Differential Sensitivity of Cones to Iron-Mediated Oxidative Damage

Abstract

purpose. In this study, the hypothesis that increased intraocular levels of iron cause oxidative damage to the retina was tested.

methods. Adult C57BL/6 mice were given an intravitreous injection of saline or 0.10, 0.25, or 0.50 mM FeSO₄. Scotopic electroretinograms (ERGs) were performed 3, 7, and 14 days after injection, and photopic ERGs were performed on day 14. Hydroethidine was used to identify superoxide radicals and lipid peroxidation was visualized by staining for hydroxynonenal (HNE). Retinal cell death was evaluated by TUNEL and measurement of inner nuclear layer (INL) and outer nuclear layer (ONL) thickness. Levels of rhodopsin and cone-opsin mRNA were measured by quantitative real time RT-PCR. Cone density was assessed by peanut agglutinin staining and confocal microscopy.

results. Compared with retinas in saline-injected eyes, retinas from eyes injected with FeSO₄ showed greater fluorescence after intravenous injection of hydroethidine due to superoxide radicals in photoreceptors, greater photoreceptor staining for HNE, a marker of lipid peroxidation, and increased expression of Heme oxygenase 1, an indicator of oxidative stress. ERG b-wave amplitudes were reduced (photopic > scotopic) in FeSO₄-injected eyes compared with those in saline-injected eyes. Numerous TUNEL-stained nuclei were seen along the outer border of the ONL, the location of cone cell nuclei, at 1 and 2 days after injection of FeSO₄. In FeSO₄-injected eyes, the thickness of the ONL, but not the INL, was significantly reduced, and 17 days after injection, there were 3.8- and 2.6-fold reductions in the mRNAs for M-cone and S-cone opsin, respectively, whereas there was no significant difference in rhodopsin mRNA. Confocal microscopy of peanut agglutinin–stained sections showed dose-dependent FeSO₄-induced cone drop out.

conclusions. Increased intraocular levels of FeSO₄ cause oxidative damage to photoreceptors with greater damage to cones than rods. This finding suggests that the oxidative defense system of cones differs from that of rods and other retinal cells, and that cones are more susceptible to damage from the type of oxidative stress imposed by iron.

Intraocular foreign bodies containing iron are toxic to the retina and, if left in the eye for prolonged periods, result in an extinguished electroretinogram (ERG) and blindness. This condition is referred to as siderosis bulbi, because there is progressive deposition of iron in ocular tissues. ¹ Dense collections of ferritin particles are seen in the cytoplasm and organelles of ocular cells, and it has been hypothesized that the large accumulations cause physical damage that kills retinal cells. ² It has also been hypothesized that iron liberated from intraocular hemorrhage causes damage by inducing inflammation. ^{3 4} A third hypothesis relating to iron-induced retinal damage is that iron liberated from hemoglobin after intraocular hemorrhage or from other sources causes oxidative damage in the retina. ⁵  

The effects of iron in the eye are complicated, because there are multiple forms of iron with different reactivity and several proteins that modulate reactivity and/or levels. Ceruloplasmin is a copper-binding serum protein that oxidizes highly reactive Fe²⁺ ions to less reactive Fe³⁺ ions. ⁶ This effect not only alters the nature of iron, but also promotes its export from tissues by allowing it to bind to transferrin, the major iron transporter that only binds the Fe³⁺ form. Patients with aceruloplasminemia have defective transport of iron out of some tissues including the retina and some brain regions, resulting in retinal degeneration and dementia usually occurring in an age range of 40 to 60 years. ^{7 8 9} At 5 months of age and older, mice deficient in ceruloplasmin and hephaestin, another protein that oxidizes Fe²⁺ to Fe³⁺, develop high levels of iron in the retina and retinal pigmented epithelium (RPE), and beyond 6 months of age they show areas of RPE hyperplasia and atrophy, with some areas of photoreceptor atrophy and occasional subretinal neovascularization. ¹⁰ Thus, increased levels of iron in the eye, whether from a metallic foreign body, high levels of hemoglobin, or defective export of iron, cause retinal degeneration. As noted earlier, there are several hypotheses as to the mechanism of iron-induced damage, but since iron is capable of reacting with oxygen to generate reactive oxygen species, ¹¹ recent studies seem to assume oxidative damage plays a role, although this has not been conclusively proven. In this study we sought to test this hypothesis and determine whether the damage is homogeneous or localized to a particular cell type. 

Materials and Methods

Intravitreous Injections

Six to 8-week-old C57BL/6 mice (Jackson Laboratories, Bar Harbor, ME) were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and the U.S. National Institutes of Health Guide for the Care and Use of Laboratory Animals. Intravitreous injections were given with a Harvard pump apparatus (Harvard Apparatus, Holliston, MA) and pulled-glass micropipettes, as previously described. ¹² Micropipettes were calibrated for delivery of 1 μL of vehicle on depression of a foot switch. Mice were anesthetized with ketamine and xylazine for general anesthesia and, under a dissecting microscope, the beveled tip of a micropipette was passed through the sclera just behind the limbus into the vitreous cavity, and the foot switch was depressed. Experimental eyes were injected with various concentrations of FeSO₄ (Sigma-Aldrich, St. Louis, MO) diluted in 0.9% NaCl and control eyes were injected with 0.9% NaCl. 

Recording of Electroretinograms

Electroretinograms (ERGs) were recorded (Espion ERG; Diagnosys LLL, Littleton, MA), as previously described. ¹³ For scotopic recordings, mice were dark adapted overnight for 12 hours, and for photopic recordings, mice were adapted for 15 minutes to a background of white light at an intensity of 30 cd/m². Both scotopic and photopic ERGs were performed under dim red illumination. The mice were anesthetized with intraperitoneal injection of Avertin (Aldrich, Milwaukee, WI) and topical administration of 0.5% proparacaine hydrochloride (Alcon Laboratories, Fort Worth, TX). Pupils were dilated with 1% tropicamide (Alcon Laboratories). The mice were placed on a pad heated to 39°C, and platinum loop electrodes were placed on each cornea after application of gonioscopic prism solution (Alcon Laboratories). A reference electrode was placed subcutaneously in the anterior scalp between the eyes, and a ground electrode was inserted into the tail. The head of the mouse was held in a standardized position in a Ganzfeld bowl illuminator that ensured equal illumination of the eyes. Recordings for both eyes were made simultaneously with electrical impedance balanced. Scotopic recordings were made at 11 intensity levels of white light ranging from −3.00 to 1.40 log cd-s/m². Six scotopic measurements were taken at each flash intensity. Photopic recordings were made at three intensity levels of white light ranging from 0.60 to 1.40 log cd-s/m². Five photopic measurements were taken at each flash intensity. The ERG apparatus measures the ERG response multiple times at each flash intensity and records the average value. 

Tissue Processing

The mice were euthanatized at different time points after intravitreous injections, depending on what was being investigated (detailed later), and eyes were immediately frozen in embedding compound (Tissue Tek; Miles Diagnostics, Elkhart, IN). Ten-micrometer sections were cut, dried with cold air for 5 minutes, fixed in freshly prepared 4% paraformaldehyde in 0.05 M phosphate-buffered saline (PBS) at 4°C for 30 minutes, and rinsed with Tris-buffered saline (TBS) for 10 minutes. Outer and inner nuclear layer thickness measurements were obtained as previously described. ¹³  

Immunohistochemical Staining for 4-Hydroxy-2-nonenal

Rabbit anti-HNE antibody was used for indirect immunofluorescent staining (Alpha Diagnostic International, Inc., San Antonio, TX). The frozen sections were dried at room temperature and postfixed in 4% paraformaldehyde for 15 minutes. The sections were washed with phosphate-buffered saline (PBS), blocked with 10% normal goat serum in PBS for 30 minutes at room temperature, and incubated overnight at 4°C with 1:100 rabbit anti-HNE diluted with PBS and containing 2% goat serum. After they were washed three times with PBS/0.1% Triton X-100, the sections were incubated for 1 hour at room temperature in 1:800 donkey anti-rabbit IgG conjugated with Cy3 (Jackson ImmunoResearch Laboratory, West Grove, PA). They were washed three times with PBS/0.1% Triton X-100 and then incubated at room temperature for 4 minutes with the nuclear dye Hoechst 33258 (1:1200; Sigma-Aldrich). The slides were viewed with a fluorescence microscope (Nikon Instruments Inc., New York, NY). Images were captured at the same exposure time for each section with a digital camera (Nikon) and software (SPOT RT 3.4; Diagnostic Instruments, Sterling Heights, MI). 

Terminal Deoxynucleotidyl Transferase-Mediated dUTP-Biotin Nick-End Labeling

Apoptotic cells were detected by TUNEL with an apoptosis detection kit (Apoptag Red In Situ Apoptosis Detection Kit; Chemicon International Inc., Temecula, CA). Sections were counterstained with Hoechst 33258 (1:1200; Sigma-Aldrich) and examined with a fluorescence microscope (Nikon). 

In Vivo Detection of Superoxide Radicals with Hydroethidine

Superoxide radicals were detected in the retina using hydroethidine, which is taken up by cells and reacts with superoxide to form ethidium. Ethidium binds DNA and emits red fluorescence. ¹⁴ Three days after an intravitreous injection of FeSO₄ or saline, mice were given a tail vein injection of 500 μL of 0.5 mg/mL of hydroethidine in PBS. After 45 minutes, the mice were euthanatized and eyes were removed and frozen. Ocular frozen sections were fixed in 2% paraformaldehyde for 5 minutes, rinsed in PBS, and stained for 3 minutes at room temperature with Hoechst 33258. The slides were examined with a fluorescence microscope (Nikon) at 510-nm excitation and >580-nm emission for detection of ethidium and at 360-nm excitation and >460-nm emission for detection of Hoechst 33258. All images were captured using the same exposure time (SPOT RT 3.4 software; Diagnostic Instruments). 

Staining with Peanut Agglutinin and Confocal Microscopy

The mice were euthanatized and eyes were fixed in 4% paraformaldehyde for 1 to 2 hours. Twenty-micrometer ocular frozen sections were dried, placed in 10% normal goat serum in PBS for 30 minutes at room temperature, and incubated for 1 hour at room temperature in 1:30 rhodamine-conjugated peanut agglutinin (PNA; Vector Laboratories, Burlingame, CA) and Hoechst nuclear dye (Sigma-Aldrich) in PBS containing 1% normal goat serum. The sections were viewed with a confocal microscope (LSM 510 META; Carl Zeiss Meditec, Oberkochen, Germany, with a 20×/0.75 numeric aperture objective; Plan-Apochromat; Carl Zeiss Meditec) using excitation wavelengths of 543 nm (PNA) or 405 nm (Hoechst). Images were acquired in the frame-scan mode, and stack size was 230.3 × 230.3 μm (512 × 512 pixels). 

Quantitative Real-Time RT-PCR

Retinas were dissected and snap frozen individually. Isolation of RNA was accomplished by lysis in RLT buffer, homogenization (QIAshredder columns), purification on mini silica-gel spin-column (RNeasy), DNase I digestion on the column and elution in 25 μL water using the manufacturer’s protocols (all products, Qiagen, Valencia, CA). Eight microliters of the RNA solution were used to synthesize cDNA using MMLV reverse transcriptase (iScript; Bio-Rad, Hercules, CA) and an optimized blend of oligo(dT) and random primers. Real-time PCR was performed in 20-μL reaction volumes (iQ SYBR Green SuperMix; Bio-Rad), with 0.5 μL cDNA template and primers at a concentration of 0.5 μM (Chromo4 machine; Bio-Rad). Amplification was performed with a polymerase activation step of 95°C for 3 minutes followed by 43 cycles of 95°C for 10 seconds, 59°C for 5 seconds, and 70°C for 15 seconds. At the end of every cycle the fluorescence in each well was measured at 80°C. All amplifications were performed in duplicate. Relative quantification of transcripts of genes of interest with respect to transcripts for the ribosomal protein S16 was performed using the equation 2^(−ΔΔCt), as previously described. ¹⁵ The following primers were used in the real-time PCRs: S16 5′ primer 5′-CACTGCAAACGGGGAAATGG-3′, 3′ primer 5′-TGAGATGGACTGTCGGATGG -3′; Opn1mw 5′ primer 5′-TCATTTCCTGGGAGAGATGG-3′, 3′ primer 5′-AGGCCATAAGGCCAGTACCT-3′; Opn1sw 5′ primer 5′-GCCTCAGTACCACCTTGCTC-3′, 3′ primer 5′-CTGGCGATGAAGACTGTGAA-3′; Rho 5′ primer 5′-GTCAGCCACCACTCAGAAGG-3′, 3′ primer 5′-CTGGCTGGTCTCCGTCTTG -3′; and Hmox1 5′ primer 5′-CACGCATATACCCGCTACCT-3′, 3′ primer 5′-AAGGCGGTCTTAGCCTCTTC-3′. 

Results

Effect of Intraocular Injection of FeSO₄ on Superoxide Radicals and Lipid Peroxidation in Photoreceptors

We tested whether increased Fe²⁺ in the retina results in generation of superoxide radicals by injecting FeSO₄ into the vitreous and after 3 days injecting hydroethidine intravenously. Hydroethidine is taken up into cells and in the presence of superoxide radicals is converted to ethidium, which binds DNA and emits red fluorescence. ¹⁴ Mice were euthanatized 45 minutes after infusion of hydroethidine, and eyes were rapidly removed and frozen. Fluorescence microscopy of postfixed ocular frozen sections showed that eyes injected with 1 μL of 0.25 mM FeSO₄ showed strong fluorescence in photoreceptor inner segments and mild diffuse fluorescence elsewhere in the retina (Fig. 1A) . Fellow eyes that had been injected with saline showed only mild diffuse fluorescence throughout the retina (Fig. 1B) . Eyes injected with 1 μL of 0.5 mM FeSO₄ showed intense fluorescence in photoreceptor inner segments and scalloping of the ONL (Fig. 1C) , which suggests that elevated levels of Fe²⁺ in the eye cause an increase in the generation of superoxide radicals predominantly in photoreceptors, and the irregularity of the ONL suggests photoreceptor damage. Intraocular injection of 1 μL of 0.5 mM FeSO₄ may also result in substantial systemic exposure, because the contralateral eyes showed fluorescence in photoreceptor inner segments (Fig. 1D)that appeared greater than that in eyes contralateral to those injected with 0.25 mM FeSO₄ (Fig. 1B)and clearly greater than the fluorescence in mice that did not receive any FeSO₄ (Figs. 1E 1F) . 

Immunofluorescent staining for 4-hydroxynonenal (HNE) in eyes injected with 0.25 mM FeSO₄ showed strong staining in photoreceptors indicating substantial lipid peroxidation in photoreceptors (Fig. 1G) . There was mild staining elsewhere in the retina. Contralateral saline-injected eyes showed only mild fluorescence throughout the retina (Fig. 1H) . Sections from eyes injected with 0.25 mM FeSO₄ that were treated only with secondary antibody showed almost no background fluorescence (Fig. 1I) . 

Effect of Intraocular Injection of FeSO₄ on Heme Oxygenase 1

The Hmox1 gene has two known transcriptional enhancer sequences both of which contain multiple antioxidant response elements (AREs), ¹⁶ and its expression is consistently elevated by many types of oxidative stress in many different settings. ¹⁷ The other major categories of stimuli that induce Hmox1 expression are nitrosative stress, thiol-reactive substances, and cytokines. Quantitative real time RT-PCR showed a 20.4-fold increase in Hmox1 mRNA in the retina 1 day after injection of FeSO₄ and a 9.5-fold increase at 3 days (Table 1 ; P < 0.001 at both time points). 

Effects of Increased Levels of Fe²⁺ in the Retina on Apoptosis in Cone Photoreceptors and the Thickness of the ONL

One day after injection of 1 μL of 0.25 mM FeSO₄, there were several TUNEL-positive cells localized in a fairly thin band in the retina (Fig. 2A) . When the TUNEL staining was combined with Hoechst 33245 staining of the same section to visualize the nuclei, it was shown that TUNEL-stained cells occurred along the outer edge of the ONL (Fig. 2B) , which is the location of cone cell nuclei. Two days after injection of FeSO₄ there were many more TUNEL-positive cells in a somewhat wider band in the outer portion of the ONL (Figs. 2C 2D) . There was no TUNEL staining in the retinas of saline-injected fellow eyes (not shown). 

Retinal morphology was examined 14 days after injection. The retina was normal in saline-injected eyes (Fig. 3A)and fairly normal in eyes injected with 0.10 mM FeSO₄, except that the ONL seemed to be thinner (Fig. 3B)than that in saline-injected eyes; however, measurement of ONL area by image analysis did not show a significant reduction (Fig. 3D) . Injection of 0.25 mM FeSO₄ caused a significant thinning of the ONL (Figs. 3C 3D) , but the inner nuclear layer was normal (Fig. 3E) . 

Effect of Intraocular Injections of FeSO₄ on Cone Opsin Transcripts

One day after injection of 0.25 mM FeSO₄, the mRNAs for M- and S-cone opsins were reduced by 8- and 5.8-fold, respectively, and rhodopsin mRNA was reduced by 2.5-fold (Table 2) . At 17 days after injection, the cone opsin transcripts were still significantly reduced (3.8- and 2.6-fold), whereas rhodopsin transcripts were not significantly different from those in vehicle-injected eyes. 

Effect of Increased Levels of Fe²⁺ on ERG Amplitudes

When ERGs are performed in the dark-adapted state, they are referred to as scotopic ERGs and provide a valuable assessment of retinal function driven primarily by rod photoreceptors. Eyes injected with 0.10 mM FeSO₄ showed moderate reductions in average a-wave (Figs. 4A 4C)and b-wave (Figs. 4B 4D)amplitudes at 3 and 7 days, which had recovered by 14 days (Figs. 4E 4F) . Eyes injected with 0.25 mM or 0.50 mM FeSO₄ showed marked decreases in average a- and b-wave amplitudes at all three time points. At day 14, representative waveforms from eyes injected with 0.25 mM (Fig. 4I)or 0.50 mM FeSO₄ (Fig. 4J)were markedly diminished in size. 

When ERGs are performed in the light-adapted state, they are referred to as photopic ERGs and provide a means of assessing retinal function driven primarily by cone photoreceptors. C57BL/6 mice were given an intravitreous injection of 1 μL of 0.10, 0.25, or 0.50 mM FeSO₄ in one eye and 1 μL of saline in the fellow eye. After 14 days, photopic ERGs showed a significant reduction in average b-wave amplitudes in eyes injected with 0.50 or 0.25 mM FeSO₄ compared with those injected with saline or 0.10 mM FeSO₄ (Fig. 5A) . The dose-dependent reduction in b-wave amplitudes is illustrated by the appearance of the representative waveforms (Figs 5B 5C 5D 5E) . 

Table 3shows the reduction in photopic and scotopic ERG b-waves as a percent of control 14 days after intraocular injection of 0.1, 0.25, or 0.5 mM FeSO₄. For scotopic ERGs, comparisons were made by using data obtained at the third flash intensity, −2.2 log cd-s/m², and for photopic ERGs, comparisons were made by using data obtained with background illumination and at the highest flash intensity, 1.4 log cd-s/m². The reduction in the photopic b-wave was significantly greater than the reduction in scotopic b-wave for each of the three doses. These intensities were chosen because the first one is below the cone threshold, and the b-wave is derived from rod activity, whereas the second intensity is well above the cone threshold, and the background intensity saturates rods and makes their contribution minimal. Eyes injected with iron showed large b-wave reductions for three intensities near the upper end of the scale recorded under photopic conditions, and modest b-wave reductions for low intensities that primarily stimulate rods. This indicates that high levels of intraocular iron cause greater loss of cone function than rod function. 

Iron-Induced Reduction in Cone Density

Fourteen days after injection of saline or 0.1 or 0.25 mM FeSO₄, ocular frozen sections were cut parallel to the horizontal meridian starting at the superior pole of the retina and were stained with rhodamine-conjugated PNA for detection of cone inner and outer segments and Hoechst for detection of nuclei. The sections midway between the optic nerve and the superior (superior midperipheral retina) or inferior (inferior midperipheral retina) poles of the retina were examined by confocal microscopy, because topographic differences have been noted with various types of retinal degenerations. The cone density assessed by staining with PNA was normal in both the superior and inferior retina in eyes injected with saline and appeared similar to that seen in noninjected eyes (Fig. 6 , first two columns). Eyes that had been injected with 0.1 mM FeSO₄ showed reduced cone density in the superior retina (Fig. 6 , third column, top) and less noticeable change in the inferior retina (Fig. 6 , third column, bottom). Eyes injected with 0.25 mM FeSO₄ showed more severe reduction in cone density, particularly in the superior retina (Fig. 6 , fourth column, top). In contrast, the ONL thickness showed little difference, suggesting that rod photoreceptors were much less affected. 

Discussion

In this study, intravitreous injection of FeSO₄ caused increased generation of superoxide radicals and oxidative damage in photoreceptors. This effect was associated with reduced retinal function determined by ERGs, with the greatest reduction in the photopic signaling pathway, which is initiated by cones. There was loss of photoreceptors by apoptosis, resulting in thinning of the ONL and no significant change in the INL. Apoptotic nuclei were concentrated along the outer border of the ONL, the location of cone cell nuclei, and whereas the mRNAs for M- and S-cone opsins were reduced 17 days after injection, there was not a significant reduction in rhodopsin mRNA. Staining of cone matrix sheaths with PNA showed reduced cone density, somewhat greater in the superior midperiphery of the retina than in the inferior midperiphery. These data suggest that FeSO₄ causes oxidative damage to photoreceptors and that cones are more susceptible than rods. 

This is an intriguing finding, because it indicates that with respect to the endogenous antioxidant defense system, photoreceptors differ from other retinal cells and cones differ from rods. The oxidative stress posed by FeSO₄ is different from that posed by paraquat, an herbicide that generates reactive oxygen species, because the inner retina is more susceptible to paraquat than it is to FeSO₄. ¹⁸ We hypothesize that different components of the endogenous antioxidant defense system may provide different levels of protection against various types of oxidative stress and that the cell types in the retina differ with respect to their antioxidant defense system. The results of this study are consistent with this hypothesis, but to prove the hypothesis, it is necessary to define how well components of the antioxidant defense system protect against specific types of oxidative stress and determine their roles in various cell types. We are currently defining which components of the endogenous antioxidant defense system protect retinal cells against paraquat-induced damage and have found that SOD1 provides some protection, ¹⁹ whereas SOD3 does not (Dong A, Campochiaro PA, unpublished data). It will be useful to determine whether protection against FeSO₄ is provided by different components of the defense system than those that protect against paraquat. 

Intraocular injection of FeSO₄ provides a model of oxidative damage-induced retinal degeneration that may have relevance to particular ocular diseases in which Fe overload has been demonstrated. Intraocular implantation of Fe particles in rats ²⁰ or rabbits ²¹ provides similar models, but requires a surgical procedure, and the Fe dose is unknown. Intraocular Fe wires caused selective damage to photoreceptors and after implantation of Fe particles, rats showed apoptosis only in the ONL and not in any other layers of the retina. Our findings confirm those previous observations and extend them to show that cones are more susceptible than rods. 

The FeSO₄-injection model is relevant to aceruloplasminemia in which increased levels iron in the retina may be associated with midperipheral atrophic changes in the retina and RPE ²² or drusenlike yellow spots and atrophic changes in the macula. ²³ Agents that protect against FeSO₄-induced retinal damage could provide benefit in patients with aceruloplasminemia. Mice deficient in ceruloplasmin and hephaestin provide a highly relevant model for aceruloplasminemia, but the retinal degeneration occurs slowly over several months, making it impractical for screening a large number of agents, but reasonable to use for confirmatory testing of agents found to protect against intraocular FeSO₄-induced retinal damage. 

Iron overload has also been shown in a few patients with age-related macular degeneration. ^{24 25} Although it is unlikely that Fe overload is a primary problem in most patients with AMD, it may occur secondary to photoreceptor damage from another cause and contribute to ongoing oxidative damage. Iron overload has been seen in association with several neurodegenerative disorders, including the RCS rat model of retinitis pigmentosa, in which the primary defect is known to be a mutation that has nothing to do with Fe metabolism or transport. ²⁶ It appears that damaged neurons may accumulate Fe. One possible mechanism is loss of transferrin on damaged cells, compromising export of Fe. Whatever the source of Fe overload, the present study supports the hypothesis that its presence leads to exacerbation of oxidative stress and the nature of that oxidative stress differs from that due to other sources of oxidative stress, such as paraquat. Defining the components of the oxidative defense system that best protect against Fe-induced damage to photoreceptors and RPE cells may help in the development of treatments for retinal degenerations associated with iron overload. 

²  Contributed equally to the work and therefore should be considered equivalent authors.

 Supported by the Macula Vision Research Foundations and Dr. and Mrs. William Lake. PAC is the George S. and Dolores Dore Eccles Professor of Ophthalmology.

 Submitted for publication May 13, 2006; revised September 1 and 21, 2006; accepted November 14, 2006.

 Disclosure: B.S. Rogers, None; R.C.A. Symons, None; K. Komeima, None; J.-K. Shen, None; W. Xiao, None; M.E. Swaim, None; Y.Y. Gong, None; S. Kachi, None; P.A. Campochiaro, 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.

 

Intravitreous injection of FeSO₄ results in superoxide radical generation and lipid peroxidation in the retina. Adult C57BL/6 mice were given an intravitreous injection of 0.25 mM (A) or 0.50 mM FeSO₄ (C) and saline in the fellow eye (B, D). Mice that received an injection of saline in one eye (E) and no FeSO₄ in the fellow eye and mice that received no injections (F) served as additional controls. Three days after injection, the mice were used to assess for superoxide radical generation (A–F) or lipid peroxidation (G–I). To assess for superoxide radicals in the retina, mice were euthanatized 45 minutes after injection of 250 mg of hydroethidine into the tail vein. Eyes were rapidly removed and frozen, and ocular frozen sections were examined by fluorescence microscopy. Retinas from eyes injected with 0.25 mM FeSO₄ showed mild diffuse fluorescence throughout the retina with focal areas of strong fluorescence in the region of the photoreceptor inner segments and cell bodies (A). The outer border of the ONL was irregular. Retinas from saline-injected eyes contralateral to those injected with 0.25 mM FeSO₄ showed mild diffuse fluorescence throughout the retina (B). Retinas from eyes injected with 0.50 mM FeSO₄ showed focal areas of intense fluorescence on a background of moderately intense fluorescence in photoreceptor inner segments and cell bodies (C). Eyes contralateral to those injected with 0.50 mM FeSO₄ (D) showed mild fluorescence throughout the retina, but particularly in photoreceptor inner segments, that appeared greater than that seen in the eye contralateral to those injected with 0.25 mM FeSO₄ (B) and much greater than that seen in retinas from mice that did not receive an injection of FeSO₄ in either eye (E, F). To assess for lipid peroxidation, retinas were immunofluorescently stained with an antibody directed against 4-hydroxynonenal (HNE). Retinas from eyes injected with 0.25 mM FeSO₄ showed intense fluorescence in the photoreceptor inner and outer segments with mild fluorescence elsewhere in the retina (G). Contralateral saline-injected eyes showed only mild fluorescence throughout the retina (F). Sections from eyes injected with 0.25 mM FeSO₄ that were treated only with secondary antibody showed almost no background fluorescence (I).

Table 1.

 

Gene Expression Analysis for Hmox1 in Mouse Retina after Intravitreous Injection of Normal Saline or 0.25 mM FeSO₄

Gene	Time (days)	Eyes (n)			Average Δ(CtS16 − CtHmox1)			SD (ΔCt)			ΔΔCt	Increase Due to FeSO4 (x-fold)	P (t-test)
				Control	FeSO4	Control	FeSO4	Control	FeSO4
Hmox1	1	4	4	−8.6	−4.2	0.2	0.3	4.4	20.4	<0.001
Hmox1	3	8	8	−9.2	−6.0	1.0	0.8	3.2	9.5	<0.001

 Real-time RT-PCR was performed in duplicate for each sample.

Figure 2.

 

Intravitreous injection of FeSO₄ caused apoptosis of photoreceptors. C57BL/6 mice were given an intravitreous injection of 1 μL of 0.25 mM FeSO₄ in one eye and 1 μL of saline in the fellow eye. One day after injection of 0.25 mM FeSO₄ there were several TUNEL-positive cells localized in a fairly thin band in the retina (A, B). Two days after injection of FeSO₄ there were many more TUNEL-positive cells in a somewhat wider band (C, D). Superimposition of the TUNEL staining on nuclei stained with Hoechst 33245 showed that apoptotic cells were concentrated along the outer border of the ONL, the location of cone cell bodies. There was no TUNEL staining in the retinas of saline-injected fellow eyes (not shown).

Figure 3.

 

Intravitreous injection of FeSO₄ caused selective thinning of the ONL. C57BL/6 mice were euthanatized 14 days after an intravitreous injection of 1 μL of 0.10 or 0.25 mM FeSO₄ or saline. The area of the ONL was measured on 10-μm serial frozen sections stained with hematoxylin and eosin. Retinas appeared normal after injection of saline (A), and although the ONL appeared somewhat thin in eyes injected with 0.10 mM FeSO₄ (B), measurements by image analysis showed that it was not significantly different from that in saline-injected eyes (D). In eyes injected with 0.25 mM FeSO₄ the ONL appeared quite thin (C), and in this case, image analysis confirmed that there was a statistically significant reduction compared with saline-injected eyes (D). In contrast, there was no difference in the thickness of the inner nuclear layer in eyes injected with 0.25 mM FeSO₄ compared with those injected with saline (E). *P < 0.0001 for the difference from saline-injected eyes, by ANOVA with the Bonferroni correction for multiple comparisons.

Table 2.

 

Gene Expression Analysis for Cone Opsins and Rhodopsin in Mouse Retina after Intravitreous Injection of Normal Saline or 0.25 mM FeSO₄

Gene	Time (d)	Eyes (n)			Average Δ(CtS16 − Ctgene)			SD (ΔCt)			ΔΔCt	Decrease Due to FeSO4 (x-fold)	P (t-test)
				Control	FeSO4	Control	FeSO4	Control	FeSO4
Opn1mw	1	4	4	−2.5	−5.6	0.7	1.1	−3.0	8.0	<0.004
Opn1mw	17	12	12	−3.8	−5.8	0.5	1.3	−1.9	3.8	<0.001
Opn1sw	1	4	4	−0.8	−3.3	0.6	1.9	−2.5	5.8	<0.05
Opn1sw	17	12	12	−2.0	−3.3	0.6	1.1	−1.4	2.6	<0.001
Rho	1	4	4	4.7	3.4	0.7	0.6	−1.3	2.5	<0.05
Rho	17	12	12	4.4	4.2	0.3	0.5	−0.2	1.2	0.19

 Real-time RT-PCR was performed in duplicate for each gene for each sample.

Figure 4.

 

Intravitreous injection of FeSO₄ caused dose-dependent reductions in scotopic ERG amplitudes. C57BL/6 mice were given an intravitreous injection of 1 μL of 0.10 mM, 0.25 mM, or 0.50 mM FeSO₄ in one eye and 1 μL of saline in the fellow eye. Scotopic ERGs were performed 3, 7, and 14 days after injections. Eyes injected with 0.10 mM FeSO₄ showed moderate reductions in average a-wave (A, C) and b-wave (B, D) amplitudes at 3 and 7 days, which had recovered by 14 days (E, F). Eyes injected with 0.25 or 0.50 mM FeSO₄ showed marked decreases in average a- and b-wave amplitudes at all three time points (A–F). Representative wave forms 14 days after injection of each of the doses are shown in (G–J) Compared to waveforms from eyes injected with saline (G) or 0.10 mM FeSO₄ (H), those injected with 0.25 mM (I) or 0.50 mM FeSO₄ (J) were markedly diminished in size.

Figure 5.

 

Intravitreous injection of FeSO₄ caused dose-dependent reductions in photopic ERG amplitudes. C57BL/6 mice were given an intravitreous injection of 1 μL of 0.10, 0.25, or 0.50 mM FeSO₄ in one eye and 1 μL of saline in the fellow eye. Fourteen days after the injections, there was a significant reduction in average photopic b-wave amplitudes in eyes injected with 0.50 or 0.25 mM FeSO₄ compared with those injected with saline or 0.10 mM FeSO₄ (A). Representative waveforms also show the dose-dependent reduction in b-wave amplitudes caused by FeSO₄ (B–E).

Table 3.

 

Reduction in Photopic and Scotopic ERG b-Waves 14 Days after Injection of Various Concentrations of FeSO₄

Dose	Wave	Percent of Control	n	P
0.10 mM	Photopic b-wave	53.6 ± 6.7	8
0.10 mM	Scotopic b-wave	119.6 ± 28.7	8	0.042
0.25 mM	Photopic b-wave	30.2 ± 7.1	8
0.25 mM	Scotopic b-wave	52.8 ± 3.8	8	0.014
0.50 mM	Photopic b-wave	29.8 ± 5.0	13
0.50 mM	Scotopic b-wave	64.6 ± 11.5	13	0.010

 Statistical comparisons were made by Student’s two-tailed t-test.

Figure 6.

 

Selective drop-out of cones after intraocular injection of FeSO₄. Adult C57BL/6 mice received intravitreous injections of saline or 0.1 or 0.25 mM FeSO₄ in one eye, with no injection in the fellow eye. Fourteen days after injections, mice were euthanatized and 20-μm frozen sections were cut parallel to the horizontal meridian starting at the superior pole of the retina. Sections were stained with rhodamine-conjugated PNA for detection of cone inner and outer segments and Hoechst for nuclei detection. Sections midway between the optic nerve and the superior (superior midperipheral retina) inferior (inferior midperipheral retina) poles of the retina were examined by confocal microscopy. The cone density was normal in both superior and inferior retina in eyes injected with saline and appeared similar to that in noninjected eyes. Eyes that had been injected with 0.1 mM FeSO₄ showed reduced cone density in the superior retina and less noticeable change in the inferior retina. Eyes injected with 0.25 mM FeSO₄ showed more severe reduction in cone density, particularly in the superior retina. In contrast, the ONL thickness showed little difference, suggesting that rod photoreceptors were much less affected.