California, San Francisco). Breeding pairs of these rats were kept in plastic cages in the animal care facility, and births were monitored daily. All rats were kept on the 12-h dark/12-h light cycle.

RPE were isolated from 10-day-old rats (27). Cells were seeded into six-well plastic tissue culture plates at a density of approximately 10⁵ cells/well or into 60-mm plastic tissue culture dishes (2.5 × 10⁵ cells/dish) and cultured in minimum essential medium (MEM) (Life Technologies, Inc.) supplemented with 10% fetal bovine serum (Life Technologies, Inc.), 100 units/ml of both penicillin and streptomycin, and 2 mM glutamine. The cells were used when they reached confluence, usually after 7-10 days of growth at 37 °C in 95% air, 5% CO₂.

ROS were isolated from adult Long Evans rats by centrifugation in a continuous sucrose gradient (28) and were resuspended in MEM without supplements at a final concentration of 2 × 10⁸/ml. The ROS suspension (100 µl) was layered on top of the RPE cell monolayer, and the medium was mixed with gentle rocking. For controls, an equal volume of vehicle (MEM) was added to RPE cells. The cells were incubated in a humidified incubator at 37 °C in 5% CO₂, 95% air for the time indicated in each experiment before RNA was isolated or nuclear protein extract (NPXT) was prepared. As a morphological control, some cell cultures from the same experiments were double immunostained using anti-ROS antiserum as described before (28), and phagosomes were visualized with a Odyssey XL confocal laser scanning microscope (Noran Instruments, Middleton, WI). To study the specificity of gene expression for phagocytosis of ROS, Polybead polystyrene latex spheres (1.053-µm outer diameter, Polysciences, Warrington, PA) or fluorescent yellow-green latex microspheres (1.0-µm outer diameter, Molecular Probes, Eugene, OR), were used in the place of ROS (2 × 10⁸/ml in 100 µl of MEM layered on top of the RPE cell monolayer). In representative cell cultures, phagocytized fluorescent beads were observed directly with the confocal scanning microscope to ensure, as with ROS, that phagocytosis of latex beads took place.

After incubation with ROS or latex beads, the plates were put on ice and washed three times with ice-cold phosphate-buffered saline. The cells were kept on ice until guanidine-thiocyanate extraction buffer was added to isolate total cell RNA (29). Gel electrophoresis of RNA (5 µg/lane) was performed under denaturing conditions on a 1.2% agarose-formaldehyde gel. RNA was transferred to Hybond-N nylon membrane (Amersham Corp.) followed by hybridization at 42 °C with [³²P]DNA probes for zif-268 (16, 30), c-fos, tis-1 (18, 31), or cox-2. (20). [³²P]DNA probe for the glyceraldehyde-3-phosphate dehydrogenase gene (gapdh) (32) was used in each hybridization as a reference. Autoradiography and quantification of Northern blots were performed using GS-250 Molecular Imager (Bio-Rad) or Instant Imager (Packard, Meriden, CT).

NPXTs and cytoplasmic protein extracts (CPXTs) were prepared from RPE cells (33, 34, 35). DNA sense and mutant oligonucleotides, 21-30 base pairs in length, containing the consensus and mutant binding sequences for AP-1 (consensus, 5-CGCTTGATGACTCAGCCGGAA-3; mutant, 5-CGCTTGATGACTGCCGGAA-3), AP-2 (consensus, 5-GATCGAACTGACCGCCCGCGGCCCGT-3; mutant, 5-GATCGAACTGACCGCGCGGCCCGT-3), zif-268 (consensus, 5-GGATCCAGCGGGGGCGAGCGGGGGCGA-3; mutant, 5-GGATCCAGCGGGCGAGCGGGCGA-3), and tis-1 (consensus, 5-GAGTTTTAAAAGGTCATGCTCAATTTGGAT3) (36) were obtained from either Santa Cruz Biotechnology (Santa Cruz, CA) or Promega Corporation (Madison, WI), or were synthesized at the LSU Medical Center Core Facility. NPXT and CPXT concentrations were determined by the method of Bradford (37), using whole histones as standards. To calibrate the EMSA system further, purified transcription factors (AP-1) from Promega were used. Protein-DNA complexes were formed in 10-15 µl at 0-4 °C by the tandem addition of 0-10 µg of each NPXT, CPXT, or purified transcription factor into an assembly buffer consisting of 50 mM KCl, 50 mM Tris-HCl, pH 7.5, 6 mM magnesium acetate, 1 mM EDTA, 1 mM -mercaptoethanol, 5% glycerol, 0.01% Nonidet P-40, and 0.1-1 µg/µl poly(dI-dC), followed by the addition of 10,000-50,000 cpm of end-radiolabeled sense or mutant oligonucleotide DNA. Samples were incubated for 30 min in 500-µl GeneAmp reaction tubes (Perkin Elmer) at 0-4 °C. Protein-DNA complexes were analyzed using two gel systems to detect both weak and strong protein-DNA interactions: a high porosity composite gel type consisting of 4.5% polyacrylamide (acrylamide:bisacrylamide, 80:1), 2.5% glycerol using either 0.25 × TBE (1 × TBE = 90 mM Tris, pH 8.4, 90 mM boric acid, 1 mM EDTA) or TGEM buffer (25 mM Tris, pH 8.3, 190 mM glycine, 1 mM EDTA, 5 mM magnesium acetate) for both gel matrix and running buffer or a low porosity gel type consisting of 5% polyacrylamide (acrylamide:bisacrylamide 40:1) using 0.5 TBE as running buffer. Both gel types were run at 23 °C on a Bio-Rad Mini Protean II vertical electrophoresis apparatus. After electrophoresis, gels were overlaid with Saran Wrap (Dow), dried onto Whatman no. 1 filter paper, and exposed to Kodak Biomax autoradiographic film for 0.5-3 h, or they were scanned using either a GS-250 Molecular Imager or Instant Imager. Band shift intensities were quantitated using the analytical package provided with each instrument.

RESULTS AND DISCUSSION

Primary cultures of rat RPE cells readily phagocytize freshly isolated rat ROS. This was monitored by direct microscopic observation of immunostained phagosomes inside RPE cells (data not shown). Kinetic studies using ROS, compared with nonspecific latex particles, strongly suggest that RPE cells, in this in vitro system, phagocytize ROS through a specific receptor-mediated mechanism (11, 38). Resting RPE cells in culture revealed low levels of zif-268 mRNA; c-fos, tis-1, and cox-2 mRNA were undetectable. Because ERGs are extremely sensitive to experimental manipulations, ROS were added in small volumes (100 µl) of serum-free medium because larger volumes (e.g. 1 ml) or the addition of serum enhanced ERG expression. To control for other factors that contributed to nonspecific induction of ERGs (e.g. exposure to room air, shaking, low temperature, and/or high pH due to manipulations carried out outside the CO₂ tissue culture incubator), separate six-well plates were used for each time point, three of which received ROS, and the other three wells served as controls.

c-fos, zif-268, and tis-1 mRNA abundance in normal Long Evans rat RPE cells was rapidly and transiently increased during ROS phagocytosis (Fig. 1). After 30 min of incubation with ROS, the c-fos, zif-268, and tis-1 mRNA expression was greatly enhanced compared with controls. Maximum mRNA abundance was observed at 30 min for c-fos, at 45 min for zif-268, and at 60 min for tis-1. The amplitude of responses differed: c-fos and tis-1 mRNA accumulation ranged from 6- to 20-fold, whereas ROS induction of zif-268 was smaller, ranging from 2- to 3-fold, which is similar to the relative induction of these genes in cultured rat astrocytes by mitogens (39). By 2 h, mRNA levels for all three genes decreased but still remained higher than controls. RPE cells from RCS-rdy⁺p⁺ rats, which are a normal congenic control for the dystrophic RCS-rdy/rdy⁺p⁺ rats, showed similar gene induction during phagocytosis of ROS (data not shown).

cox-2 is induced in cells in response to diverse stimuli, such as tetradecanoylphorbolacetate, platelet-derived growth factor, lipopolysaccharide, or platelet-activating factor (40). However, no expression of cox-2 was observed in RPE cells in response to ROS phagocytosis or to platelet-activating factor (100 µM), although both 50 nM tetradecanoylphorbolacetate and 10 ng/ml platelet-derived growth factor did induce this gene (Fig. 2a).

EMSA using zif-268, AP-1, and TIS-1 consensus oligonucleotides and RPE nuclear extracts showed the appearance of shifted bands during ROS phagocytosis (Figs. 1 and 2, b and c). Most of the zif-268-shifted bands were observed in nuclear extracts compared with extracts prepared with RPE cytosol. There was some enhancement in the shifted band in RPE cytosol 45 min after the addition of ROS. However, this was much lower than the corresponding change observed in nuclear extracts. Using zif-268 mutant oligonucleotides, no shifted bands were observed under these experimental conditions (Fig. 2c), suggesting that the shifted bands specifically recognized the zif-268 consensus sequence. The maximum intensity of gel-shifted bands occurred at 30 min for AP-1, 45 min for zif-268, and 60 min for TIS-1. This suggests that c-fos, zif-268, and tis-1 mRNAs are translated very rapidly into transcription factor proteins. The short half-life of c-fos mRNA (t_1/2 ~ 15 min) suggests that, indeed, the cellular mechanism for translation of c-fos mRNA into c-FOS gene product was very rapid (41).

Detailed analysis of the kinetics of induction of DNA binding activity of transcription factor proteins compared with their respective mRNA levels revealed a seeming disparity between these two patterns of expression. For example, a significant increase in zif-268 protein appears at 15 min after incubation with ROS, well before there is a demonstrable increase in zif-268 protein levels (30 min). Similarly, increased mRNA expression for c-fos and tis-1 is evident after 2 h of ROS phagocytosis, whereas increased AP-1 DNA binding activity can only be seen at 30 min and TIS-1 binding activity at 60 and 75 min. This difference in the kinetics of expression suggests the involvement of post-transcriptional and translational control in the expression of these genes. Since the constitutive level of zif-268 mRNA expression is relatively high in REP cells, regulation of zif-268 protein expression may involve the existence of a translational block. This type of zif-268 modulation has been shown in other cells (42). The initial induction of zif-268 by ROS phagocytosis may relieve the translational constraint by specific phagocytosis stimuli and then transcriptionally activate zif-268 mRNA synthesis. The possible involvement of post-transcriptional regulation is even more evident in the case of AP-1 transcription factor. AP-1 is formed as either a homodimer or as a heterodimer by different members of the fos and jun proto-oncogene families. It is well established that AP-1 is not only regulated by transcription of fos and jun genes, but also at the translational level (43) and by post-translational phosphorylation of Jun and Fos proteins (14). The very narrow peak of DNA binding activity of AP-1 protein coexisting with longer induction of c-fos mRNA expression suggests this post-transcriptional regulation in RPE cells.

The transcription factor AP-2 was also rapidly and transiently expressed in RPE nuclear extracts during ROS phagocytosis with a prominent peak at 15-30 min after incubation with ROS (Fig. 3). However, using the NFB consensus oligonucleotide (5-AGTTGAGGGGACTTTCCCAGGC-3 in EMSA, no evidence was found of an enhanced expression in the RPE nuclear extract during phagocytosis (Fig. 3). Tate et al. (45) found activation of NFB DNA binding in cultured human RPE cells after prolonged (4-h) incubation with bovine ROS. This response was not specific for ROS as the same level of induction was found during phagocytosis of latex beads. The foregoing results suggest that initial NFB induction is not an early event in ROS phagocytosis.

EMSA of RPE nuclear extracts with AP-2 and NFB consensus oligonucleotides after 0-60 min of incubation of Long Evans RPE cells with ROS.

Rat RPE cells in primary culture display specificity toward phagocytosis of ROS compared with that of nonspecific particles (11). However, these cells can also ingest large quantities of latex beads (10, 46) presumably through a nonspecific mechanism (44). Therefore, to see if the induction of ERGs by ROS phagocytosis also occurred with latex bead particles, RPE cultures were fed with either ROS or the same number of latex particles that had a diameter similar to ROS. Even though RPE cells phagocytized large quantities of latex beads (as confirmed by visualization of phagocytized fluorescent latex beads in a confocal scanning microscope; data not shown), only ROS-mediated phagocytosis stimulated expression of c-fos, zif-268, and tis-1 mRNAs (Figs. 2d and 4), as well as AP-1, AP-2, and zif-268 nuclear transcription factor proteins (Fig. 4) in RPE cells.

RPE cells are the primary target of a genetic defect in dystrophic RCS rats, which makes them unable to phagocytize ROS effectively. Although mutant RPE cells in culture bind ROS in the same manner as normal RPE cells, only a small fraction of bound ROS will subsequently be ingested (10, 28). In RCS rat RPE cells, the genetic defect is not yet known but may involve a putative receptor for RPE phagocytosis and/or the signal transduction mechanism activated by this receptor. The hypothesis that ROS-induced activation of ERGs in dystrophic RPE cells is impaired was therefore tested. Our results show that RPE cells from dystrophic RCS rats did not display enhanced expression of c-fos, zif-268, or tis-1 mRNAs with the ROS challenge, nor was there induction of AP-1, AP-2, or zif-268 nuclear transcription factor proteins (Fig. 4). The inability of RPE cells from dystrophic RCS rats to respond to phagocytotic stimuli at the ERG level supports the idea that the defect may involve a signal transduction event associated with a putative RPE membrane receptor for ROS phagocytosis.

In conclusion, selective induction of certain immediate-early transcription factors took place at the onset of phagocytosis in the RPE. The expression of the transcription factors followed specific patterns, starting with AP-2 (15 min after the addition of ROS), followed by AP-1 (30 min), zif-268 (45 min), and TIS-1 (60 min). Gene induction was specific for ROS phagocytosis because (a) it occurred only while RPE cells phagocytized outer segments and not when they ingested latex beads; (b) some transcription factors were induced while others were unaffected; and (c) RPE cells from RCS dystrophic rats did not display activation of immediate-early transcription factors upon the addition of ROS. This selective activation of ERGs in RPE cells supports the notion that phagocytosis of ROS by RPE cells is indeed a receptor-mediated process, different from nonspecific phagocytosis of particles such as latex beads. Also, for the first time, this study provides a qualitative assay (ERG expression) to distinguish between specific and nonspecific phagocytosis by RPE cells.

During phagocytosis of ROS, a receptor-mediated event may trigger a signal in RPE cells which, in turn, activates certain ERGs in a specific pattern. Gene cascades may then be opened by activating the subsequent expression of genes that encode proteins necessary for the completion of phagocytosis, preparation of the cell for the next cycle, and for the fulfillment of other functions. Furthermore, signals engaged on this receptor-gene activation cascade may be critical for the survival of photoreceptors, particularly in the aging retina, and may be relevant to the understanding of events crucial to senile macular degeneration.