Glutathione distribution in normal and oxidatively stressed cells

Available online at www.sciencedirect.com R

Experimental Cell Research 285 (2003) 9 –14

www.elsevier.com/locate/yexcr

Glutathione distribution in normal and oxidatively stressed cells Jeffrey G. Ault* and David A. Lawrence Division of Molecular Medicine, Wadsworth Center, New York State Department of Health, Albany, NY 12201-0509, USA Received 13 May 2002, revised version received 18 September 2002

Abstract Glutathione is the most abundant of the low-molecular-mass molecules that provide reducing equivalents that protect cells from oxidative stress. We used immunoelectron microscopy to investigate glutathione distribution in normal and oxidatively stressed cells. Here, for the first time, we show that reduced glutathione is distributed relatively evenly throughout the cell, with the exception of the lumen of the rough endoplasmic reticulum, where little is detected. Oxidant exposure, either to 0.1 mM diamide or ethycrinic acid, eventually caused cellular glutathione depletion. However, despite entering a cell within seconds, both oxidants required hours to dramatically affect glutathione levels in the majority of cells in a population. Interestingly, cells within a homogeneous cell line population lost glutathione at different rates. Structural changes associated with oxidative stress, such as increased vacuolization and membrane blebbing, were correlated with glutathione depletion. Oxidant-exposed cells that appeared normal had higher glutathione levels than those within the same population that appeared stressed. The last reserves of cellular glutathione were found within mitochondria. © 2003 Elsevier Science (USA). All rights reserved. Keywords: Glutathione; Immunoelectron microscopy; Oxidative stress

Introduction Glutathione (GSH) is a tripeptide thiol found in millimolar concentrations in virtually all cells [1]. It is a cofactor in many enzymatic reactions and is important in intracellular cysteine storage. GSH is the reduced form, which can be oxidized to a disulfide form (GSSG) or to form mixed disulfides with other thiol-containing reactants (GSSR). A high GSH-to-GSSG ratio is maintained within cells, which has been shown to be important in the structural integrity and functional processes of membranes, the maintenance and polymerization of microtubules, the conformation of proteins and modulation of their activities, the initiation and elongation of proteins, and the metabolism of electrophilic agents [1]. GSH is also important in both DNA and RNA synthesis [2,3]. Maintaining a high GSH-to-GSSG ratio contributes to

* Corresponding author. Division of Molecular Medicine, Wadsworth Center, New York State Department of Health, Empire State Plaza, P.O. Box 509, Albany, NY 12201-0509, USA. Fax: ⫹1-518-474-7992. E-mail address: [email protected] (J.G. Ault).

the redox homeostasis of a cell, which provides an antioxidant defense mechanism against ionizing radiation, reactive oxygen species, free radicals, and toxic xenobiotics [4]. Exposure to oxidant species, such as superoxide radical (O䡠2⫺), hydrogen peroxide (H2O2), hydroxyl radical (HO䡠), and lipid peroxides (LOOH), occurs during cellular metabolism and environmental interactions. As oxidant levels increase, the cell experiences oxidative stress, which causes molecular changes that can result in cell toxicity and death. Cumulative oxidative effects contribute to aging [5], and chronic oxidative stress contributes to the pathology of numerous diseases, such as diabetes, atherosclerosis, rheumatoid arthritis, and cancer. Oxidants can damage the lung, kidney, eye, skin, joint, heart and cardiovascular system, gastrointestinal tract, and brain, nervous, and neuromuscular systems [6]. GSH and other less abundant low-molecular-mass molecules provide reducing equivalents that, along with antioxidant enzymes, nullify oxidant species. Knowing the cellular location of GSH and how its distribution is affected by exposure to oxidant species is important in understanding its role in defending the cell against oxidative stress. By using immunoelectron microscopy on three hu-

0014-4827/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0014-4827(03)00012-0

10

J.G. Ault, D.A. Lawrence / Experimental Cell Research 285 (2003) 9 –14

man leukocytic cell lines, we determined the cellular distribution of GSH and observed how this distribution changes during prolonged oxidant exposure.

Materials and methods Cell culture conditions and treatments Jurkat R (T-cell), THP-1 (monocyte-like), and CCRF-SB (B-cell) cells were used and grown in RPMI 1640 medium supplemented with penicillin, streptomycin, and 10% fetal bovine serum at 37°C in a humidified atmosphere of 5% CO2 in air. Oxidative stress was induced by exposing cells to either 0.1 mM diamide [diazenedicarboxylic acid bis(N,N⬘-dimethylamide)] or 0.1 mM ethycrinic acid in medium. Various treatment times were tried before settling for a 2-h treatment time with diamide and a 5-h treatment time with ethycrinic acid. These times produced GSH depletion, but not signs of necrosis, in the majority of cells. Many cells had structural changes associated with oxidative stress, such as vacuoles and membrane blebbing. The location of the last cellular GSH reserves during oxidative stress were determined in select cells at these times and in select cells of longer treatment times. The location of the last cellular GSH reserves was also determined by inhibiting GSH synthesis for 24 h with 1.0 mM L-buthionine-(S,R)-sulfoximine (BSO) in medium. BSO is a specific inhibitor of ␥-glutamylcysteine synthetase (also known as glutamate cysteine ligase, G-CL), the enzyme that catalyzes the first step in GSH synthesis [7]. Oxidant exposure during GSH synthesis inhibition was performed only on Jurkart R cells. Cells were exposed to 0.1 mM diamide, 1.0 mM BSO in medium for 30 min. 8.1-GSH antibody Because GSH is found ubiquitously in mammalian tissue, it is necessary to modify GSH to produce specific antibodies against it and to ensure reactivity toward GSH and not toward GSSG or GSSR. We used a mouse monoclonal IgG1 antibody (8.1-GSH; StressGen, Victoria, BC, Canada) against the glutathione adduct with N-ethylmaleimide (GS-NEM) [6]. N-Ethylmaleimide (NEM) is a lowmolecular-mass molecule that penetrates cells easily and reacts with sulfhydryl groups. The 8.1-GSH antibody is specific for GS-NEM and does not recognize NEM adducts with other thiols (R-S-NEM) [8]. Immunoelectron microscopy Experimental and control cells were sequentially treated as follows: fixed with cold 4% paraformaldehyde in phosphate-buffered saline (PBS) (pH 7.4) for 20 min, treated with cold 10 mM NEM in PBS (pH 7.2) for 20 min, postfixed with cold 1% glutaraldehyde in PBS (pH 7.4) for

Table 1 Subcellular GSH distribution in Jurkat cells before and after oxidative treatments Treatment

Nucleus

Cytoplasm

Mitochondria

No treatment BSOb (24 h) Diamide (2 h) Normal looking cells Stressed cells Ethycrinic acid (5 h) Normal looking cells Stressed cells BSO/diamide (0.5 h)

95 ⫾ 2.1 1.2 ⫾ 0.2

105 ⫾ 2.9 1.5 ⫾ 0.2

86 ⫾ 5.1 19 ⫾ 2.0

74 ⫾ 1.1 29 ⫾ 2.4

81 ⫾ 1.9 31 ⫾ 2.7

66 ⫾ 2.8 52 ⫾ 3.7

65 ⫾ 5.9 3.6 ⫾ 0.5 0.3 ⫾ 0.1

87 ⫾ 6.2 7.6 ⫾ 0.7 0.4 ⫾ 0.1

87 ⫾ 16.7 27 ⫾ 4.3 3.2 ⫾ 1.2

a

a Mean number (⫾ S.E.M.) of gold particles per ␮m2 in micrographs of Jurkat cells before and after various treatments. b BSO, L-buthionine-(S,R)-sulfoximine.

1 h, washed twice in PBS (pH 7.4), dehydrated in an ethanol series, and embedded in LR White. Semithin (0.20 – 0.25 ␮m) sections were cut using a Diatome diamond knife on a Reichert Ultracut E ultramicrotome. The sections were blocked for 30 min in a blocking solution containing 50 ␮g/mL goat sera, 50 ␮g/mL rabbit IgG, 20 ␮g/mL bovine serum albumin (BSA) in Tris-buffered saline (TBS) (containing 1 ␮g/mL BSA, 0.05% Tween 20, 0.5 M NaCl, 20 mM NaN3, pH 7.4), exposed to 28 ␮g/mL 8.1-GSH antibody in TBS (pH 7.4) with 20 ␮g/mL rabbit IgG for 1 h, washed four times in TBS, labeled with a 1:100 dilution of 10-nm gold particles conjugated to goat anti-mouse IgG antibodies (Ted Pella, Inc., Redding, CA) in TBS with 10 ␮g/mL goat sera for 1 h, washed five times in TBS, and stained with uranyl acetate for 20 min and lead for 1 min. Negative controls for the antibody were cells not treated with NEM and cells treated with 1.0 mM L-buthionine(S,R)-sulfoximine (BSO) for 24 h, which dramatically reduces GSH levels. The sections were viewed at 80 kV with a Zeiss 910 transmission electron microscope.

Results Immunolocalization of glutathione using the 8.1-GSH antibody on three human leukocytic cell lines (Jurkat R, THP-1, and CCRF-SB) showed that GSH is distributed throughout the cell. Besides confirming cytosolic and mitochondrial reserves of GSH determined by other methods [9], immunoelectron microscopy demonstrated unequivocally that GSH is in the nucleus as well. A nuclear GSH reserve has long been implied from GSH analysis of cellular fractions containing isolated nuclei [10] and fluorescent microscopy using thiol-reacting fluorescent probes [11–14] and from the assumption that GSH helps protect DNA from oxidation [15]. Generally, GSH appeared to be evenly distributed within the cytoplasm, nucleus, and mitochondria. Slightly, but significantly, higher GSH levels were observed in the cytoplasm (Table 1). However, relatively few gold

J.G. Ault, D.A. Lawrence / Experimental Cell Research 285 (2003) 9 –14

Fig. 1. GSH is distributed throughout the cell. Immunolabeling of a THP-1 cell shows GSH (labeled by the gold particles) in the nucleus (N), mitochondria (M), and cytoplasm. The lumen of the rough endoplasmic reticulum (arrows), by comparison, has almost no gold label; the few particles observed suggest that there is only a small amount of GSH in the lumen. Bar indicates 0.5 ␮m.

labeling particles were observed over the narrow lumen of the rough endoplasmic reticulum (ER), where disulfide bonds form during protein folding (Fig. 1). The 8.1-GSH antibody recognizes the glutathione adduct with GS-NEM [8]. NEM is a low-molecular-mass molecule that penetrates cells easily and reacts with sulfhydryl groups. To stabilize cellular GSH distribution at the time points of interest, cells were fixed with cold 4% paraformaldehyde in PBS prior to, and postfixed with cold 1% glutaraldehyde in PBS immediately after, NEM treatment, which was also done on ice. This sequence of treatments was the most optimum to prevent the efflux of the GS-NEM conjugate out of the cell or its possible redistribution into the nucleus [16]. Cells that were fixed after the NEM treatment or those fixed with glutaraldehyde prior to NEM showed lesser amounts of immunolabel (data not shown), the former because of GS-NEM efflux and the latter because glutaraldehyde inhibits the formation of the GS-NEM conjugate. Cells not treated with NEM prior to antibody exposure were used as the negative controls and had no label (data not shown). The specificity of the 8.1-GSH antibody against the GS-NEM adduct was further demonstrated in cells treated with 1.0 mM BSO for 24 h. BSO is a specific inhibitor of ␥-glutamylcysteine synthetase, the enzyme that catalyzes the first step in GSH synthesis [7]. BSO causes GSH levels to fall [17]. After a 24-h treatment, cells were significantly depleted of label (Table 1). The last GSH reserves remaining in BSO-treated cells were those in mitochondria, where the oxidative effects of H2O2 need to be controlled (Fig. 2).

11

Cells exposed to an oxidant, either diamide or ethycrinic acid, did not show signs of oxidative stress initially. Despite the oxidant entering the cell quickly, in the case of diamide within seconds [1], cells appeared normal and GSH levels, as indicated by immunolabeling, remained high. However, after a 2-h exposure to 0.1 mM diamide or a 5-h exposure to 0.1 mM ethycrinic acid, cells showed significantly lower GSH levels compared to untreated cells (Table 1, Fig. 3). Cells with most of their GSH depleted also showed structural effects of oxidative stress (i.e., many vacuoles and membrane blebbing). Adjacent cells with label indicating higher GSH levels appeared normal, with no structural signs of oxidative stress (Fig. 3C). As expected, longer treatment times produced a higher proportion of cells with GSH depletion and signs of oxidative stress as well as cells that had undergone necrosis. In most necrotic cells, GSH was still detected in mitochondria, demonstrating that, as in BSOtreated cells, the last reserves of GSH in oxidatively stressed cells are those in the mitochondria (Fig. 3D). THP-1 cells in particular had extensive vacuolization when exposed to oxidants (Fig. 4). The many vacuoles in these monocyte-like cells are typical of those found in activated monocytes and macrophages. They are the result of exocytosis and have been shown to contain the inflammatory protein MRP-14 [18]. We exposed Jurkat R cells to both 1.0 mM BSO and 0.1 mM diamide to test the effect of GSH synthesis inhibition on GSH levels during oxidant exposure. Our rationale was that, if the upregulation of GSH synthesis is the main mechanism in maintaining GSH levels during oxidant exposure, then inhibiting GSH synthesis during oxidant exposure would cause cells to lose GSH quickly. However, if GSH levels are maintained mainly by glutathione reductase activity converting GSSG back to GSH, then GSH may still be present for sometime during this combined treatment. We examined cells after 30 min, which was 90 min sooner than when diamide alone caused the majority of cells in a population to show significantly lower GSH levels (Table 1),

Fig. 2. The last reserves of cellular GSH are in mitochondria. Immunolabeling of a Jurkat R cell after L-buthionine-(S,R)-sulfoximine treatment shows the majority of the gold particles are in the mitochondria (M) and not in the cytoplasm or nucleus (N). Bar indicates 0.5 ␮m.

12

J.G. Ault, D.A. Lawrence / Experimental Cell Research 285 (2003) 9 –14

Fig. 3. GSH distribution during oxidative stress. Jurkat R cells were exposed to 0.1 mM ethycrinic acid for 5 h. (A) An untreated control cell showing GSH (labeled by the gold particles) distributed throughout the nucleus (N), mitochondria (M), and cytoplasm. (B) An oxidatively stressed cell showing membrane blebbing (arrowhead), numerous vacuoles (V), and swelling (arrows) of the endoplasmic reticulum that is directly surrounding the nucleus (N). Immunolabeling shows the cell is heavily depleted of GSH. (C) Two oxidant-exposed cells. Whereas one cell appears normal (asterisk), the other has vacuoles (V) and membrane blebbing (arrowheads) associated with oxidative stress. Compared to the levels in the normal cell and in the control cell in A, GSH (gold particles) is almost depleted in the nucleus (N) and cytoplasm, but not in the mitochondria (M), of the oxidatively stressed cell. (D) Mitochondria (M) are where the last reserves of cellular GSH (gold particles) are found in oxidatively stressed cells. Note the extraction of the cytoplasm in this necrotic cell. Bars indicate 0.5 ␮m.

with many cells still appearing normal with no signs of oxidative stress. With the combined treatment, all cells were almost completely depleted of GSH, even in their mitochondria (Table 1) and showed extensive oxidative damage (Fig. 5). Damage was so extensive that both the cytoplasm and nucleoplasm were heavily extracted (Fig. 5), indicating that oxidant-induced changes occurred that caused permeability of both the plasma membrane and the nuclear envelope. This experiment supported the view that cells maintain GSH levels by the upregulation of GSH synthesis.

Discussion All types of biological molecules, including DNA, lipids, proteins, and carbohydrates, are susceptible to oxidation. At

millimolar concentrations, GSH provides the cell with substantial amounts of reducing equivalents that can be used to nullify oxidant species. We show that the cellular distribution of GSH is indicative of its role in protecting the cell from oxidation, in that it is everywhere within the cell to protect against molecular changes caused by oxidative stress. On-section immunoelectron microscopy allowed us to investigate GSH distribution within and between cellular compartments. Relative comparisons between the nuclear, cytoplasmic, and mitochrondrial GSH pools are shown in Table 1. Though the cytoplasm had slightly, but significantly, higher amounts of GSH in untreated cells, the nuclear/cytoplasmic GSH gradiant is close to 1. This is similar to the results from fractionation studies using nonaqueous conditions [10], but much different from the results from

J.G. Ault, D.A. Lawrence / Experimental Cell Research 285 (2003) 9 –14

Fig. 4. Activated THP-1 cells producing many large vacuoles (asterisks). The nucleus (N) of one cell is shown. Bar indicates 1.0 ␮m.

fluorescent light microscopy using thiol-reacting fluorescent probes, which show a nuclear/cytoplasmic GSH gradient of about 3-fold [11,12]. It has been demonstrated however that this high nuclear/cytoplasmic GSH ratio observed in unfixed cells with fluorescent microscopy is the result of redistribution of the GS-fluorophore conjugate to the nucleus [16]. In our study, we fixed the cells to stabilize cellular GSH distribution prior to NEM treatment, thus preventing the redistribution or efflux of the GS-NEM conjugate. A few gold-labeling particles were observed lying within the narrow lumen of the rough endoplasmic reticulum (RER), where disulfide bonds form during protein folding. These few particles may be an artifact of the radial distance within which a gold particle can be separated from the antibody complex to which it is tethered (i.e., a gold particle may settle over the narrow lumen, though the primary antibody/antigen complex is outside of it) or they may represent a small amount of GSH within the lumen. The ratio of GSH to GSSG is approximately 30- to 100-fold less in the RER than in the cytosol [19]. Interestingly, compared to GSSG, GSH is preferentially transported into the RER lumen [20]. Cuozzo and Kaiser [21] have elegantly demonstrated in yeast that GSH competes with protein thiols for disulfide-bond formation by the major oxidation pathway, defined by the protein Ero1, and that the net effect of glutathione (GSH ⫹ GSSG) in the ER is to reduce protein disulfide bonds. They show that the net reducing equivalents, provided by glutathione, buffer the ER against transient hyperoxidizing conditions, such as during oxidative stress. GSH may be involved in correcting improper, relatively unstable, disulfide bonds that occur by mistake during protein folding, either by directly reducing them or by reducing oxidized protein disulfide isomerase to the dithiol state, thus renewing its disulfide-bond-reshuffling activity [21]. Mitochondria are essential for cell life, by providing

13

energy for cellular functions. While producing energy necessary to convert ADP to ATP, mitochondria also produce reactive oxygen species as byproducts of the consumption of molecular oxygen in the electron transport chain [22]. The accumulation of superoxide anion is eliminated by manganese superoxide dismutase, which generates hydrogen peroxide. GSH within mitochondria aids the metabolism of hydrogen peroxide, via GSH peroxidase, thus removing it from the chain of reactions that generate hydroxyl radical and other reactive species, all of which can damage macromolecules associated with energy production and other mitochondrial functions. Therefore, it is not surprising that, through cellular evolution, mitochondria have the last reserves of cellular GSH in stressed cells. Interestingly, GSH is not synthesized within the mitochondria [17] but is transported in through a multicomponent transport system [23]. The high affinity with which GSH is transported into the mitochondrial matrix allows mitochondria to take up GSH even when the cytosolic GSH levels are low [23], thus assuring that residual amounts of cellular GSH are used in the protection of mitochondrial functions. Cellular GSH levels are not dramatically affected by low oxidant concentrations initially. This is most likely due to the millimolar amounts of GSH in the cell (3.5–5.0 mM in the case of leukocytes [1]) and to the efficiency of the enzymatic pathways that replenish GSH. Cellular GSH levels are maintained by glutathione reductase activity, which converts GSSG back to GSH [24], and the upregulation of GSH synthesis [25]. Continuous exposure to low oxidant concentrations, however, eventually causes GSH depletion, most likely due to the exhaustion of a particular substrate used to replenish GSH. What was surprising was the heterogeneity of GSH depletion in homogeneous cell lines. Cells

Fig. 5. GSH levels are maintained during oxidative stress by the upregulation of GSH synthesis. GSH is completely depleted in this Jurkat R cell exposed to 1.0 mM L-buthionine-(S,R)-sulfoximine and 0.1 mM diamide for 30 min. The last cellular GSH reserves in the mitochondria (M) are gone, and the cytoplasm and nucleoplasm (N) are heavily extracted, suggesting that oxidative damage caused permeability of both the plasma membrane and the nuclear envelope. Bar indicates 0.5 ␮m.

14

J.G. Ault, D.A. Lawrence / Experimental Cell Research 285 (2003) 9 –14

lost GSH at different rates. There may be a number of reasons for the differential depletion. Sensitivity to oxidantinduced stress may be cell cycle stage dependent, with certain stages being more sensitive than others. GSH levels are known to fluctuate throughout the cell cycle [26]. Also, heterogeneity may be caused by the indiscriminate randomness of oxidant-induced molecular changes and the impact that each change has to the GSH supply. Some changes lead to chain reactions that produce additional oxidant species and highly reactive chemicals, which cause greater cellular damage [27]. Such events could deplete GSH reserves faster. In particular, changes that increase membrane permeability could decrease cellular GSH levels faster, by letting more oxidants in faster and GSH out, as would changes that impact— either directly or indirectly—the cell’s ability to replenish GSH. Indeed, we showed in this investigation that GSH reserves during oxidant exposure are depleted much faster when GSH synthesis is inhibited. Cells have evolved defense mechanisms to resist oxidative stress. GSH, in millimolar concentrations that provide a large reserve of reducing equivalents to counter oxidation, is one of them. In this investigation, we have shown that GSH is located throughout the cell to provide protection and that only when GSH is heavily depleted do structural signs of oxidative stress occur.

Acknowledgments We thank James Kirkwood for culturing the cells used in this study and the Wadsworth Center’s Photography/Illustrations Unit for help with illustrations. This research was done at the Wadsworth Center’s EM Core Facility and was supported in part by grant NIH ES03778.

References [1] N.S. Kosower, E.M. Kosower, The glutathione status of cells, Int. Rev. Cytol. 54 (1978) 109 –160. [2] M. Suthanthiran, M.E. Anderson, V.K. Sharma, A Meister, Glutathione regulates activation-dependent DNA synthesis in highly purified normal T lymphocytes stimulated via the CD2 and CD3 antigens, Proc. Natl. Acad. Sci. USA 87 (1990) 3343–3347. [3] F.J. Staal, M. Roederer, L.A. Herzenberg, L.A. Herzenberg, Intracellular thiols regulate activation of nuclear factor ␬B and transcription of human immunodeficiency virus, Proc. Natl. Acad. Sci. USA 87 (1990) 9943–9947. [4] B.A. Arrick, C.F. Nathan, Glutathione, metabolism as a determinant of therapeutic efficacy: a review, Cancer Res. 44 (1984) 4224 – 4232. [5] T. Finkel, N.J. Holbrook, Oxidants, oxidative stress and the biology of ageing, Nature 408 (2000) 239 –247. [6] B. Halliwell, Oxidants and human disease: some new concepts, FASEB J. 1 (1987) 358 –364. [7] O.W. Griffith, L-Buthionine-SR-sulfoximine: mechanism of action, resolution of diastereomers and use as a chemotherapeutic agent, in: N. Taniguchi, T. Higashi, Y. Sakamoto, A. Meister (Eds.), Glutathione Centennial. Molecular Perspectives and Clinical Implications, Academic Press, San Diego, 1989, pp. 285–299.

[8] J.P. Messina, J. Mazurkiewicz, D.A. Lawrence, Production and characterization of monoclonal antibodies to thiol-modified glutathione, in: P.A. Cerutti, O.F. Nygaard, M.G. Simic (Eds.), Anticarcinogenesis and Radiation Protection, Plenum, New York, 1987, pp. 407– 412. [9] M.J. Meredith, D.J. Reed, Status of the mitochondrial pool of glutathione in the isolated hepatocyte, J. Biol. Chem. 257 (1982) 3747– 3753. [10] M.A. Tirmenstein, D.J. Reed, The glutathione status of rat kidney nuclei following administration of buthionine sulfoximine, Biochem. Biophys. Res. Commun. 155 (1988) 956 –961. [11] G. Bellomo, M. Vairetti, L. Stivala, F. Mirabelli, P. Richelmi, S. Orrenius, Demonstration of nuclear compartmentalization of glutathione in hepatocytes, Proc. Natl. Acad. Sci. USA 89 (1992) 4412– 4416. [12] G. Bellomo, M. Vairetti, G. Palladini, F. Mirabelli, L. Stivala, P. Richelmi, A nuclear pool of glutathione in hepatocytes, in: G. Poli, E. Albano, M.U. Dianzani (Eds.), Free Radicals: From Basic Science to Medicine, Birkha¨ user Verlag, Basel, 1993, pp. 400 – 411. [13] D.W. Voehringer, D.J. McConkey, T.J. McDonnell, S. Brisbay, R.E. Meyn, Bcl-2 expression causes redistribution of glutathione to the nucleus, Proc. Natl. Acad. Sci. USA 95 (1998) 2956 –2960. [14] J.M. Hansen, K.K. Harris, M.A. Philbert, C. Harris, Thalidomide modulates nuclear redox status and preferentially depletes glutathione in rabbit limb versus rat limb, J. Pharmacol. Exptl. Therapeutics 300 (2002) 768 –776. [15] B.E. Sandstrom, S.L. Marklund, Effects of variation in glutathione peroxidase activity on DNA damage and cell survival in human cells exposed to hydrogen peroxide and t-butyl hydroperoxide, Biochem. J. 271 (1990) 17–23. [16] K. Briviba, G. Fraser, H. Sies, B. Ketterer, Distribution of the monochlorobimane-glutathione conjugate between nucleus and cytosol in isolated hepatocytes, Biochem. J. 294 (1993) 631– 633. [17] O.W. Griffith, A. Meister, Origin and turnover of mitochondrial glutathione, Proc. Natl. Acad. Sci. USA 82 (1985) 4668 – 4672. [18] V.P. Nacife, M. deN.C. Soeiro, T.C. Arau´ jo-Jorge, H.C.C.-F. Neto, M.deN.L. Meirelles, Ultrastructural, immunocytochemical and flow cytometry study of mouse peritoneal cells stimulated with carrageenan, Cell Struct. Funct. 25 (2000) 337–350. [19] C. Hwang, A.J. Sinskey, H.F. Lodish, Oxidized redox state of glutathione in the endoplasmic reticulum, Science 257 (1992) 1496 –1502. [20] G. Ba´ nhegyi, L. Lusini, F. Puska´ s, R. Rossi, R. Fulceri, L. Braun, V. Mile, P. di Simplicio, J. Mandl, A. Benedetti, Preferential transport of glutathione versus glutathione disulfide in rat liver microsomal vesicles, J. Biol. Chem. 274 (1999) 12213–12216. [21] J.W. Cuozzo, C.A. Kaiser, Competition between glutathione and protein thiols for disulphide-bond formation, Nat. Cell Biol. 1 (1999) 130 –135. [22] J.C. Ferna´ ndez-Ceca, C. Garcı´a-Ruiz, A. Colell, A. Morales, M. Mari, M. Miranda, E. Ardite, Oxidative stress: role of mitochondria and protection by glutathione, BioFactors 8 (1998) 7–11. [23] J. Mårtensson, J.C.K. Lai, A. Meister, High-affinity transport of glutathione is part of a multicomponent system essential for mitochondrial function, Proc. Natl. Acad. Sci. USA 87 (1990) 7185–7189. [24] N.S. Kosower, E.M. Kosower, B. Wertheim, W.S. Correaa, Diamide, a new reagent for the intracellular oxidation of glutathione to the disulfide, Biochem. Biophys. Res. Commun. 37 (1969) 593–596. [25] J.S. Woods, T.J. Kavanagh, J. Corral, A.W. Reese, D. Diaz, M.E. Ellis, The role of glutathione in chronic adaptation to oxidative stress: studies in a normal rat kidney epithelial (NRK52E) cell model of sustained upregulation of glutathione biosynthesis, Toxicol. Appl. Pharmacol. 160 (1999) 207–216. [26] H. Ohara, T. Terasima, Variations of cellular sulfhydryl content during cell cycle of HeLa cells and its correlation of cyclic change of x-ray sensitivity, Exp. Cell Res. 58 (1969) 182–185. [27] J.D. Hayes, L.I. McLellan, Glutathione and glutathione-dependent enzymes represent a co-ordinately regulated defence against oxidative stress, Free Radic. Res. 31 (1999) 273–300.