Expression and Role of the Early-Response Gene Oxr1 in the Hyperoxia-Challenged Mouse Retina

Abstract

purpose. To examine the response of mouse retina to sustained hyperoxia. Hyperoxia is toxic to photoreceptors after sustained exposure (7–14 days in the C57BL/6J mouse) but has been reported to enhance photoreceptor function after short-term exposure.

methods. Retinas from the hyperoxia-vulnerable C57BL/6J mouse and from the hyperoxia-resistant BALB/cJ mouse were examined after 0, 3, 7, 14, and 35 days’ exposure to 75% oxygen. Quantitative PCR, TUNEL, and immunohistochemical techniques were used to trace the regulation and site of expression of the early-response, potentially protective gene Oxr1.

results. In the C57BL/6J retina, Oxr1 was upregulated at 3 days of exposure, matching the early period of resistance to hyperoxia in this strain, and fell below control levels at 14 days, when photoreceptor degeneration had begun. By contrast, the stress-related gene GFAP was upregulated only at 7 to 14 days. Immunohistochemistry showed a concentration of Oxr1 in the inner part of photoreceptor outer segments, but, as photoreceptors underwent apoptosis, Oxr1 concentrated in the nucleus, confirming earlier reports that photoreceptors were resistant to hyperoxia until 14 days in the BALB/cJ mouse and, correspondingly, that the upregulation of Oxr1 in outer segments was sustained until 14 days.

conclusions. The patterns of Oxr1 expression observed suggest that the gene is associated with resistance to hyperoxic challenge and that it acts at the level of the outer segment. The retinal response to hyperoxia may constitute acute and chronic phases in which photoreceptors are first resistant, and then vulnerable, to oxidative damage. Understanding this biphasic response may be important in understanding the role of oxygen in the progress of retinal dystrophy.

The normal adult retina is vulnerable to damage by environment-related stress, even in conditions that are environmentally normal. Examples of this normal-environment vulnerability include the hyperoxia-related erosion of the anterior edge of the retina, ^{1 2} the depletion of the photoreceptor layer associated with increases in daily ambient light, ³ the vulnerability of photoreceptors to light during the night phase of the normal daily cycle (Organisciak D, et al. IOVS 1998;39:ARVO Abstract 4571), ⁴ and the progressive loss of retinal sensitivity in the last decades of human life, associated with the loss of rods throughout life. ^{5 6} One factor contributing to this vulnerability is the absence of autoregulation of the choroidal circulation, which delivers oxygen to photoreceptors. Any factor that decreases the consumption of oxygen by photoreceptors, including light adaptation ⁷ or the depletion of the photoreceptor population by disease ^{8 9} or age, ¹⁰ causes a sustained rise in oxygen tension in the outer retina. Rises in retinal oxygen sustained over days and weeks are known to be specifically toxic to photoreceptors ^{11 12 13 14 15} and may contribute to the progressive degeneration of the retina seen in advanced age or in the later stages of retinal dystrophy. ¹⁰  

However, the effect of hyperoxia on the retina is not simply toxic. Earlier reports provide evidence that relatively short periods of hyperoxia (hours rather than days) can enhance the rodent electroretinogram (ERG). ¹⁶ In humans, hyperbaric oxygen therapy, which involves repeated, brief (2-hour) episodes of intense hyperoxia, can enhance the ERG and significantly slow retinal dystrophy. ^{17 18} We have, therefore, examined the response of the C57BL/6J mouse retina to hyperoxia, in which photoreceptor death begins with a delay of 7 to 14 days, ^{13 14 15} to test whether significant regulation of gene expression could be detected before the onset of photoreceptor death. That is, we sought to establish the basis for the acute enhancing and chronically toxic effects of hyperoxia on retinal function. 

We report that in the C57BL/6J mouse retina, the expression of the mouse homologue (C7) ¹⁹ of the human protective (oxidation-resistant) gene Oxr1 ²⁰ goes through a biphasic course, first upregulated when photoreceptors are resistant to hyperoxia and later downregulated as photoreceptors succumb. Oxr1 has been shown to be involved in free radical scavenging of hydrogen peroxide, with the protein located in the mitochondria of human cells, ²¹ and in the nucleoli ¹⁹ of human, mouse, and rat cell lines. It seems possible that Oxr1, acting as a free radical scavenger, plays a protective role in the early stages of hyperoxia-induced death of photoreceptors. 

Methods

Rearing Conditions

Animals were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Mice (C57BL/6J, BALB/cJ) were raised in dim (5 lux) cyclic illumination (12 hours light/12 hours dark). These levels of lighting were maintained during exposure to hyperoxia. Three-month-old adult mice (C57BL/6J, BALB/cJ) were placed in a clear acrylic plastic chamber and exposed to an atmosphere of 75% oxygen for 0 (control), 3, 7, 14, or 35 days. Oxygen levels were maintained using a controlled feedback device (Oxycycler; Biospherix, Lacona, NY). 

Tissue Preparation

Animals were killed by cervical dislocation, and the eyes were removed. For histologic processing, eyes were immediately immersion-fixed in 4% paraformaldehyde for 3 hours. For RNA extraction, eyes were placed on ice without fixation. 

Histologic Processing

After fixation, eyes were cryoprotected in 30% sucrose overnight. Eyes were embedded (Tissue-Tek; Sakura Finetek, Torrance, CA), snap frozen in liquid nitrogen, cryosectioned at 14 μm, mounted on gelatinized poly-L-lysine–coated glass slides, stored at −20°C, and used for immunohistochemistry and TUNEL procedures. 

RNA Extraction, Microarrays, and Real-Time PCR

Tissue from six retinas was pooled, and RNA was extracted using reagent (Trizol; Invitrogen, Carlsbad, CA) and RNA purification (RNAqueous-Micro kit; Ambion, Austin, TX). Genomic DNA was removed using DNase treatment according to the purification kit (RNAqueous-Micro kit; Ambion) protocol. Quality control of RNA was performed using the 2100 Bioanalyzer; Agilent, Palo Alto, CA), and only RNA with an RNA integrity number greater than 9.0 was used. Retinal mouse RNA was used as the template with the qRT-PCR system (Superscript III Platinum One-Step; Invitrogen) according to the standard protocol. A region of mouse Oxr1 was amplified using forward 5′-CAG GGA GTG GGA GGT AGT GT-3′ and reverse 5′-ATG GTT CTT GGT GGA AGG TG-3′ primers. Mouse GFAP was amplified using forward 5′-TCCTGGAACAGCAAAACAAG-3′ and reverse 5′-CAGCCTCAGGTTGGTTTCAT-3′ primers (The Harvard NeuroDiscovery Center: http://www.neurodiscovery.harvard.edu/). GAPDH was used as a control gene and was amplified using forward 5′-TGAAGCAGGCATCTGAGGG-3′ and reverse 5′-CGAAGGTGGAAGAGTGGGAG-3′ primers (RT-Database: http://medgen.ugent.be/rtprimerdb/). 

Real-time PCR was performed using a real-time system (Rotorgene 3000; Corbett Research, Sydney, Australia) and appropriate software (Rotor-Gene 6; Corbett Research). Analysis for discovery of fold change was performed using the Pfaffl method ²² with GAPDH primers as control primers and no oxygen treatment as control. Relative fold change was expressed as a percentage fold change relative to control (0 days) and normalized to a control gene (GAPDH). 

For miocroarray (GeneChip; Affymetrix, Santa Clara, CA) analysis, RNA was processed, labeled, and hybridized to a mouse genome 430 2.0 array (GeneChip; Affymetrix) according to the standard protocol. This work was performed at the John Curtin School of Medical Research (Biomolecular Resource Facility; http://brf.jcs.anu.edu.au/index.htm). Analysis was performed using GCOS V1.4, and sorting for robust changes was performed according to the Affymetrix guidelines (www.affymetrix.com). Comparisons were performed with RNA extracted from control age-matched animals (no hyperoxia) to animals exposed to hyperoxia for 3, 7, and 14 days. Experiments were performed with biological replicates. Only genes with a significance of P < 0.05 were considered in this study. 

Immunohistochemistry and TUNEL

Cryosections were labeled with a rabbit polyclonal antibody against mouse Oxr1, ¹⁹ kindly provided by Michael Volkert. ²¹ The Oxr1c antibody was diluted 1:200 in PBS, and labeling was visualized with a secondary antibody to rabbit IgG conjugated with Alexa 488 (Molecular Probes, Eugene, OR), or it was biotinylated and then bound to streptavidin-Alexa 488. Some sections were also labeled with a mouse monoclonal antibody to rod opsin (1:100, Rho4D2; a gift from Robert Molday, Vancouver, BC, Canada), or with a mouse monoclonal antibody to cytochrome oxidase IV subunit I (1:200; catalog no. MS404; Mitoscience, Eugene, OR). Labeling with these latter antibodies was visualized with a secondary antibody to mouse IgG conjugated to Alexa Fluor 594 (1:1000; Molecular Probes). Some sections were labeled with the TUNEL technique for apoptotic nuclei ²³ using published protocols, ^{24 25} with Alexa 594 (Molecular Probes) as the reporting chromophore. Sections were also labeled with bisbenzamide using published protocols ²⁵ (this labels nuclear DNA and is effective in showing general cytology). 

Optical Densitometry

Images for optical densitometry were collected using an inverted fluorescence microscope (Axiovert 200; Carl Zeiss, Thornwood, NY) and a digital camera (AxioCam MRc5; Carl Zeiss). During the acquisition of images from animals in different treatment groups, the gain, contrast, and color balance of the scanning system were held constant. Images were converted to gray scale and were analyzed for densitometry using ImageJ (developed by Wayne Rasband, National Institutes of Health, Bethesda, MD; available at http://rsb.info.nih.gov/ij/index.html). 

Results

Hyperoxia-Induced Photoreceptor Death in the C57BL/6J Retina

Hyperoxia caused photoreceptor-specific cell death in the C57BL/6J retina. For the first 7 days of exposure, no increase in photoreceptor death could be detected with the TUNEL technique (Figs. 1A 1B) , but the death of photoreceptors became evident by 7 days of exposure as TUNEL⁺ cells in the ONL (Fig. 1C) , concentrated in the retina inferior to the optic disc. By 14 days of exposure, photoreceptor death was evident as TUNEL⁺ cells in the ONL and as thinning of the ONL, confirming earlier reports. ^{13 14 15}  

After 14 days of exposure to hyperoxia, thinning of the ONL was detectable only in a region of the retina inferior and close to the optic disc (Fig. 2) . In mice kept for 35 days in hyperoxia, however, the thinning had spread both inferiorly, toward the inferior edge of the retina, and superiorly, near the optic disc. 

Expression of Oxr1, GFAP, and Mitochondria-Related Genes during Exposure to Hyperoxia

The expression of Oxr1 showed a biphasic pattern over the14-day period examined. Assessed by quantitative qPCR (Fig. 3) , the expression of Oxr1 mRNA was raised 80% over control levels at 3 days of exposure, fell back to near-control levels at 7 days of exposure, and fell to −20% of control levels at 14 days of exposure. As assessed by microarray (GeneChip; Affymetrix) analysis (Table 1) , the expression of Oxr1 was 3.2-fold above control levels at 3 days of exposure, fell to control levels at 7 days of exposure, and fell to −3.2-fold at 14 days. 

For comparison, we traced the expression of an established retinal stress gene with the same techniques. Assessed by qPCR, the expression of the stress-induced intermediate filament protein GFAP (expressed by astrocytes and, under stress, by Müller cells ^{26 27 28} ) was steady up to 3 days of exposure and rose 210% at 7 days and 620% at 14 days (Fig. 3) . Microarray (GeneChip; Affymetrix) analysis showed the same trend (Table 1) ; GFAP expression remained within ±1.5-fold of control levels to 3 days and then rose to 14.34-fold above control at 14 days. In addition, we traced the expression of genes for three mitochondrial proteins known to be involved in the oxidative stress response (Gpx1, Txn1, MnSOD) from the microarray data. The expression of these three genes was raised only in the later stages of exposure, at 14 days. The expression of GAPDH, which was used as a reference gene to normalize the qPCR data, remained close to control levels during the 14-day period of exposure. 

Immunohistochemistry for Oxr1

Immunolabeling of control C57BL/6J retina for Oxr1 (Fig. 4)showed a band of labeling near the junction of inner and outer segments. To identify the locus of labeling more closely, we examined material double-labeled for Oxr1 and rod opsin at higher resolution. At higher magnification (Figs. 5E 5F 5G) , rod opsin labeling and Oxr1 labeling colocalized at the inner end of the outer segment. 

The more global labeling of the retina by the Oxr1 antibody is shown in Figures 5A 5B 5C 5D . Each of these images shows a section of a C57BL/6J retina from the optic disc (at right) to the inferior edge of the retina (at left). The intensity of labeling rose from control levels to 3-day exposure and then fell toward control levels. At each of the three sites (near the anterior edge, midway between this edge and the optic disc, and near the disc) in each section, we measured the intensity (optical density) of Oxr1 labeling in the inner (ganglion cell and inner plexiform), middle (inner nuclear and outer plexiform), and outer (photoreceptor) layers of the retina. 

Overall (Fig. 6) , labeling tended to be higher in the outer photoreceptor layers than in the middle or inner layers. In all layers, however, Oxr1 labeling was raised at 3 days of exposure before falling back to or below control levels. Oxr1 protein levels in the superior retina followed the trends illustrated for the inferior retina in Figures 5 and 6 , rising at 3 days and then falling to or below control levels (data not shown). 

Localization of Oxr1 Labeling

The higher magnification images in Figures 5E 5F 5Gsuggest that the Oxr1 labeling colocalizes with the inner end of outer segments. To test this we examined the junction of inner and outer segments in retinas double-labeled for Oxr1 and rod opsin (Fig. 7A)and for Oxr1 and mitochondria-specific enzyme cytochrome oxidase (CO; Fig. 7B ). Oxr1 labeling was consistently located at the inner end of the outer segments, overlapping the rod opsin label. The density of labeling across these two micrographs is shown in Figures 7C and 7D . Peak labeling for Oxr1 (asterisks) colocalized with strong opsin labeling at the inner end of outer segments (Fig. 7C)but was external to peak labeling for CO in mitochondria (Fig. 7D) . Thus, peak Oxr1 labeling localized to the inner end of the outer segment. Oxr1 and opsin labeling both extended into the region of the inner segment, declining with distance from the inner ends of the outer segment. 

Late Shift in Oxr1 Labeling

A noticeable change occurred in the structures labeled by the Oxr1 antibody as the degeneration of photoreceptors proceeded. In the vulnerable region of inferior retina, at 7 days and, more markedly, at 14 days, Oxr1 labeling shifted from outer segments, which were by then shortened and disorganized, to nuclei of the ONL (Fig. 5H) , where Oxr1and TUNEL labeling colocalized (Fig. 8) . TUNEL⁺ labeling of nuclei consistently colocalized with Oxr1 labeling; some Oxr1⁺ nuclei were not, however, TUNEL⁺, suggesting that Oxr1 may translocate to the nuclei shortly before the nuclear DNA begins to fragment. This shift in labeling was observed only in the ONL, thus only in photoreceptors, and was observed only where photoreceptor nuclei were becoming TUNEL⁺ (i.e., where they were dying). 

Oxr1 in the Hyperoxia-Challenged BALB/cJ Mouse Retina

Confirming a previous report, ¹³ hyperoxia for 14 days did not cause detectable photoreceptor death in the BALB/cJ retina, assessed by the thickness of the ONL (Fig. 9)or by TUNEL labeling (data not shown). Nevertheless, the expression of Oxr1 protein in the outer segment increased from control to day 7 and then fell toward control levels (Fig. 9) . Even at 14 days, there was no evidence of Oxr1 concentrating in nuclei of the ONL, as observed in the C57BL/6J mouse. 

Discussion

Summary of Findings

The present results address the apparent discrepancy between facilitatory and toxic effects of hyperoxia on the retina. They show that in the C57BL/6J mouse retina, the expression of the early response protein Oxr1 is biphasic during the 14-day period of hyperoxia examined. Oxr1 is upregulated at 3 days, when photoreceptors are resistant to hyperoxia, and are downregulated by 14 days, when the photoreceptors have begun to die. Oxr1 was one of a number of genes to show this time course, in a microarray study of the C57BL/6J retina (preliminary report by Natoli and Stone ²⁹ ). By contrast, numerous other genes, exemplified here by GFAP, Gpx1, Txn1, and MnSOD, showed a delayed, monotonic regulation, with the change in expression coinciding with the onset of photoreceptor death. Quantitative PCR confirmed the time course of Oxr1 mRNA expression detected by microarray, and immunohistochemistry showed that the expression of the protein also followed this time course. 

Immunohistochemistry also showed that Oxr1 protein is strongly expressed by photoreceptors in a band at the inner part of the outer segment. These correlations suggest that Oxr1 may play a role in the initial resistance of C57BL/6J of photoreceptors to hyperoxic stress. Oxr1 expression was not confined to photoreceptors. Expression in the middle and inner layers of retina was also noted and also varied during the 14-day period of hyperoxia exposure examined, following the same pattern as in photoreceptors. The expression of Oxr1 was similar in retina superior and inferior to the optic disc. Photoreceptor vulnerability was greatest in the part of the retina close to and inferior to the disc, although photoreceptor death spread to the superior retina if the exposure to hyperoxia persisted. The source of this localized vulnerability to hyperoxia is unknown. 

The shortest period of hyperoxia examined in this study was 3 days. The possibility that significant regulation of Oxr1 may occur within this period deserves further study, given evidence of the significant upregulation of Oxr1 in cultured cells, within hours of the start of hyperoxia. ²¹  

Cellular Location of Oxr1

Previous studies have identified Oxr1 as localized to nucleoli in mouse tissue ¹⁹ and to mitochondria in yeast. ^{20 21} Elliott and Volkert ²¹ stressed that the anti-oxidant activity for which they screened in identifying the gene required Oxr1 to be located in mitochondria. Therefore, we anticipated that the protein Oxr1 would be located principally in mitochondria, and immunolabeling localized intense labeling to the junction of inner and outer segments, close to where mitochondria concentrate in the outer part of the inner segment. At higher power, however, peak immunolabeling for Oxr1 appeared localized to the inner part of the outer segments, which does not contain mitochondria, and to be separate from mitochondria in the inner segment. Further, we could not locate Oxr1 in the mitochondria prominent at the other end of the photoreceptor, in the axon terminal. ³⁰ Our results do not, however, preclude the presence of lower levels of Oxr1 in the mitochondria inner segment. 

Late in the process of photoreceptor degeneration in our model, Oxr1 was differently located, in the nuclei of photoreceptors in the ONL. This nuclear labeling occurred only in regions of retina in which photoreceptors were dying; in many instances, Oxr1 and TUNEL labeling colocalized. In such areas, Oxr1 antibodies labeled all TUNEL⁺ nuclei and additional nuclei that were not TUNEL⁺. This suggests that the nuclei may become Oxr1⁺ before their nuclear DNA fragments, attracting the TUNEL label. This nuclear location of Oxr1 in the nuclei of cells undergoing apoptosis has not previously been noted, and its functional significance is unknown. 

This late movement of Oxr1 to the nuclei of dying cells indicates that the protein can move within the cell. The observation ²¹ that Oxr1 is localized to mitochondria was made in cultured cells after short exposures (hours); it is possible that localization changed by the third time point, which we monitored. 

Early and Late Phases in the Response of the Retina to Hyperoxia

Vingolo et al. ¹⁸ have recently summarized a long-term trial of hyperbaric oxygen therapy for retinal dystrophy. Their summary indicates, confirming earlier suggestions, ^{17 18 31} that repeated brief periods of intense hyperoxia are not toxic to the retina and, by contrast, result in significant therapeutic benefit. Acceptance of empirically developed therapies is often and reasonably delayed until a mechanism for their effects and the limiting conditions for their use are defined. This study provides novel evidence that the retinal response to hyperoxia is biphasic. In the acute stage, the upregulation of genes such as Oxr1 is in phase with photoreceptor resistance to hyperoxia. In the later, chronic stage, early-response genes such as Oxr1 are downregulated, other stress-related genes are expressed, and photoreceptor death begins. 

In the C57BL/6J mouse the acute phase lasts less than 1 week. Data at 7 days appear to be transitional; by 14 days, the pattern of gene expression has changed and photoreceptor death has begun. The difference between C57BL/6J and BALB/cJ mice in the vulnerability of photoreceptors to hyperoxia ¹³ could lie in the dynamics of expression of genes such as Oxr1. In BALB/cJ mice, resistance of photoreceptors continues until at least 14 days of exposure. The expression of Oxr1 is also more persistent, showing a maximum at 7 days rather than at 3 days, and is still upregulated at 14 days. 

Further work is needed to better define these acute and chronic stages for different species, such as humans and mice, and to test, for example, whether the effects of repeated 2-hour episodes of intense hyperoxia used in hyperbaric oxygen therapy can be explained in these terms. The present suggestion of acute and chronic phases in the retina’s response to hyperoxia gives the first outline of how these different effects might be understood rather than dismissed as contradictory. 

 Supported by grants from Retina Australia, the National Health and Medical Research Council of Australia, and the Australian Research Council.

 Submitted for publication January 10, 2008; revised April 8 and May 21, 2008; accepted August 14, 2008.

 Disclosure: R. Natoli, None; J. Provis, None; K. Valter, None; J. Stone, 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.

 

Sections of oxygen-exposed retina showing the INL and ONL labeled for normal DNA (blue) and for DNA fragmenting in apoptotic nuclei (red, TUNEL). TUNEL labeling becomes prominent at 7 days (C), and the ONL is significantly thinned by 14 days (D). (A–D) Inferior midperiphery of the retina.

Figure 2.

 

Thickness of the ONL, relative to the INL, over the experimental period. Thinning of the ONL is apparent at 14 days inferior and close to the optic disc. In retinas exposed for an additional 3 weeks (35 days total), the thinning had extended toward the inferior edge of the retina and into the superior retina. Error bars denote SEM.

Figure 3.

 

Variation in expression of Oxr1 and GFAP, assessed by quantitative PCR, over the 14-day period of exposure used. Error bars denote SEM.

Table 1.

 

Fold Change of Oxrl, GFAP, and Mitochondrial Proteins (Gpxl, Txn1, and SOD1) during Hyperoxic Exposure

Gene	3 Days	7 Days	14 Days
Oxr1	3.3	NC	−2.6
GFAP	NC	3.06	14.34
Gpx1	NC	NC	2.1
Txn1	NC	1.5	1.9
MnSOD	−1.86	NC	1.7
GAPDH	NC	NC	NC

 NC, no change to baseline levels (fold change, <1.5 or >−1.5).

Figure 4.

 

Immunolabeling of Oxr1 (green) and rod opsin (red) in control C57BL/6J retina. The section was counterstained with bisbenzamide (blue), which labels nuclear DNA. Oxr1 labeling is maximal between the inner segment (IS) and outer segment (OS) layers.

Figure 5.

 

Immunolabeling for Oxr1 (green) in the hyperoxia-challenged mouse retina. (A–D) Sections from the optic disc region (right), to the edge of the retina, across the retina inferior to the optic disc. Labeling for Oxr1 increased from 0-day (control) to 3-day exposure to hyperoxia and then decreased to 7 days and 14 days. Regions were sampled for optical densitometry from the periphery (approximately 0.5 mm from the edge), midperiphery (halfway between the disc and the edge), and near (approximately 0.5 mm from) the optic disc. (E–H) Regions inferior to but near (approximately 400 μm from) the optic disc, at higher power (40×), with rod opsin also immunolabeled (red). Note the shift in Oxr1 labeling from the inner part of the outer segments (E–G) to cell bodies in the ONL (H).

Figure 6.

 

Optical densitometry of Oxr1 immunolabeling of the sections in Figure 5 . Measurements were made in each section at the three locations shown in Figure 5 , over the 14-day period of exposure. At each location and time, the density of Oxr1 labeling was measured over the ganglion cell and inner plexiform layers (inner layers, A); over the inner nuclear and outer plexiform layers (middle layers, B), and over the outer nuclear layer, inner segment, and outer segment layer (outer layers, C). Expression was highest in the outer photoreceptor layers. At all sites, Oxr1 labeling density peaked at 3 days’ exposure.

Figure 7.

 

Subcellular localization of Oxr1. Arrowheads: inner segment (IS) and outer segment (OS). The outer nuclear layer lies to the right. (A) Region of the junction of OS (left) and IS (right), from normal mouse retina. Labeling for rod opsin (red) shows the extent of the OS. Green labeling shows Oxr1 protein. (B) The same region labeled for Oxr1 (green) and the mitochondrial enzyme CO; mitochondria concentrate in the outer part of the inner segment. Peak Oxr1 labeling is external to CO labeling. (C, D) Optical densitometry of the images in (A) and (B). The peaks of labeling coincide in (C, asterisk) and are separate in (B, asterisks), with the CO peak lying internal to the Oxr1 peak.

Figure 8.

 

TUNEL and Oxr1 labeling in the vulnerable inferior retina, at 14 days’ exposure to hyperoxia. (A) Many nuclei in the ONL were TUNEL⁺ (red). (B) Many nuclei in the same region of the section were Oxr1⁺ (green). (C) When the images were superimposed, most of the Oxr1⁺ nuclei were also TUNEL⁺. Arrowheads: colocalized cell. Arrows: cell with Oxr1 nuclei labeling only.

Figure 9.

 

Labeling of DNA (blue) and immunolabeling of Oxr1 (green) in the inferior retina of BALB/cJ mice exposed to 75% oxygen for 0 days (A), 3 days (B), 7 days (C), and 14 days (D). gcl, ganglion cell layer; inl, inner nuclear layer; onl, outer nuclear layer; os, outer segment.