IOVS
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kortuem, K.
Right arrow Articles by Levin, L. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kortuem, K.
Right arrow Articles by Levin, L. A.
(Investigative Ophthalmology and Visual Science. 2000;41:3176-3182.)
© 2000 by The Association for Research in Vision and Ophthalmology, Inc.

Differential Susceptibility of Retinal Ganglion Cells to Reactive Oxygen Species

Kim Kortuem, Lynette K. Geiger and Leonard A. Levin

From the University of Wisconsin Medical School, Department of Ophthalmology and Visual Sciences, Madison.


   Abstract
Top
 Abstract
Introduction
 Methods
 Results
 Discussion
 References
 
PURPOSE. Retinal light exposure is a source of oxidative stress, and retinal cells contain molecules that scavenge or inactivate reactive oxygen species (ROS). Yet, ROS also play a role in signal transduction, and some retinal cells (e.g., neurotrophin-dependent retinal ganglion cells, RGCs) may use ROS as part of the signaling process for cell death. RGCs might therefore have specialized mechanisms for regulating ROS levels. The hypothesis that RGCs might regulate ROS differently from other retinal cells was tested by studying their differential response to oxidative stress in vitro.

METHODS. RGCs were retrogradely labeled by injecting the fluorescent tracer DiI into the superior colliculi of postnatal day 2 through 4 Long–Evans rats. At postnatal days 7 through 9 the retinas were dissociated with papain and cultured with and without specific ROS-generating systems and/or scavengers. RGCs were identified by their DiI positivity using rhodamine filters. Living cells, determined by metabolism of calcein–AM viewed with fluorescein filters, were counted in triplicate. Degenerate reverse transcription–polymerase chain reaction (RT–PCR) using primers specific to peroxidase homology regions was used to survey for novel peroxidases expressed within normal retinas.

RESULTS. Compared with other retinal cells, RGCs were remarkably resistant to cell death induced by superoxide anion, hydrogen peroxide, or hydroxyl radical. Catalase counteracted the effect of each ROS-generating system on retinal cells, consistent with damage occurring via a hydrogen peroxide intermediate. Aminotriazole, L-buthionine sulfoximine, and sodium azide partly abrogated the RGC resistance to oxidative stress, suggesting that this resistance may be mediated by catalase and/or glutathione peroxidase. A limited expression survey within the retina using degenerate RT–PCR did not demonstrate novel peroxidases.

CONCLUSIONS. These data suggest a role for one or more endogenous peroxidases within RGCs, which could possibly be protective under conditions of axonal damage. Exploration of the unique characteristics of RGC resistance and susceptibility to injury may help in better understanding the pathophysiology of diseases associated with primary axonal damage.


   Introduction
Top
 Abstract
 Introduction
Methods
 Results
 Discussion
 References
 
Reactive oxygen species (ROS) are ubiquitous molecules involved in a variety of cell processes, including signal transduction,1 2 defense against infective organisms,3 regulation of gene expression,4 and the signaling of cell death.5 ROS are therefore generated under a variety of physiological and pathologic conditions but can also be by-products of the inherent "leakiness" of the mitochondrial electron transport system,6 particularly superoxide anion (O2-).

Within the nervous system and under specialized situations, ROS are able to transduce signals leading to cell death. For example, there is a burst of superoxide anion when sympathetic neurons are deprived of nerve growth factor.7 Similar signaling of neuronal cell death by specific ROS has subsequently been demonstrated in central nervous system neurons, including hippocampal8 neurons and cerebellar granule cells.9 Therefore, aberrantly elevated ROS levels could interfere with normal cellular physiology, and, hence, multiple mechanisms exist for regulating their levels. These include the superoxide dismutases, reduced glutathione, catalase, thioredoxin peroxidase, and glutathione peroxidase. Yet, signal transduction and other physiological processes that rely on ROS must necessarily coexist with cellular defenses against ROS. It would therefore stand to reason that there could be differences between cell types in the nature of the defenses against various ROS, corresponding to the differences in requirements for ROS involved in cellular function.

Unlike most parts of the nervous system, the retina is unusual in that it is vulnerable to potentially high levels of oxidative stress as a result of light exposure.10 11 It is not surprising that certain pathologic conditions have been hypothesized to be due to excessive oxidative damage, for example age-related macular degeneration.12 Retinal cells contain multiple ROS scavengers,13 14 15 presumably to protect them from oxidative stress. Yet, some retinal cells (e.g., neurotrophin-dependent developing retinal ganglion cells, RGCs) presumably require ROS as part of the signaling process for cell death, analogous to that seen in other neurotrophin-dependent neurons.7 16 To accomplish these two conflicting goals, RGCs could have specialized mechanisms for handling ROS. We tested this hypothesis by studying the differential response of RGCs to oxidative stress in vitro.


   Methods
Top
 Abstract
 Introduction
 Methods
Results
 Discussion
 References
 
Animals
All experiments were performed in accordance with ARVO, institutional, federal, and state guidelines regarding animal research.

Materials
Cell culture reagents were obtained from GIBCO (Grand Island, NY). The retrograde fluorescent tracers 1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine (DiIC18) and 1,1'-dihexdecyl-3,3,3',3'-tetramethylindocarbocyanine (DiIC16), and the fluorescent viability agent calcein-AM were obtained from Molecular Probes (Eugene, OR). Papain was obtained from Worthington Biochemical (Freehold, NJ). Unless noted, all other reagents were obtained from Sigma (St. Louis, MO).

RGC Labeling and Culture
RGCs were labeled and cultured using previously described methods.17 Briefly, ganglion cells were retrogradely labeled by stereotactic injection of the fluorescent tracer DiI dissolved in dimethylformamide into the superior colliculi of anesthetized postnatal day 2 through 4 Long–Evans rats. DiIC18 was used for most experiments, and DiIC16 was also used for experiments studying the effects of the tracer itself. At postnatal days 7 through 9 the animals were killed by decapitation, the eyes enucleated, and the retinas dissected free in Hanks’ balanced salt solution (HBSS). After two incubations in HBSS containing papain (12.5 U/ml), each for 30 minutes at 37°, the retinas were gently triturated with a Pasteur pipette and plated on poly-L-lysine–coated 96-well flat-bottomed tissue culture plates (0.32 cm2 surface area/well) at a density of approximately 2000 cells/mm2. The cells were cultured for 24 hours in Eagle’s minimal essential medium (MEM) with methylcellulose (0.7%), glutamine (2 mM), gentamicin (1 µg/ml), glucose (22.5 mM final concentration), and prescreened fetal calf serum (5%). In some experiments the defined serum supplement B2718 (GIBCO) was substituted for fetal calf serum.

Ganglion Cell Identification and Counting
RGCs were identified by the presence of retrogradely transported cytoplasmic DiI, which appears reddish orange when viewed with rhodamine filters under epifluorescence. Cell viability was determined by metabolism of calcein–AM, producing green fluorescence when viewed with fluorescein filters. Briefly, cells were incubated in a 1 µM solution of calcein-AM in phosphate-buffered saline (PBS) for 20 minutes, after which the medium was replaced with fresh PBS. Survival of RGCs was determined by identifying the percentage of DiI+ cells that were also calcein+ in 5 low-power fields. Survival of non-RGCs was determined by identifying the percentage of phase-visible cells that were also calcein+ in 5 high-power fields. Although using phase-positivity to identify non-RGCs would also include some cells that were DiI+, the percentage of RGCs in the retina is so low (approximately 1%) that this would not significantly affect our counts. Wells were counted in triplicate. Results are expressed as mean ± SEM, based on counts of all fields per condition.

Treatments
Standard biochemical systems were used to generate ROS. Concentrations were chosen on the basis of initial dose response experiments, to give 10% to 25% survival after treatment. Menadione was used to generate intracellular O2- by redox cycling, whereas xanthine/xanthine oxidase was used to generate extracellular O2-. Hydrogen peroxide was prepared as dilutions from a 30% stock. To generate hydroxyl radical via the Fenton reaction, a combination of copper(II) sulfate and 1,10 phenanthroline with 450 µM ascorbic acid, previously neutralized to pH 7.4, was used.19 Although Fe is the predominant metal leading to the generation of OH· via the Fenton reaction, the copper system was chosen because it had been well standardized. ROS scavengers or peroxidase inhibitors were used at concentrations described in the Results section and were added to retinal cultures simultaneous with the ROS-generating systems.

Degenerate Primer Design
Two sets of degenerate primers were designed from known peroxidase sequence motifs. The first set of primers was designed using catalase sequences and the second set using sequences common to catalase and thioredoxin peroxidase. Primers were obtained from Integrated DNA Technologies (Coralville, IA).

The following genes were used in designing the primers (GenBank Accession Nos. are in parentheses): rat liver catalase (M11670), human catalase (E01497), maize catalase-1 (X12538), maize catalase-2 (X54819), maize catalase-3 (X12539), mouse thioredoxin peroxidase (U51679), and human thioredoxin peroxidase (U25182). Sequences were aligned with ClustalW 1.7 and regions of homology identified with BoxShade.

The first set of degenerate primers was designed from an alignment of rat liver catalase, human catalase, maize catalase-1, maize catalase-2, and maize catalase-3. Two motifs were used, for forward and reverse primers, respectively. The forward primer was the 25-bp sequence 5'–GYGGKTTYGCHGTSAARTTYTACAC–3' and was 192-fold degenerate. The reverse primer was the 20-bp sequence 5'–CKSSHCTGVAGCAKYTTRTC–3' and was 576-fold degenerate.

The second set of primers was designed from an alignment of catalase and thioredoxin peroxidase gene sequences. These primers were chosen from sequences of rat liver catalase, human catalase, mouse thioredoxin peroxidase, and human thioredoxin peroxidase. The forward primer was the 22-bp sequence 5'–GTYYYCWTYYTYTAYCCAYWKS–3' and was 4096-fold degenerate. The reverse primer was the 20-bp sequence 5'–KTYAMKRCCRGKYTDCCARC–3' and was 1536-fold degenerate.

Degenerate RT–PCR
An adult, Long–Evans female rat was killed by exposure to CO2, and the retinas from both eyes were dissected. RNA was prepared with the guanidinium thiocyanate–phenol–chloroform method20 and reverse-transcribed using an oligo-dT primer.21 Degenerate PCR was then performed with conditions optimized for each set of primers. For the catalase primers the conditions were 3.5 mM MgCl2, 2.5 µM primers, and 56°C annealing temperature. For the catalase/thioredoxin peroxidase primers, the conditions were 4.0 mM MgCl2, 20 µM primers, and 54°C annealing temperature. In all reactions there were 40 cycles (94°C x 15 seconds, annealing temperature x 30 seconds, 72°C x 30 seconds) followed by a 10-minute extension at 72°C.

Transformation and Sequencing
PCR products were run on a 1% low melting point agarose gel. Bands were excised, melted, and the amplimers purified and ligated into pST-Blue1 (Novagen, Madison, WI). After transformation of competent cells and plating, positive clones were grown, purified, and sequenced by automated fluorescence sequencing.

Statistical Analysis
Mean values were compared with Student’s unpaired t-test. ANOVA followed by Student–Newman–Keuls post hoc comparison was used to analyze the effects of multiple independent treatments on cell survival.


   Results
Top
 Abstract
 Introduction
 Methods
 Results
Discussion
 References
 
Resistance of RGCs to Oxidative Stress
Mixed retinal cultures were prepared from dissociated neonatal rat retinas containing RGCs that had previously been retrogradely labeled with the fluorescent dye DiI. Cultures were incubated with either menadione (which increases intracellular superoxide anion levels by redox cycling), xanthine/xanthine oxidase (which raises extracellular superoxide levels), hydrogen peroxide, or the combination of copper sulfate, phenanthroline, and ascorbate (which produces hydroxyl radical via the Fenton reaction). Live cells were identified by the presence of fluorescent staining with calcein-AM. The percentage of live and dead cells was calculated separately for DiI+ cells (i.e., RGCs) and DiI- cells (retinal cells). For each oxidative stress, there were concentrations that reduced RGC survival significantly less than the effect on other retinal cells (Fig. 1) . For example, menadione reduced survival of RGCs to 59% ± 4% of control, compared with 4% ± 2% for all retinal cells (P = 0.00003). Similar findings were seen with xanthine/xanthine oxidase (97% ± 9% versus 30% ± 2%; P = 0.0002), H2O2 (102% ± 1% versus 18% ± 1%; P = 0.00001), and copper sulfate, phenanthroline, and ascorbate (66% ± 5% versus 5% ± 2%; P = 0.00001).



View larger version (18K):
[in this window]
[in a new window]
  		Figure 1. Effect of ROS-generating systems on RGC and all retinal cell (All Cells) survival. Retinal cells were cultured in triplicate for 24 hours in the presence of diluent (balanced salt solution) control, 20 µM menadione, 10 µM xanthine with 2 mU/ml xanthine oxidase (X/XO), 0.0001% H2O2, or 200 nM CuSO4/200 nM phenanthroline/410 µM ascorbate (Cu/P/Asc). Percentage relative survival is calculated relative to survival in diluent control at 24 hours.

 
Resistance of RGCs to Oxidative Stress Mediated by Peroxides
To clarify the nature of the ROS to which RGCs were relatively resistant (compared with other retinal cells), ROS-generating systems were combined with specific ROS scavengers. Catalase (a peroxidase) completely inhibited the ability of hydrogen peroxide and xanthine/xanthine oxidase to kill retinal cells and was able, in part, to inhibit the effects of menadione (Fig. 2) . There was a significant difference in the survival rate of retinal cells in the absence or presence of catalase for menadione (9% ± 2% versus 36% ± 3%; P = 0.00008), H2O2 (107% ± 5% versus 2% ± 1%; P < 0.00001) and xanthine/xanthine oxidase (149% ± 5% versus 8% ± 1%; P < 0.00001) but not in the survival rate of RGCs in the presence or absence of catalase (all comparisons P > 0.1). These results are consistent with H2O2, xanthine/xanthine oxidase, and menadione primarily causing cell death through a peroxide intermediate, with RGCs being relatively protected. The decreased rescue by catalase of menadione, which increases intracellular O2-, suggested that the superoxide anion may also be neurotoxic.



View larger version (18K):
[in this window]
[in a new window]
  		Figure 2. Catalase inhibits cell death induced by various ROS-generating systems. Retinal cells were cultured in triplicate for 24 hours with or without catalase (500 U/ml) in the presence of diluent (balanced salt solution) control, 5.4 µM menadione, 0.0001% H2O2, or 10 µM xanthine with 2 mU/ml xanthine oxidase (X/XO). Percentage relative survival is calculated relative to survival in diluent control at 24 hours.

 
Although these results are consistent with a peroxide causing the death of retinal cells, it does not prove it. Mammalian cells have at least three superoxide dismutases that convert O2- into H2O2. An alternative explanation for our results showing that catalase rescued retinal cells under the conditions described above would be that the toxic intermediate is O2-; and by driving the reaction to the right (with catalase), we decreased the levels of O2-. To explore this possibility, we used the superoxide dismutase mimic CuDIPS alone and in combination with catalase. We found that CuDIPS was significantly more toxic to retinal cells (survival 5% ± 1% of control) than RGCs (survival 80% ± 4% of control; P < 0.00001), similar to the results seen with H2O2 (Fig. 3) . A similar result was seen when cells were cultured with another superoxide dismutase mimic, MnTMPyP (data not shown). The toxicity of CuDIPS was inhibited by catalase. Together, these results are consistent with cellular O2- only becoming toxic when dismutated to peroxide and supports the hypothesis that the neurotoxic intermediate is a peroxide.



View larger version (17K):
[in this window]
[in a new window]
  		Figure 3. The superoxide dismutase mimic CuDIPS is toxic to retinal cells other than RGCs (All Cells), and this toxicity is ameliorated by catalase. Retinal cells were cultured in triplicate for 24 hours in the presence of diluent (balanced salt solution) control, 25 µM CuDIPS, or CuDIPS with 500 U/ml catalase. Percentage relative survival is calculated relative to survival in diluent control at 24 hours.

 
Although the combination of iron or copper and ascorbate can generate hydroxyl radical, ascorbate alone can both lead to the generation of H2O222 23 and potentiate the effect of trace levels of H2O2 by reducing Cu2+ or Fe3+ alone.24 Therefore, to better understand the activity of the hydroxyl radical–generating system that we used (copper sulfate, phenanthroline, and hydrogen peroxide), we studied the activities of each of the three components. When mixed retinal cultures were incubated with copper, phenanthroline, or ascorbate, alone or in all possible combinations, only the presence of ascorbate was associated with the death of retinal neurons other than RGCs (P = 0.0001 by ANOVA; Fig. 4 ), consistent with its ability to generate H2O2.



View larger version (22K):
[in this window]
[in a new window]
  		Figure 4. Determination of which component(s) in the hydroxyl radical–generating system is responsible for the difference in survival between RGCs and other retinal cells (All Cells). Retinal cells were cultured in triplicate for 24 hours with or without copper (200 nM), phenanthroline (200 nM), or ascorbate (410 µM). Percentage relative survival is calculated relative to survival in diluent control at 24 hours.

 
Ascorbate alone would be insufficient in vivo to kill retinal neurons, because it is present in millimolar concentrations within the eye. We hypothesized that trace copper might be present within the fetal calf serum used for cell culture and that this copper would lead to production of hydroxyl radical in the presence of ascorbate. To study this, we cultured retinal neurons in defined medium containing the serum supplement B27 instead of fetal calf serum. By adding physiologically relevant concentrations of copper in the presence of ascorbate, we were able to establish that the combination of copper and ascorbate was sufficient for death of retinal neurons and RGCs (Fig. 5) .



View larger version (22K):
[in this window]
[in a new window]
  		Figure 5. Copper is necessary for the toxicity induced by ascorbate when retinal cells are cultured under serum-free conditions. Retinal cells were cultured in triplicate for 24 hours in MEM with 2% B27 with varying amounts of chelated copper, in the presence or absence of ascorbate (410 µM). Percentage relative survival is calculated relative to survival in diluent control at 24 hours.

 
Finally, to test our prediction that the toxic effect of the copper, phenanthroline, ascorbate system was due to a peroxide intermediate, we studied the effects of the peroxide scavenger catalase on cells incubated in the presence of ascorbate, using H2O2 as a positive control. As expected, although it has no significant effects on RGCs, catalase significantly abrogated the neurotoxic effects of ascorbate on all retinal cells compared with control (148% ± 8% with catalase versus 16% ± 6% without; P < 0.00001), similar to its effect when coincubated with H2O2 (174% ± 8% with catalase versus 6% ± 4% without; P < 0.00001; Fig. 6 ). Furthermore, the survival of all retinal cells other than RGCs was significantly (P = 0.0002 for ascorbate and P = 0.0001 for H2O2) higher than control survival in the presence of catalase, suggesting that a peroxide intermediate was responsible, in part, for the cell death associated with experimental cell culture.



View larger version (19K):
[in this window]
[in a new window]
  		Figure 6. The toxicity of ascorbate to retinal cells is mediated by a peroxide intermediate. Retinal cells were cultured in triplicate for 24 hours with ascorbate (Asc; 410 µM) or H2O2 (0.0001%) in the presence or absence of catalase (Cat; 500 U/ml). Percentage relative survival is calculated relative to survival in diluent control at 24 hours.

 
Resistance of RGCs to Oxidative Stress Not an Artifact of Cell Labeling
It is conceivable that the highly RGC-specific protection from oxidative stress that we observed could have resulted from an artifact of the RGC-labeling procedure. The RGCs were all retrogradely labeled with DiIC18, a carbocyanine dye. It is known that certain carbocyanine dyes, including DiIC18, block the mitochondrial electron transport system, via a rotenone-like effect.25 This inhibitory activity could potentially interfere with the generation of one or more toxic ROS in RGCs and confound the observed results. To examine this possibility, we compared the effects of oxidative stress on RGCs that had been labeled with DiIC18 to those labeled with DiIC16, a carbocyanine dye that does not possess appreciable inhibition of electron transport activity.26 There was no significant difference between the two dyes in the survival of either ascorbate-treated RGCs (60% ± 5% versus 58% ± 8%; P = NS) or ascorbate-treated retinal cells (3% ± 2% versus 2% ± 1%; P = NS; Fig. 7 ).



View larger version (19K):
[in this window]
[in a new window]
  		Figure 7. The difference in survival of RGCs compared with other retinal cells (All Cells) is not due to the specific effects of DiIC18. RGCs were retrogradely labeled with either DiIC18 or DiIC16, and mixed retinal cells cultured in triplicate for 24 hours in the presence or absence of 410 µM ascorbate. Percentage relative survival is calculated relative to survival in diluent control at 24 hours.

 
As another method for testing whether the dye itself was responsible for the increased survival of RGCs after oxidative stress, we designed a paradigm whereby all retinal cells were labeled with DiIC18, and then tested whether they were protected from oxidative stress compared with nonlabeled retinal cells. We prepared mixed retinal cultures from animals that had not previously had their retinas retrogradely labeled. We then incubated the cultured cells with DiIC18 for 24 hours, concurrent with ascorbate or control. Robust DiI labeling was confirmed by red fluorescence of all cells under epifluorescence microscopy. As predicted, there was no protective (or toxic) effect of DiIC18 on survival of retinal neurons (Fig. 8) .



View larger version (13K):
[in this window]
[in a new window]
  		Figure 8. Ascorbate is toxic to retinal cells even when labeled with DiI. RGCs were not retrogradely labeled. Mixed retinal cells were cultured in triplicate for 24 hours with diluent control or 5 µM DiIC18, in the presence or absence of 410 µM ascorbate (Asc). Percentage relative survival is calculated relative to survival in diluent control at 24 hours.

 
Resistance of RGCs to Peroxides Due to an Endogenous Peroxidase
Two well-characterized intracellular peroxidases are catalase and glutathione peroxidase. To assess their possible contribution to the resistance of RGCs to oxidative stress, cultures were incubated with three different peroxidase inhibitors. 3-Amino-1,2,4-triazole is a moderately specific inhibitor of catalase. L-buthionine sulfoximine (BSO) is a specific {gamma}-glutamylcysteine synthetase inhibitor, which leads to glutathione depletion and thereby impairs glutathione peroxidase activity. Sodium azide (NaN3) is a broad-spectrum peroxidase inhibitor. To assess the effect of inhibiting peroxidase activity, retinal cultures were incubated with these agents in the presence of H2O2. Compared with the relatively high survival (92% ± 9%) of RGCs in the presence of H2O2 alone, there was significant toxicity with aminotriazole (6% ± 4%; P = 0.0001), BSO (5% ± 3%; P = 0.00001), and NaN3 (14% ± 5%; P = 0.0003; Fig. 9 ). These results are consistent with an endogenous RGC peroxidase protecting against oxidative stress.



View larger version (15K):
[in this window]
[in a new window]
  		Figure 9. Peroxidase inhibitors abrogate the resistance of RGCs to oxidative stress. Mixed retinal cells (All Cells) were cultured in triplicate for 24 hours with diluent control, 20 mM aminotriazole (AT), 20 µM buthionine sulfoximine, or 1 mM NaN3 in the presence or absence of 0.0001% H2O2. Percentage relative survival is calculated relative to survival in diluent control at 24 hours.

 
The fact that all 3 peroxidase inhibitors led to RGC toxicity is not helpful in determining whether there is one (or more) RGC-specific peroxidases, or whether this putative peroxidase is novel. To survey peroxidases within the retina, we prepared degenerate PCR primers to areas of homology common to catalase (first primer set) or catalase and thioredoxin peroxidase (second primer set). Catalase and thioredoxin peroxidase are two well-characterized peroxidases within mammalian cells and contain strong regions of homology. By performing degenerate PCR between these homology regions, we hoped to determine whether novel peroxidases exist within the retina. However, all 100 clones (using the first primer set) and all 50 clones (using the second primer set) from RT–PCRs of rat retinal cDNA demonstrated only either the peroxidases used in designing the primers or genes that were falsely primed (i.e., did not contain both homology regions). The inability to demonstrate new peroxidases with these initial experiments implies that if a novel RGC-specific peroxidase exists, it either is expressed at low levels in the retina or does not contain the targeted homology regions.


   Discussion
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
References
 
These results demonstrate that neonatal RGCs are comparatively more resistant to oxidative stress than other retinal cells (as previously suggested by Armstrong et al.27 ) and that this resistance is most likely due to an increased resistance to peroxide(s). The finding that catalase increased survival of cells other than RGCs is also consistent with this hypothesis, in that the initial in vitro cell death of retinal cells is likely dependent in part on oxidative stress, and that reducing H2O2 levels reduces this death. RGCs could have an inherent protection against cell death due to certain types of oxidative stress, and we hypothesize that this protection involves the possession of sufficient constitutive levels of one or more peroxidases.

There are several cautions in interpreting our findings. First, we cannot be certain that the lessened susceptibility is due to a peroxidase within RGCs. Hydrogen peroxide, being an uncharged species, diffuses fairly freely across cell membranes. Therefore, if there is an equilibrium between H2O2 and another ROS, then any pharmacological intervention that decreases H2O2 levels extracellularly (e.g., with catalase) might be expected to decrease the intracellular concentration of that ROS. As an example of this concept, xanthine/xanthine oxidase raises extracellular superoxide levels, and being a charged species, poorly crosses cell membranes, instead probably acting via conversion to H2O2 (which does cross cell membranes). The ability of catalase to block the toxic effect of xanthine/xanthine oxidase is therefore consistent with the latter being due to indirectly increasing levels of H2O2, even though the xanthine/xanthine oxidase produces extracellular O2-. A similar argument applies to the experiments with ascorbate, which leads to the generation of H2O2.22 23

Nonetheless, it is unlikely that differences in scavenging O2- within RGCs are responsible for their increased resistance to oxidative stress, because the experiments with the superoxide dismutase mimic CuDIPS demonstrated decreased viability in the presence of CuDIPS, opposite to what would be expected if O2- were the toxic ROS. Furthermore, this decreased viability was abrogated by catalase, consistent with a peroxide intermediate. On the other hand, hydroxyl radical is in equilibrium with H2O2 via the Fenton reaction, and it is possible that differences between RGCs and other cells in the scavenging of OH· could be responsible for the results that we observed.

A second caution is that the nature of the increased resistance to oxidative stress was not defined by our studies. It is possible that it is due to increased levels of expression of known peroxidases (e.g., catalase, glutathione peroxidase, or thioredoxin peroxidase). It is equally possible that RGCs contain a novel peroxidase and that our degenerate PCR–based strategy for detecting novel peroxidases was inefficient. For example, RGCs make up less than 1% of the retinal cell population. If a novel peroxidase was RGC-specific, then it may show up in less than 1% of clones of a degenerate PCR library. Strategies for identifying novel peroxidases may require other techniques, such as RGC cDNA libraries or suppression PCR. Another explanation for the increased RGC resistance to oxidative stress is that RGCs do not contain the molecule(s) with which the increased peroxide (or alternate ROS) react or that the oxidized molecule(s) do not have the same downstream effect.

Our study was restricted to neonatal retinal cells, and, thus, we cannot apply our findings to adult animals. Young rats were used because their RGCs undergo a wave of target deprivation-induced cell death from approximately postnatal days 5 through 10 and because adult RGCs are extremely difficult to culture. Levels of antioxidant enzymes in the retina are in part a function of age,14 28 and our failure to find novel peroxidases using degenerate PCR could reflect differences in expression between adult and neonatal retinas. The ontogeny of resistance to oxidative stress is of great interest, and appropriate study of ROS-scavenging enzymes in the retina awaits a future investigation. We also did not differentiate between the various retinal cell types other than RGCs. These include not only retinal neurons (photoreceptors, bipolar, amacrine, and horizontal cells) but also glial cells (Müller cells and astrocytes) and endothelial cells. However, if another cell type were as resistant to oxidative stress as RGCs, it would have to be present in small number, because in several experiments in which H2O2 was used RGCs were virtually the sole surviving cell type. For example, if another cell type were as resistant to oxidative stress as RGCs, and present in the retina in the same number as RGCs, then we could expect equal numbers of DiI+ and DiI- live cells. Yet in most experiments almost all the live cells were RGCs, implying that any other "resistant" cell type would make up less than 1% of the retina.

What could explain the increased resistance of RGCs to peroxide-mediated oxidative stress? One thought is that RGCs in the inner retina of diurnal animals are exposed to high levels of ascorbate within the vitreous and that this could serve as a local oxidative stress. We found that ascorbate was toxic in the presence of Cu2+, whether added exogenously to the culture medium or contained within the fetal calf serum used for our studies, which is similar to other studies showing the need for a metal ion for the formation of hydroxyl radical.29 Under normal circumstances there is minimal free copper within the vitreous, although intraocular foreign bodies containing copper may cause varying degrees of electroretinographic dysfunction or intraocular inflammation.30 Iron is the predominant metal participating in the Fenton reaction in biological systems, and although we used a copper system because it had previously been characterized, it is likely that iron-dependent reactions are more important within the retina.

Although RGCs may express unique defenses against extracellular ROS based on their anatomic milieu, it is also possible that they may have particular requirements for regulating levels of specific ROS to accomplish the physiological processes that rely on those ROS. We studied postnatal day 7 through 9 RGCs, which are completing the process of developmental programmed cell death. As part of this process, approximately 50% of RGCs die as a result of competition for target-derived neurotrophic factors.31 32 In addition, RGCs are necessarily transected as part of the retinal dissociation procedure. If, indeed, ROS are involved in the signaling process of apoptosis after growth factor deprivation or axotomy signals,7 8 16 then it is conceivable that tight regulation of ROS levels may be necessary to avoid aberrant death signals that could result in apoptosis of inappropriate cells. Understanding how ROS levels are controlled could therefore provide insight into the mechanisms of cell death after axonal damage and, possibly, lead to methods for preventing RGC loss in optic nerve diseases.


   Footnotes
 
Supported by the Retina Research Foundation, the Glaucoma Foundation, NIH EY00340, and an unrestricted departmental grant from Research to Prevent Blindness. LAL is a Research to Prevent Blindness Dolly Green scholar.

Submitted for publication January 13, 2000; revised April 14, 2000; accepted April 24, 2000.

Commercial relationships policy: N (KK, LKG); C3,7,8 (LAL).

Corresponding author: Leonard A. Levin, University of Wisconsin Medical School, 600 Highland Avenue, Madison, WI 53792.


   References
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 

  1. Finkel, T. (1998) Oxygen radicals and signaling Curr Opin Cell Biol 10,248-253[Medline][Order article via Infotrieve]
  2. Goldschmidt–Clermont, PJ, Moldovan, L. (1999) Stress, superoxide, and signal transduction Gene Exp 7,255-260
  3. Clark, RA (1999) Activation of the neutrophil respiratory burst oxidase J Infect Dis 179(suppl 2),S309-S317
  4. Dalton, TP, Shertzer, HG, Puga, A. (1999) Regulation of gene expression by reactive oxygen Annu Rev Pharmacol Toxicol 39,67-101[Medline][Order article via Infotrieve]
  5. Jabs, T. (1999) Reactive oxygen intermediates as mediators of programmed cell death in plants and animals Biochem Pharmacol 57,231-245[Medline][Order article via Infotrieve]
  6. Du, G, Mouithys–Mickalad, A, Sluse, FE (1998) Generation of superoxide anion by mitochondria and impairment of their functions during anoxia and reoxygenation in vitro Free Radic Biol Med 25,1066-1074[Medline][Order article via Infotrieve]
  7. Greenlund, LJ, Deckwerth, TL, Johnson, EM (1995) Superoxide dismutase delays neuronal apoptosis: a role for reactive oxygen species in programmed neuronal death Neuron 14,303-315[Medline][Order article via Infotrieve]
  8. Chan, SL, Tammariello, SP, Estus, S, Mattson, MP (1999) Prostate apoptosis response-4 mediates trophic factor withdrawal-induced apoptosis of hippocampal neurons: actions prior to mitochondrial dysfunction and caspase activation J Neurochem 73,502-512[Medline][Order article via Infotrieve]
  9. Schulz, JB, Weller, M, Klockgether, T. (1996) Potassium deprivation-induced apoptosis of cerebellar granule neurons: a sequential requirement for new mRNA and protein synthesis, ICE-like protease activity, and reactive oxygen species J Neurosci 16,4696-4706[Abstract/Free Full Text]
  10. Mainster, MA (1987) Light and macular degeneration: a biophysical and clinical perspective Eye 1,304-310
  11. Organisciak, DT, Darrow, RM, Barsalou, L, et al (1998) Light history and age-related changes in retinal light damage Invest Ophthalmol Vis Sci 39,1107-1116[Abstract/Free Full Text]
  12. Winkler, BS, Boulton, ME, Gottsch, JD, Sternberg, P. (1999) Oxidative damage and age-related macular degeneration Mol Vis 5,32[Medline][Order article via Infotrieve]
  13. Armstrong, D, Santangelo, G, Connole, E. (1981) The distribution of peroxide regulating enzymes in the canine eye Curr Eye Res 1,225-242[Medline][Order article via Infotrieve]
  14. Castorina, C, Campisi, A, Di Giacomo, C, Sorrenti, V, Russo, A, Vanella, A (1992) Lipid peroxidation and antioxidant enzymatic systems in rat retina as a function of age Neurochem Res 17,599-604[Medline][Order article via Infotrieve]
  15. De La Paz, MA, Zhang, J, Fridovich, I. (1996) Antioxidant enzymes of the human retina: effect of age on enzyme activity of macula and periphery Curr Eye Res 15,273-278[Medline][Order article via Infotrieve]
  16. Estevez, AG, Spear, N, Manuel, SM, et al (1998) Nitric oxide and superoxide contribute to motor neuron apoptosis induced by trophic factor deprivation J Neurosci 18,923-931[Abstract/Free Full Text]
  17. Levin, LA, Clark, JA, Johns, LK (1996) Effect of lipid peroxidation inhibition on retinal ganglion cell death Invest Ophthalmol Vis Sci 37,2744-2749[Abstract/Free Full Text]
  18. Brewer, GJ, Torricelli, JR, Evege, EK, Price, PJ (1993) Optimized survival of hippocampal neurons in B27-supplemented Neurobasal: a new serum-free medium combination J Neurosci Res 35,567-576[Medline][Order article via Infotrieve]
  19. Jacobson, MD, Raff, MC (1995) Programmed cell death and Bcl-2 protection in very low oxygen Nature 374,814-816[Medline][Order article via Infotrieve]
  20. Chomczynski, P, Sacchi, N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction Anal Biochem 162,156-159[Medline][Order article via Infotrieve]
  21. Levin, LA, Schlamp, CL, Spieldoch, RL, Geszvain, KM, Nickells, RW (1997) Identification of the bcl-2 family genes in the rat retina Invest Ophthalmol Vis Sci 38,2545-2553[Abstract/Free Full Text]
  22. Sestili, P, Brandi, G, Brambilla, L, Cattabeni, F, Cantoni, O. (1996) Hydrogen peroxide mediates the killing of U937 tumor cells elicited by pharmacologically attainable concentrations of ascorbic acid: cell death prevention by extracellular catalase or catalase from cocultured erythrocytes or fibroblasts J Pharmacol Exp Ther 277,1719-1725[Abstract/Free Full Text]
  23. Peterkofsky, B, Prather, W. (1977) Cytotoxicity of ascorbate and other reducing agents towards cultured fibroblasts as a result of hydrogen peroxide formation J Cell Physiol 90,61-70[Medline][Order article via Infotrieve]
  24. Buettner, GR, Jurkiewicz, BA (1996) Catalytic metals, ascorbate and free radicals: combinations to avoid Radiat Res 145,532-541[Medline][Order article via Infotrieve]
  25. Anderson, WM, Chambers, BB, Wood, JM, Benninger, L. (1991) Inhibitory effects of two structurally related carbocyanine laser dyes on the activity of bovine heart mitochondrial and Paracoccus denitrificans NADH-ubiquinone reductase: evidence for a rotenone-type mechanism Biochem Pharmacol 41,677-684[Medline][Order article via Infotrieve]
  26. Anderson, WM, Trgovcich–Zacok, D. (1995) Carbocyanine dyes with long alkyl side-chains: broad spectrum inhibitors of mitochondrial electron transport chain activity Biochem Pharmacol 49,1303-1311[Medline][Order article via Infotrieve]
  27. Armstrong, D, Ueda, T, Ueda, T, et al (1998) Dose dependent mechanisms of lipid hydroperoxide induce retinal pathology Yagi, K Yagi, KS eds. Pathophysiology of Lipid Peroxides and Related Species ,57-76 Karger Basel, Switzerland.
  28. Organisciak, DT, Darrow, RM, Darrow, RA, Lininger, LA (1998) Environmental light and age-related changes in retinal proteins Williams, TP Thistle, AB eds. Photostasis and Related Phenomena ,79-92 Plenum Press New York.
  29. Winterbourn, CC (1981) Hydroxyl radical production in body fluids: roles of metal ions, ascorbate and superoxide Biochem J 198,125-131[Medline][Order article via Infotrieve]
  30. McGahan, MC, Bito, LZ, Myers, BM (1986) The pathophysiology of the ocular microenvironment, II: copper-induced ocular inflammation and hypotony Exp Eye Res 42,595-605[Medline][Order article via Infotrieve]
  31. Perry, VH, Henderson, Z, Linden, R. (1983) Postnatal changes in retinal ganglion cell and optic axon populations in the pigmented rat J Comp Neurol 219,356-368[Medline][Order article via Infotrieve]
  32. Ma, YT, Hsieh, T, Forbes, ME, Johnson, JE, Frost, DO (1998) BDNF injected into the superior colliculus reduces developmental retinal ganglion cell death J Neurosci 18,2097-2107[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 IOVSHome page
J. C. O'Connor, D. M. Wallace, C. J. O'Brien, and T. G. Cotter
A Novel Antioxidant Function for the Tumor-Suppressor Gene p53 in the Retinal Ganglion Cell
Invest. Ophthalmol. Vis. Sci., October 1, 2008; 49(10): 4237 - 4244.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
M. L. Chalasani, V. Radha, V. Gupta, N. Agarwal, D. Balasubramanian, and G. Swarup
A Glaucoma-Associated Mutant of Optineurin Selectively Induces Death of Retinal Ganglion Cells Which Is Inhibited by Antioxidants
Invest. Ophthalmol. Vis. Sci., April 1, 2007; 48(4): 1607 - 1614.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
C. R. Schlieve, C. J. Lieven, and L. A. Levin
Biochemical activity of reactive oxygen species scavengers do not predict retinal ganglion cell survival.
Invest. Ophthalmol. Vis. Sci., September 1, 2006; 47(9): 3878 - 3886.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
S. Roth, J. C. Dreixler, A. R. Shaikh, K. H. Lee, and V. Bindokas
Mitochondrial Potassium ATP Channels and Retinal Ischemic Preconditioning
Invest. Ophthalmol. Vis. Sci., May 1, 2006; 47(5): 2114 - 2124.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
K. I. Swanson, C. R. Schlieve, C. J. Lieven, and L. A. Levin
Neuroprotective Effect of Sulfhydryl Reduction in a Rat Optic Nerve Crush Model
Invest. Ophthalmol. Vis. Sci., October 1, 2005; 46(10): 3737 - 3741.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
G. Tezel, X. Yang, and J. Cai
Proteomic Identification of Oxidatively Modified Retinal Proteins in a Chronic Pressure-Induced Rat Model of Glaucoma
Invest. Ophthalmol. Vis. Sci., September 1, 2005; 46(9): 3177 - 3187.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
G. Tezel and X. Yang
Caspase-Independent Component of Retinal Ganglion Cell Death, In Vitro
Invest. Ophthalmol. Vis. Sci., November 1, 2004; 45(11): 4049 - 4059.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
S. Aijaz, L. Erskine, G. Jeffery, S. S. Bhattacharya, and M. Votruba
Developmental Expression Profile of the Optic Atrophy Gene Product: OPA1 Is Not Localized Exclusively in the Mammalian Retinal Ganglion Cell Layer
Invest. Ophthalmol. Vis. Sci., June 1, 2004; 45(6): 1667 - 1673.
[Abstract] [Full Text] [PDF]

Home page
 Arch OphthalmolHome page
C. D. Alldredge, C. R. Schlieve, N. R. Miller, and L. A. Levin
Pathophysiology of the Optic Neuropathy Associated With Friedreich Ataxia
Arch Ophthalmol, November 1, 2003; 121(11): 1582 - 1585.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
J. P. Vrabec, C. J. Lieven, and L. A. Levin
Cell-Type-Specific Opening of the Retinal Ganglion Cell Mitochondrial Permeability Transition Pore
Invest. Ophthalmol. Vis. Sci., June 1, 2003; 44(6): 2774 - 2782.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
A. Ohira, M. Tanito, S. Kaidzu, and T. Kondo
Glutathione Peroxidase Induced in Rat Retinas to Counteract Photic Injury
Invest. Ophthalmol. Vis. Sci., March 1, 2003; 44(3): 1230 - 1236.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
C. K. M. Choy, P. Cho, W.-Y. Chung, and I. F. F. Benzie
Water-Soluble Antioxidants in Human Tears: Effect of the Collection Method
Invest. Ophthalmol. Vis. Sci., December 1, 2001; 42(13): 3130 - 3134.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
X. Luo, G. N. Lambrou, J. A. Sahel, and D. Hicks
Hypoglycemia Induces General Neuronal Death, Whereas Hypoxia and Glutamate Transport Blockade Lead to Selective Retinal Ganglion Cell Death In Vitro
Invest. Ophthalmol. Vis. Sci., October 1, 2001; 42(11): 2695 - 2705.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kortuem, K.
Right arrow Articles by Levin, L. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kortuem, K.
Right arrow Articles by Levin, L. A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS