Induction of Glutathione S-Transferase hGST 5.8 Is an Early Response to Oxidative Stress in RPE Cells

Abstract

purpose. To delineate the role of the glutathione S-transferase (GST) isozyme hGST 5.8 in protection mechanisms against oxidative stress, the effect of low-level transient exposure of H₂O₂ to retinal pigmented epithelial (RPE) cells on hGST 5.8 and other enzymes involved in defense against oxidative stress was examined.

methods. Cultured human RPE cells were exposed to 50 μM H₂O₂ for 20 minutes. Subsequently, the cells were washed and resuspended in the culture media. The cells were pelleted and lysed, and the levels of lipid peroxidation products including thiobarbituric acid–reactive substances (TBARS), glutathione (GSH), glutathione peroxidase (GPX), glucose 6-phosphate dehydrogenase, glutathione reductase, GST, catalase (CAT), and superoxide dismutase (SOD) were determined and compared with levels in control cells. Total GSTs were purified by GSH-affinity chromatography, and the isozymes were separated by isoelectric focusing, characterized, and quantitated. hGST 5.8 was quantitated by an immunologic method as well as by determining activity toward its preferred substrate, 4-hydroxynonenal (4-HNE). Kinetic constants of hGST 5.8 purified from H₂O₂-treated cells were also determined and compared with those of control cells.

results. Exposure of RPE cells to 50 μM H₂O₂ for 20 minutes showed a significant increase in TBARS (1.8-fold) andγ -glutamyl cysteine synthetase (γ-GCS) activity (1.6-fold). A significant increase (1.2-fold) was also observed in GPX activity toward cumene hydroperoxide, but CAT and SOD activities remained unchanged. There was no significant increase in GST activity toward 1-chloro-2, 4-dinitrobenzene but GST activity toward 4-HNE was increased by 1.4- to 1.8-fold. The increase in GST activity toward 4-HNE was associated with a 2.8-fold increase in protein of the isozyme hGST 5.8, which uses 4-HNE as the preferred substrate.

conclusions. Results of these studies show that the induction of hGST 5.8, which is involved in the detoxification of the lipid peroxidation products 4-HNE and hydroperoxides, may be an early adaptive response of RPE cells exposed to low levels of transient oxidative stress. It is suggested that this isozyme may be crucial for protecting the RPE from low levels of chronic oxidative stress. Observed increases in GPX and γ-GCS activities are consistent with this idea, because GPX activity is also expressed by hGST 5.8, and γ-GCS is the rate-limiting enzyme in biosynthesis of GSH, the substrate for hGST 5.8.

Reactive oxygen intermediates (ROIs) are continually generated in cells through processes such as the univalent reduction of O₂, metabolism of xenobiotics by mono-oxygenases, and phagocytosis. ROIs indiscriminately attack DNA, proteins, lipids, and other cellular components. The role of ROIs in initiating lipid peroxidation may be viewed as more damaging to the cell because of the propagation of the autocatalytic chain of lipid peroxidation. ¹ Toxicity of lipid peroxidation products such as 4-hydroxynonenal (4-HNE) is well documented, ² and lipid peroxidation has been implicated in the cause of disorders such as atherogenesis, ^{3 4} cataractogenesis, ^{5 6} neurodegenerative diseases, ^{7 8} retinopathy, ^{9 10} and diabetic complications. ¹¹ Adducts of 4-HNE and proteins have been identified in atherosclerotic plaques, ¹² and lipid peroxidation products have been shown to cause lens opacity ^{5 6} and retinal degeneration. ¹⁰ Aerobic organisms have evolved sophisticated mechanisms to combat the potential deleterious effects of ROIs in the form of a multitier defense system consisting of enzymes such as superoxide dismutase (SOD), catalase (CAT), glutathione peroxidase (GPX), and glutathione S-transferases (GSTs). In addition, reductants and free radical scavengers such as glutathione (GSH), urate, and tocopherols provide protection against ROIs through nonenzymatic mechanisms. 

GPX and free radical scavengers such as tocopherols are perceived as the major deterrents for the propagation of the lipid peroxidation cascade initiated by ROIs. GSTs, comprising a multigene family of xenobiotic metabolizing enzymes, also act as a defense against lipid peroxidation because of their GPX activity toward lipid hydroperoxides. ^{13 14} We have identified a group of GSTs with exceptionally high activity for conjugating 4-HNE to GSH, in addition to their GPX activity. ^{15 16 17} We believe this group of GST isozymes plays an important role in protection against low levels of chronic oxidative stress that may exert its toxic effects by amplification through the autocatalytic lipid peroxidation cascade. ¹⁸ The GST isozymes belonging to this group are present in rat, ⁶ human, ¹⁹ and bovine ²⁰ ocular tissues. The protective role of these GST isozymes against lipid peroxidation in ocular tissues is suggested by our studies showing that their induction in lens epithelium attenuates the opacities of lenses in organ cultures caused by 4-HNE. ^{5 6}  

The presence of these GST isozymes in retina ^{19 20} perhaps underscores their physiological role in this tissue that is unusually rich in polyunsaturated fatty acids. Retinal pigment epithelial (RPE) cells comprise a single layer of cells between the sensory retina and choroid. This group of GSTs may be particularly important for RPE cells because of their role in phagocytosis and degradation of spent rod and cone outer segment membranes, the processes known to cause oxidative stress. Therefore, in the present studies we have examined the protective role of the antioxidant defense enzymes, including GSTs, against relatively low levels of oxidative stress in RPE cells. Human RPE cells in culture were exposed to short-term low levels of H₂O₂, and the effect on the levels of SOD, CAT, GPX, GST, glutathione reductase (GR), glucose 6-phosphate dehydrogenase (G-6PD), γ-glutamyl cysteine synthetase (γ-GCS), and GSH was determined. To evaluate the role of specific GST isozymes involved in the detoxification of lipid peroxidation products, these isozymes were purified from the control and H₂O₂-treated cells, quantitated, and characterized for their kinetic properties toward the products of lipid peroxidation. Furthermore, the effect of H₂O₂ exposure on specific isozymes of GSTs was determined. In the results of these studies, the preferential induction of GST isozymes metabolizing the toxic products of lipid peroxidation appeared to be an early adaptive response of RPE cells to oxidative stress. 

Materials and Methods

Dulbecco’s modified Eagle’s medium (DMEM), penicillin-streptomycin solution (P-S), phosphate-buffered saline (PBS), HEPES, trypsin, and MEM nonessential amino acid solution were purchased from Gibco (Grand Island, NY); fetal bovine serum (FBS) from Biocell Laboratories (Carson, CA); GSH, 1-chloro-2,4-dinitrobenzene (CDNB), epoxy-activated Sepharose 6B, cumene hydroperoxide, Coomassie brilliant blue R-250, and peroxidase-conjugated goat anti-rabbit IgG from Sigma ( St. Louis, MO); and 4-HNE from Cayman Chemical (Ann Arbor, MI). Ampholines were procured from Pharmacia/LKB (Piscataway, NJ) and the reagents for sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and transblotting from Bio-Rad (Richmond, CA). Dilinoleoyl phosphatidylcholine was purchased from Avanti Polar Lipid (Birmingham, AL), and its hydroperoxides were synthesized as described by us previously. ²¹ The polyclonal antibodies raised by us previously against the α-, μ-, and π-class human GSTs ^{21 22} and the polyclonal antibodies against the recombinant mGSTA4-4 raised in rabbit ¹⁷ were used in these studies. The IgG fractions from each of the antibodies were first purified by diethylaminoethyl cellulose and subsequently by protein A–column chromatography. The specificities of these antibodies toward their respective antigens have been stringently established by us previously. ^{17 21}  

Cell Culture

Simian virus (SV) 40–transformed fetal male RPE cells (Coriell Institute, Camden, NJ) were chosen as a suitable model to investigate the effect of low levels of oxidative stress. These cells exhibit epithelioid morphology and are able to phagocytize rod outer segments, a characteristic feature of RPE cells. Previous studies have shown that these cells respond to oxidative stress in a fashion similar to that of primary cultures of RPE cells. ^{23 24 25 26} Thus, these cells were chosen over the primary cultures that not only provide a limited number of cells but also may not provide cellular consistency because of possible variability associated with donor sources. The SV 40–transformed fetal male RPE cells used in these studies were grown in DMEM supplemented with 10% FBS, 1% P-S solution, 10 mM HEPES (pH 7.4), and 10 mM MEM nonessential amino acids in a humidified incubator under 95% air and 5% CO₂. The cells were trypsinized and passaged every 5 days. 

Exposure of Cells to H₂O₂

The cells were exposed to 50 μM H₂O₂ at 37 °C for 20 minutes in phenol red–free MEM medium. Thereafter, the cells were washed twice with cold PBS and allowed to recover in the original plating media for 3 hours. The cells were then centrifuged, washed, lysed, and subjected to analyses for biochemical parameters and purification of GST isozymes. The average time lag between the initial contact by cells to H₂O₂ and the analysis for biochemical parameters was approximately 6 hours. 

Enzyme Assays

GR and G-6PD activities were assayed according to the procedure described by Beutler, ²⁷ and γ-GCS activity was determined by the method described by Seelig and Meister. ²⁸ GSH was measured by the method of Beutler et al. ²⁹ Aliquots of homogenates used for GSH determinations did not contain β-mercaptoethanol. GST activity toward CDNB was determined according to the method of Habig et al., ³⁰ and the activity toward 4-HNE was determined spectrophotometrically at 224 nm, according to the procedure described by Alin et al. ³¹ One unit of GST activity was defined as the amount of enzyme catalyzing the conjugation of 1 micromole of the electrophilic substrate with GSH per minute at 25°C for CDNB and at 30°C for 4-HNE. CAT activity was assayed according to the procedure described by Holmes and Masters, ³² and SOD activity was determined by the method described by Paoletti and Mocali. ³³ One unit of CAT activity was defined as the amount of enzyme required to decompose 22.94 micromole of peroxide per milliliter per minute at 30°C. One unit of SOD activity was defined as the amount of enzyme necessary to decrease the reference rate to 50% of maximum inhibition at room temperature. GPX activities of the GST isozymes toward lipid hydroperoxides were determined as described by us previously. ³⁴ Protein content was measured using the method of Bradford ³⁵ with bovine serum albumin as standard. The K _m and k _cat values were calculated using software to calculate nonlinear regression (Hyper, Algor Inc., Pittsburgh). Lipid peroxidation in the homogenates of control and H₂O₂-treated RPE cells was measured by the method described previously. ³⁶ Three separate experiments with triplicate measurements in each were performed to determine the effect of H₂O₂ on antioxidant defenses in RPE cells (n = 9), and groups were compared using Student’s two-tailed t-test. 

Electrophoresis and Western Blot Analysis

SDS-β-mercaptoethanol–polyacrylamide gel electrophoresis was performed using the buffer system described by Laemmli. ³⁷ The resolving and stacking gels contained 12.5% and 7.1% (wt/vol) acrylamide, respectively. Western blot analyses was performed according to the method of Towbin et al. ³⁸ with slight modification, as described by us previously. ²²  

Purification of GST Isozymes

The cells (control and H₂O₂-treated) used for GST purification were harvested from culture by scraping with a rubber policeman and frozen at −80°C after washing with PBS, until enough cells were obtained for the experiments. The cells were thawed, lysed in 10 mM potassium phosphate buffer (pH 7.0) containing 1.4 mMβ -mercaptoethanol (buffer A), and sonicated with a cell disrupter (Sonifier Model W185D; Heat Systems–Ultrasonics, Plainview, NY). The purification of total GSTs from cells was performed according to the method described by us previously. ^{19 20} All purification steps were performed at 4°C, and enzyme activity during the purification was monitored with CDNB and 4-HNE as the substrates. The cell homogenate prepared in buffer A, was centrifuged at 28,000g for 45 minutes, and the supernatant was dialyzed overnight against buffer A (100 volumes, three changes). The dialyzed supernatant was subjected to GSH-affinity chromatography ^{19 20} to obtain total purified GSTs. The GSH-affinity column was pre-equilibrated with 22 mM potassium phosphate buffer (pH 7.0) containing 1.4 mM β-mercaptoethanol (buffer B) at a flow rate of 5 ml/h, and this flow rate was maintained throughout the affinity chromatography. The unbound proteins were thoroughly washed out of the column with buffer B. Total GSTs were eluted with 10 mM GSH in 50 mM Tris-HCl (pH 9.6) containing 1.4 mM β-mercaptoethanol and were thoroughly dialyzed against buffer A. The individual GST isozymes were separated by isoelectric focusing (IEF) in a column (LKB-8100; LKB; Gaithersburg, MD) using ampholines in the pH range of 3.5 to 10, and a 0% to 50% (wt/vol) sucrose density gradient, as described by us previously. ³⁹ After IEF at 1600 V for 24 hours, 0.8-ml fractions were collected and monitored for pH and GST activity with CDNB as the substrate. The isozyme peaks obtained during IEF were pooled separately and subjected to structural, kinetic and immunologic characterization. The enzyme preparations used for kinetic studies were dialyzed against buffer A, and those used for structural studies were dialyzed against 0.1% aqueous acetic acid. 

Results

Results of the constitutive activities of antioxidant defense enzyme and GSH levels in RPE cells (Table 1) indicate that these cells contain CAT, SOD, the major GSH-linked defense enzymes, and the enzymes for GSH synthesis and recycling. Malonaldehyde (MDA) content (measured as thiobarbituric acid–reactive substances [TBARS]) of the cells exposed to H₂O₂ was 1.8-fold higher than that of control cells. These results confirmed that even though no notable cell death was observed during H₂O₂ exposure in these experiments (data not presented), the cells sustained substantial oxidative stress as indicated by increased lipid peroxidation. CAT and SOD activities were not affected by H₂O₂ treatment under the condition specified in this study. There was no notable change in the steady state GSH levels in H₂O₂-treated RPE cells, despite a significant increase of approximately 1.6-fold in the activity of γ-GCS, which is the rate-limiting enzyme for GSH biosynthesis. No notable changes were observed in the activities of GR or G-6PD. However, GPX activity of the treated cells toward phosphatidylcholine hydroperoxide and cumene hydroperoxide increased by 1.4- and 1.2-fold, respectively. There was no change in GST activity toward CDNB, but a significant increase in GST activity toward 4-HNE (1.4-fold) was observed. These results indicated a selective induction or activation of GST isozymes that use 4-HNE as the preferred substrate. This was consistent with the increase in GPX activity, because 4-HNE–metabolizing GST isozymes are known to express GPX activity toward lipid hydroperoxides. ¹⁵ A concordant increase in the activity of γ-GCS, the rate-limiting enzyme in biosynthesis of GSH, which is a substrate of GSTs, supports the idea of a protective role of these enzymes against lipid peroxidation. 

Our previous studies ^{19 40} have shown that the major classes of human GSTs (α, μ, π) show only minimal activity toward 4-HNE, but the specific activity of the minor isozyme hGST 5.8 toward 4-HNE is approximately 100 times higher than that of the α, μ, andπ classes of GSTs. Although our results suggest that hGST 5.8 in RPE cells is induced or activated, the extent of its induction cannot be assessed by measuring only activity toward 4-HNE, because the major GST isozyme of RPE cells, GST π, also expresses some activity toward 4-HNE. Therefore, to determine whether H₂O₂ exposure causes selective induction of 4-HNE–metabolizing GST isozymes (hGST 5.8) we purified the total GST isozymes from equal amounts (4.4 × 10⁸) of control and H₂O₂-treated RPE cells. The results showed that the total amounts of GST protein in the control and treated cells was not altered significantly (Table 2) . This was consistent with the idea that only the minor GST isozyme hGST 5.8, which constitutes approximately 4% of total GST protein, was primarily affected by H₂O₂ exposure and that there was little effect on the major GST isozymes of RPE. Total GST activity toward CDNB purified from equivalent amounts of control and treated cells was similar. However, total GST activity toward 4-HNE in H₂O₂-treated cells was approximately 1.8 times higher than that in the control cells. 

The results of SDS-PAGE and western blot studies presented in Figures 1 A and 1B show that the protein of the 4-HNE–metabolizing isozyme was selectively increased in H₂O₂-treated cells. The 4-HNE–metabolizing GST isozymes of mammalian tissues show interspecies immunologic similarities but are immunologically distinct from the other α, μ, π, or θ classes of GSTs. ¹⁵ The results presented in Figure 1B (lane 4) showed that the GST isozyme recognizing mGSTA4-4 antibodies (mGSTA4-4 is the mouse orthologue of hGST 5.8) was selectively increased after exposure to H₂O₂. The densitometric scan of the bands on an imager (Model T^M2000; Alpha, San Leandro, CA) revealed a 2.8-fold induction of a protein cross-reacting with antibodies against mGSTA4-4 in H₂O₂-treated cells. These results further confirmed that hGST 5.8, which specifically recognizes antibodies against mGSTA4-4, was selectively induced by exposure to low levels of H₂O₂. A 2.8-fold increase in hGST 5.8 protein and only a 1.4-fold and 1.8-fold increase in the GST activity toward 4-HNE in the cytosol (Table 1) and purified GSTs (Table 2) , respectively, may appear to be inconsistent at first glance. However, it may be that GST π, which constitutes more than 90% of the total GSTs of RPE cells, also has minimal activity toward 4-HNE (Table 2) . Because GST π was not increased by H₂O₂ exposure, a linear correlation between the increase of hGST 5.8 protein and the activity toward 4-HNE may not be expected. 

To determine whether hGST 5.8 of RPE has high activity for 4-HNE and other products of lipid peroxidation, the GST isozymes of RPE cells were separated by IEF. When aliquots of total GST isozymes obtained by GSH-affinity chromatography (Table 2) were subjected to IEF, the enzymes from control and H₂O₂-treated cells showed identical IEF profiles. The two peaks (Fig. 2) at isoelectric point values of 4.6 (major) and 5.8 (minor) were identified as GST π and hGST 5.8 by western blot analysis. The minor peak was recognized only by the antibodies against mGSTA4-4 and not by those against GST π, GST α, or GST μ (Fig. 3) , whereas the major peak was recognized only by the antibodies against GST π. These results show that only GST π and hGST 5.8 were expressed in RPE cells and that the μ and α classes were absent in these cells. Kinetic properties of GST π were similar to those reported for human GST π from other tissues and its activity toward the commonly used substrate CDNB (18.3 U/mg protein) was highest among all the substrates used in this study. In contrast, the activity of hGST 5.8 toward 4-HNE (32.9 U/mg protein) was 4.6-fold higher than that toward CDNB (7.14 U/mg protein). These results are consistent with our previous observations ^{19 20} on the substrate preferences of this isozyme for 4-HNE. 

The kinetic parameters of the isozymes presented in Table 3 showed that the catalytic efficiency (k _cat/K _m) of hGST 5.8 toward 4-HNE was almost two orders of magnitude higher than that of GST π, particularly in H₂O₂-treated cells. The reason for the observed higher catalytic efficiency of hGST 5.8 isolated from H₂O₂-treated cells compared with that isolated from control cells (Table 3) is not clear. It is possible that H₂O₂ may also activate hGST 5.8. Further studies are needed to explore this possibility. The K _m of hGST 5.8 for 4-HNE was found to be low but was probably in the expected range of physiological concentration of 4-HNE generated locally in cells during oxidative stress. A selective increase in this isozyme after exposure to low levels of H₂O₂ suggests that it plays an important role in terminating the lipid peroxidation cascade (by reducing hydroperoxides) as well as in detoxifying the toxic end products of lipid peroxidation. 

Discussion

The results presented in this study show that transient exposure of apparently nontoxic, low levels of H₂O₂ cause selective induction of GST isozyme highly reactive toward 4-HNE, a toxic end product of lipid peroxidation. Although there was no significant increase in total GST activity measured with CDNB as the substrate, a marked induction of hGST 5.8 protein was observed. Our previous studies ^{19 20} have shown that hGST 5.8, which has at least an order of magnitude higher catalytic efficiency (k _cat/K _m) toward 4-HNE compared with that of other GST isozymes, is constitutively expressed in bovine and human retina. Present studies demonstrate its presence in human RPE. Induction of this GST isozyme in absence of any notable increase in SOD or catalase in RPE cells exposed to low levels of H₂O₂ for a short time suggest that induction of this enzyme is an early adaptive response of these cells to oxidative stress. 

Tate et al. ⁴¹ have shown that the treatment of RPE cells with 250 and 500 μM H₂O₂ for 18 hours causes approximately 1.5-fold and 4-fold increases in CAT activity, respectively. Similarly, it has been shown that CAT activity increases by approximately twofold when RPE cells phagocytize bovine rod outer segments every other day for 2 weeks. ⁴² Absence of any increase in CAT activity after short-term exposure of H₂O₂ during the present studies suggests that CAT induction may be a later event in the adaptive response of RPE cells to sustained high levels of oxidative stress. In RPE cells exposed to 50 μM H₂O₂ only for 20 minutes, a significant increase in MDA formation was observed, because lipid peroxidation initiated by a single ROI molecule can lead to an autocatalytic cascade of reactions resulting in accumulation of the toxic and stable products of lipid peroxidation. hGST 5.8 can limit the amplification of ROI toxicity in the lipid peroxidation cascade by its GPX activity. It also can detoxify 4-HNE, which is reported to disrupt cellular functions by inhibiting the key enzymes, affecting Ca²⁺ homeostasis, and inhibiting protein and DNA synthesis. ² Induction of 4-HNE–metabolizing GST isozyme in the epithelial cells of rat lens protects the lens from 4-HNE–induced opacity. ⁶ The results of the present studies suggest that this GST isozyme may also act as a defense mechanism in RPE against the deleterious effect of lipid peroxidation. 

Based on the results of present studies, we speculate that hGST 5.8 may be relevant to the cause and defense mechanisms against age-related macular degeneration (AMD). AMD is a degenerative disease of the macular photoreceptors, RPE, and Bruch’s membrane that is the leading cause of irreversible blindness in people more than 50 years of age in the United States. ⁴³ Although the cause of AMD is not completely understood, recent studies suggest that oxidative stress may be a key factor in the pathogenesis of this disorder. ⁴⁴ It has been suggested that the RPE layer may play a critical role in the genesis and progression of AMD. ⁴⁵ The RPE cell is particularly susceptible to oxidative damage caused by high oxygen pressure from the underlying choriocapillaris, light exposure, and high concentration of polyunsaturated fatty acids in photoreceptor outer segments. ^{44 46} The protective mechanisms in RPE against oxidative stress are therefore crucial to detoxifying ROIs, which include the superoxide radical (O₂·−), hydrogen peroxide (H₂O₂), the hydroxyl free radical (OH·−), and the stable end products of lipid peroxidation, particularly 4-HNE and its homologous alkenals. Although CAT and SOD may be important to the protection mechanisms against O₂·− and H₂O₂, these mechanisms would be ineffective against the cumulative toxicity of the lipid peroxidation products because of the continuous generation of ROIs in RPE cells. Thus, the enzymes involved in the detoxification of lipid peroxidation products should be important in the hierarchy of antioxidant defense mechanisms. Previous studies ⁴⁷ showing the selective induction of 4-HNE–metabolizing GST isozyme in rat liver during enhanced lipid peroxidation are consistent with this idea. 

Our earlier studies have shown that administration of low doses of dietary antioxidants such as curcumin and t-butylated hydroxytoluene induce hGST 5.8 orthologues in rat lens epithelium ^{5 6} and their induction protects from the notable toxic effect (i.e., lens opacity) of 4-HNE. If 4-HNE and other lipid peroxidation products are also involved in the mechanisms of AMD, it is possible that modulation of hGST 5.8 in retina is a strategy to delay AMD. However, it must be noted that the direct relevance of these in vitro findings to clinical AMD is only suggested, not proven. Future studies comparing these enzymes in RPE from clinical AMD and age-matched control specimens are therefore required. 

The significance of these results can also be discussed relative to the possible physiological role of 4-HNE. Although it is known that at higher concentrations (>20 μM) 4-HNE is toxic, at low concentrations (0.1–10 μM) it may involve the signal transduction pathways. ^{48 49} At the estimated physiological concentrations, 4-HNE affects cell proliferation and differentiation. It modulates activities of phospholipase-C and ornithine decarboxylase and has been suggested to act as a signaling molecule. ^{50 51} Constitutive expression of hGST 5.8 in RPE cells observed in present studies appears to be important for maintaining physiological levels of 4-HNE formed because of the ever-present oxidative stress. Enhanced oxidative stress in RPE caused by H₂O₂ exposure in the present studies, evident from the increased MDA formation, perhaps necessitates the induction of 4-HNE–metabolizing isozyme hGST 5.8 to prevent the accumulation of 4-HNE to toxic levels. The RPE cell model described in the present studies may therefore be useful for delineating the physiological role of 4-HNE. 

 Supported in part by Grant EY04396 (YCA) from the National Eye Institute, National Institutes of Health; Minority Investigator Supplement EY04396 (BFG) from the National Institutes of Health; the Paul Beeson Physician Faculty Scholars in Aging Research Program; and Research to Prevent Blindness (BFG).

 Submitted for publication January 21, 1999; revised April 23, 1999; accepted June 3, 1999.

 Commercial relationships policy: N.

 

Table 1.

 

Effect of H₂O₂ on the Levels of MDA, GSH, and Antioxidant Defense Enzymes

Biochemical Parameters	Control	50-μM H2O2-Treated
Lipid peroxide (nanomoles MDA/mg protein)*	0.60 ± 0.14	1.08 ± 0.20**
GSH (nanomoles/mg protein)	14.4 ± 1.5	15.23 ± 1.4
Glutathione S-transferases (micromoles/min · mg protein)
With CDNB	0.13 ± 0.01	0.133 ± 0.01
With 4-HNE	0.056 ± 0.004	0.078 ± 0.01, †
Glutathione peroxidase (micromoles/min · mg protein)
With CU-OOH	0.047 ± 0.007	0.054 ± 0.003, †
With PC-OOH	0.044 ± 0.007	0.06 ± 0.01, †
Catalase (micromoles/min · mg protein)	4.94 ± 1.1	5.2 ± 0.74
Superoxide dismutase (micromoles/min · mg protein)	12.1 ± 1.5	12.0 ± 1.8
Glutathione reductase (micromoles/min · mg protein)	0.055 ± 0.007	0.052 ± 0.006
Glucose-6-phosphate dehydrogenase (micromoles/min · mg protein)	0.066 ± 0.006	0.074 ± 0.006
γ-Glutamyl cysteine synthetase (micromoles/min · mg protein)	0.052 ± 0.009	0.082 ± 0.01, †

 Values are means ± SD; n = 9. Both groups (control and H₂O₂-treated) were compared by Student’s t-test. CU-OOH, cumene hydroperoxide; PC-OOH, phosphatidylcholine hydroperoxide; MDA, malonaldehyde.

*  MDA was estimated as TBARS.

†   P < 0.01 was considered significant.

Table 2.

 

Purification of GST Isozymes of RPE Cells

Fraction	Total GST Activity* (U)			Total Protein (mg)	Specific Activity (U/mg protein)			% Yield
		with CDNB	with 4-HNE			CDNB	4-HNE	CDNB	4-HNE
Control cells
28,000g Supernatant	3.1 ± 0.18	1.21 ± 0.09	23.4 ± 1.4	0.132	0.052	100	100
GSH-affinity chromatography	2.92 ± 0.09	0.97 ± 0.10	0.199 ± 0.024	14.7	4.9	94	80
Isoelectric focusing
GST 5.8 (hGST 5.8)	0.05 ± 0.002	0.23 ± 0.011	0.007 ± 0.001	7.14	32.86	73, †	25, †
GST 4.6 (GST π)	2.2 ± 0.12	0.07 ± 0.004	0.12 ± 0.006	18.33	0.58
H2O2-treated cells
28,000g Supernatant	3.4 ± 0.18	1.8 ± 0.18	25.4 ± 1.1	0.134	0.071	100	100
GSH-affinity chromatography	3.2 ± 0.04	1.7 ± 0.24	0.196 ± 0.01	16.3	8.7	94	94
Isoelectric focusing
GST 5.8 (hGST 5.8)	0.08 ± 0.006	0.45 ± 0.03	0.009 ± 0.001	8.9	50.0	74, †	29, †
GST 4.6 (GST π)	2.43 ± 0.12	0.07 ± 0.003	0.12 ± 0.003	20.25	0.58

 Each value represents means ± SD of three determinations; 440 × 10⁶ cells were starting material for purification. Enzyme activities during the purification were monitored using CDNB and 4-HNE as substrates.

*  One unit of activity was defined as 1 micromole CDNB and 4-HNE used per minute at 25°C and 30°C, respectively.

†  Represents the combined yield of the activity towards CDNB and 4-HNE for all the IEF Peaks.

Figure 1.

 

(A) SDS-β-mercaptoethanol–polyacrylamide gel electrophoresis of GSH affinity-purified total GSTs from control and 50-μM H₂O₂-treated RPE cells. Lane 1 contained protein standards; lanes 2 and 3 contained total purified GST protein from control and H₂O₂-treated RPE cells, equivalent to 3.3 × 10⁶ cells, respectively. The gel was stained by Coomassie brilliant blue R-250. (B) Western blot analysis of purified total GST from the control and 50-μM H₂O₂-treated RPE cells using antibodies raised against rec-mGSTA 4-4. Lane 1, Prestained molecular weight markers; lane 2, mGSTA 4-4 used as a positive control; lanes 3 and 4, total GSTs purified from control and H₂O₂-treated RPE cells, equivalent to 36 × 10⁶ cells, respectively. The blot was developed by enhanced chemiluminescence (Amersham, Arlington Heights, IL).

Figure 2.

 

IEF profiles of purified total GSTs from control (A) and 50-μM H₂O₂-treated (B) RPE cells. Ampholines in the pH range 3.5 to 10.0 were used. (x), pH gradient; (•) GST activity toward CDNB as substrate. In both profiles, the enzyme purified from the same amount of cells (440 × 10⁶ cells) was used.

Figure 3.

 

Western blot analysis of control and 50-μM H₂O₂-treated RPE cells GST isozymes separated by IEF, using the antibodies raised against the α (A), μ (B), π (C), and mGSTA 4-4 (D) GST isozymes. Lane 1 represents respective positive control samples. Lanes 2 and 3 contained GST 4.6 (GST π), and lanes 4 and 5 contained GST 5.8 from control and H₂O₂-treated cells, respectively. The blot was developed using HRP-conjugated goat anti-rabbit secondary antibody with 4-chloro-1-naphthol as substrate.

Table 3.

 

Kinetic Constants of GSTs of RPE Cells with CDNB and 4-HNE

Enzyme	With CDNB					With 4-HNE
		Km (M × 10−3)	kcat * (S−1)	kcat/Km (103 · M−1 · S−1)	Km (M × 10−3)	kcat * (S−1)	kcat/Km (103 · M−1 · S−1)
Control cells
GST 4.6	0.56 ± 0.01	19.6 ± 1.5	35.0	0.041 ± 0.001	0.62 ± 0.03	15.1
GST 5.8	0.81 ± 0.02	8.5 ± 0.34	10.5	0.037 ± 0.001	28.6 ± 1.5	773
H2O2-treated cells
GST 4.6	0.56 ± 0.01	22.5 ± 1.9	40.2	0.041 ± 0.001	0.70 ± 0.04	17.1
GST 5.8	0.74 ± 0.03	10.5 ± 0.46	14.2	0.032 ± 0.001	48.5 ± 4.3	1515.6

 Values are means ± SD of three determinations.

*  Calculated on the basis of the molecular weights of the purified enzyme, (i.e., pI 4.6, 45,000; pI 5.8, 50,000).