Delayed Loss of Cone and Remaining Rod Photoreceptor Cells due to Impairment of Choroidal Circulation after Acute Light Exposure in Rats

Abstract

purpose. To examine the long-term effects of acute photooxidative stress in the retina, retinal pigment epithelium (RPE), and choroid.

methods. Albino rats injected with either the protective antioxidant phenyl-N-tert-butylnitrone (PBN) or saline 30 minutes before exposure to 5 klx white fluorescent light for 6 hours were kept for up to 3 months in 5 lux cyclic light. Electroretinograms were recorded, and the outer nuclear layer (ONL) and the choroidal thickness and area were measured after hematoxylin-eosin (H&E) staining. The expression of rod, cone, and RPE cell markers was detected by Western blotting, and apoptosis was analyzed by TUNEL staining. Oxidative stress was analyzed by immunohistochemistry against 4-hydroxynonenal (4-HNE)–modified proteins. Retinal and choroidal ultrastructures were observed by transmission electron microscopy (TEM). Choroidal circulation was analyzed by in vivo staining of the choroidal layer by trypan blue.

results. In the saline-injected animals, TUNEL- and 4-HNE–labeling in the ONL, RPE, and choroid were higher 24 hours and 7 days after light exposure, and ERG amplitude, ONL and choroidal thickness and area, and rhodopsin and RPE65 expression were lower 7 or more days after light exposure than in phenyl-N-tert-butylnitrone (PBN)–injected animals. In the saline-injected animals, the expression of mid-wavelength opsin and the presence of cone cells in the ONL and the choroidal circulation were preserved for 7 days after light exposure but started to decrease by 1 month and continued to decrease for 3 months after light exposure. An increase in TUNEL-positive cells was observed in the ONL at the inferior peripheral retina, just behind the iris, by 3 months after light exposure. Delayed loss of cone cells, remaining rod cells, and choroidal circulation were counteracted by PBN treatment.

conclusions. Although cone cells are resistant to cell damage induced by acute photooxidative stress, progressive loss of cone cells continued for up to 3 months after light exposure. Impaired choroidal circulation is likely to be involved in the mechanism of delayed photoreceptor cell death after light exposure. Preserving choroidal circulation may provide a novel target for preserving the cone and the remaining rod cells in patients with retinal degeneration such as retinitis pigmentosa.

Epidemiologic studies have suggested that excessive light may enhance the progression and severity of age-related macular degeneration (AMD) and some forms of retinitis pigmentosa. ^{1 2} The light from the operating microscope used in ophthalmic practice can cause photic maculopathy. ³ Acute light exposure causes photoreceptor and retinal pigment epithelial (RPE) cell damage, ⁴ and apoptosis is the main pathway of light-induced cell death. ⁵ Exposure of the retina to intense light causes lipid peroxidation of retinal tissues, ^{6 7 8} and lipid peroxidation is propagated by free radicals, especially lipid radicals. ⁹ We recently reported that an increase in the retinal proteins modified by reactive aldehydes such as 4-hydroxynonenal (4-HNE), which is an end product of nonenzymatic oxidation of n-6 polyunsaturated fatty acids, ¹⁰ is a molecular event that precedes retinal degeneration caused by acute light exposure. ¹¹ Thus, the detection of 4-HNE protein modifications is thought to be a biomarker for lipid peroxidation induced by light exposure. Retinal damage caused by light exposure can be reduced by various types of antioxidants, such as ascorbate, ¹² dimethyl thiourea, ⁷ thioredoxin ^{13 14} and phenyl-N-tert-butylnitrone (PBN). ^{15 16} Accordingly, oxidative stress is likely to be involved in the pathogenesis of light-induced retinal damage. 

In the experimental settings of previous studies, acute light–induced retinal damage was examined 12 to 96 hours after light exposure for assessing apoptosis and 96 hours to 2 weeks after light exposure for assessing outer nuclear layer (ONL) thickness or number of ONL nuclei. ^{7 16 17} Thus, the profile of retinal damage after acute light exposure is known for the short term. The differential effects of gene mutations on rod and cone cells in animal models of hereditary retinal degeneration have been shown. ^{18 19} In addition, the delayed loss of cone cells in the animal models of rod cell degeneration has been reported. ^{20 21} Although it has been clearly shown that acute light exposure causes cell death in rod and RPE cells, ^{5 17} the susceptibility to light exposure in cone cells remains unclear. Recently, the induction of oxidative stress by acute light exposure was reported in choroidal endothelial cells in vivo, ²² but light-induced damage in the choroidal layer remains to be tested. 

We examined the long-term (up to 3 months) effects of photooxidative stress against retinal cells in albino rats exposed to acute damaging light. Profiles of cell or tissue damage in the retina, RPE, and choroid after light exposure were longitudinally assessed by electroretinography, histology, expression analyses for rod, cone, and RPE cell-specific markers, and a newly established method for in vivo staining of choroidal layer. 

Materials and Methods

Antibodies

The rabbit polyclonal anti–RPE65 antibody was kindly provided by Jian-Xing Ma (Departments of Cell Biology and Medicine, University of Oklahoma Health Sciences Center, Oklahoma City, OK). ²³ Other antibodies were purchased, as follows: mouse monoclonal anti–rhodopsin antibody (MA 1–722; Affinity BioReagents, Golden, CO); rabbit polyclonal anti–mid-wavelength cone opsin (m-opsin) antibody and mouse monoclonal anti–glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody (sc-30022 and sc-32233, respectively; Santa Cruz Biotechnology, Santa Cruz, CA); peroxidase-linked anti–mouse IgG and anti–rabbit IgG antibodies (Amersham Biosciences, Buckinghamshire, UK); and mouse monoclonal anti–4-HNE-modified protein antibody (anti–4-HNE antibody; NOF Corporation, Tokyo, Japan). ²⁴ Normal mouse IgG was also purchased (Dako, Carpinteria, CA). 

Animal Care

All procedures were carried out in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and the University of Oklahoma Health Sciences Center (OUHSC) Guidelines for Animals in Research. All protocols were reviewed and approved by the Institutional Animal Care and Use Committees of the OUHSC and the Dean A. McGee Eye Institute. Sprague–Dawley (Harlan Sera-Laboratory; Indianapolis, IN) rats were born and raised in our vivarium and were kept under dim cyclic light (5 lux, 12 hours on/off, 7 am-7 pm, Central Time) before experimentation. In this study, 93 rats were used for experiments. 

Light Exposure

Rats (5–6 weeks of age) were injected intraperitoneally with PBN (Sigma, St. Louis, MO; 50 mg/kg of 20 mg/mL dissolved in saline), a potent free-radical trapping agent, or an equivalent volume of saline 30 minutes before damaging light exposure, as described previously. ^{15 16 25} All light exposures began at 8:00 am. Briefly, unanesthetized rats were exposed to 5 klx diffuse, cool, white fluorescent light for 6 hours in clear plastic cages with wire tops. Each cage contained one rat. Drinking water was supplied by a bottle attached to the side of the cage so that there was no obstruction between the light and the animal. The rats were returned to the dim cyclic light environment for 1 day, 7 days, 1 month, and 3 months after acute light exposure. The animals were humanely killed, and the eyes were enucleated (light animals). Rats injected with saline or PBN and kept in dim cyclic light without damaging light exposure were age-matched to the light-exposed animals and used as controls (dim animals). 

Electroretinography

Flash electroretinograms (ERGs) were recorded 7 days, 1 month, and 3 months after light exposure using a Ganzfeld-type ERG recording system (UTAS-E3000; LKG Technologies Inc., Gaithersburg, MD), as reported previously. ¹¹ All animals were dark adapted for 16 hours before they were. After anesthesia was induced by intramuscular injection of a mixture of ketamine (120 mg/kg) and xylazine (6 mg/kg), the pupils were dilated with 0.5% tropicamide and 0.5% phenylephrine hydrochloride eyedrops (Santen Pharmaceutical, Osaka, Japan). Gold electrodes were placed on both eyes. An identical reference electrode was placed in the mouth, and a ground electrode was placed on the tail. A single flashlight (25 dB for 10 msec) from a halogen source was used as the light stimulus. The a-wave amplitude was measured as the difference in voltage between baseline just before the flash and the peak of a-wave, and the b-wave was measured as difference in voltage between peaks of a- and b-waves. The a- and b-wave amplitudes obtained from the right and left eyes were averaged in each animal. 

Preparation of Whole Retinal Samples for Western Blotting

Whole retinal samples were prepared as previously described. ¹¹ Briefly, after deep anesthesia was induced by the intramuscular injection of a mixture of ketamine (120 mg/kg) and xylazine (6 mg/kg), rats were perfused through the left cardiac ventricle with ice-cold PBS (pH 7.4) to wash out the blood, and the eyes were then enucleated. The cornea and the lens were removed, and the retina was separated from each eyecup. After cardiac perfusion with ice-cold PBS, the adhesion between the photoreceptor cell layers and retinal pigment epithelial (RPE) cell layers was weak, so they were easily separated. After the retina was removed, the remaining eyecups were analyzed as an RPE cell fraction. Accordingly, this fraction also contained choroid and sclera. Samples were sonicated in radioimmunoprecipitation buffer (Upstate Biotechnology, Lake Placid, NY) containing a protease inhibitor cocktail (Upstate Biotechnology) and were centrifuged at 10000g for 15 minutes at 4°C to collect the supernatants. Two retinas were pooled for each sample. 

Western Blotting

Western blotting was performed as previously described, with slight modifications. ^{11 17} After protein concentrations were determined by the DC protein assay kit (BioRad, Hercules, CA), equal aliquots (20 μg) of protein samples from the retina or the RPE fraction were applied to 4% to 15% gradient sodium dodecylsulfate polyacrylamide gel (BioRad) and were electrophoretically separated. Gels were electrophoretically transferred to polyvinylidene difluoride membranes (Millipore, Bedford, MA) and blocked with 10% nonfat dry milk for 1 hour at room temperature. Membranes containing retinal samples were incubated with anti–rhodopsin (1:2000), anti–m-opsin (1:100), or anti–GAPDH (1:2000, loading control) antibodies for 1 hour at room temperature. Membranes containing samples of the RPE fraction were incubated with the anti–RPE65 (1:500) or anti–GAPDH (1:2000) antibodies for 1 hour at room temperature. Membranes were then incubated with the peroxidase-linked anti–mouse IgG antibody (1:5000; for rhodopsin and GAPDH) or with the peroxidase-linked anti–rabbit IgG antibody (1:5000; for m-opsin and RPE65) for 1 hour at room temperature. Chemiluminescence signals were developed (SuperSignal West Dura Extended Duration Substrate; Pierce, Rockford, IL) and were detected with a digital imaging system (IS4000R; Kodak, New Haven, CT). Care was taken to ensure that the intensities of detected bands were within the linear range of the camera and that no pixels were saturated. Intensities of protein bands were determined with the use of Image J 1.32j software (developed by Wayne Rasband, National Institutes of Health, Bethesda, MD, and available at http://rsb.info.nih.gov/ij/). 

Preparation of Retinal Tissue Sections

The preparation of retinal tissue sections was performed as described previously. ^{25 26 27} After CO₂-induced unconsciousness and cervical dislocation, eyes were enucleated from the dim and light-exposed animals. For paraffin embedding, enucleated eyes were fixed with 4% paraformaldehyde containing 20% isopropanol, 2% trichloroacetic acid, and 2% zinc chloride for 24 hours at room temperature and embedded in paraffin. Sections (4-μm thick) containing the whole retina, including the optic disc, were cut along the vertical meridian of the eyeball. For epoxy-resin embedding, enucleated eyes were fixed with 2% paraformaldehyde and 2% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) for 48 hours at 4°C. The cornea and lens were removed from the eyeballs, and the eyecups were radially cut into six pieces concentric to the optic nerve head (ONH). Pieces of eyecups from the superior retina were postfixed with 2% osmium tetroxide for 2 hours, dehydrated in a graded series of ethanol, and embedded in epoxy resin. For light microscopy, semithin sections (approximately 900-nm thick) were cut and stained with toluidine blue. For electron microscopy, ultrathin sections (approximately 90-nm thick) were cut and stained with uranyl acetate and lead citrate. 

Measurement of the Outer Nuclear Layer or Choroidal Layer Thickness and Area

Paraffin-embedded retinal sections were stained with hematoxylin and eosin (H&E). For each section, digitized images of the entire retina were captured with a digital imaging system (Eclipse E800; Nikon, Tokyo, Japan) at 4× magnification with 1300 × 1030 pixels. To cover the entire retina, five images were obtained from each section. The ONL and the choroidal thickness were measured at 0.5, 1.0, 1.5, 2.0, 2.5, and 3.0 mm superior and inferior to the ONH and at the periphery, 100 μm from the inferior and superior edges of the retina, with Image J 1.32j software. The ONL and the choroidal areas were calculated by integrating the area under the thickness histogram from 3 mm superior and 3 mm inferior to the ONH (Kaleida Graph software version 3.6; Synergy Software, Reading, PA). ²⁷ Values obtained from the right and the left eyes were averaged for each rat. 

TUNEL

TUNEL was performed on paraffin-embedded sections (ApopTag Peroxidase In Situ Apoptosis Detection Kit; Chemicon, Temecula, CA) in accordance with the manufacturer’s instructions. The chromogen 3′,3′-diaminobenzidine (Dako) was used. Terminal deoxynucleotidyl transferase (TdT) enzyme was omitted in the negative controls. 

Immunohistochemistry for 4-HNE–Modified Proteins

Methods of immunohistochemistry were described previously. ²⁵ Paraffin-embedded sections were deparaffinized, and endogenous peroxidase activity was inactivated with 3% H₂O₂ for 5 minutes. After blocking with serum-free blocking reagent (Dako) for 30 minutes at room temperature, the sections were incubated with the anti–4-HNE (1:200) antibody diluted with antibody diluent (Dako) for 2 hours at 37°C and then were incubated with the peroxidase-linked anti–mouse IgG polymer (EnVision+ System; Dako) for 1 hour at 37°C. Signals were developed with 3′,3′-diaminobenzidine (Dako) as chromogen for 2 minutes Sections were incubated with normal mouse IgG instead of anti–4-HNE antibody for the negative controls. 

Transmission Electron Microscopy

Ultrathin sections stained with uranyl acetate and lead citrate were photographed using an transmission electron microscope (EM-002B; Topcon, Tokyo, Japan). ²⁶  

Choroidal Layer Staining In Vivo

After deep anesthesia was induced by the intramuscular injection of a mixture of ketamine (120 mg/kg) and xylazine (6 mg/kg), rats were slowly injected with 1 mL of 0.4% trypan blue solution (Invitrogen, Grand Island, NY) into the left ventricle of the heart, and then the right atrium was cut to drain the blood. The eyes were enucleated and fixed with 4% paraformaldehyde containing 20% isopropanol, 2% trichloroacetic acid, and 2% zinc chloride for 24 hours at room temperature. After fixation, the cornea and lens were removed with the assistance of a dissecting microscope, and the eyecup was radially cut concentric to the ONH (Fig. 1A) . After fixation, the retinal layer became opaque (Fig. 1A , retina), and the injected blue dye was only seen in the major retinal vessels on the surface of retina (Fig. 1A , arrowheads). After careful removal of the retinal layer, the blue-stained choroidal layer appeared (Fig. 1A , choroid). Under high magnification, the major choroidal vessels were easily recognized (Fig. 1B , white arrows). When the layer stained by blue dye was peeled, the transparent tissue indicative of sclera was seen (Fig. 1A , sclera), and in this field only remnants of choroidal tissue were seen in blue (Fig. 1A , arrows). To assess the distribution of blue dye among the retinal layers, the pieces of eyecups before (Figs. 1C 1D)and after (Figs. 1E 1F 1G)peeling of the retinal layer were embedded (OCT Tissue-Tek; Sakura Finetek, Torrance, CA), and frozen sections were cut (15-μm thick) and observed. In the retinal section from the piece of eyecup before the peeling of the retinal layer, blue dye was located in the choroidal layer (Fig. 1D) . After peeling of the retinal layer, some RPE cells attached to the surface of the choroidal layer (Fig. 1E , arrowheads) were free of blue dye (Fig. 1G , arrows). Accordingly, this method specifically stained the choroidal layer in vivo, leaving adjacent RPE and scleral layers unstained. 

Results

Lipid Peroxidation after Light Exposure

Because the retinal damage caused by light exposure is thought to be initiated by oxidative stress, ^{6 7 8} we initially examined lipid peroxidation by immunohistochemistry for 4-HNE protein modification. Weak staining was observed throughout the retinal layers, the RPE, and the choroid in the dim animals (Fig. 2A) . Twenty-four hours after light exposure, staining intensity was remarkably greater in the retinal layers, the RPE, and the choroidal layers in the saline-injected animals (Fig. 2B , arrowhead and arrows, respectively) compared with the dim animals. Staining of these layers was less remarkable in the PBN-injected animals (Fig. 2C) . Thus, free radical–mediated lipid peroxidation occurred in the retina, the RPE, and the choroid early after damaging light exposure. 

Retinal and Choroidal Damage after Light Exposure

Representative ERG tracings from animals injected with saline or PBN are shown in Figure 3A . ERG a- and b-wave amplitudes were significantly lower in the saline-injected animals than in the PBN-injected animals for up to 3 months after light exposure (Figs. 3B 3C , light). There was no difference in the a- or b-wave amplitudes in the saline- or PBN-injected dim animals at 3 months (Figs. 3B 3C , dim). In addition, the appearance of the ONL was identical between the saline- and the PBN-injected animals not exposed to acute light damage (Fig. 3B , dim). A remarkable loss of photoreceptor nuclei was observed in the saline-injected animals 7 days after light exposure, and the whole ONL disappeared by 3 months after light exposure (Fig. 4A , light and PBN (–)), whereas the ONL nuclei in the PBN-injected animals were well preserved for 3 months (Fig. 4A , light and PBN (+)). Red blood cells were consistently observed in the choroidal layer from the dim animals (Fig. 4A , dim) and the light-exposed, PBN-injected animals (Fig. 4A , light and PBN (+)). Red blood cells were still observed in the choroidal layer from the light-exposed, saline-injected animals 7 days and 1 month after light exposure; however, the vascular structure in the choroidal layer was replaced by fibrotic tissues, and the red blood cells disappeared by 3 months after the light exposure (Fig. 4A , light and PBN (–)). ONL and choroidal thickness (Figs. 4B 4C , respectively, light) and area (Figs. 4D 4E , respectively, light) were significantly smaller in the light-exposed, saline-injected animals than the light-exposed, PBN-injected animals for up to 3 months. There were no differences in ONL choroidal thickness (Figs. 4B 4C , respectively, dim) or area (Figs. 4D 4E , respectively, dim) in the saline- or PBN-injected dim animals at 3 months. ERG amplitudes (Figs. 3B 3C)and the ONL area (Fig. 4D)in the saline-injected animals tended to decrease from 1 to 3 months after light exposure. This was accompanied by a progressive decrease of the choroidal area for 3 months after light exposure (Fig. 4E) . 

Rod, Cone, and RPE Cell Damage after Light Exposure

To assess the types of cells that may be damaged by light exposure, the expression of rod and cone cell markers in retinal samples and the expression of an RPE cell marker in the RPE fraction were assessed in samples from the dim and the light-exposed animals (Fig. 5) . In the retinal samples, the level of rhodopsin decreased significantly 7 days after light exposure in the saline-injected animals compared with the dim animals and remained low level for 3 months (Figs. 5A 5C) . On the other hand, the level of m-opsin in the saline-injected animals started to decrease 1 month after light exposure and decreased further by 3 months (Figs. 5A 5D) . In the RPE fraction, the level of RPE 65 decreased significantly 7 days after light exposure in the saline-injected animals compared with the dim animals and remained low for 3 months (Figs. 5B 5E) . By 3 months after light exposure, the levels of all three markers analyzed were significantly higher in PBN–injected animals than in the saline-injected animals. Thus, the results indicate that susceptibility to cell damage induced by acute light exposure may be different among cell types and that a mechanism of cell death other than acute cell damage is likely to be involved in the delayed cell loss in cone cells after acute light exposure. 

TUNEL after Light Exposure

Apoptosis after light exposure was assessed by TUNEL staining (Fig. 6) . By 7 days after light exposure, the nuclei of the ONL, the RPE (arrowheads), and the vascular endothelial cells in the choroid (arrows) in the saline-injected animals were TUNEL positive (Fig. 6B) , whereas fewer cells stained positively in the PBN-injected animals (Fig. 6C) . Some nuclei of the vascular endothelial cells in the saline-injected animals showed positive staining by 1 month after light exposure (Fig. 6D , arrows). Inferior peripheral retinas just behind the iris in the saline-injected animals had only a few ONL nuclei that stained positively by 7 days after light exposure (Fig. 6E) , whereas a substantial number of ONL nuclei stained positively by 3 months after light exposure (Fig. 6F) . In the age-matched dim animals, few ONL nuclei stained positively by 3 months (Fig. 6G) . The results suggest the presence of sustained apoptosis in the choroidal vascular endothelium after light exposure and the presence of a delay in apoptosis of the remaining rod cells in the peripheral retina by mechanisms other than acute damage by light or aging. 

Histologic Analyses

Morphologic changes in the retina and the choroid were further analyzed by light and electron microscopy (Fig. 7) . One month after light exposure, a single cell layer typically remained in the ONL in the superior retina from the saline-injected animals. This single layer predominantly consisted of cells with polymorphic clumps of heterochromatin surrounded by a substantial amount of lightly stained euchromatin (Fig. 7A , arrows). Electron microscopy revealed that the ONL contained some cells with a central clump of heterochromatin and a small amount of peripheral euchromatin, cells with shrunken chromatin (Fig. 7B , arrowhead), and primarily cells with lobulated heterochromatin and large amounts of euchromatin (Fig. 7B , arrows), indicative of mature rod cells, apoptotic cells, and cone cell nuclei, respectively. ¹⁸ Thus, cone cells survived much later than most of the rod cells after light exposure. One month after light exposure, the ONL was relatively well preserved when red blood cells were still observed in the choroidal layer just beneath the RPE (Fig. 7C , left, black arrowheads), whereas the ONL was severely thinned when the choroidal layer collapsed, leaving only the large outer vessels of the choroid perfused (Fig. 7C , right, white arrowheads). Interestingly, rosette formation was frequently observed in the ONL at the border between the perfused and the collapsed choroid (Fig. 7C , arrow). By 3 months after light exposure, the RPE and the ONL had totally disappeared, but the inner nuclear layer (INL) was well preserved when the red blood cells were observed in the outer part of the choroidal layer (Fig. 7D , left, black arrowheads), whereas the INL also disappeared when the choroidal layer collapsed completely (Fig. 7D , right, white arrowheads). By 3 months, the choroid and retina were relatively well demarcated by Bruch membrane in most sections. Even with the complete loss of the RPE and the ONL, small breaks in Bruch membrane were seen at 3 months (Fig. 7E , white square). Prolapse of choroidal tissues was frequently observed in such breaks; however, in none of the sections were any choroidal vessels infiltrating the retinal layer (Fig. 7F) . Collectively, the results suggest that choroidal circulation is involved in the loss of retinal tissue 1 and 3 months after light exposure. 

Damage to Choroidal Perfusion after Light Exposure

We assessed the status of choroidal circulation by in vivo staining of the choroidal layer using trypan blue (Fig. 8) . The choroidal layer in the dim animals was homogeneously stained by blue dye throughout the 3 months (Fig. 8C , dim). When the eyecups were flipped, the blue dye located in the choroidal layer was clearly seen through the transparent sclera (Fig. 8A , dim); thus, a back view of the eyecup is another way to observe the choroid stained by trypan blue. The choroid in saline-injected, light-exposed animals showed a patchy pattern of staining 7 days after light exposure (Figs. 8A 8C , light and 7 d). Staining diminished remarkably by 1 month (Figs. 8A 8C , light and 1 m), and no staining was observed in most parts of the choroid by 3 months (Figs. 8A 8C , light and 3 m). In contrast, choroidal staining was well preserved in PBN-injected, light-exposed animals 3 months after light exposure (Figs. 8A 8C , light PBN (+) and 3 m). None of the eyecups showed apparent leakage of blue dye into the retinal layer (Fig. 8B) . 

Discussion

By 24 hours after light exposure, a remarkable increase of 4-HNE labeling was observed in the retinas of saline-injected animals, whereas this increase was inhibited in PBN-injected animals (Fig. 2) . By 7 days after light exposure, significant decreases in ERG amplitudes and in ONL thickness and area were observed in the saline-injected animals. These decreases were counteracted by pretreatment with PBN, a free-radical scavenger (Figs. 3 4) . The production of 4-HNE is the result of lipid peroxidation and cleavage in response to oxidative stress and aging. Recently, several molecular targets of protein modification by 4-HNE were reported in retina and RPE in vivo. ^{28 29} Accordingly, the results suggest the involvement of free radicals in the pathogenesis of retinal light damage, which is consistent with previous reports. ^{6 7 8} By 7 days after light exposure, significant downregulation was observed in rhodopsin expression in the retina and in RPE65 expression in the RPE fraction in the saline-injected animals (Fig. 5) . Substantial labeling by TUNEL in the ONL and RPE nuclei (Fig. 6)was observed in these animals, clearly suggesting that rod and RPE cells were removed by apoptosis early after light exposure, as previously reported. ^{5 17} By 7 days or 1 month after light exposure, it was frequently seen that only a single layer remained in the ONL 1 to 1.5 mm superior to the ONH, where the most severe damage occurred after light exposure (Fig. 4) . This single layer of the ONL predominantly consisted of cone cells (Figs. 7A 7B) , reported to be approximately 3% of the original population of the ONL nuclei in the dim animals. ¹⁸ In addition, the level of m-opsin expression 7 days after light exposure was similar to its expression in the dim animals (Fig. 5D) , clearly suggesting that cone cells were more resistant to acute light damage than rod and RPE cells. 

By 7 days after light exposure, significant decreases in choroidal thickness and area (Fig. 4)and an increase of TUNEL labeling in the choroidal layer (Fig. 6B)were observed; these changes were counteracted by PBN pretreatment (Figs. 4 6C) . In addition, the up-regulation of 4-HNE labeling was observed in choroidal cells by 24 hours after light exposure (Fig. 2B) . We previously reported ³⁰ that the upregulation of 8-hydroxy-2-deoxyguanosine (8-OHdG), a marker for oxidative stress–induced DNA damage, ³¹ and the nuclear translocation of NF-κB, a redox-sensitive transcription factor, ³² are molecular events involved in photoreceptor cell damage after light exposure. Recently, the upregulation of 8-OHdG and the activation of NF-κB were also measured in choroidal endothelial cells after light exposure in mice. ²² Results from the present study and previous studies suggest that photooxidative stress damages choroidal cells during the acute phase after light exposure. ^{22 30}  

By 3 months after light exposure, further downregulation of m-opsin expression was observed (Fig. 7D) ; thus, a delayed loss of cone cells occurred after acute light exposure. An increase in TUNEL-positive cells was observed in the ONL at the inferior peripheral retina, just behind the iris, by 3 months after light exposure (Fig. 6E) , indicating the apoptosis of remaining rod cells, whereas only a few TUNEL-positive cells were observed by 7 days after light exposure (Fig. 6F) . This delayed increase in TUNEL staining is not likely to be an effect of aging because age-matched dim animals did not show such an increase in staining (Fig. 6G) . Accordingly, the results strongly suggest that another mechanism of delayed cell death of cone and remaining rod cells is activated after acute light exposure. The presence of red blood cells in the choroid 1 and 3 months after light exposure is likely to be related to the severity of retinal damage (Figs. 7C 7D) . By in vivo staining of the choroidal layer, the choroidal circulation appeared in most areas 7 days after light exposure and was followed by a progressive loss of choroidal circulation (Fig. 8) . Importantly, the loss of choroidal circulation coincided with the downregulation of m-opsin expression. It was reported that nutritional and oxygen supplies in the outer retina depend on the choroidal circulation. ³³ Accordingly, our results strongly suggest that delayed loss of cone cells and of the remaining rod cells in peripheral retina relates to the progressive loss of choroidal circulation. This is supported by the loss of the total outer retina simultaneously with a complete absence of red blood cells in the choroidal layer (Fig. 7D) . 

The reason for a progressive loss of choroidal circulation is still unclear. Our results suggest that acute photooxidative stress-mediated choroidal damage had something to do with the decrease in the choroidal blood supply (Fig. 8)and resulted in the decrease of the choroidal thickness (Fig. 4C) , but it was not severe enough to cause a complete collapse of the choroidal vasculature complexes (Fig. 8) . It has been reported that obstruction of the choroidal vasculature occurred 1 to 7 days after laser exposure during photodynamic therapy. ^{34 35} Thus, choroidal damage resulting from photooxidation is an acute event and is not likely to explain the reason for the progressive loss of the choroidal circulation. In clinical situations, photocoagulation therapy for proliferative diabetic retinopathy has been widely used to involute the neovasculature through decreases in oxygen and nutritional requirements of the retina by thinning of the retinal cells. The involution of choroidal vasculature resulting from decreases in the number of rod and RPE cells may explain the delayed loss of choroidal circulation in this study. Previous studies reported the delayed cone cell death in pig and mouse models of retinitis pigmentosa; however choroidal circulation was not examined. ^{18 19 21} It will be interesting to determine whether choroidal circulation in these animal models is affected. 

In summary, cone cells are resistant to cell damage induced by photooxidative stress, but progressive loss of cone cells occurred for up to 3 months after acute light exposure. We hypothesize that impaired choroidal circulation is likely to be involved in the mechanism of delayed cone cell death after light exposure. Preserving choroidal circulation may be a novel target for preserving cone and remaining rod cells in patients with retinal degenerations such as retinitis pigmentosa. 

 Supported by National Eye Institute Grants EY04149, EY00871, and EY12190; National Center for Research Resources Grant RR17703; Research to Prevent Blindness, Inc.; and the Foundation Fighting Blindness. MT is a recipient of a Research Fellowship from the Japan Society for the Promotion of Science (JSPS) for Young Scientists.

 Submitted for publication September 7, 2006; revised December 17, 2006, and January 5, 2007; accepted February 19, 2007.

 Disclosure: M. Tanito, None; S. Kaidzu, None; R.E. Anderson, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

In vivo staining of the choroidal layer by trypan blue. (A, B) Dissected eyecup from an eyeball perfused with trypan blue. H&E–stained (C, E) and unstained (D, F, G) frozen sections from eyecup tissues before (A, retina) and after (A, choroid) removal of the retinal layers. SC, sclera; CH, choroid; ROS, rod outer segment; ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer.

Figure 2.

 

Immunohistochemistry for 4-HNE. Rats were injected with saline or PBN 30 minutes before 6-hour damaging light exposure. After 24 hours, eyes were enucleated from the dim rats (A, D) and the light-exposed (B, C) rats. Representative images of retinal sections stained with anti–4-HNE antibody (A–C) or normal mouse IgG as a staining control (D) are shown (four eyes from four rats were analyzed in each group). CH, choroid; RPE, retinal pigment epithelium; OS, outer segment; IS, inner segment; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer; GCL, ganglion cell layer; PBN, phenyl-N-tert-butylnitrone.

Figure 3.

 

Electroretinography amplitudes of a- and b-waves. Rats were injected with saline or PBN 30 minutes before 6-hour damaging light exposure. Age-matched rats that were injected with saline or PBN were kept in dim light and used as controls. ERGs were recorded 7 days, 1 month, and 3 months after light exposure. Representative ERG recordings (A) and the means (± SD) of the a-waves (B) and b-waves (C) are shown. Asterisks indicate significant differences (P < 0.01) between the saline- and PBN-injected animals at each time point using a paired t-test. The number of animals in each group is indicated in parentheses.

Figure 4.

 

H&E staining and quantification of the ONL or choroidal thickness and area. (A) Representative sections from rats injected with saline or PBN 7 days, 1 month, and 3 months after exposure to damaging light. Sections stained with H&E were 1.0 to 1.5 mm superior from the optic nerve head (representative of 6–12 eyes from 3–6 rats analyzed in each group). CH, choroid; ONL, outer nuclear layer; INL, inner nuclear layer. (B–E) Mean (± SD) for the ONL thickness (B) and the area (D); and the choroidal thickness (C) and the area (E) of rats in (A) are shown. Differences were considered significant (*P < 0.05; **P < 0.01) between saline- and PBN-injected animals; paired t-test. The number of animals in each group is indicated in parentheses.

Figure 5.

 

Western blotting using rhodopsin, m-opsin, and RPE65. Representative blots for rhodopsin, m-opsin, and GAPDH (loading control; A) and for RPE65 and GAPDH in the RPE fraction (B) using retinal samples from rats injected with saline or PBN 7 days, 1 month, and 3 months after exposure to damaging light are shown (representative of four independent experiments of four rats analyzed in each group). Densitometric analyses of Western blot analysis for rhodopsin (C), m-opsin (D), and RPE65 (E). Mean (± SD) densities standardized to GAPDH are shown (n = 4 rats in each group). Differences were considered significant (*P < 0.05; **P < 0.01) between groups using a one-way ANOVA followed by Scheffé post hoc test.

Figure 6.

 

TUNEL staining of retinal sections. (A–H) Representative TUNEL staining of retinal sections 1.0 to 1.5 mm superior from the optic nerve head (A–D, H) and in the peripheral retina adjacent to the inferior ciliary body (E–G) are shown (representative of four eyes from four rats analyzed in each group). Rats were injected with saline or PBN 30 minutes before 6-hour damaging light exposure. Age-matched rats injected with saline or PBN were kept in dim light and used as controls. CH, choroid; RPE, retinal pigment epithelium; ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer; peri, peripheral retina.

Figure 7.

 

Light and electron microscopy of retinal sections. Eyes were enucleated from rats 1 month (A–C) and 3 months (D–F) after the light exposure. Retinal sections were stained with H&E (A, D) or toluidine blue (C, E) and were observed with light microscopy or transmission electron microscopy (B, F), respectively. Representative sections are shown. CH, choroid; BM, Bruch membrane; RPE, retinal pigment epithelium; ONL, outer nuclear layer; INL, inner nuclear layer; IPL, inner plexiform layer.

Figure 8.

 

Choroidal layer staining with trypan blue. Rats were injected with saline or PBN 30 minutes before 6-hour damaging light exposure. Age-matched rats injected with saline or PBN were kept in dim light and used as controls. Cardiac injection of trypan blue was performed before enucleation of the eyes. Representative images of the back view of the eyecup (A), the retinal layer (B), and the choroidal layer (C) are shown (representative of at least six eyes from three rats analyzed in each group). S, superior; N, nasal; I, inferior; T, temporal.

The authors thank Mark Dittmar (Dean A. McGee Eye Institute) for maintaining the animal colonies used in this study and Louisa J. Williams and Linda S. Boone (Dean A. McGee Eye Institute) for their excellent retinal section preparation.