Effect of Thioltransferase (Glutaredoxin) Deletion on Cellular Sensitivity to Oxidative Stress and Cell Proliferation in Lens Epithelial Cells of Thioltransferase Knockout Mouse

Abstract

purpose. To examine the physiological function of the thioltransferase (TTase)/glutathione (GSH) system in the lens using TTase knockout mouse (TTase ^−/−) lens epithelial cells (LECs) as a model.

methods. Primary LEC cultures were obtained from wild-type (TTase ^+/+) and TTase ^−/− mice. Characterization and validation of the cells were determined by immunoblotting for TTase and α-crystallin proteins and by immunohistochemistry for glutathionylated proteins. Cell proliferation was examined by 3-(4,5-dimethyl-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium and BrdU analysis, and cell apoptosis after H₂O₂ stress was assessed by fluorescence-activated cell sorter analysis. Reloading of TTase protein into the TTase ^−/− cells was achieved with reagent.

results. Primary LEC cultures obtained from wild-type (TTase ^+/+) and TTase ^−/− mice were characterized and found to contain lens-specific α-crystallin protein. Western blot analysis confirmed the absence of TTase protein in the TTase ^−/− cells and its presence in the wild-type cells. TTase ^−/− LECs had significantly lower levels of glutathione (GSH) and protein thiols with extensive elevation of glutathionylated proteins, and they exhibited less resistance to oxidative stress than did TTase ^+/+ cells. These cells were less viable and more apoptotic, and they had a reduced ability to remove H₂O₂ after challenge with low levels of H₂O₂. Reloading of purified TTase into the TTase ^−/− cells restored the antioxidant function in TTase ^−/− cells to a near normal state.

conclusions. These findings confirm the importance of TTase in regulating redox homeostasis and suggest a new physiological function in controlling cell proliferation in the lens epithelial cells.

Thioltransferase (TTase), also known as glutaredoxin, is a thiol-disulfide oxidoreductase and has cytosolic (TTase-1 or Grx1) and mitochondrial (TTase-2 or Grx2) isoforms. ¹ TTase-1 is a ubiquitous heat-stable cytosolic protein with a molecular mass of 11.8 kDa. ^{2 3 4} TTase-1 receives reducing equivalents directly from glutathione (GSH) and catalyzes thiol/disulfide exchange reactions. ^{2 3} The active site of the enzyme contains two redox-active cysteine residues, Cys-Pro-Tyr(Phe)-Cys, and is conserved from Escherichia coli to mammals. ² TTase-1 is a multifunctional enzyme involved in several important physiological processes in eukaryotic and prokaryotic cells, ^{5 6} including the donation of reducing equivalents to ribonucleotide reductase for DNA biosynthesis, ⁷ the dethiolation of protein-thiol mixed disulfides, ^{8 9} the regeneration of oxidatively damaged key glycolytic and oxidation defense enzymes, ^{8 10} the deiodination of thyroxin, ¹¹ the reduction of oxidized ascorbate, ^{12 13} and the regulation of cellular signal transduction. ¹⁴  

Oxidative stress has been implicated in age-related cataract formation. ^{15 16 17} Excessive generation of reactive oxygen species (ROS) molecules from either the environment or from mitochondria of the lens epithelial cells can damage cellular macromolecules such as proteins, DNA, and lipids, leading to opacification of the lens and compromising lens transparency. Hence, lens cells are equipped with a variety of antioxidants, oxidation defense, and repair systems that can effectively remove ROS from cells and repair damaged macromolecules. Several known protein thiol oxidation damage repair systems are present in lens cells and other cells in mammals. These include the GSH-dependent TTase system for reducing protein-thiol mixed disulfides ¹⁸ and the NADPH-dependent thioredoxin/thioredoxin reductase system for reducing protein-protein disulfides. ¹⁹ Additionally, oxidized methionine in the protein can be reduced by methionine sulfoxide reductase. ²⁰ All three systems contribute individually or synergistically to keep the lens in an overall reduced state, thus helping to maintain lens transparency. 

In the present study, we investigated the cytosolic TTase/GSH system in lens cells as a protein repair system using TTase ^−/− mouse lens epithelial cells (LECs) as a model in which TTase-1 has been deleted without disrupting the mitochondrial Grx2 isoform. We studied cell integrity, cell proliferation, oxidant accumulation, and various types of oxidative damages that take place in TTase ^−/− LECs when challenged with oxidative stress conditions. Our results show that the deletion of cytosolic TTase caused a significant decrease in mouse LEC proliferation and increased sensitivity toward H₂O₂-induced oxidative stress. For simplicity, throughout this article TTase will be referred to as cytosolic TTase or TTase-1. 

Materials and Methods

Materials

2′, 7′-Dichlorodihydrofluorescein diacetate (DCFH-DA) was obtained from Molecular Probes (Eugene, OR). Protein transfection reagent was from Gene Therapy Systems (BioPORTER; San Diego, CA). Rabbit polyclonal antibody for αA-crystallin was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). All other chemicals and reagents were standard commercial products of analytical grade. 

Generation of TTase ^−/− Mice

DNA sequencing data showed that the mouse TTase gene consists of three exons. The protein coding region is located in the first two exons, whereas exon 3 contains the 3′ untranslated region of the gene. Briefly, a genomic clone containing the mouse TTase gene was isolated from the strain 129SV mouse genomic library. A gene-targeting vector was constructed by replacing exon 2 and some of the sequences in introns 1 and 2 with a neomycin-resistance cassette. ²¹ The targeting vector was linearized with NotI enzyme and then transfected into R1 embryonic stem cells ²² derived from (129SV × 129SVJ) F1 mice (a generous gift from Andras Nagy, Mount Sinai Hospital, Toronto, ON, Canada). Of the 298 clones screened, 51 were identified to contain the targeted TTase allele. One targeted clone (clone 89) was microinjected into blastocysts from C57BL/6 mice according to the standard protocol. ²³ Chimeric mice derived from microinjection transmitted 129SV chromosomes into the offspring. Homozygous TTase knockout mice (TTase ^−/−) were derived from breeding of heterozygous knockout mice (TTase ^+/−) in a mixed genetic background between (129SV × 129SVJ) F1 and C57BL/6mice. Homozygous knockout mice appeared normal and healthy at 15 months of age. More details of the generation and characterization of the TTase ^−/− mice model will be published elsewhere. 

Northern and Western Blot Analyses of Tissue from Wild-Type and TTase Knockout Mice

Brain, heart, and kidney tissue from wild-type (TTase ^+/+), heterozygous TTase (TTase ^+/−), and homozygous TTase knockout (TTase ^−/−) mice were homogenized in guanidinium isothiocyanate solution, and total RNA was isolated according to the method described by Chirgwin et al. ²⁴ Thirty micrograms of total RNA were denatured with glyoxal and subjected to blot analysis according to the procedures described by Thomas. ²⁵ Eye tissue from TTase ^+/+ and TTase ^−/− mice was analyzed by the same procedure for the Northern blot of TTase and G3PD (control). For protein analysis, the tissue was homogenized in 50 mM potassium phosphate buffer (pH 7.8), containing 0.1% Triton X-100, 3% glycerol, and 1 mM phenylmethylsulfonyl fluoride with a homogenizer (Polytron; Glen Mills, Clifton, NJ), followed by sonication. The homogenates were clarified by centrifugation at 20,000g for 15 minutes and were stored at −70°C. Protein concentrations of tissue homogenates were determined by the use of a bicinchoninic acid protein assay kit (Pierce, Rockford, IL). Thirty micrograms of tissue protein were separated on a SDS-polyacrylamide gel for protein blot analysis. The protein blot membrane was reacted with polyclonal antibodies against the whole human TTase protein generated in rabbits (Lou MF, unpublished results, 2007) and goat (American Diagnostica, Greenwich, CT). 

Purification of Recombinant Human TTase

Human cDNA for TTase was cloned into pET21d (+) vector, expressed in E. coli, and purified using His-bind column (Novagen, Madison, WI) according to the manufacturer’s protocol. 

Preparation of Anti–TTase Antibody

Affinity-purified rabbit anti–TTase antibody was prepared using purified whole recombinant human TTase protein (Bethyl Laboratories, Montgomery, TX), as reported previously. ²⁶  

Primary Mouse LEC Culture

Primary LEC cultures were prepared from four or five 2-week old wild-type (TTase ^+/+) and TTase ^−/− mice in a 129SV × C57BL/6 mixed background. Mouse lens capsules with attached epithelial layers were obtained under sterile conditions and were placed in a 96-well plate containing 75 μL MEM (with 20% FBS and 50 μg/mL gentamicin)/well. This tissue was incubated for 2 to 3 weeks in a humid atmosphere with 5% CO₂ at 37°C. Medium was changed every week, and the cell plate was observed under an inverted light microscope to detect cell growth. Primary LECs that grew out of the lens capsules were subdivided and seeded at 5 × 10⁵ cells in 60-mm tissue culture dishes. 

Enzyme and Other Assays

TTase activity was assayed using a previously described method. ²⁶ Protein concentrations in cell lysates were determined by the BCA method according to the manufacturer’s protocol (Pierce Chemical, Rockford, IL), with bovine serum albumin as the standard. Lactate dehydrogenase activity was measured as described earlier. ²⁷ GSH concentrations in cell lysates were determined using Ellman reagent. ²⁸ Hydrogen peroxide concentration in cell culture medium was measured using the method of Hildebrant et al. ²⁹ Glyceraldehyde-3-phosphate dehydrogenase (G3PD) activity was assayed according to the method of Bergmeyer et al. ³⁰  

Western Blot Analysis

Proteins in cell lysates or tissue homogenates were separated by 12% SDS-PAGE and transferred to membrane (TransBlot; Bio-Rad, Hercules, CA) that was probed with anti–TTase antibodies diluted in TBST buffer (10 mM, pH 7.5, Tris-HCl, 100 mM NaCl, 0.1% Tween 20) followed by goat anti–rabbit IgG horseradish peroxidase (Santa Cruz Biotechnology). Immunodetection was performed with chemiluminescent reagent (Santa Cruz Biotechnology). The immunoblot was analyzed with an imaging system of Fluor-S MAX MultImager (Bio-Rad). 

Loading of TTase ^−/− Mouse LECs with Purified Recombinant TTase

Purified human TTase was loaded into TTase ^−/− mouse LECs using a protein transfection reagent according to the manufacturer’s protocol (BioPORTER; Gene Therapy Systems). This protein transfection reagent is a lipid-mediated protein delivery system that delivers protein in a functionally active form into the cytoplasm of cells. With the use of an FITC-tagged antibody (provided with the kit), the delivery efficiency of the reagent was determined by flow cytometry to be 78%. Briefly, 20 μg recombinant TTase mixed with the reagent (two BioPORTER [Gene Therapy Systems] tubes, each containing 10 μg purified recombinant TTase) was added to serum-deprived TTase ^−/− mouse LECs seeded in six-well plates. A control experiment was conducted using a nonrelated protein, β-galactosidase, provided with the protein transfection kit in accordance with the manufacturer’s instructions. After 4 hours of incubation, the cells were washed and used for experiments. 

Flow Cytometric Quantification of DCF Fluorescence in H₂O₂-Treated Wild-Type, TTase ^−/−, and Recombinant TTase-Loaded TTase ^−/− Mouse Primary LECs

The membrane-permeable fluorescent dye DCFH-DA crosses the cell membrane and undergoes deacetylation by intracellular esterases, producing the nonfluorescent compound DCFH, which is trapped inside the cells. Oxidation of DCFH by ROS, including H₂O₂, produces the highly fluorescent DCF. Externally added H₂O₂ freely diffuses into the cells and oxidizes DCFH to give DCF fluorescence. Hence, DCF fluorescence intensity inside the cells is proportional to the intracellular H₂O₂ level. H₂O₂ levels in cells can be quantified by flow cytometric determination of the cellular DCF fluorescence. ³¹ Wild-type, TTase ^−/−, and recombinant TTase-loaded TTase ^−/− mouse LECs were removed by trypsinization, washed with Ca²⁺- and Mg²⁺-free phosphate-buffered saline (PBS; Sigma-Aldrich, St. Louis, MO), and resuspended in PBS containing 50 μM DCFH-DA. After 5 minutes the cells were again washed with PBS before the basal DCF fluorescence level was determined by flow cytometry (FACScan; Becton Dickinson, San Jose, CA). After a baseline was acquired, 50 μM H₂O₂ was added to all cell groups, and DCF fluorescence levels were determined at given time points. Because this 50 μM H₂O₂ did not cause oversaturation of fluorescence in the cells at the beginning of the experiment, this concentration was chosen for the flow cytometry study. Excitation was 488 nm, and green emission was measured at 525 ± 10 nm on 10,000 gated cells using log amplification. The arithmetic mean fluorescence channel was derived using flow cytometry software (CellQuest; Becton Dickinson, San Jose, CA). 

Cell Proliferation and Viability Assays

Cell proliferation was determined by manual cell counting using a hemacytometer and by MTS tetrazolium reagent using assay kit (CellTiter 96 AQ_ueous One Solution Cell Proliferation; Promega, Madison, WI) according to the manufacturer’s protocols. MTS color change was monitored (EL_X 800 Universal Microplate Reader; Bio-Tek Instruments, Winooski, VT) at 492 nm absorbance. Cell viability of wild-type and TTase ^−/− mouse LECs was determined after treatment with a bolus of 100 μM H₂O₂ followed by the MTS assay. Cells without H₂O₂ treatment were used as controls. 

Determination of Cell Proliferation by 5-Bromodeoxyuridine Incorporation Assay

Cell proliferation of TTase ^+/+ and TTase ^−/− LECs was determined with the use of a cell proliferation ELISA 5-bromodeoxyuridine (BrdU) chemiluminescence kit (Roche Applied Sciences, Indianapolis, IN). BrdU, a chemical analog of thymidine, was used in BrdU incorporation assay for cell proliferation, in accordance with the manufacturer’s instructions. Briefly, cells were seeded onto black 96-well plates overnight, and medium was replaced with 2% FBS-containing MEM for another overnight. BrdU-labeling solution (final concentration, 10 μM) and PDGF (1 ng/mL) were added to serum-starved cells and incubated for 1 hour. After labeling, cells were fixed and incubated with anti–BrdU-POD. Excess antibody was removed by washing the cells with 1× PBS, and chemiluminescence was determined with a luminometer equipped with automatic substrate injectors (FLUOstar OPTIMA; BMG Labtech GmbH, Offenburg, Germany). Specific chemiluminescence was expressed as RLU/s. 

Comparison of the Effect of H₂O₂ on G3PD Activity in Wild-Type, TTase ^−/−, and Recombinant TTase-Loaded TTase ^−/− Mouse LECs

Wild-type, TTase ^−/−, and recombinant TTase-loaded TTase ^−/− mouse LECs were gradually serum starved by incubation overnight in medium containing 1% FBS and then in serum-free medium for 30 minutes before exposure to a bolus of 150 μM H₂O₂. Cells were removed by trypsinization at 15, 30, 60, and 120 minutes, washed with PBS, and lysed with a lysis buffer. Cell lysates were centrifuged at 13,000 rpm for 10 minutes, and G3PD activity in the supernatant was determined as described. 

Effect of H₂O₂ on Apoptosis in Wild-Type and TTase ^−/− Mouse LECs

Wild-type and TTase ^−/− mouse LECs at 60% to 70% confluence were incubated for 1 hour in serum-free medium with and without 50 μM H₂O₂, a concentration that did not cause severe damage to the cells, which could be recultured for apoptosis studies. Cells were then incubated for another 16 hours in medium containing 20% FBS before analysis for apoptosis. To quantify apoptosis, an Annexin V apoptosis detection kit was used (Biovision, Mountain View, CA). Annexin V detects the translocation of membrane phospholipid phosphatidylserine from the inner face of the plasma membrane to the cell surface after initiating apoptosis. ³² H₂O₂-treated cells were trypsinized and were washed once with 20% FBS-containing MEM and once with PBS. The cells were then stained with Annexin V⁻, and 10,000 cells were analyzed by flow cytometry. 

Effect of H₂O₂ on Cell Membrane Integrity

The integrity of the cell membrane was measured by following the extent of cytosolic lactate dehydrogenase (LDH) release into the medium. Wild-type and TTase ^−/− mouse LECs were gradually serum starved, as described, before exposure to a bolus of 150 μM H₂O₂ for 0, 15, 30, 60, and 90 minutes. The cell lysates were centrifuged, and the supernatant was analyzed for LDH activity. 

Immunohistochemistry

For immunohistochemistry, cells were grown on culture slides (BD Biosciences, Bedford, MA) and fixed in freshly prepared 4% paraformaldehyde for 30 minutes. Then the cells were washed three times with PBS for 10 minutes and were blocked overnight at 4°C with 3% gelatin in PBS. Cells were covered with a monoclonal antibody (1:100 dilution) against protein-GSH mixed disulfide (Virogen, Watertown, MA) and were incubated for 2.5 hours at room temperature in a humid chamber. After this, cells were washed three times with PBS containing 0.1% Triton-X-100 for 10 minutes and were incubated with Cy5-conjugated fluorescent secondary antibodies (Molecular Probes) for 2 hours at room temperature in a dark, humid chamber. Finally, the cells were washed three times with PBS containing 0.1% Triton-X-100 and were mounted with a coverslip using anti–fade agent (MOI Biomedia, Foster City, CA). Fluorescent and phase-contrast pictures of cells were acquired with a confocal laser scanning microscope (MRC1024ES; Bio-Rad) and a 60× oil immersion objective. 

Statistical Analysis

All tests were two-sided, and P < 0.05 was considered significant; one-way ANOVA and two-way ANOVA with replication were used to analyze the data. Subsequent to significant ANOVAs, pairwise comparisons between the TTase ^+/+ cells and TTase ^−/− cells were performed by using Dunnett modified t-test. 

Results

Northern and Western Blot Analyses of Tissue from Wild-Type and TTase Knockout Mice

As shown in Figure 1A , brain, heart, kidney, and eye tissue from a wild-type mouse showed a clear mRNA band for TTase, whereas the TTase ^+/− tissue had a considerably weaker mRNA band. Interestingly, a smaller TTase mRNA band was expressed in the TTase ^−/− mouse. For some reason, this smaller TTase was more intense in the TTase ^−/− eye tissue, even though G3PD control indicated that an equal amount of RNA was used in each sample (Fig. 1A) . Because exon 2 of the TTase allele was deleted in the knockout mice, the mRNA expressed in the tissue of the TTase ^−/− mouse is likely a fused product of exon 1, with downstream sequences such as exon 3. However, Western blot analysis (Fig. 1B)of the brain, heart, and kidney tissue observed only TTase-positive bands in the tissue of wild-type and TTase ^+/− animals; no TTase full-sized protein was seen in tissue of TTase ^−/− mouse, using two sources of anti–human TTase antibody, one custom made by Bethyl Laboratories (data not shown) and another from a commercial source (American Diagnostics). 

Characterization and Validation of the TTase ^+/+ and TTase ^−/− Primary Mouse LECs

Primary cell cultures from TTase ^+/+ (129SV) and TTase ^−/− mouse lens epithelia (in a 129SV and C57BL/6 mixed background) were established and used for characterization and validation. Western blot analysis of both cell types showed the presence of the lens-specific αA-crystallin protein (Fig. 2A) , confirming that the cells obtained were indeed LECs.³³ Western blot analysis also showed that TTase was present in TTase ^+/+ cells but was not detectable in TTase ^−/− cells (Fig. 2B) . TTase activity in TTase ^−/− LECs was approximately 80% less than that of TTase ^+/+ LECs (Fig. 2C) . However, loading TTase ^−/− LECs with human recombinant TTase using protein transfection reagent (BioPORTER; Gene Therapy Systems) resulted in increased TTase activity in TTase ^−/− cells (Fig. 2D) . This increased TTase activity in TTase-loaded TTase ^−/− cells was shown to be nearly fivefold greater than in TTase ^−/− cells and 25% higher than normal cellular TTase activity. In contrast, TTase activity was not changed in TTase ^−/− cells when β-galactosidase, a non-TTase protein, was delivered by the same procedure (BioPORTER; Gene Therapy Systems; Fig. 2D ). 

Cell Proliferation and Morphology

Lens epithelial cells from the TTase ^+/+ mouse were elongated, but TTase ^−/− cells were round and had more volume (data not shown). These cells were tightly adherent to the culture plate and were very slow in spreading, indicating impaired cell migration. Cell proliferation assessed by manual cell counting showed a significant difference (P < 0.01) between the TTase ^+/+ and TTase ^−/− LEC cells (Fig. 3A) . Deleting TTase had a negative effect on cell proliferation in which the approximate cell doubling time was increased from 13 hours for TTase ^+/+ cells to 33 hours. Cell proliferation determined by the BrdU method for DNA biosynthesis also showed a significant decrease (*P < 0.05 and **P < 0.02) in TTase ^−/− cells (Fig. 3B) . 

Cell Viability and Membrane Integrity in the Presence of H₂O₂

Figure 4Ashows a comparison of cell viability of TTase ^+/+ and TTase ^−/− LECs after treatment with a bolus of 100 μM H₂O_2. Hydrogen peroxide caused an initial 50% decrease in cell viability 1 hour after treatment in both cell types, as determined by the MTS method. However, monitoring cell viability up to 6 hours after treatment showed a significant decrease (P < 0.05) in cell viability in TTase ^−/− cells compared with TTase ^+/+ cells. 

Cell membrane integrity was determined in TTase ^+/+ and TTase ^−/− LECs after H₂O₂ challenge using the released LDH enzyme in the cell culture medium as an indicator. TTase ^+/+ and TTase ^−/− LECs were treated with a bolus of 150 μM H₂O₂, and LDH activity in the cell culture medium was assayed at given time points (Fig. 4B) . Significant LDH activity was detected in culture medium 15 minutes after the H₂O₂ challenge; thereafter, a progressive increase in LDH release was observed in TTase ^+/+ and TTase ^−/− mouse LECs up to 90 minutes. However, LDH released from TTase ^−/− mouse LECs was much higher (P < 0.05) than from TTase ^+/+ cells, indicating that the lack of TTase in cytosol affected cell membrane integrity. 

GSH, Protein Thiols, and Protein-GSH Mixed Disulfides in TTase ^+/+, TTase ^−/−, and TTase-Loaded TTase ^−/− Cells

Cells that lacked TTase appeared to affect the redox status in the cells. As shown in Figure 5A , GSH content in TTase ^−/− cells was only 40% that of TTase ^+/+ cells. It is well documented that oxidized GSH (GSSG) forms mixed disulfides with protein thiols; thus, protein thiol contents in TTase ^+/+ and TTase ^−/− LECs were measured. As shown in Figure 5B , the protein thiol level in TTase ^−/− cells was only 70% that in TTase ^+/+ cells (P < 0.05). Protein-GSH mixed disulfide levels in TTase ^+/+and TTase ^−/− LECs were measured by immunohistochemistry with the use of mouse anti–monoclonal antibody (Virogen, Watertown, MA) and Cy5-labeled second antibody and were visualized by confocal microscopy. As shown in Figure 5C , fluorescence intensity was significantly higher in TTase ^−/− cells than in the TTase ^+/+ cells, indicating that the lack of TTase resulted in an accumulation of protein-GSH mixed disulfides in the cells. 

The effect of oxidative stress on the level of GSH in both cell types was compared by exposing cells to a bolus of 150 μM H₂O₂. The oxidant had no effect on GSH level in TTase ^+/+ cells during the first 30 minutes. After that, there was a gradual decrease with less than 10% GSH loss at 90 minutes. However, the same condition caused a marked (approximately 40%) decline in GSH level in TTase ^−/− cells within 15 minutes. This decline continued throughout the experimental period of 90 minutes; results are summarized in Figure 5D . 

H₂O₂ Detoxification in TTase ^+/+, TTase ^−/−, and TTase-Loaded TTase ^−/− Cells

ROS marker DCF fluorescence in TTase ^+/+ and TTase ^−/− LECs was measured by fluorescence-activated cell sorter (FACS). H₂O₂ at 50 μM was chosen for this detoxification study because this low level did not generate fluorescence overload in the cells that would interfere with the FACS analysis. As shown in Figure 6 , neither type of cell showed endogenous ROS (DCF fluorescence) at the time. After the addition of a bolus of H₂O₂ (50 μM), the DCF fluorescence gradually increased in both groups. Mean DCF fluorescence in the TTase ^−/− group was progressively elevated to 120 mean value within 5.5 minutes, whereas TTase ^+/+ cells showed only 20 to 30 mean fluorescence value, which was maintained throughout the 5.5-minute experimental period, indicating that TTase ^+/+ LECs have an ability to detoxify H₂O₂ for cellular protection. Purified human recombinant TTase was loaded into TTase ^−/− LECs using the protein transfection reagent to examine whether reloading TTase into the TTase ^−/− LECs would restore the H₂O₂ detoxification efficiency of TTase ^−/− cells. As shown in Figure 2D , loading purified TTase into TTase ^−/− LECs successfully increased TTase activity in those cells. TTase ^−/− cells reloaded with purified TTase behaved similarly to TTase ^+/+ cells and showed little difference in accumulated ROS (Fig. 6) . This strongly suggested that the depletion of TTase from LECs impaired the ability of the cells to effectively detoxify H₂O₂ and probably all ROS. 

Effect of H₂O₂ on Glyceraldehyde 3-Phosphate Dehydrogenase Activity in TTase ^+/+ and TTase ^−/− LECs

Glyceraldehyde 3-phosphate dehydrogenase (G3PD) is easily oxidized and inactivated by forming glutathionylated G3PD; the inactivated G3PD activity can be restored by TTase. ³⁴ Therefore, we examined the status of G3PD activity in TTase ^+/+ and TTase ^−/− cells after challenge with oxidative stress. Treating both TTase ^+/+ and TTase ^−/− LECs with a bolus of 150 μM H₂O₂ caused significant reduction (P < 0.05) in G3PD activity in both cell types (Fig. 7A) , but the activity loss (30% at 30 minutes) in TTase ^+/+ cells was gradually restored to nearly 90% after 120 minutes (Fig. 7A) . However, in TTase ^−/− cells, the G3PD activity loss was earlier and more severe (approximately 70% loss within 15 minutes) with much less reactivation (65% of the original activity) than in TTase ^+/+ cells (Fig. 7A) . These data suggest that the depletion of TTase caused a significant decrease in the cell’s ability to protect G3PD from H₂O₂-mediated inactivation. Figure 7Bsummarizes the percentage of remaining G3PD activity (compared with untreated control) in TTase ^+/+ and TTase ^−/− cells with and without reloading of pure TTase 15 minutes after H₂O₂ treatment. The reintroduction of TTase into TTase ^−/− cells significantly improved the ability of these cells to protect G3PD from H₂O₂ inactivation (Fig. 7B) . 

Effect of H₂O₂ on Apoptosis in Wild-Type and TTase ^−/− Mouse LECs

Under normal growth conditions, approximately 10% TTase ^+/+ cells underwent early stage of apoptosis, whereas cells lacking TTase showed nearly 30% apoptosis (Fig. 8) , indicating that TTase may have some protective function against early-stage apoptosis. We used a 50-μM bolus of H₂O₂ to induce apoptosis in TTase ^+/+ and TTase ^−/− LECs because this level of stress still allows the cells to reattach 60 minutes after treatment. Under this stress condition, there was no change in apoptosis in TTase ^+/+ cells (Fig. 8) . However, treating TTase ^−/− cells with a 50-μM bolus of H₂O₂ induced nearly 75% of the cells to enter an early stage of apoptosis (Fig. 8) , further suggesting that TTase protects cells from H₂O₂-induced apoptosis. 

Discussion

The primary lens epithelial cell cultures used in this study were established from TTase ^+/+ and TTase ^−/− mouse lens epithelia and were validated to contain the lens-specific αA-crystallin protein. ³³ Because the tissue of TTase ^−/− mice, including the eye, expresses a truncated form of TTase mRNA from the fusion of exon 1 and exon 3 sequences, it is conceivable that the same TTase mRNA is also expressed in LECs isolated from these mice. The upregulation of truncated TTase mRNA in the null eye probably reflects a compensatory upregulation of gene activity in response to loss of TTase function. The mouse TTase is 107 amino acids long, and its C-terminal 38 amino acids are encoded by exon 2. The exon 1–exon 3 fusion mRNA expressed from the targeted TTase allele would allow extension of the reading frame into the noncoding region of exon 3 to include 17 more amino acids, resulting in a protein that is 21 amino acids shorter (because of the deletion of exon 2) than in the wild-type protein. The active site of TTase protein (23-Cys-Pro-Tyr-Cys-26) is encoded by exon 1. However, 13 of 14 amino acids in the GSH-binding site, which are encoded by the exon 2 sequence, would be missing in the mutant protein. This may greatly affect the structure and function of the mutant TTase protein in the knockout mice. Toward this end, we examined the expression of TTase protein in various tissue and LECs from knockout mice using two different preparations of antibodies against the human TTase (both antibodies were made against the entire length of TTase protein). These studies show that no TTase protein could be detected in tissue or LECs of homozygous knockout mice (Figs. 1B 2B) . We also could not detect any truncated TTase protein that might be expressed from the mutated TTase gene, suggesting that if the mutant TTase protein were produced, because of its extreme short half-life, the protein would have been degraded in the process. Even if it were not degraded, the phenomenon of oversensitivity to oxidation exhibited by the TTase ^−/− cells could not have been caused by the truncated TTase proteins because TTase ^−/− cells could be normalized when pure TTase protein was delivered into the TTase ^−/− cells. This strongly suggests that the abnormal property of the TTase ^−/− cells is caused by a lack of TTase-1, not by possible toxicity from the truncated TTase protein in the lenses of TTase knockout mice. 

The residual TTase activity shown in the TTase ^−/− cells likely resulted from a mitochondrial isoform of TTase-2 (glutaredoxin 2), a recently identified member of the TTase family. ³⁵ However, TTase-2 is too low to be seen by Western blot analysis under our current conditions. These experiments established that the primary cell cultures we used were mouse LECs and that the cells obtained from TTase ^−/− mouse lenses were devoid of TTase. Because wild-type mice (TTase ^+/+) with the same genetic background as that of TTase ^−/− mice were used for isolation of LECs, the increased susceptibility of TTase ^−/− cells to oxidative stress observed in our study is believed to have resulted from the deficiency of TTase but not from the differences of the mouse genetic background. 

TTase depletion in lens epithelial cells showed a wide range of effects to the function and survival of the cells, including slower cell proliferation, decreased cell viability, and increased sensitivity to oxidative stress-induced damage. In comparison with the wild-type LECs, TTase ^−/− cells were rounder and tightly attached to the culture plate, and they migrated very slowly (data not shown). It has been reported that actin deglutathionylation plays a key role in growth factor–mediated actin polymerization, translocation, and reorganization near the cell periphery, which are important steps in cell migration. ³⁶ The slow cell migration we observed in TTase ^−/− cells may be the result of impaired deglutathionylation of actin. As assessed by manual cell counting and BrdU incorporation, deleting TTase extensively lowered the rate of cell proliferation. DNA biosynthesis was low in TTase ^−/− cells, as measured by BrdU incorporation (Fig. 3B) . It is likely that the impaired cell proliferation observed in TTase ^−/− cells might have been the result of decreased DNA biosynthesis, which requires TTase as a hydrogen donor for ribonucleotide reductase. ⁷  

Earlier studies carried out in our laboratory have shown that inhibiting TTase activity in HLE B3 cells by cadmium before H₂O₂ treatment caused a marked decrease in cell viability. ³⁴ TTase ^−/− cells also exhibited impaired cell viability and cell membrane integrity under conditions of H₂O₂ stress (Figs. 4A 4B) , confirming the importance of TTase in maintaining cell viability and protecting the cell membrane from oxidant-induced damage. GSH and protein thiol levels were lower in TTase ^−/− cells than in TTase ^+/+ cells (Figs. 5A 5B) . We hypothesized that those low levels of GSH and protein thiols resulted from the accumulation of glutathionylated proteins in TTase ^−/− cells because deglutathionylation requires TTase. With the use of monoclonal anti–GSH antibody, which specifically recognizes glutathionylated proteins, we showed that the glutathionylated protein level was much higher in TTase^−/− cells than in TTase ^+/+ cells (Fig. 5C) . Furthermore, exposing both cell types to H₂O₂ stress showed a marginal GSH depletion effect only in TTase ^+/+cells but more extensive and prolonged GSH loss in the TTase ^−/− cells, further indicating that protein deglutathionylation was indeed severely impaired by the absence of TTase in the cells. Progressive loss of GSH and protein thiols and the accumulation of glutathionylated proteins are the hallmark of aging and cataractous lens epithelial cells. ¹⁶ In this regard, TTase ^−/− cells appear to resemble the epithelial cells from aging and cataractous lenses. 

Compared with TTase ^+/+ cells, TTase ^−/− cells were inefficient in decomposing H₂O₂. Interestingly, loading TTase ^−/− cells with purified recombinant TTase resulted in normalization of the ability of these TTase knockout cells to decompose H₂O₂ (Fig. 6) . This inefficiency might have resulted from low GSH levels in TTase ^−/− cells given that GSH is essential for the function of glutathione peroxidase in cellular H₂O₂ detoxification. Even though TTase itself has no peroxidase activity, it has been shown to enhance the activity of certain peroxiredoxins in plant cells. ³⁷ Therefore, the effect of TTase on cellular H₂O₂ removal, as we observed in this study, was probably contributed by a similar TTase property in mammalian cells. 

Protection and regeneration of G3PD activity from H₂O₂-induced oxidative stress was also markedly reduced in TTase ^−/− cells compared with TTase ^+/+cells. Loading TTase ^−/− cells with recombinant TTase, however, significantly improved cellular ability to protect G3PD from H₂O₂-mediated inactivation (Figs. 7A 7B) . Reactivation of oxidatively inactivated G3PD is a known function of TTase. G3PD is a key glycolytic enzyme for ATP supply in the cells with critical cysteine residues highly sensitive to oxidants such as H₂O₂. The molecular nature of the H₂O₂-mediated inactivation of purified G3PD in vitro has been shown to be the oxidation of the critical cysteine residues to monothiol-oxidation products such as sulfenic acid. ³⁸ In many cell types, however, including the lens epithelial cells, it has been demonstrated that H₂O₂ inactivated G3PD by inducing the formation of protein-GSH mixed disulfide ^{34 39} and that TTase can effectively reactivate G3PD by removing the protein-GSH mixed disulfide at the active site. ³⁴ The observation that TTase ^−/− cells had impaired cell viability and cell membrane integrity might have been the result of impaired H₂O₂ decomposition and protein/enzyme reactivation. 

Furthermore, previously published data implicate that TTase is an antiapoptotic protein. TTase obtained from E. coli protected cerebellar granule neurons from dopamine-induced apoptosis by activating Ras-phosphoinositide 3-kinase and jun N-terminal kinase pathways. ⁴⁰ Conversely, TTase inhibition by cadmium led to the initiation of apoptosis, ⁴¹ suggesting an important role for TTase in protecting cells against apoptosis. Our studies using TTase ^−/− cells confirm those previous findings and establish TTase as an important cellular enzyme that confers protection against oxidant-induced apoptosis. 

Ho et al. ²³ recently reported that a TTase knockout mouse line with deleted exons 1 to 3 did not display any extra susceptibility to acute oxidation-induced injury in heart and lung when the animal was subjected to ischemia/reperfusion or hyperoxia. Cells isolated from embryonic fibroblasts of the knockout mouse were selectively sensitive to oxidants such as diquat and paraquat but not to H₂O₂ or diamide. These contrasting results suggest that different tissue and cell types may depend differently on TTase protection. Because of the unique location of the eye, the lens in particular is more sensitive to oxidative stress and consequently is more sensitive to the loss of any redox regulating enzymes, such as TTase. Our current findings further substantiate the unique oxidative-sensitive nature of the lens tissue. 

In summary our study provides strong evidence that TTase is an important enzyme in protecting LECs from oxidative damage through its involvement in reducing H₂O₂ level, regenerating oxidatively damaged proteins/enzymes, and acting as an antiapoptotic agent. 

³  These authors contributed equally to the work presented here and should therefore be regarded as equivalent authors.

⁴  Present affiliation: St. Erik’s Eye Hospital, Karolinska Institutet, Stockholm, Sweden.

 Presented at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, April 2005.

 Supported by American Heart Association Grant 0655631Z (Y-SH), National Institutes of Health Grant RO1 EY10595 (MFL), and a postdoctoral fellowship from the Swedish Medical Research Council (SL).

 Submitted for publication October 30, 2007; revised April 2, 2008; accepted August 20, 2008.

 Disclosure: S. Löfgren, None; M.R. Fernando, None; K.-Y. Xing, None; Y. Wang, None; C.A. Kuszynski, None; Y.-S. Ho, None; M.F. Lou, 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.

 

Northern and Western blot analyses of the tissue from TTase ^+/+, TTase ^+/−, and TTase ^−/− mouse. (A) Northern blot analysis of TTase mRNA expression in eye, brain, heart, and kidney tissue of wild-type and TTase knockout mice. The Northern blot membrane was hybridized with a full-length mouse TTase cDNA. G3PD mRNA analysis was used as a control for equal sample application on gel. (B) Western blot analysis of TTase in brain, heart, and kidney tissue of wild-type and TTase knockout mice. The protein blot membrane was initially reacted with rabbit anti–human TTase antibodies and then re-reacted with rabbit copper-zinc superoxide dismutase (CuZnSOD) anti–human antibodies. A duplicate protein blot membrane has also been reacted with a goat TTase-1 anti–human antiserum. (A, B) +/+, +/−, and −/− represent wild-type, heterozygous TTase knockout, and homozygous TTase knockout mice, respectively.

Figure 2.

 

Characterization and validation of TTase ^+/+ and TTase ^−/− mouse LECs. Immunoblot analysis of TTase ^+/+ and TTase ^−/− mouse LECs. Mouse LEC layers were obtained from wild-type and TTase ^−/− mouse and homogenized. These homogenates were analyzed by SDS-PAGE and subjected to immunoblot analysis using either anti–α-A crystalline antibody or anti–human TTase antibody. (A) Detection of α-A crystallin in wild-type and TTase ^−/− LECs. (B) Detection of TTase in wild-type and TTase ^−/− LECs. Lane 1: TTase ^−/− LECs. Lane 2: TTase ^+/+ LECs. (C) TTase enzyme activity in TTase ^+/+ and TTase ^−/− LEC lysates. Error bars indicate SD. n = 5. *Significant difference from the TTase ^+/+ cells (P < 0.05). (D) TTase activity in TTase ^+/+, TTase ^−/−, TTase-loaded TTase ^−/−, and β-galactosidase–loaded TTase ^−/− mouse LECs. Human recombinant TTase (20 μg) and β-galactosidase (20 μg) were introduced into TTase ^−/− cells using protein transfection reagent. Error bars indicate SD, n = 5. *P < 0.05; **P < 0.05.

Figure 3.

 

Comparison of cell proliferation of TTase ^+/+ and TTase ^−/− mouse LECs in culture. (A) TTase ^+/+ (▪) and TTase ^−/− (⧫) LEC proliferation determined by manual cell counting. Error bars indicate SD; n = 5. *P < 0.01. (B) Comparison of TTase ^+/+ (▪) and TTase ^−/− (□) LEC proliferation by BrdU incorporation. *P < 0.05; **P < 0.02.

Figure 4.

 

Effect of H₂O₂ on cell viability and LDH release in TTase ^+/+ and TTase ^−/− LECs. (A) Determination of TTase ^+/+ (▪) and TTase ^−/− (□) LEC viability by MTS reagent. TTase ^+/+ and TTase ^−/− LECs were treated with a bolus of 100 μM H₂O₂ for indicated times, and then MTS color change was measured using an ELISA plate reader at 490 nm. MTS reagent was added to the cells 1 hour before absorbance measurements. (B) LDH release of TTase ^+/+ (⧫) and TTase ^−/− (▪) LECs. Cells were treated with a bolus of 150 μM H₂O₂ for the indicated times, and the cell culture medium was collected for LDH activity measurements. LDH activity was determined as described and expressed as a percentage of total LDH activity in the cells. Error bars indicate SD. n = 5. *Significant difference from the TTase ^+/+ cells (P < 0.05).

Figure 5.

 

GSH, protein thiol, and protein-GSH mixed disulfide levels and effect of H₂O₂ on GSH level of TTase ^+/+ and TTase ^−/− LECs. (A) GSH level of TTase ^+/+ and TTase ^−/− LECs. (B) Protein thiol level of TTase ^+/+ and TTase ^−/− LECs. (C) Comparison of protein-GSH mixed disulfide level of TTase ^+/+ and TTase ^−/− LECs using immunohistochemistry. Original magnification, ×60. (D) Effect of a bolus of 1.5 mM H₂O₂ treatment on GSH level of TTase ^+/+ and TTase ^−/− LECs. TTase ^+/+ and TTase ^−/− LECs were treated with H₂O₂, and GSH concentration was determined spectrophotometrically, as described, at indicated time points and expressed as a percentage of control. Results are based on the average of five determinations. Error bars indicate SEM. *Significant difference from the TTase ^+/+ cells (P < 0.05).

Figure 6.

 

H₂O₂ detoxification in TTase ^+/+, TTase ^−/−, and TTase-loaded TTase ^−/− mouse LECs. FACS analysis of the time course of mean DCF fluorescence intensity of TTase ^+/+ (▪), TTase-loaded TTase ^−/− (○), and TTase ^−/− (•) LECs treated with 50 μM H₂O₂. Error bars indicate SD. n = 5. *Significant difference from the TTase ^+/+ and TTase-loaded TTase ^−/− cells (P < 0.05).

Figure 7.

 

Effect of H₂O₂ on G3PD activity in TTase ^+/+, TTase ^−/−, and TTase-loaded TTase ^−/− LECs. (A) Comparison of G3PD activity of TTase ^+/+ and TTase ^−/− LECs after oxidative stress. Cultured TTase ^+/+ (▪) and TTase ^−/− (□) LECs were treated with a 150-μM bolus of H₂O₂ for indicated times, and then cell lysates were prepared and G3PD activity was assayed as described. Results are based on the average of five determinations. Error bars indicate SEM. (B) The percentage of G3PD activity remained in the TTase ^+/+, TTase ^−/−, and TTase-loaded TTase ^−/− LECs 15 minutes after treatment with a 150-μM bolus of H₂O₂. Each cell group without H₂O₂ treatment was considered the control. Results are based on the average of 5 determinations. Error bars indicate SEM. *P < 0.05.

Figure 8.

 

Effect of H₂O₂ on apoptosis of TTase ^+/+ and TTase ^−/− LECs. TTase ^+/+ and TTase ^−/− LECs were treated with 50 μM H₂O₂ in serum-free medium for 1 hour, and then medium was replaced with new MEM with 10% FBS and incubated at 37°C for 16 hours. The cells were trypsinized, washed with PBS, stained with Annexin V-FITC, and analyzed by flow cytometry. Ten thousand cells were analyzed, and the mean fluorescence channel was derived. Results are based on the average of five determinations. Error bars indicate SEM. *Significant difference from the TTase ^+/+ control cells (P < 0.05).

The authors thank You Zhou of the Center of Biotechnology, University of Nebraska-Lincoln, for conducting the immunohistochemical study.