Downregulation of ATP Synthase Subunit-6, Cytochrome c Oxidase-III, and NADH Dehydrogenase-3 by Bright Cyclic Light in the Rat Retina

Abstract

purpose. Retinas of albino rats born and raised in bright cyclic light (300–800 lux) are less susceptible to light-induced apoptosis than retinas of animals born and raised in dim cyclic light (5 lux). In this study, the objective was to study the mechanisms of neuroprotection in the bright cyclic light–reared retina by identification of differentially expressed genes with differential display (DD)-PCR.

methods. Albino rats were born and raised in 5- or 400-lux cyclic light (12 hours on/off). At 6 to 8 weeks of age, animals were either killed to harvest retinas or exposed to 1700 lux illumination for 12 or 24 hours. The neuroprotection of 400-lux cyclic light rearing was evaluated by DNA fragmentation and quantitative histology. The differentially expressed candidate genes were identified by DD-PCR. Northern blot analysis was used to quantitate differential expression of selected genes. Differential expression of protein was determined by Western blot and enzyme activity analysis. Cellular localization of transcripts was determined by in situ hybridization.

results. DNA fragmentation and quantitative histology results indicated that 400-lux cyclic light rearing protected the retina from light-induced apoptosis compared with 5-lux cyclic light rearing. DD-PCR analysis showed that a 283-bp expressed sequence tag (EST) was downregulated in retinas of rats raised from birth in 400-lux cyclic light. A BLAST search identified the EST as the 3′-terminal sequence of mitochondria-encoded NADH dehydrogenase subunit 3 (ND-3). Northern blot analysis showed that the EST hybridized to two mRNA transcripts, the larger of which was confirmed to encompass the adenosine triphosphate (ATP) synthase subunit 6 (ATPase-6), cytochrome c oxidase subunit III (CO-III), ND-3, and tRNA-Gly. Northern blot analysis demonstrated that CO-III and ATPase-6 were downregulated 1.8- and 2.3-fold by 400-lux cyclic light compared with 5-lux cyclic light, respectively; however, there was no change in cytochrome c oxidase subunit I and II (CO-I and -II) or in 12S ribosomal RNA (12S rRNA), a mitochondrial housekeeping gene. Western blot analysis using anti-CO-III antibody showed more CO-III protein in retinal mitochondria from dim-light–raised rats. The enzyme activity of CO was two times higher in retinal homogenates from dim-light–raised rats than those from bright-light–raised rats. In situ hybridization using a ³⁵S-labeled CO-III probe showed that the CO-III transcript was present and downregulated in most of the retinal layers of bright-light–reared animals.

conclusions. Rearing in cyclic light at 400-lux downregulates the expression of ATPase-6, CO-III, and ND-3 compared with rearing in 5-lux cyclic light. The authors hypothesize that these changes are adaptive responses to light stress that provide neuroprotection to retinal cells by elevating the level of stress-related factors and reducing the level of oxidized cytochrome c, the form that activates the apoptotic cascade of cell death.

The retinas of albino rats that are raised in relatively bright cyclic light (300–800 lux) are resistant to constant light-induced apoptosis, whereas those from animals raised in the darkness or dim cyclic light (5 lux) can be severely damaged. ^{1 2 3 4 5 6} Adult rats raised in dim cyclic light rapidly acquire this neuroprotective capability when placed in bright cyclic light before constant light challenge, ^{7 8} whereas animals born and raised in bright cyclic light lose neuroprotection when returned to dim cyclic light. ⁶ These studies demonstrate that the retina has a remarkable ability to respond to environmental light cues by up- or downregulating neuroprotective mechanisms that modulate the severity of light-induced retinal apoptosis. Several stress-related biochemical alterations have been implicated in the retinal protection: (1) The components of the phototransduction pathway, rhodopsin and transducin, are decreased by bright light, and arrestin is increased ⁵ ; (2) antioxidants such as vitamins E and C and activity of three glutathione enzymes are elevated by bright light ⁴ ; (3) polyunsaturated fatty acid such as docosahexaenoic acid, which is a substrate of lipid peroxidation, is decreased. ⁸ However, the molecular mechanisms underlying this protection remain to be elucidated. 

To gain better understanding of the molecular mechanisms of the retinal adaptive neuroprotection caused by light-exposure history, we initiated studies to identify the differentially expressed genes or proteins in retinas of rats raised from birth in dim (5 lux) or bright (400 lux) cyclic light, using the technique of differential display (DD)-PCR. Herein, we report that three mitochondrial enzyme subunits, cytochrome c oxidase CO-III, adenosine triphosphatase (ATPase)-6, and ND-3, are downregulated by bright cyclic light in the rat retina. 

Materials and Methods

A Northern blot analysis kit (Northern Max), a nonisotopic detection kit (BrightStar BioDetect), and a psoralen-biotin nonisotopic labeling Kit (BrightStar Psoralen-Biotin Nonisotopic Labeling) were purchased from Ambion (Austin, TX). An RNA extraction kit (RNAimage) was from Genhunter (Nashville, TN). Extraction reagent (TRIzol), a vector (pCRII-TOPO), an RT-PCR system (Superscript First-Strand Synthesis System), single-strand RNA and protein standard, and DNA polymerase (Platinum Taq) were bought from Invitrogen (Carlsbad, CA). α-[³³P]dATP (2000 Ci/mmol) was from Perkin Elmer Life Sciences (Boston, MA). EDTA-free protease inhibitors were from Roche Molecular Biochemicals (Basel, Switzerland). Anti-CO-III mouse monoclonal antibody was from Molecular Probes (Eugene, OR). The bicinchoninic acid (BCA) protein assay reagent was from Pierce (Rockford, IL). Chemiluminescence (Enhanced Chemiluminescence; ECL) reagents were from Amersham Pharmacia Biotech (Buckinghamshire, UK). Positive-charged nylon and 0.45-μm nitrocellulose membrane were from Ambion and Bio-Rad (Hercules, CA), respectively. CO assay kit and all other analytical chemical reagents were from Sigma-Aldrich (St. Louis, MO). 

Animals

Albino Sprague-Dawley rats were born and raised in 5 lux or 400 lux cyclic light (12 hours on/off). Retinas were always harvested at 10:00 AM (3 hours into the light cycle) except for rats used in Figure 5C , which were harvested at 10:00 PM (3 hours into the dark cycle). At 6 to 8 weeks of age, rats were killed with an overdose of CO₂, and their retinas were dissected rapidly and frozen immediately in liquid nitrogen. The animal care strictly conformed to 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. 

Light Illumination and DNA Fragmentation

To produce light-induced retinal damage, albino rats (200–300 g) born and raised in 5 or 400 lux cyclic light were exposed to light of 1700 lux provided by two 40-W cool white fluorescent tube bulbs for 12 or 24 hours. During exposure, the rats had free access to food and water. After exposure, the animals were rested in the dark for 12 or 24 hours. DNA laddering experiments were performed as described previously by Chen et al. ⁹ with some modification. Briefly, retinas were homogenized in lysis buffer (50 mM Tris-HCl [pH 8.0]; 10 mM EDTA; 0.5% SDS; 0.5 mg/mL proteinase K). The resultant homogenates were extracted twice with phenol-chloroform. Any contaminating RNA was digested by incubating with 20 μg/mL of RNase A for 2 hours at 37°C. The concentration of genomic DNA was determined by measuring the absorbance at 260 nm, and 10 μg of DNA was run in 2% agarose gel containing 0.5 μg/mL ethidium bromide. 

Morphologic Evaluation of Photoreceptor Protection

According to our previously published procedures, ^{8 10} animals were killed by an overdose of CO₂. The eyes were enucleated, fixed, and embedded in paraffin, and 5-μm-thick sections were taken along the vertical meridian to allow comparison of all regions of the retina. Outer nuclear layer (ONL) thickness was measured in eight defined areas in each of the superior and inferior hemispheres. Measures were made at 480-μm intervals. Five eyes were measured for each light treatment. 

RNA Isolation

Retinas were dissected in ice-cold phosphate-buffered saline (PBS) under a microscope to remove potential contaminants such as vitreous humor, lens, and cornea. Retinal RNA was isolated by extraction reagent according to the manufacturer’s instructions (TRIzol; Invitrogen). The RNA was determined by measuring the absorption at 260 nm, and the integrity was assessed by the fluorescence intensity ratio of 28s/18s rRNA resolved in 1.2% denaturing-agarose gel. 

Differential Display Polymerase Chain Reaction

Differential display (DD)-PCR was performed as described by the instructions provided with the kit (RNAimage; GenHunter) and was theoretically capable of amplifing all expressed sequence tags (ESTs) of a cell or tissue by 240 combinations of three 1-base anchored oligo-dT primers (H-T₁₁A, C, G) and 80 random 13-mer primers (AP1-80). Briefly, 0.2 μg DNA-free total RNA was reverse transcribed by oligo-dT primer to produce the first-strand cDNA. One-tenth of the first-strand reaction mixture was used as a template for PCR, which was performed according to the following conditions: 94°C for 3 minutes; 40 cycles at 94°C for 30 seconds, 40°C for 2 minutes, and 72°C for 1 minute; and 72°C for 5 minutes. α-[³³P]dATP was incorporated into the PCR products, which were displayed on a 6% denatured sequencing gel and visualized by autoradiography. The differentially expressed ESTs in the gel were cut and recovered by boiling for 2 minutes in distilled water. The recovered ESTs were used as templates of reamplification under the same conditions as just described, with the exception that α-[³³P]dATP was omitted. The candidate differentially expressed ESTs were cloned into the vector (pCRII-TOPO; Invitrogen). 

Reverse Transcription–Polymerase Chain Reaction

The gene-specific cDNA fragments were amplified from rat retina by RT-PCR. The primers were designed based on the rat mitochondrial genomic sequence in GenBank (accession number X14848; http://www.ncbi.nlm.nih.gov/Genbank; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD). Their nucleotide sequences and the size of the products are shown in Table 1 . Five micrograms of DNA-free total retinal RNA were used to synthesize the first-strand cDNA by oligo-dT_12-18, which was conducted as described by the instructions for the first-strand synthesis system (Superscript; Invitrogen) for RT-PCR. PCR was performed in 50 μL of reaction mixture containing 200 μM dNTP, 200 nM of each primer, and 1 U of DNA Polymerase (Platinum Taq; Invitrogen), under the following conditions: 95°C for 2 minutes; 35 cycles at 95°C for 30 seconds, 58°C for 30 seconds, and 72°C for 45 seconds; and 72°C for 10 minutes. The cDNA fragments were cloned into the vector (pCRII-TOPO; Invitrogen). 

Northern Blot Analysis

Northern blot analysis performed (Northern Max; Ambion), using the procedures from the provider’s instructions with some modifications. Briefly, 15 μg of total RNA were run on a 1.2% formaldehyde-denaturing agarose gel and transferred to a positively charged nylon membrane. The RNAs were cross-linked to nylon membranes by ultraviolet light. The prepared RNA blots were prehybridized in hybridization solution (ULTRAhyb; Ambion) at 42°C for 1 hour and then hybridized with probe (8 ng/mL of hybridization solution) overnight. After posthybridization washing, the signal was detected with a nonisotopic dectection kit (BrightStar BioDetect; Ambion). The membrane was stripped, and then hybridized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or 18S ribosomal RNA probes (18S rRNA; gift from Wei Cao, Dean A. McGee Eye Institute, Oklahoma City, OK) for normalization of RNA loading and to 12S rRNA for quantification of mitochondria. We used a psoralen-biotin nonisotopic labeling kit (BrightStar Psoralen-Bioten; Ambion) to prepare nonisotopic probes of the EST, CO-III, ATPase-6, 12S rRNA, GAPDH, and 18S rRNA cDNA fragments. The purified PCR products were denatured at 100°C for 10 minutes and then combined with psoralen-biotin and irradiated at 365 nm for 45 minutes. 

Quantification of Gene Expression

Northern blot membranes were scanned, and the optical density was determined on computer (one-Dscan program; Scanalytics, CSP Inc., Billerica, MA). The quantification of gene expression was determined by the ratio of optical density between gene of interests and control gene GAPDH. 

In Situ Hybridization

In situ hybridization with ³⁵S-labeled CO-III probe was performed as described by Robinson et al. ¹¹ To generate antisense and sense transcriptions of CO-III in vitro, the recombinant vectors were linearized by BamHI or NotI, and then transcribed with T7 and SP6 RNA polymerase, respectively. After they were dewaxed, the paraffin-embedded retina sections were hybridized with ³⁵S-labeled riboprobe (10,000 cpm/retina section). To visualize the location of CO-III transcripts in the retina, silver grains in the dark-field images were colored red in image management software (Photoshop; Adobe Systems, Mountain View, CA) and overlaid onto the bright-field images. The intensity and number of silver grains indicate the relative amount and location of CO-III transcripts in the retina. 

Isolation of Mitochondria and Western Analysis

Retinal mitochondria were isolated as previously described by He et al. ¹² with slight modification. All reagents were precooled in ice and four retinas were homogenized in isolation buffer (20 mM Tris-HCl [pH 7.6]), 1 mM EDTA, 5 mM EGTA, and protease inhibitors). The homogenates were centrifuged at 1,000g for 10 minutes, and the supernatant was centrifuged at 11,400g for 1 hour to produce a mitochondria-enriched pellet. Western blot analysis was performed according to standard protocols. Mitochondrial proteins (20 μg per lane) were run on 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred onto a 0.45-μm nitrocellulose membrane, which was washed with TTBS buffer (20 mM Tris-HCl [pH 7.4]), 0.4 M NaCl, and 0.1% Tween). A monoclonal antibody raised in the mouse using yeast CO-III as the antigen (1:1000; Molecular Probes, Eugene, OR) and a goat anti-mouse secondary antibody (1:5000) were used to probe the membrane, which was then developed by chemiluminescence reagent. 

CO Activity in Retinal Homogenates

The CO activity of retinal homogenates was analyzed by a CO assay kit, using the procedures described in the manual. Briefly, two retinas from each rat were pooled and homogenized in isolation buffer (250 mM sucrose, 10 mM Tris-HCl [pH 7.0]), 1 mM EDTA, and protease inhibitors). Approximately 20 to 30 μg of retinal crude protein was added to 1.1 mL of reaction solution, which contained 50 μL of 0.22 mM ferrocytochrome c (fully reduced by sodium hydrosulfite), 10 mM Tris-HCl (pH 7.0), and 120 mM KCl. The decrease of absorbance at 550 nm was recorded during a 1-minute reaction time and used to determine the activity of CO by the formula (milliunits per milligram total protein) = (ΔA/min_sample − ΔA/min_blank) × 1.1/mg protein × 21.84. Retinal homogenates of six independent experiment groups from each light condition were measured in duplicate. The probability was determined by Student’s t-test (P < 0.05). 

Results

Retinal Neuroprotection of Bright Cyclic Light Rearing

Retinal neuroprotection provided by bright cyclic light rearing has been well documented in our laboratory. ^{4 6 8 10 13 14} In this study, we further evaluated the neuroprotective effect of 400-lux cyclic light compared to 5-lux cyclic light rearing. The experimental paradigm used in the present study is shown in Figure 1A . Two groups of rats were raised from birth in 5- or 400-lux cyclic light. At 6 to 8 weeks of age, animals were either used for experiments or exposed to 1700-lux constant light. As shown in Figure 1B , 400-lux cyclic light rearing prevented DNA fragmentation, a typical characteristic of apoptotic cell death. Figure 1B also indicates that illumination of 1700 lux for 12 hours did not cause DNA fragmentation in retinas from either group. It is likely that this level and length of illumination was below the threshold dose of light-induced retinal apoptosis in our animals. 

Morphologic evaluation of photoreceptor cell rescue was performed by quantitative histology. After exposure to fluorescent light (1700 lux) for 24 hours, animals were returned to their original light environments for an additional 5 days before morphologic analysis. Figure 2A shows that the ONL thickness along the vertical meridian of the eye of rats raised in 5-lux cyclic light was approximately 45 μm in the superior and inferior regions of the eye. Exposure to constant light caused a dramatic loss of photoreceptor cells, as evidenced by thinning of the ONL in both the superior and inferior hemispheres. However, significant resistance of photoreceptor cells to light-induced apoptosis was observed in the animals raised in 400-lux cyclic light (Fig. 2B) . 

Identification of the Differentially Expressed Sequence

We used DD-PCR of 13-mer random primers and 1-base anchored oligo-dT primers ^{15 16} to identify genes differentially expressed in retinas of rats raised in dim or bright cyclic light. As shown in Figure 3A , DD-PCR products of AP₁ and H-T₁₁A (their sequences are shown in Table 1 ) had a differentially expressed EST, which was downregulated by bright (400 lux) cyclic light. The EST was sequenced and found to have 283 nucleotides (Fig. 3C) and was subjected to BLAST homology query against GenBank. The BLAST results identified the 283-bp EST as the 3′-terminal sequence of mitochondrial ND-3 gene. Northern blot analysis was used to confirm the differential expression from DD-PCR, and showed that a small and a ∼1.9-kb transcript were hybridized to the EST (Fig. 3B) . The small signal, which was approximately 400 bp, was presumed to be the mature ND-3 gene. The ∼1.9-kb transcript appears to encompass ATPase-6, CO-III, ND-3, and tRNA-Gly between CO-III and ND-3, based on two findings: (1) the nucleotide sum of these four adjacent genes is ∼1.9 kb; and (2) the EST from DD-PCR indicated that the ND-3 gene was the end gene of the transcript. 

Confirmation of Components in the ∼1.9-kb Mitochondrial Transcript

To confirm that the ATPase-6, CO-III, and tRNA-Gly were encompassed in the ∼1.9-kb transcript, four gene fragments were amplified from rat retinas (Figs. 4A 4B 4C) . Two fragments were amplified from a single gene, 483-bp from ATPase-6 and 713-bp from CO-III. The other two fragments spanned more than one gene. The 492-bp spanned ATPase-6 and CO-III, and the 515-bp spanned CO-III, tRNA-Gly, and ND-3. The organization of the ∼1.9-kb mitochondrial transcript and the localization of the cloned cDNA fragments are shown in Figure 4D . The nonredundant nucleotide sequences derived from the overlaying four gene fragments and EST are 1872 (GenBank accession number AF504920). Among the known nucleotide sequences, 10 nucleotides varied compared with the rat mitochondrial genomic sequence in the GenBank (accession number X14848). We have not yet ruled out the possibility that the variation of 10 nucleotides resulted from RT-PCR or sequencing artifacts, because these respiratory enzymes are highly conserved. We did not clone the 5′-terminal nucleotides of this transcript. However, it is reasonable to assume that there is no other gene at the 5′-end, because the size on the Northern blot (∼1.9kb) is very near the calculated value of the four genes (1881 bp). 

Expression Analysis of CO-III and ATPase-6

Northern blot analysis indicated that CO-III and ATPase-6 hybridized to two distinct transcripts (Fig. 5) . The top ∼1.9-kb transcript was common for CO-III, ATPase-6, and EST, which implied that the three genes were also encompassed within it. The ∼800-bp transcript of CO-III and ATPase-6 are the mature protein-encoding mRNA species. ¹⁷ Quantitative analysis of the 0.8-kb transcript indicated the expression level of CO-III and ATPase-6 mRNA was downregulated 1.8- and 2.3-fold (GAPDH gene as internal reference), respectively, by bright cyclic light–rearing compared with dim cyclic light rearing (Figs. 5A 5B) . Figure 5C shows that the expression level of CO-III was not different at 10:00 AM (3 hours into light cycle) and 10:00 PM (3 hours into dark cycle) in dim-raised rats, implying that the decreased expression of CO-III in bright-raised rats is regulated by light intensity rather than time of day. 

To examine whether the downregulation of CO-III by bright cyclic light is due to the loss of mitochondria, we further conducted Northern blot analysis of the expression of a mitochondrial housekeeping gene 12S rRNA, which was not decreased by bright cyclic light rearing under conditions in which the expression of CO-III and ATPase-6 was decreased (Figs. 6A 6B) . These results demonstrate that the downregulation of CO-III and ATPase-6 by bright cyclic light was not due to the loss of mitochondria. Northern blot analysis results (Figs. 6C 6D) show the expression of CO-I and -II was not reduced at the level of mRNA by bright cyclic light rearing. 

Western Blot Analysis of CO-III and CO Activity in Retina

As CO is the last and key enzyme in the mitochondrial electron transport chain, we determined whether the retinal CO-III protein level and the CO enzyme activity were decreased in retinas of rats raised in bright light compared with those raised in dim light. Western blot analysis (Fig. 7 , top) showed that the expression of CO-III protein was downregulated by bright light. The 28-kDa protein corresponds to the complex of CO-II/III. ^{18 19} The CO enzyme activity in the retinas of dim-light–reared rats was 257 mU/mg total protein, compared with 127 mU/mg total protein in bright-light–reared rats (Fig. 7 , histogram). 

In Situ Localization of CO-III

We used in situ hybridization to locate CO-III transcripts in the retina. The signals from the antisense riboprobe were easily detected (the red color signal in Figs. 8A 8C ). In contrast, signals with the sense riboprobe of CO-III were relatively sparse and uniformly distributed (Figs. 8B 8D) . The CO-III transcripts appeared in most retina layers, including photoreceptor inner segments (IS), outer plexiform layer (OPL), inner nuclear layer (INL), inner plexiform layer (IPL), and ganglion cell layer (GCL). However, labeling of the photoreceptor outer segments (OS) and outer nuclear layer (ONL) was weak. The cellular distribution of CO-III transcripts in retina is consistent with the report of Chen et al. ²⁰ for rat and suggests the downregulation of CO-III transcript by bright cyclic light in most retinal layers (compare Figs. 8A , dim; 8C bright). However, because there was only a ∼twofold difference in the Northern blot quantification (Fig. 5) and enzyme activity (Fig. 7) between dim- and bright-light–raised rats, the in situ differences are not statistically significant. 

Discussion

We used a differential screen to identify several respiratory enzymes subunits that have reduced expression in the retinas of rats reared in bright cyclic light, which are resistant to acute-light–induced apoptotic retinal cell death. The downregulation is due to light levels, because there was no diurnal variation in the transcript levels in retinas of rats raised in dim cyclic light. Furthermore, the downregulation of expression of these genes by bright cyclic light was not due to the loss of mitochondria, as the expression of mitochondrial housekeeping gene 12S rRNA between two groups of animals. The ∼1.9-kb transcript that encodes four genes is essential for respiration, including ND-3, CO-III, and ATPase-6. These genes encode NADH dehydrogenase (complex I), CO (complex IV), and F₀F₁-ATPase, respectively. ²¹ CO (CO, EC1.9.3.1) is the terminal enzyme of the mitochondrial respiratory chain, catalyzing the transfer of electrons from reduced cytochrome c to molecular oxygen and is therefore a key enzyme for the production of ATP. In retina, CO has been localized by histochemistry and immunohistochemistry to all retinal layers where mitochondria are present. ^{20 22} By in situ hybridization (Fig. 8) , we found CO-III throughout the rat retina. Northern blot analysis revealed a reduced expression of the mitochondrial transcript in retina in response to bright- versus dim-light rearing. This is in agreement with the reduction in CO-III protein levels and CO enzyme activity. 

The three genes CO-III, ATPase-6 and ND-3 are linked closely in the mitochondrial genome, ¹⁷ and posttranscriptional processing is necessary to produce mature protein-encoding genes. ²³ The precursor mitochondrial transcript containing CO-III, ATPase-6, and ND-3 has been demonstrated in human tissue. ²⁴ In our study, CO III, ATPase-6, and ND-3 were shown to be contained in a precursor mRNA species and downregulated by 400-lux light rearing. However, expression of the two other subunits of CO, CO-I, and CO-II, which are distant from CO-III in the mitochondrial genome, was not changed by bright cyclic light rearing (Fig. 6) . Their expression at the level of protein remains to be elucidated. The expression of CO-III and ATPase-6 genes has been shown to change in response to environmental factors such as reactive oxygen species. ²⁵ In addition, the finding that the CO-III and its adjacent genes such as ND-3 and ATPase-6/8 are much more vulnerable to environmental effects than genes distant from CO-III such as ND-4, CO-I, and CO-II, have been reported by You et al. ²⁶ in vitro. 

The decreased expression of these three important mitochondrial genes implies reduced efficiency of electron transport, oxidative phosphorylation, and ATP/energy production in retinas of rats reared in bright cyclic light. Reduced mitochondrial efficiency would require the retina to have lower oxygen utilization under bright light. Consistent with this hypothesis, studies have previously shown that retina cells use less oxygen in light than in the dark. ^{27 28 29} Retina cell metabolism has substantial energy requirements to maintain the “dark current” and the high level of glutamate release at the synaptic terminals, both of which are more active in the dark than in the light. ³⁰  

Our study, in combination with previous studies, demonstrates a correlation between reduced metabolism in mitochondria and increased photoreceptor survival. The downregulation of mitochondrial efficiency by light could participate in this protection at several levels, including elevating the level of stress-related factors and reducing production of oxidized cytochrome c. We have shown that retinas in rats born and raised in bright cyclic light have higher levels of small antioxidant molecules (vitamins A and C) and greater activity of three antioxidant enzymes (glutathione-peroxidase, -reductase, and -S-transferase) compared with those raised in dim cyclic light. ^{4 6} Less active mitochondria would be likely to precondition the retina to light stress before acute-light–induced oxidative damage, which could upregulate retinal stress-related proteins and antioxidants and be an effective strategy against light-induced apoptosis. 

Mitochondria in bright cyclic light–reared rats have lower CO activity and therefore may be less likely to activate the apoptotic pathway through the release of oxidized cytochrome c. Release of cytochrome c from mitochondria is the key factor of mitochondria-mediated cell death, ³¹ and only the oxidized form is able to induce apoptosis. ^{32 33} As CO activity in bright cyclic light–reared animal is decreased, more cytochrome c is almost certainly in the reduced state, whereas the higher CO activity in dim cyclic light–reared animal is likely to maintain more cytochrome c in the oxidized state. This suggests that the retinal mitochondria in dim cyclic light contain higher steady state levels of oxidized cytochrome c and are primed to induce apoptosis in response to stress signals, whereas the mitochondria in bright cyclic light–reared animals contain lower levels of oxidized cytochrome c and are less capable of initiating cell death. 

Our model predicting that mitochondria in retinal cells sense oxidant stress due to bright cyclic light and, in response, reduce their oxidative metabolism by reducing gene expression is supported by some recent reports. Steroid hormone receptors have been localized to mitochondria, suggesting that mitochondrial gene expression may be regulated by retinoids and steroid hormones. ³⁴ Upregulation of endogenous steroids or injection of dexamethasone protects against light-induced apoptosis. ^{35 36} The hormone 17 β-estradiol was recently shown to attenuate significantly the oxidative neuronal death induced by 24-hour exposure to hydrogen peroxide. ³⁷ Although the mechanism for protection is not known, it is possible that estradiol could affect mitochondrial gene expression, given that its receptors are in the mitochondria as well as the nucleus. ^{38 39} Cytokines such as FGF2 have been shown to modulate CO activity in the mitochondria. ⁴⁰ Several studies have shown that bright light can induce the expression of cytokines, including ciliary neurotrophic factor (CNTF) and FGF2, ⁷ and that administration of these cytokines protects photoreceptors from cell death from inherited mutations ⁴¹ and constant bright light exposure. ^{42 43} The mechanisms by which these cytokines act are not well characterized. However, it is possible that they bind cell surface receptors and activate signaling mechanisms that lead to reduced expression of essential mitochondrial genes. 

In summary, the results of this study as well as those of previous studies suggest that the retinas of bright cyclic light–reared animals have at least two mechanisms of protection that involve mitochondria. First, there is a reduction in mitochondrial metabolism that is likely to precondition the retina to light stress before acute light-induced oxidative damage, which elevates the level of stress-related factors, including stress proteins and antioxidants. Second, the retinas are further protected from apoptosis by the reduction in the enzyme activity of CO, which is likely to maintain cytochrome c in the reduced state. 

 Supported by National Institute of Health Grants EY00871, EY04149, EY06973, EY12190, and RR17703; Research to Prevent Blindness, Inc., New York, New York; The Foundation Fighting Blindness, Baltimore, Maryland; Oklahoma Center for the Advancement of Science and Technology; Samuel Roberts Nobel Foundation, Inc., Ardmore, Oklahoma; and Presbyterian Health Foundation, Oklahoma City, Oklahoma.

 Submitted for publication September 30, 2003; revised December 17, 2003 and March 25, 2004; accepted April 17, 2004.

 Disclosure: H. Huang, None; F. Li, None; R.A. Alvarez, None; J.D. Ash, 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.

 

Table 1.

 

Nucleotide Sequence of Primers and the Size of their Products

Gene	Forward Primer	Reverse Primer	Products (bp)
CO-I	CCCATTTCAACTTGGCTTACA	TGAGCCGCAAATTTCAGAG	681
CO-II	AGCCGGGGTGTCTTCTATCT	CAAAGTGGGCTTTTGCTCAT	598
CO-III	CCACCAAACCCATGCATACC	CAGTATCATGCTGCGGCTTC	713
ATPase-6	CCTCTTTCATTACCCCCACA	GGCCTGCTGTAATGTTTGCT	483
ND-3 or EST	AAGCTTGATTGCC (API)	AAGCTTTTTTTTTTTA (H-T11A)	283
ATPase-6 to CO-III	TTCAACCGATAGCACTAGCAGTACG	ACAATAAACAGGATTATTCCGTATCGG	492
CO-III to ND-3	TTCTATCTCAGACGGAATTTACGGC	GAAGTAGTAAGGCGATTTCTAGGTCGA	515
12s rRNA	AGGTTTGGTCCTGGCCTT	GGCGGTGTGTGCGTACTTCATT	840

 Orientation: 5′→3′.

Figure 1.

 

Retinal neuroprotection by bright cyclic light rearing. (A) Schematic diagram of rat rearing paradigm. Two groups of rats were raised in bright or dim cyclic light (12 on/off). Some animals were killed to harvest retinas and others were exposed to 1700 lux for 12 or 24 hours. (▵) Time point for DD-PCR experiments before constant light illumination; (▴) time point for DNA laddering after light illumination and resting in the dark for 24 hours. (B) DNA laddering. The light history and illumination were shown at the top of gel image. Lane M: 100 bp DNA marker.

Figure 2.

 

Measurements of outer nuclear layer thickness. Rat retinal ONL thickness was measured along the vertical meridian, and the results are expressed as mean ONL thickness ± SD (n = 5) at 480-μm intervals. (A) Animals raised in dim (5 lux) cyclic light. (B) Animals raised in 400 lux cyclic light. Rats were either exposed or not exposed to constant light for 24 hours at illumination of 1700 lux.

Figure 3.

 

Identification of differentially EST. (A) Differential display (DD)-PCR. The DD-PCR products were amplified with primer AP₁ and H-T₁₁(A). α-[³³P]dATP was incorporated into the DD-PCR products, which were displayed in 6% denaturing sequencing gel, and then exposed to x-ray film overnight at −80°C. The differentially EST was recovered from gel and used as the template for reamplification. Arrowhead: differentially EST in (D) dim and (B) bright-raised rats. (B) Northern blot analysis. The differentially EST from DD-PCR was amplified by PCR and the products were labeled by psoralen-biotin; 80 ng of labeled probe was used to hybridize to the total RNA of retina from rats born and raised in dim or bright cyclic light. (C) The nucleotides of EST. The EST was cloned into the PCRII-TOPO vector and sequenced by the Sequence Facility of Oklahoma Medical Research Foundation. The two primers for DD-PCR, AP₁ and H-T₁₁(A), are in bold type.

Figure 4.

 

Cloning of cDNA fragments in the ∼1.9-kb mitochondrial transcript. (A) Lane 1: 713-bp CO-III fragments. The two primers are located within the CO-III. (B) Lane 2: the 483-bp ATPase-6 fragments. The two primers are located within the ATPase-6. (C) The two fragments span more than one gene. Lane 3: the 492-bp fragment spans ATPase-6 and CO-III. The forward primer is located within ATPase-6; the reverse primer is located within CO-III. Lane 4: the 515-bp fragment spans CO-III, tRNA-Gly, and EST (or ND-3). The forward primer is located within CO-III; the reverse primer is located within EST (or ND-3). Lane M: 100-bp DNA ladder. (D) Schematic diagram of the ∼1.9-kb mitochondrial transcript. Thick line: shows the four genes in the transcript: ATPase-6, CO-III, tRNA-Gly, and ND-3. Arrow: direction (5′→3′). The five thin lines represent the cloned gene fragments, respectively. 1, 483-bp ATPase-6 fragment; 2, 713-bp CO-III fragment; 3, 283-bp EST; 4, 492-bp fragment spanning ATPase-6 and CO-III; and 5, 515-bp fragment spanning CO-III, tRNA-Gly, and EST (or ND-3).

Figure 5.

 

Quantification of CO-III and ATPase-6 expression. The top panels are representative Northern blot results using CO-III probe (A, C) or ATPase-6 (B). The Northern blot analysis were stripped and then hybridized to GAPDH probe for RNA normalization and gene relative expression analysis. The histograms show the quantification of gene expression from three independent experiments. (A, B) Lane D: dim cyclic light; lane B, bright cyclic light. (C) The retina of rats raised in dim cyclic light environment, but harvested at the different times of the light cycle. Lane 1: the retinas were harvested at 10:00 AM (3 hours into the light cycle); lane 2: the retinas were harvested at 10:00 PM (3 hours into the dark cycle).

Figure 6.

 

Northern blot analysis of mitochondrial enzymes. (A) Northern results. The same membrane was stripped and hybridized with the CO-III, ATPase-6, 12S rRNA, and 18S rRNA probes, respectively. The ∼0.8-kb bands were shown in the blots from CO-III and ATPase-6 probe. (B) The quantification of Northern blot analysis shown in (A). (C) Northern results. The same membrane was stripped and hybridized with the three CO-I, CO-II, and 12S rRNA probes, respectively. (D) The quantification of Northern blot analysis shown in (C). 12S rRNA was used for mitochondrial control gene to measure the level of mitochondria. 18S rRNA was used to quantify RNA gel loading and transfer efficiency. The error bars were based on two duplicate Northern blot analysis. The genes for each probe are shown on the right side. Left lane: dim cyclic light; right lane: bright cyclic light.

Figure 7.

 

Western blot analysis of CO-III and CO enzyme activity in retinal homogenates. Top: Western blot analysis of 20 μg of mitochondrial protein run on 10% SDS-PAGE and immunoblotted with the anti-CO-III monoclonal antibody (1:1000). Lane D: dim cyclic light; lane B: bright cyclic light; Histogram: CO enzyme activity of retinal homogenates. The values are milliunits per milligram total protein; 1 U is defined as the oxidation of 1.0 micromole of ferrocytochrome c per minute at pH 7.0 at 25°C. The results are expressed as mean ± SD (n = 6 for each group). *P < 0.001 versus retina raised in bright cyclic light.

Figure 8.

 

In situ hybridization of CO-III in retina. (A, B) Central retinal sections from dim cyclic light raised rats; (C, D) central retinal sections from bright cyclic light raised rats. (A, C) Antisense riboprobe; (B, D) sense riboprobe. The riboprobe was specific to the CO-III sequence. Each retina section was hybridized to 10,000 cpm ³⁵S-labeled riboprobes at 65°C overnight. The red signals were derived from silver grains in the dark-field images, and then overlaid onto the bright-field images. The red signals in (A) and (C) demonstrate the relative amount and localization of CO-III. IS, photoreceptor inner segment; ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer.

The authors thank Mark Dittmar for excellent care and husbandry of the animals used in this study and Donald Fox for critical evaluation of the manuscript and helpful suggestions.