Protective Effect of Thioredoxins 1 and 2 in Retinal Ganglion Cells after Optic Nerve Transection and Oxidative Stress

Abstract

purpose. Oxidative stress has been implicated in retinal ganglion cell (RGC) death pathways after optic nerve transection (ONT) and during glaucomatous neuropathy. The authors investigated the expression and cell-protective roles of thioredoxins (cytosolic Trx1 and mitochondrial Trx2), important regulators of the cellular redox state, on RGCs after ONT and pharmacologic oxidative stress induction.

methods. ONT was performed on adult Wistar rats. Trx1 and Trx2 quantitative and spatial expression were examined with Western blot and immunohistochemistry, respectively. Electroporation and calcium phosphate-mediated procedures were used to deliver Trx1 and Trx2 expression constructs to RGCs in vivo and to cultured RGC-5 cells, respectively. Cell-protective effects of Trx1 and Trx2 overexpression on RGCs after ONT and on RGC-5 cells treated with glutamate/buthionine sulfoximine (BSO) were determined by RGC density analysis and cell viability assay, respectively.

results. Upregulation of Trx1 and Trx2 was observed in RGCs at different times after ONT and in RGC-5 cells after glutamate/BSO treatment. Trx1 and Trx2 overexpression in RGC-5 cells increased their survival rate by approximately twofold and threefold 24 and 48 hours after glutamate/BSO treatment, respectively. A neuroprotective effect of Trx1 and Trx2 overexpression on RGCs was also observed in vivo; the survival rate of RGCs was increased by 35% and 135%, respectively, 1 and 2 weeks after ONT.

conclusions. These findings provide evidence for in vitro and in vivo cell-protective effects of Trx1 and Trx2 on RGCs against oxidative stress–induced neurodegeneration.

Oxidative stress has been implicated in a variety of neurodegenerative diseases such as Alzheimer, ¹ Parkinson, ² and Huntington diseases ³ and amyotrophic lateral sclerosis. ⁴ Oxidative stress has also been proposed to be a common factor in retinal ganglion cell (RGC) death during glaucomatous neurodegeneration and after optic nerve transection (ONT). ^{5 6 7}  

Oxidative stress occurs when the generation of reactive oxygen species (ROS) exceeds cellular ability to neutralize and eliminate them. ROS excess can damage cellular constituents and initiate the apoptotic signaling pathway. ^{8 9} Protection from ROS is mediated by superoxide dismutase and by the glutathione (GSH) ¹⁰ and thioredoxin (Trx) systems. ¹¹  

Trx is a small, ubiquitous protein containing a conserved active site, Trp-Cys-Gly-Pro-Cys. Trx participates in various redox reactions by reversible oxidation of its active center dithiol to a disulfide and catalyzing dithiol–disulfide exchange reactions in multiple substrate proteins. Trx proteins are classified as two major forms: cytosolic thioredoxin-1 (Trx1) and mitochondrial thioredoxin-2 (Trx2). ¹² Trx1 plays an important defensive role against oxidative stress and performs a wide variety of biological functions, such as regulation of gene expression, control of growth, and apoptosis. ¹³ In the retina, current information has suggested that the expression of Trx1 is enhanced in the neural retina, especially the outer nuclear layer (ONL), and the retinal pigment epithelium (RPE) after exposure to light. ¹⁴ It is suggested that geranylgeranylacetone (GGA)-induced intracellular Trx1 is responsible for photoreceptor cell protection from light damage. ¹⁵ However, little is known about the expression and function of Trx1 in RGCs after oxidative stress. Trx2 regulates the mitochondrial redox state and plays an important role in cell proliferation. ¹⁶ It has been shown that cells conditionally deficient in Trx2 undergo apoptosis in the absence of exogenous stress, accompanied by an accumulation of intracellular ROS, the activation of caspase 9 and caspase 3, and the release of cytochrome c into the cytosol. ¹⁷ In the retina, the expression profile and the role of Trx2 have not yet been defined. 

In this study, we analyzed the expression of endogenous Trx1 and Trx2 in RGCs after ONT and evaluated cell-protective effects of these proteins on RGCs in a rat optic nerve axotomy model and on RGC-5 cells after the induction of oxidative apoptotic cell death with glutamate and buthionine sulfoximine (BSO). 

Methods

Animals

The use of animals for this study was approved by the Animal Research Committee of the University of California at Los Angeles and was performed in compliance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Adult male Wistar rats (250–300 g) were housed with standard food and water provided ad libitum. The animal room was lit with fluorescent lights (330 lux) automatically turned on at 3 am and off at 3 pm, and the temperature was kept at 21°C. The animals were maintained for at least 1 week in this environment before surgical procedures. 

To generate the ONT model, animals were anesthetized by inhalation of isoflurane (1.5%–3.5%) in oxygen, and the optic nerve sheath was incised with a needle knife 2 mm longitudinally, starting 3 mm behind the globe to expose the optic nerve. A cross-section of the optic nerve was made with the needle knife without damaging the adjacent blood supply. ¹⁸ The conjunctival incision was sutured, and tobramycin ophthalmic ointment (Tobrex; Alcon, Fort Worth, TX) was applied topically. ONT was performed on one eye of each rat, and the contralateral eye was used as an untreated control. 

RGC Counting after ONT

Retrograde labeling to identify RGCs was performed by placing a small piece of sterile compressed sponge (Gelfoam; Pfizer, New York, NY) soaked with 6% fluorogold (FG; Fluorochrome, Denver, CO) to the proximal cut surface of the optic nerve after ONT. Rats were humanely killed 1, 3, 5, 7, and 14 days after ONT. The eyes were enucleated and immersed in 4% paraformaldehyde in 0.1 M phosphate buffer for 1 hour. Retinas were dissected from the eyeballs and mounted on glass slides. RGCs were counted at 1, 2, and 3 mm from the center of the optic nerve in each retinal quadrant under a fluorescence microscope (LSM510; Carl Zeiss, Oberkochen, Germany) at 200× magnification. Distinguishable glial cells (bright and small cells) were not counted. Quantification was performed in a masked manner. 

RGC Isolation

RGC isolation was performed with magnetic beads, as described previously. ¹⁹ Briefly, retinas isolated from adult Wistar rats (250–300 g each, approximately 3 months old) were incubated in D-phosphate buffered saline (D-PBS; without Ca²⁺ and Mg²⁺) containing 20 U/mL papain, 1 mM l-cysteine, and 0.005% DNase I (Worthington, Lakewood, NJ) at 37°C for 30 minutes. The retinas were gently triturated with a 1-mL pipette in a D-PBS solution containing 0.15% trypsin inhibitor, 0.15% bovine serum albumin (BSA), and 0.005% DNase I. Cells were precipitated and resuspended in D-PBS containing 0.1% BSA. Macrophages and adherent cells were removed by attachment to CD 11b/c monoclonal antibody (BD PharMingen; San Diego, CA)–coated beads. Cells were then selected with magnetic beads coated with Thy-1 monoclonal antibody (Chemicon, Temecula, CA). After washing, the attached Thy-1 cells were released by incubation in DNase buffer, and RGCs were seeded on extracellular matrix–coated glass slides. The purity of RGC isolation was examined with Thy-1 immunofluorescence labeling, and 4′6-diamino-2-phenylindole dihydrochloride (DAPI; Molecular Probes, Eugene, OR) was used for nuclear staining. 

Western Blot Analysis

Each freshly dissected retina was homogenized separately by ultrasonic disruption in a lysis buffer (50 mM phosphate buffer, 0.25 M sucrose, 0.1 mg/mL phenylmethylsulfonyl fluoride in isopropanol, and 30 μg/mL aprotinin). Protein concentration was determined (BCA Protein Assay Kit; Pierce, Rockford, IL). Western blot analysis was carried out as described previously. ^{20 21} Briefly, 2 to 5 μg protein was separated on a 12.5% SDS-polyacrylamide gel (Bio-Rad, Hercules, CA) and transferred to the polyvinylidene (PVDF) membrane (Millipore, Bedford, MA). After blocking with 5% nonfat milk, the membranes were incubated with primary antibodies against Trx1 (Chemicon), Trx2 (Santa Cruz Biotechnology, Santa Cruz, CA), or β-actin (Sigma, St. Louis, MO) overnight at 4°C, followed by incubation with peroxidase-conjugated secondary antibodies. Signals were visualized (ECL Plus Detection Kit; Amersham/GE Healthcare, Piscataway, NJ), and densitometry of the bands was performed with NIH Image software (National Institutes of Health, Bethesda, MD). 

Immunohistochemistry

RGCs were retrogradely labeled as described (RGC counting after ONT) by placement of a sterile compressed sponge (Gelfoam; Pfizer) soaked with 6% FG to the proximal cut surface of the optic nerve after ONT. Three days after FG labeling, animals were deeply anesthetized with intramuscular injections of 0.8 mL/kg of a cocktail containing ketamine (100 mg/mL), 2.5 mL xylazine (20 mg/mL), 1.0 mL aceptomazine (10 mg/mL), and 1.5 mL normal saline and were transcardially perfused with 4% paraformaldehyde in 0.1 M phosphate buffer. Enucleated eyes were immersed in fixative for 1 hour, bisected, and postfixed for 3 hours. Eyecups were incubated with 30% sucrose at 4°C overnight and embedded in OCT compound (Sakura Finetek, Torrance, CA). Ten-micrometer–thick sections were obtained along the vertical meridian through the optic nerve head. The sections were washed three times with PBS in Triton-X (T-PBS) for 5 minutes and incubated with blocking buffer (T-PBS containing 5% bovine serum albumin) at room temperature for 30 minutes. After blocking, they were incubated with primary antibodies against Trx1 and Trx2 overnight at 4°C. The sections were incubated with rhodamine-conjugated anti–rabbit IgG antibody (Cappel Research Products, Durham, NC) at room temperature for 60 minutes. They were then mounted with anti–fade mounting solution. Photomicrographs of the sections at 1 mm from the center of the optic nerve were taken with a fluorescence microscope (LSM510; Carl Zeiss). Negative-control sections were incubated without primary antibodies. 

Cell Culture and Transfection

RGC-5 cells were maintained in Dulbecco modified Eagle medium (DMEM; Mediatech, Herndon, VA) containing 10% fetal bovine serum (Invitrogen, Carlsbad, CA), 100 U/mL penicillin, and 100 μg/mL streptomycin (Sigma) in a humidified atmosphere of 95% air and 5% CO₂ at 37°C, as described previously. ²² Twenty-four hours before transfection, exponentially growing cells were harvested by trypsinization and replated at a density of 1 × 10⁵ to 2 × 10⁵ cells/cm² with appropriate medium. The calcium phosphate-mediated transfection procedure ²³ was used to introduce pEGFP-C1-Trx1 ¹⁵ and pCMV-hTrx2-EGFP ²⁴ plasmid DNAs into the RGC-5 cells. These constructs were prepared by cloning of human Trx1 or Trx2 cDNA into pEGFP-C1 or pEGFP-N3, respectively. Transcription of the Trx genes in both vectors is controlled by cytomegalovirus (CMV) promoter. After 3 to 5 hours of exposure to the calcium phosphate-DNA coprecipitate, the cells were briefly treated (30 seconds) with 15% glycerol to increase the efficiency of transformation. The efficiency of transfection was determined by the ratio of EGFP-positive cells compared with the total number of cells stained with DAPI. 

Cell Viability Assay

Cell viability was determined (Cell Viability Assay Kit; BioVision, Mountain View, CA). Cells were seeded on a 96-well plate (5 × 10³ cells/well) and were treated with 5 mM or 10 mM glutamate (Fisher Scientific, Chino, CA) and 0.5 mM BSO (Sigma), which inhibits glutamate cysteine ligase, the rate-limiting enzyme in GSH biosynthesis. ²⁵ Twenty-four and 48 hours after treatment, cells were incubated with 10 μL water-soluble tetrazolium salt (WST)-1/electrocoupling solution for 3 hours, and absorbance was measured at 450 nm with a microplate reader (Microplate Manager; Bio-Rad). Wells without cells and cells without treatment with experimental agents were used as a blank and a control, respectively. 

Subcellular Localization of Trx

For subcellular localization, EGFP-tagged Trx1- and Trx2-expressing RGC-5 cells were incubated with 500 nM marker for mitochondria (Mitotracker Red CMXRos; Invitrogen) for 30 minutes at 37°C. Cells were fixed with 4% paraformaldehyde for 30 minutes and incubated with 4′-6-diamidino-2-phenylindole (DAPI) for nuclear staining. Images were acquired with fluorescence confocal microscopy (Leica, Bannockburn, IL). 

Gene Transfer into RGCs with In Vivo Electroporation

Trx1- and Trx2-expressing plasmid DNA was delivered to RGCs by electroporation (ELP), as described previously, with minor modifications. ^{26 27} In brief, animals were anesthetized with intramuscular injections of 0.8 mL/kg of a cocktail containing 5 mL ketamine (100 mg/mL), 2.5 mL xylazine (20 mg/mL), 1.0 mL acepromazine (10 mg/mL), and 1.5 mL normal saline. Plasmid DNA (4 μL; 2.5 μg/μL) was injected into the vitreous cavity with a 34-gauge needle 0.5 mm posterior to the limbus under a stereomicroscope. After 10 minutes, the cathodal electrode was placed on the cornea, and an 18-gauge needle with an attached anodal electrode was inserted subcutaneously at the middle of the forehead. Electric pulses were generated by a pulse generator (ECM 830; BTX, San Diego, CA). Parameters were as follows: electric field strength of 6 V/cm, pulse duration of 100 ms, stimulation pattern of five pulses at a frequency of 1 pulse/s. After a 10-minute pause, five more pulses with the same parameters were delivered. To determine the efficiency of ELP-mediated RGC transfection, 7 days after ELP, the optic nerve was transected as described, and RGCs were retrogradely labeled by placing a small piece of sterile compressed sponge (Gelfoam; Pfizer) soaked with dextran tetramethylrhodamine (DTMR; Molecular Probes) to the proximal cut surface of the optic nerve. Transfected RGCs were counted 2 days after ONT. To determine the effect of Trx overexpression on RGC survival, ONT was performed 3 days after ELP, when RGCs were labeled by placement of a DTMR-soaked sterile compressed sponge (Gelfoam; Pfizer) to the proximal cut surface of the optic nerve. RGCs were counted 7 and 14 days after ONT. 

Statistical Analysis

Data are presented as the mean ± SD. Differences among groups were analyzed by one-way ANOVA, followed by the Scheffé or Mann–Whitney U test. P < 0.05 was considered statistically significant. 

Results

RGC Loss after ONT

The number of surviving RGCs was determined by counting retrogradely labeled cells in flat-mount retinal preparations (Fig. 1A) . Significant RGC loss was observed from day 5 after ONT (Fig. 1B) . The number of RGCs 7 days after ONT was reduced by approximately 50% compared with controls. At 14 days after ONT, most of RGCs were lost, and the remaining population fell to less than 10% of controls. ONT is known to cause rapid and specific degeneration of RGCs, and several published studies have reported 90% or greater loss of RGCs by the second week after axotomy. ^{28 29}  

Expression of Trx1 and Trx2 in the Retina after ONT

Immunohistochemistry of Trx1 showed expression primarily in the inner nuclear layer, ganglion cell layer (GCL), and photoreceptor cells, whereas immunostaining for Trx2 was present in all retinal layers (Figs. 2A 2B) . In the GCL, Trx1- and Trx2-positive cells were colocalized with FG-labeled RGCs (Figs. 2A 2B) . Levels of Trx protein were analyzed in whole retinal extracts and in isolated RGCs. Western blot analysis of whole retinal extracts showed that the levels of Trx1 and Trx2 proteins appeared to be slightly increased after ONT compared with controls. However, these changes were not significant (Figs. 3A 3B) . Axotomy primarily affects RGCs; consequently, changes in Trx levels in response to ONT are expected to take place primarily in RGCs. Because RGCs constitute a small percentage of retinal cells, Western blot analysis of the whole retinal extract may be not sufficiently sensitive to detect modulation in Trx expression in these cells. Therefore, we analyzed Trx1 and Trx2 protein levels in isolated RGCs. Consistent with previous studies, the purity of isolated RGCs was approximately 95% (Fig. 4A) . ^{19 30} As expected, the changes in Trx1 and Trx2 protein levels in isolated RGCs were more evident than in whole retinas. Approximately 1.3- and 2.0-fold increases in Trx2 expression were observed 1 and 3 days after ONT, respectively, whereas the Trx1 level was elevated approximately 1.4-fold 7 days after ONT (Figs. 4B-1 4B-2) . 

Trx1 and Trx2 Expression in RGC-5 Cells in Response to Oxidative Stress

The effect of oxidative stress on the level of Trx expression in RGC-5 cells was determined by Western blot analysis. Oxidative stress in RGC-5 was induced by combination treatment with BSO and glutamate. Glutamate is known to modulate redox status, whereas BSO treatment reduces the level of cellular GSH, which is followed by an increase in ROS production that consequently induces several apoptotic signaling pathways, including the mitochondrial pathway. ^{31 32} Similar to the results obtained from isolated RGCs after ONT, an increase in Trx2 (1.7-fold) and Trx1 (1.4-fold) expression was observed 12 and 18 hours after glutamate/BSO treatment, respectively. (Figs. 5A 5B) . 

Overexpression of Trx1 and Trx2 in RGC-5 Cells with Induced Oxidative Stress

Several pharmacologic agents, including GGA, NGF, sulforaphane, 17-β-estradiol, and dexamethasone, have been shown to stimulate the expression of Trx1 and Trx2. ^{15 33 34 35 36} However, their effect is not restricted to the induction of these proteins only. Therefore, Trx1- and Trx2-expressing plasmids, pEGFP-C1-Trx1, ¹⁵ and pCMV-hTrx2-EGFP ²⁴ were used to evaluate the effect of overexpression of these proteins on cell survival in response to oxidative stress induced by glutamate/BSO treatment of RGC-5 cells. The efficiencies of RGC-5 transfection were 67% for pEGFP-C1-Trx1 and 63% for pCMV-hTrx2. Diffused EGFP-tagged Trx1 staining was observed outside the nuclei, indicating that Trx1 expressed from pEGFP-C1-Trx1, similar to endogenous Trx1, is localized in the cytoplasm (Fig. 6A) . EGFP-tagged Trx2 was colocalized with the marker for mitochondria (Mitotracker Red CMXRos; Invitrogen), indicating mitochondrial localization of Trx2 expressed by pCMV-hTrx2-EGFP (Fig. 6B) . 

After the transfection of RGC-5 cells with Trx1- and Trx2-expressing plasmids, cells were treated with glutamate/BSO to induce dose-dependent oxidative cell death. The cell viability assay showed a significant increase in the survival rate of Trx1-transfected RGC-5 cells. A protective effect of Trx1 overexpression was approximately 2.0-fold 24 hours after exposure to 5 mM or 10 mM glutamate with BSO (Fig. 6C-1 ). Forty-eight hours after treatment with 5 mM and 10 mM glutamate/BSO, cell viability of Trx1-transfected RGC-5 was increased approximately 2.5- and 3.0-fold, respectively (Fig. 6C-2 ). Overexpression of Trx2 also showed a significant protective effect against 5 mM glutamate with BSO treatment after 24 and 48 hours, but not against 10 mM glutamate with BSO compared with glutamate/BSO-treated untransfected cells (Figs. 6C-1 6C-2 ). 

Effect of Trx1 and Trx2 Overexpression on RGC Survival after ONT

Delivery of the EGFP-tagged Trx1 and Trx2 expression constructs to retinal cells was performed with ELP. To evaluate the expression of Trx proteins in transfected retinas, immunoblot analysis with Trx1 and Trx2 antibodies was performed. Two bands with molecular weights of approximately 12 kDa, corresponding to endogenously expressed Trx1 and Trx2, along with approximately 40-kDa products, corresponding to Trx1- and Trx2-EGFP fusion proteins, were detected in control retinas 7 days after transfection (Fig. 7A) . Expression of the fusion proteins in the retina has been determined to last for at least 5 weeks (data not shown). To evaluate the efficiency of RGC transfection, we counted EGFP-positive cells colocalized with DTMR-labeled RGCs and the total number of DTMR-labeled RGCs at 1, 2, and 3 mm from the center of the optic nerve in four areas of each retina (n = 4). Two days after ONT, the proportions of Trx1- and Trx2-transfected RGCs were determined to be 25.0% ± 4.1% and 22.2% ± 7.5%. RGCs constituted 61.3% ± 10.1% and 67.1% ± 11.5% of all Trx1- and Trx2-EGFP transfected–cells in the GCL, respectively. 

The effect of Trx1 and Trx2 overexpression on RGC survival was evaluated 7 and 14 days after ONT. The numbers of surviving RGCs after Trx1 and Trx2 transfection were 1232 ± 43 and 1205 ± 93 cells/mm², respectively (approximately 60% of control), compared with 901 ± 39 cells/mm² after transfection with vector alone (pEGFP) 7 days after ONT (Fig. 7E-1 ). This is equivalent to an approximately 35% increase in the RGC survival rate. The proportions of RGCs expressing Trx1-EGFP or Trx2-EGFP were 34.7% ± 6.3% and 35.1% ± 5.6%, respectively. Images presented in Figs. 7B and 7Cshow that in this region of the retina, most Trx1-EGFP– and Trx2-EGFP–transfected cells were colocalized with DTMR-labeled RGCs. In 14-day ONT retinas, 419 ± 125 and 398 ± 58 cells/mm² remained after Trx1 and Trx2 transfection, respectively (approximately 20% of control), compared with 174 ± 6 cells/mm² of ONT after pEGFP transfection (Figs. 7D 7E-2 ). This corresponded to an approximately 135% increase in RGC survival rate. Approximately 50% of remaining RGCs were Trx1- or Trx2-transfected RGCs. No significant differences in RGC numbers were found between ONT retinas, ONT retinas injected with saline, or ONT retinas transfected with vector alone. These results indicated a significant neuroprotective effect of Trx1 and Trx2 overexpression on RGCs in response to axotomy-induced oxidative stress. 

Discussion

Oxidative injury from amplified ROS generation has been indicated as a pathway for RGC death after ONT. ^{37 38} In this study, we hypothesized that Trx proteins, which are known to be important regulators of the cellular redox state, may have a neuroprotective effect on RGCs against ONT-induced oxidative injury. 

The first part of this study is focused on the analysis of endogenous Trx1 and Trx2 expression after ONT. No considerable changes in protein levels were detected in whole retinal extracts. In isolated RGCs, however, Trx2 and Trx1 protein levels were significantly upregulated 3 and 7 days after ONT, respectively. Consistent with the findings of the in vivo study, a significant increase in Trx1 and Trx2 protein levels was observed in RGC-5 cells after treatment with glutamate and BSO. Upregulation of Trx1 and Trx2 indicates the involvement of these proteins in stress-induced RGC degeneration, possibly through an endogenous cell-protective mechanism activated in response to oxidative injury. 

Expression of Trx1, a marker for oxidative stress, has been shown to be elevated in response to oxidative stresses induced by different injuries. ^{39 40 41} Elevated Trx1 levels were observed in various human diseases. ^{42 43 44} Trx1 has been shown to protect neurons against ischemia–reperfusion brain injury, ⁴⁵ to mediate preconditioning-induced neuroprotection, ⁴⁶ and to suppress 1-methyl-4-phenylpyridinium–induced neurotoxicity in human SH-SY5Y cells and PC12 cells. ^{47 48} Trx1-mediated neuroprotection may be attributed to its potent antioxidative and antiapoptotic properties. ⁴⁹ Trx1 antioxidative and associated cytoprotective effects were also shown in the present study by overexpression of this protein in RGC-5 cells treated with glutamate and BSO. Compared with Trx2 overexpression, Trx1 attenuated cell death induced by 5 and 10 mM glutamate with BSO treatment, suggesting that a protective effect of Trx1 against oxidative stress by these agents is more potent than Trx2 in RGC-5 cells. 

The downregulation of Trx2 expression has been associated with an increase in mitochondrial outer membrane permeability and a consequent release of cytochrome c. ³² Cytochrome c immunoreactivity after ONT increased at 1 day, peaked at 3 days, and decreased thereafter. Cytochrome c was localized almost exclusively to RGCs, suggesting that cytochrome c release was injury related. ⁵⁰ Temporal changes in Trx2 expression in the mitochondrial fraction (data not shown) of isolated RGCs observed in our experiments coincided with reported cytochrome c translocation, suggesting the involvement of Trx2 in the regulation of cytochrome c release in RGCs after ONT. Trx2 localization to the mitochondria was also demonstrated by colocalization of Trx2-EGFP with marker for mitochondria (Mitotracker Red CMXRos; Invitrogen). Furthermore, we showed that Trx2 overexpression has cytoprotective effects against oxidative stress-induced apoptosis, most likely by decreasing the permeability of the mitochondrial membrane and consequently reducing the level of cytosolic cytochrome c. 

Based on the observed neuroprotective effects of Trx1 and Trx2 overexpression on RGC-5 cells in response to oxidative stress, we analyzed the effect of these proteins on RGC survival after ONT. Optic nerve axotomy leads the neuronal redox status toward oxidation, which may be involved in cell death through direct effects on mitochondrial functions or indirect activation of caspases. ^{51 52 53} Intracellular levels of superoxide were increased in as many as 20% of RGCs and peaked 12 hours after axotomy. ⁵⁴ Asynchronous elevation in superoxide levels after axonal injury was suggested to play a role in the delayed death of axotomized RGCs. ⁵⁴ Consistent with the hypothesis that RGC death after axotomy involves elevated ROS levels, the survival of these cells was shown to depend on redox state and was stimulated by ROS scavengers. ^{38 51 52} In the present study, Trx1 and Trx2 overexpression in the retina increased the survival rate, by approximately 35%, of RGCs 1 week after axotomy, resulting in the preservation of approximately 60% of RGCs. Two weeks after ONT, Trx-mediated RGC survival was increased by 135%, which saved 20% of these neurons. It is likely that the cell-protective effect of Trx1 and Trx2 could be even higher considering the fact that ELP-mediated RGC transfection efficiency was only 25.0% ± 4.1% and 22.2% ± 7.5%, respectively. Furthermore, we believe that the effect of oxidative stress reduction on RGC survival observed in this study could be further complemented with other neuroprotective strategies, such as supplementation with neurotrophic factors. Cell-protective roles of neurotrophic factors, and brain-derived neurotrophic factor (BDNF) in particular, have been shown in several studies. No reduction in RGC count was observed after intravitreal BDNF injection by the posterior ocular route 1 week after axotomy, but in vehicle-treated and untreated eyes, RGC rates were 77% and 57% of normal, respectively. After anterior injection of BDNF or vehicle, RGC densities were normal, suggesting that the endogenous trophic responses induced by injury associated with this injection were stronger than those triggered by the posterior route. The rates of surviving RGCs decreased in BDNF-treated retinas to 41% to 42% and in the vehicle-injected to 6% of normal after posterior injections 2 week after axotomy. ²⁸ The combination of BDNF administration and lens injury preserved approximately 71% of RGCs 2 weeks after optic nerve injury. ⁵⁵ BDNF, ciliary neurotrophic factor, and regeneration-associated factors from sciatic nerve prolonged the survival of axotomized cells 2.0- to 3.0- fold compared with controls. ²⁹ Neurotrophin-4 (NT-4) or BDNF protected 30% to 40% of the axotomized RGC population by 7 days after the lesion and increased the survival of these cells by 160% (NT-4-treated) and 200% to 300% (BDNF-treated) 2 weeks after axotomy. ⁵⁶ Adenovirus (Ad)-mediated expression of BDNF by Müller cells showed 4.5-fold increased RGC survival 16 days after axotomy compared with control. However, extended expression of BDNF by Müller cells showed no further increase in RGC survival, suggesting that the BDNF effect is limited by lesion-induced changes 10 to 16 days after ONT. ⁵⁷ Administration of BDNF combined with TrkB (BDNF receptor) gene transfer increased RGC survival by more than 66%. ⁵⁸ Ad-BDNF enhanced RGC survival by approximately 40%, whereas 2 weeks after axotomy, a combinatory treatment with Ad-BDNF and systemic administration of the free radical scavenger, N-tert-butyl-(2-sulfophenyl)-nitrone, increased RGC survival by 63%. ⁵⁹  

More pronounced rescue of RGCs achieved by the overexpression of Trx proteins 14 days compared with 7 days after ONT could be viewed as a result of the attenuation of secondary events associated with increased oxidative damage contributing to neuronal degeneration. It could be that during the second week after axonal injury, the level of ROS in injured RGCs increases because of accumulation or higher production leading to more severe oxidative damage. Furthermore, although practically all RGCs are injured after axotomy, cells that die faster may contribute to the degeneration of remaining neurons (secondary degeneration). Secondary degeneration of RGC has been well described for the optic nerve crush model, in which neurons degenerating because of direct damage were associated with damage to other cells that were not injured by the primary insult. ^{60 61 62} Similarly, RGCs dying early after axotomy may contribute to the damage of neighboring RGCs or may lead to the activation of microglial cells, which in turn could be involved in secondary RGC degeneration. Microglial activation has been suggested to be secondary to RGC death because the peak of microglial activation follows the peak of axotomized RGC death (7 days after axotomy). ^{63 64} Increased expression of inducible nitric oxide (NO) synthase by injured RGCs and glial cells and subsequent NO toxicity associated with the oxidation of cellular constituents has been implicated in RGC death after axotomy. ⁶⁵ We believe that the overexpression of Trx proteins in RGCs and in glial cells reduces the damaging effect of oxidative stress on these cells and thus contributes to RGC survival. 

Involvement of oxidative damage in RGC degeneration has become more evident in the past few years. Increased ROS levels were associated with RGC degeneration not only after ONT but also during glaucomatous neurodegeneration. ³⁰ ROS have direct neurotoxic effects on RGCs and contribute to secondary degeneration by affecting glial function. ⁶⁶ In our study, the overexpression of cytosolic Trx1 and mitochondrial Trx2 protected RGCs from oxidative stress-induced neurotoxicity, suggesting that activation of these genes may have potential for the development of new neuroprotective agents to treat neurodegenerative disease. 

 Supported by the Nagai Foundation for Pharmacy and Pharmaceutical Sciences, Tokyo, Japan (YM), Research to Prevent Blindness (JC), and Oppenheimer Family Foundation (NP).

 Submitted for publication January 9, 2008; revised March 17, 2008; accepted June 13, 2008.

 Disclosure: Y. Munemasa, None; S.H. Kim, None; J.H. Ahn, None; J.M.K. Kwong, None; J. Caprioli, None; N. Piri, 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.

 

RGC loss after ONT. (A) Representative fluorescence photographs of flat mount retina 1 day (A-1) and 14 days (A-2) after ONT. (B) Density of RGCs 1, 3, 5, 7, and 14 days after ONT, estimated by FG labeling. Significant difference in RGC number was observed from 5 days after ONT, compared with 1 day after ONT (n = 3–4; *P < 0.05). Scale bar, 50 μm.

Figure 2.

 

Colocalization of the Trx1-positive (A) or Trx2-positive (B) cells with FG-labeled RGCs. GCL, retinal ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer; PL, photoreceptor layer. Scale bar, 50 μm.

Figure 3.

 

Quantitative analysis of Trx1 and Trx2 after ONT. (A) Western blot analysis of whole retinal Trx1 and Trx2 at 1, 3, and 7 days after ONT. (B) Densitometry of the immunoreactive bands for Trx1 and Trx2 showed no significant differences between Trx1 or Trx2 and control (n = 4). Data are normalized to β-actin levels in each sample. C, control.

Figure 4.

 

Trx expression in axotomized RGCs. (A) Purity of isolated RGCs. More than 95% of DAPI-positive cells were Thy-1–positive cells. Scale bar, 100 μm. (B) Western blot analysis of Trx1 and Trx2 in isolated RGCs at 1, 3, and 7 days after ONT. Densitometry of the immunoreactive bands corresponding to Trx1 and Trx2 showed a 2.0-fold increase of Trx2 and a 1.4-fold increase of Trx1 levels 3 and 7 days after ONT, respectively, compared with the control (n = 4; *P < 0.05). Data are normalized to β-actin levels in each sample. C, control.

Figure 5.

 

Western blot analysis of Trx1 and Trx2 in RGC-5 cells 3, 6, 12, 18, and 24 hours after treatment with 5 mM glutamate and 0.5 mM BSO. Densitometries of the immunoreactive bands corresponding to Trx1 and Trx2 (A, B) revealed significant increases of Trx2 and Trx1 protein levels 12 and 18 hours after treatment (n = 5; *P < 0.05, **P < 0.01). Data are normalized to β-actin levels in each sample. C, control.

Figure 6.

 

Subcellular localization of EGFP-tagged Trx1 (A) and Trx2 (B) and the effect of Trx1 and Trx2 overexpression on RGC-5 cell viability 24 (C-1) and 48 (C-2) hours after glutamate/BSO treatment. (A) Trx1-EGFP was mainly present in the cytosol. (B) Trx2-EGFP was colocalized with a marker for mitochondria (B-2). DAPI was used for nuclear staining (A-2, B-3). Scale bar, 10 μm. (C) After transfection with Trx1- or Trx2-expressing plasmid, cells were treated with 5.0 or 10.0 mM glutamate and 0.5 mM BSO. Trx1 overexpression showed cell-protective effects 24 (C-1) and 48 (C-2) hours after treatment with 5.0 or 10.0 mM glutamate and BSO treatment (n = 5–11; *P < 0.05, **P < 0.005, ***P < 0.0005). Trx2 overexpression showed a cell-protective effect 24 and 48 hours after treatment with 5.0 mM glutamate and BSO (n = 5–11; *P < 0.05, **P < 0.005, ***P < 0.0005).

Figure 7.

 

Trx1 and Trx2 overexpression in the retina and the effect on RGC survival after ONT. (A) Immunoblot analysis of control retinas after ELP with Trx1 (A-1) and Trx2 (A-2) expression constructs. Lanes 1 and 3, nontransfected retinas. Lanes 2 and 4, retinas transfected with Trx1- and Trx2-EGFP plasmids, respectively. (B, C) Trx1- and Trx2-EGFP (B-1, C-1) were primarily colocalized with DTMR-labeled RGCs (B-2, C-2). B-3 and C-3 are merged images. (D) After transfection with Trx1- and Trx2-EGFP constructs, RGCs were counted 7 and 14 days after ONT. Representative fluorescence photographs of flat mount retina transfected with vector alone (D-1), with Trx1 (D-2), and with Trx2 (D-3) expression constructs 14 days after ONT. (E) Overexpression of Trx1 and Trx2 showed neuroprotective effect on RGCs 7 and 14 days after ONT (E-1, E-2, n = 4–7; *P < 0.05, **P < 0.001). Scale bar, 50 μm.

The authors thank Drs. Junji Yodoi and Hiroshi Masutani for providing Trx expression plasmids, and Dr. Neeraj Agarwal for providing RGC-5 cells.