The role of glutathione (GSH) in determining sensitivity to platinum drugs in vivo in platinum-sensitive and -resistant murine leukaemia and plasmacytoma and human ovarian carcinoma xenografts

Numerous studies performed in vitro have suggested a role for glutathione (GSH) in determining the sensitivity/resistance of tumour cells to various platinum-based drugs. Few studies have extended these findings into the in vivo setting. We have measured GSH levels in two murine (ADJ/PC6 plasmacytoma and L1210 leukaemia and their acquired platinum-drug-resistant sublines) and five human ovarian carcinoma (PXN/100, PXN/109T/C, SKOV3, HX/62 and OVCAR-3) tumour models of varying sensitivity to cisplatin. Results showed that relatively high GSH levels may be involved, at least partially, in determining platinum drug resistance in vivo in at least some of the tumour models studied (ADJ/PC6 carboplatin and tetraplatin resistant tumours; L1210 cisplatin and tetraplatin resistant tumours and the HX/62 and OVCAR-3 human ovarian carcinoma xenografts). However, in other tumours (e.g., the acquired cisplatin resistant ADJ/PC6 plasmacytoma) non-GSH mediated mechanisms of resistance (such as enhanced DNA repair) probably account for the resistance. Pretreatment of animals with oral buthionine sulfoximine (BSO), which resulted in approximately 70% depletion in tumour GSH levels, failed to potentiate the antitumour efficacy of either cisplatin (using the ADJ/PC6 and L1210 models) or the 1,2-diaminocyclohexane (DACH) platinum-drug, tetraplatin (using the ADJ/PC6 model). These BSO plus or minus cisplatin data suggest a limited role for such combinations in the clinic.

In vitro glutathione supplementation enhances interleukin-2 production and mitogenic response of peripheral blood mononuclear cells from young and old subjects

The effect of in vitro glutathione (GSH) supplementation on mitogenic response, interleukin-1, interleukin-2 and prostaglandin E2 production, and cellular GSH level in peripheral blood mononuclear cells (PBMC) from healthy young and old human subjects was studied. In vitro addition of GSH increased cellular GSH level (P < 0.001). Glutathione supplementation at concentrations between 2 to 10 mmol/L enhanced lymphocyte proliferation but at low concentrations (0.5 and 1 mmol/L) decreased mitogenic response. Glutathione-induced enhancement of lymphocyte proliferation due to phytohemagglutinin or concanavalin A was greater in the PBMC from old subjects than in those from young subjects. At optimal concentration (5 mmol), GSH increased interleukin-2 production (P < 0.05) and decreased prostaglandin E2 and leukotriene B4 production (P < 0.01) in both age groups. Furthermore, decreased PBMC mitogenic response by in vitro addition of prostaglandin E2 was reversed by GSH supplementation. Glutathione did not have an effect on interleukin-1 production by PBMC from young subjects; however, GSH supplementation tended (P = 0.08) to increase interleukin-1 production by PBMC from old subjects. We conclude that GSH supplementation enhances T cell-mediated mitogenic response in young and old subjects. This effect is due at least in part to decreased eicosanoid production.

Nonenzymatic reductive activation of 7-N-((2-([2-(gamma-L-glutamylamino)ethyl]dithio)ethyl))mitomycin C by thiol molecules: a novel mitomycin C derivative effective on mitomycin C-resistant tumor cells

7-N-((2-([2-(gamma-L-Glutamylamino)ethyl]dithio)ethyl))mitomycin C (KW-2149) is an analogue of mitomycin C (MMC) and has prominent activities against various tumors. We studied the antitumor effects of KW-2149 in MMC-resistant variants of human colon carcinoma HT-29 (HT-29/MMC) and mouse hepatoma Hepa-I (C4, B13NBii1) cells, which are deficient in DT-diaphorase and cytochrome P450 reductase, respectively. These enzymes mediate the reductive activation of MMC in the cells. Although HT-29/MMC and C4, B13NBii1 cells showed significant resistance to MMC, they showed sensitivity tl KW-2149 comparable to their parental tumors, indicating that DT-diaphorase and cytochrome P450 reductase could not be involved in the activation of KW-2149. In studying the activation mechanism of KW-2149, we found that glutathione (GSH) and cysteine significantly enhanced the cytotoxicity of KW-2149 in HT-29 cells. The DNA adduct of KW-2149 was increased when HT-29 cells or the isolated nuclei of the cells were incubated with KW-2149 in the presence of physiological concentrations of GSH and cysteine. KW-2149 alkylated calf thymus DNA in the presence of GSH and cysteine in vitro. These results indicate that activation of KW-2149 by thiol molecules, unlike MMC, could be an important activation mechanism of KW-2149 to form DNA adduct and to exert its cytotoxicity. This is the reason why KW-2149 is effective against MMC-resistant tumors with deficiencies in the MMC activation enzymes.

Synthesis and transport of glutathione by cultured human retinal pigment epithelial cells

PURPOSE

To characterize synthesis and transport of glutathione (GSH) by cultured human retinal pigment epithelial (RPE) cells.

METHODS

Cultured human RPE cells were depleted of glutathione, then incubated with various concentrations of cysteine, glutamate, and glycine or with glutathione. High-performance liquid chromatography was used to measure intracellular glutathione with time.

RESULTS

After depletion with diethylmaleate, intracellular glutathione was resynthesized from the amino acid precursors within 60 minutes. The addition of buthionine sulfoximine, a known inhibitor of GSH synthesis, completely eliminated the observed increase. Similar incubation with exogenous GSH failed to increase the intracellular GSH concentration.

CONCLUSION

Cultured human RPE cells are able to rapidly synthesize glutathione from exogenously administered amino acids, but they are incapable of direct GSH uptake.

Prostaglandin E2 synthesis after oxidant stress is dependent on cell glutathione content

The role of glutathione in protecting prostaglandin (PG) generation after exposure of fibroblasts to oxidant stress was investigated. Exposure of 3T3 fibroblasts to H2O2, followed by washing and then 20 microM arachidonic acid, caused a dose-dependent decrease in PG synthesis as assessed by radioimmunoassay. PGE2 production decreased from 3.7 +/- 1.1 to 0.15 +/- 0.04 pmol/microgram protein, and prostacyclin (PGI2) formation decreased from 0.56 +/- 0.03 to 0.06 +/- 0.03 pmol/microgram protein after exposure to 200 microM H2O2. Decreasing intracellular glutathione with 50 micrograms/ml 1,3-bis(chloroethyl)-1-nitrosourea (BCNU) enhanced the H2O2-induced decrease in PGE2 synthesis. Another glutathione-depleting agent, 1-chloro-2,4-dinitrobenzene (CDNB), also potentiated the H2O2-induced decrease in PGE2 formation. However, although PGI2 production was decreased by H2O2, neither BCNU nor CDNB potentiated this decrease. Without oxidant stress, extreme glutathione depletion decreased PGE2 synthesis and caused PGI2 synthesis to exceed PGE2. In summary, oxidant stress decreases both PGE2 and PGI2 formation. However, the primary effect of decreasing cell glutathione during oxidant stress is a reduction in PGE2 formation, not PGI2. This implies that the predominant effect of glutathione depletion during oxidant stress is on the PGE2 isomerase(s) and not PGH synthase or PGI2 synthase.

Relationship between cellular glutathione level and susceptibility to LAK killing in human pharyngeal carcinoma cell line

We examined the relationship between cellular glutathione (GSH) level and susceptibility to lymphokine-activated killer (LAK) cell-mediated cytolysis in KB human pharyngeal carcinoma cells. Treatment of KB cells with D,L-buthionine-S,R-sulfoximine (BSO), a gamma-glutamyl cysteine synthetase blocker, resulted in decreased total intracellular GSH levels associated with increased susceptibility to LAK killing. In contrast, treatment with oxothiazolidine-4-carboxylate (OTZ, a precursor of cysteine), which is known to increase cellular GSH level, decreased the susceptibility of KB cells to LAK killing. Both agents had no effects on binding frequency of KB cells to LAK cells. These results suggest that intracellular GSH in tumor cells play a protective role against LAK mediated cytolysis, specially in the post-binding killing phase.

Modulation of gamma-glutamylcysteine synthetase large subunit mRNA expression by butylated hydroxyanisole

Dietary 2(3)-tert-butyl-4-hydroxyanisole (BHA) treatment has been shown to increase hepatic glutathione (GSH) content in rats and mice. Subsequent studies in our laboratory have demonstrated that hepatic gamma-glutamylcysteine synthetase (GCS) activity is increased in mice treated with dietary BHA. To test whether this increase in GCS activity follows an increase in hepatic messenger RNA for the large subunit of GCS (GCS-LS mRNA), a 390-base pair fragment corresponding to a region near the 5' end of the rat GCS-LS cDNA sequence was amplified using the PCR reaction and used to detect GCS-LS mRNA on Northern blots. Hepatic GSH, GCS activity, and GCS-LS mRNA levels were determined either in mice treated with BHA in the diet for 12 days or mice injected with diethyl maleate (DEM), phorone, and/or DL-buthionine-[S,R]-sulfoximine (BSO) over a 24 hr period. BHA caused a 1.5-fold increase in GSH levels, a 1.7-fold increase in hepatic GCS activity by Day 12, and a rapid 5-fold increase in hepatic GCS mRNA levels reaching maximal levels after 2-3 days. Partial depletion of GSH with either phorone (70%) or DEM (50%) resulted in a 4- to 5-fold increase in hepatic GCS-LS mRNA levels by 9 hr and a 1.5- to 2-fold increase in hepatic GSH and GCS activity by 24 hr. Depletion of GSH with the GCS enzyme inhibitor BSO had no effect on GCS mRNA expression, even though GSH was depleted to 30%. When BSO was combined with the phorone treatment GSH levels were depleted to < 10%, but the large increase in GCS-LS mRNA seen with phorone alone was greatly attenuated. These data suggest that depletion of GSH per se, is not sufficient to induce elevation of GCS-LS mRNA levels, but that the formation of GSH conjugates may be required to trigger GCS-LS mRNA induction. The increase in GCS-LS mRNA levels may account for the increase in GCS activity and elevation of GSH observed following BHA treatment, as well as the "rebound" of GSH above control levels observed 18-24 hr following depletion of GSH by other chemicals. These results are consistent with the Michael acceptor, hypothesis by Talalay.

Protection of cultured rat gastric cells against oxidant-induced damage by exogenous glutathione

BACKGROUND/AIMS

Reduced glutathione (GSH) is an intracellular protectant against oxidants. The present study determined whether extracellular GSH protects against oxidant damage or whether an uptake system of GSH is present in cultured gastric cells.

METHODS

Hydrogen peroxide was generated by glucose oxidase and glucose. Cytotoxicity was assessed by 51Cr release. Intracellular GSH was assayed by the method of Tietze.

RESULTS

Pretreatment with extracellular GSH decreased H2O2-induced 51Cr release. Treatment with GSH enhanced cellular GSH content. Protection by pretreatment with GSH was prevented by buthionine sulfoximine (an inhibitor of gamma-glutamylcysteine synthetase). Enhancement of intracellular GSH was also prevented by buthionine sulfoximine. Acivicin (an inhibitor of gamma-glutamyl transpeptidase) prevented intracellular accumulation of GSH from extracellular GSH. Cysteine was effective in preventing damage and enhancing intracellular GSH content, whereas both glutamine and glycine were not.

CONCLUSIONS

Extracellular GSH protects cultured gastric cells from H2O2 damage by accelerating intracellular GSH synthesis; this is mediated by membrane-bound gamma-glutamyl transpeptidase acting on extracellular GSH (which supplies these cells with cysteine) and then by intracellular gamma-glutamylcysteine synthetase.

Effect of GSH depletion mediated by l-BSO on the insecticidal activity of DDT and fenitrothion in Triatoma infestans (Hemiptera)

l-BSO injected at a sublethal dose in nymph V of Triatoma infestans produces a transient GSH depletion in the abdomen. A depletion of the total content of GSH was observed in nymph II fed with a sublethal concentration of l-BSO. Either ingestion or injection pretreatments of T. infestans nymph V with sublethal concentrations of l-BSO produced a slight synergism on the toxicity of fenitrothion. Feeding administration of a sublethal concentration of l-BSO to nymphs II of T. infestans resulted in a weak potentiation of the acute toxicity of DDT.

Antisera to glutathione: characterization and immunocytochemical application to the rat cerebellum

Rabbits were immunized with reduced glutathione (gamma-glutamyl-cysteinyl-glycine) coupled to bovine serum albumin by glutaraldehyde or a mixture of glutaraldehyde and formaldehyde. The antisera that were formed were tested qualitatively, by screening them against more than 50 amino acids and peptide conjugates that had been immobilized on cellulose discs (spot test), and quantitatively, by immunogold analysis of test conjugates that had been embedded in an epoxy resin. It was shown that the antisera selectively recognized the reduced and oxidized forms of glutathione and that they did not exhibit any significant crossreactivity with glutamate, cysteine, glycine, gamma-glutamyl-cysteine or cysteinyl-glycine. Immunocytochemistry of Vibratome sections of rat cerebellum suggested that glutathione occurs in glial cells as well as in neurons. This was confirmed by electron microscopic, immunogold cytochemistry of tissue from rat cerebellum that had been freeze-substituted and embedded in Lowicryl under low temperature. Gold particles were concentrated over Golgi epithelial cells and perivascular glial processes, but also occurred over several types of neuronal profile including Purkinje and granule cell bodies, and mossy fibre terminals. At the subcellular level, glutathione-like immunoreactivity was found in the cytoplasmic matrix, mitochondria and nuclei. The immunolabelling intensity was strongly reduced in animals that had been pretreated with buthionine sulphoximine, which is known to depress the level of glutathione by inhibiting gamma-glutamyl-cysteine synthetase. The availability of antisera to glutathione is likely to further our understanding of the physiological and pathophysiological roles of this tripeptide.

Mechanism of methylmercury efflux from cultured astrocytes

To study the mechanism of methylmercury (MeHg) efflux from the central nervous system cells, cultured astroglia obtained from neonatal rats were incubated with 10 microM MeHg-cysteine (CySH) for 30 min. After being washed four times, cells were incubated in Hg-free medium, and the release of MeHg from the cells was monitored. The amount of MeHg released in the medium approached a plateau level (ca. 31% of the loaded amount) at 4 hr. Treatment of the cells with a CySH precursor, 2-oxothiazolidine-4-carboxylic acid (OTC), resulted in a significant increase of cellular levels of CySH and glutathione (GSH). OTC also increased 1.5-fold the MeHg efflux from the loaded cells. Another GSH enhancer, GSH isopropyl ester, also stimulated MeHg export from the cells. Ion-exchange column chromatography using DEAE-Sephadex revealed that the MeHg metabolite thus released was exclusively MeHg-GSH conjugate, both with and without OTC. Since the MeHg efflux was suppressed significantly by the presence of probenecid, the efflux occurred via the probenecid-sensitive organic acid transport system. Even though the cellular GSH levels were depleted drastically by treatment with L-buthionine-(S,R)-sulfoximine (BSO), a considerable level (90% of the control) of Hg efflux was detected. Since neither GSH- nor CySH-MeHg was detected in the culture medium of the BSO-treated cells, GSH depletion may trigger some other secretion system(s) in the cells. These results suggest that conjugation with GSH is the major pathway for MeHg efflux in rat astroglia, and that elevation in the cellular GSH level would possibly be a logical therapy for MeHg poisoning, promoting the accelerated elimination of MeHg from the critical tissues.

Evaluation of methods for measuring cellular glutathione content using flow cytometry

The currently available flow cytometric stains for cellular glutathione were evaluated, examining the labelling of both human and rodent cell lines under various conditions of concentration, time, and temperature. Procedures were used that depleted glutathione (GSH) while having a minimal effect on other cellular sulphydryls in order to estimate linearity and the extent of background staining. As previously reported, monochlorobimane was highly specific for GSH in rodent cells but failed to label human cells adequately because of its low affinity for human glutathione S-transferases. Higher concentrations of monochlorobimane achieved more complete labelling of the human cellular GSH pool but gave increased background fluorescence due to non-GSH binding. The analogue monobromobimane, which binds nonenzymatically to sulphydryls, reacted more readily with GSH than with protein sulphydryls and, provided that stain concentration and incubation time were controlled, gave reproducible staining of human cells with approximately 20% of total fluorescence due to background staining. Of the currently available stains for measuring GSH in human cells, monobromobimane is the agent of choice. Mercury orange also binds more readily to GSH than to protein, giving a degree of specificity, and it has the additional advantage of being excited at 488 nm. However, the reproducibility of staining with mercury orange was less consistent than that using monobromobimane, and a higher background fluorescence was seen. Two additional stains, o-phthaldialdehyde and chloromethyl fluorescein, could also be used to label cellular GSH, but both gave an unacceptably high level of background staining. It is recommended that flow cytometric GSH assays should routinely include a sample of cells that have been depleted of GSH in order to determine the extent of background labeling.

Metal-induced genotoxic adaptation in barley (Hordeum vulgare L.) to maleic hydrazide and methyl mercuric chloride

Presoaked seeds of barley, Hordeum vulgare L., were exposed for 2 h to maleic hydrazide (MH), 5 x 10(-2) M or methyl mercuric chloride (MMCl), 10(-4) M with or without a prior conditioning with MH, 5 x 10(-3) M; MMCl, 10(-5) M; cadmium sulfate (CdSO4), 10(-4) M or zinc sulfate (ZnSO4), 10(-1) M; the interexposure time was 2 h. Subsequently as the seeds germinated a number of endpoints were measured that included mitotic index, mitotic chromosome aberrations and micronuclei (MNC) in embryonic shoot cells fixed at 32, 36, 40, 44, 48 and 52 h of recovery, and seedling height on day 7. The results demonstrated that prior conditioning exposure to MH or metals induced genotoxic adaptation to the subsequent challenge exposure to MH and MMCl. Cadmium-induced genotoxic adaptation against either MH or MMCl challenge exposure was, however, significantly prevented when the presoaked seeds were pre-exposed to buthionine sulfoximine, 10(-3) M for 2 h, thereby providing evidence that the underlying mechanism of genotoxic adaptation possibly involved phytochelatins.

Buthionine sulfoximine protects the viability of adult rat Leydig cells exposed to ethane dimethanesulfonate

Exposure to (EDS) in vivo or in vitro alters adult rat Leydig cell function within 3 hr and these cells subsequently die within 24-48 hr. Previously, we determined that reducing intracellular glutathione levels with buthionine sulfoximine (BSO) protects adult rat Leydig cell function during short-term (3 hr) EDS exposures. The present study extends these observations by assessing whether BSO also can prevent Leydig cell death following EDS exposure. To assess whether BSO can prevent EDS-induced Leydig cell death, Leydig cells were cultured in the presence or the absence of 4 mM BSO (12 hr), and subsequently with increasing doses of EDS (3 hr). The effects of EDS on Leydig cell death were estimated using a cytotoxicity assay (MTT assay) and by assessing the ability of adult rat Leydig cells to maintain LH-stimulated testosterone production and [35S]methionine incorporation 24 hr following EDS exposure. This culture duration was chosen because sustained alterations in MTT reduction, LH-stimulated testosterone production, and [35S]methionine incorporation 24 hr following EDS exposure were found to reflect alterations in Leydig cell viability. The results suggest that lowering intracellular levels of glutathione with BSO prevents EDS-induced Leydig cell death. Restoring intracellular levels of glutathione with glutathione ethyl ester (8 mM) in the presence of BSO also restored the ability of EDS to kill Leydig cells. Taken together, these data suggest that glutathione mediates the mechanism by which EDS kills adult rat Leydig cells.

Buthionine sulfoximine treatment impairs rat diaphragm function

Activation of the glutathione (GSH) redox cycle with production of glutathione disulfide (GSSG) has been shown to occur in the diaphragm during inspiratory resistive loading (RB). Buthionine sulfoximine (BSO) lowers tissue GSH by irreversibly inhibiting the rate-limiting synthesis enzyme gamma-glutamylcysteine synthetase. We investigated the effects of BSO on rat diaphragm function, both at rest and after a period of RB. Rats in the RB groups underwent inspiratory RB until they were unable to sustain 70% of their maximal airway pressure. A portion of the diaphragm was analyzed for GSH and GSSG levels, and measurements of in vitro contractile properties included contraction times, maximal tetanic tension (Po), maximal twitch tension (Pt), and force frequency curves. BSO treatment produced a profound depletion of diaphragmatic GSH. Neither BSO nor RB alone significantly altered diaphragm contractile properties at this load of RB. But, in BSO-RB rats, there was a significant decrease in Pt, Po, and tetanic tension at all frequencies of stimulation compared with those in other groups. These data reveal that animals treated with BSO followed by inspiratory resistive loading exhibit marked diaphragm impairment, suggesting that GSH may play an important role in protecting the diaphragm from the stress induced by this resistive breathing protocol.

Nephrotoxicity mechanism of cis-platinum (II) diamine dichloride in mice

Male Swiss OF1 mice were injected subcutaneously with 20 mg/kg of cis-platinum (II) diamine dichloride (cis-platin). Examination of cryostat kidney sections stained for alkaline phosphatase (APP) revealed damage to about 10, 20, 40 and 50% of the proximal tubules after 7, 24, 48 and 72 h, respectively. Pretreatment with the glutathione synthesis inhibitor, buthionine sulfoximine (BSO), (i.p. 3 mmol/kg) potentiated the tubule damage of cis-platin. In contrast, pretreatment with organic anion transport inhibitor probenecid (i.p. 3 x 0.75 mmol/kg) reduced the number of damaged tubules by approximately 80% at 72 h after cis-platin injection. Pretreatment with the gamma-glutamyltranspeptidase (gamma-GT) inactivator acivicin (AT-125, 50 mg/kg p.o., plus 50 mg/kg i.p.) failed to prevent cis-platin induced renal toxicity. Pretreatment with the beta-lyase inactivator aminooxyacetic acid (AOAA, 2 x 100 mg/kg p.o.) and with the renal cysteine conjugate S-oxidase inhibitor methimazole (40 mg/kg i.p.) reduced the number of damaged tubules by approximately 40% and 75%, respectively in mice treated with cis-platin. The results suggest that the platinum-sulfhydryl group complexes formed are taken up by the kidney cells through an organic anion transport mechanism which is probenecid-sensitive. In the cells these complexes are stable for several hours, depending on the intracellular glutathione (GSH) level, and gradually undergo transformation to reactive metabolite(s) by renal intracellular beta-lyase and S-oxidase.

Development and characterization of hydrogen peroxide-resistant Chinese hamster ovary (CHO) cell variants--II. Relationships between non-protein sulfhydryl levels and the induction/stability of the oxidant-resistant phenotype

Hydrogen peroxide sensitive and resistant sublines of Chinese hamster ovary (CHO) cells were tested for their non-protein sulfhydryl (NPSH) content in an attempt to establish whether a relationship exists between resistance to growth inhibition elicited by the oxidant and the NPSH pool. Cell variants characterized by increasing levels of resistance to hydrogen peroxide displayed a significant increase in cellular NPSH (expressed on a per million cell basis). Growth of resistant cells for various lengths of time in the absence of H2O2 decreased resistance, whereas the NPSH content did not vary (at least up to 127 days of growth in peroxide-free medium). The NPSH pool returned to control levels after an additional 82 days. These changes, however, were probably related to differences in cell size/amount of total cell proteins in the sublines. Indeed, when NPSH levels were expressed on a per milligram protein basis, essentially no variations were observed in sensitive and resistant sublines. It is important to note that, even by expressing the NPSH content on a per million cell basis, no correlation was found with the degree of resistance to growth inhibition elicited by the oxidant. Further experiments have demonstrated that, under conditions of reduced NPSH content (obtained by growing the cells in the presence of a glutamylcysteine synthetase inhibitor), the cytotoxic action of hydrogen peroxide was very slightly, if at all, augmented in both wild type and resistant cells. We may therefore conclude that cellular NPSH do not afford significant protection against growth inhibition induced by hydrogen peroxide in wild type cells, and that the same lack of effect occurs in cells with an increased NPSH content and carrying the oxidant-resistant phenotype.

Effect of glutathione depletion on formation of paramagnetic chromium in Chinese hamster V-79 cells

Incubation of Chinese hamster V-79 cells with either buthionine sulfoximine (BSO) or diethylmaleate (DEM) prior to exposure to Na2CrO4 resulted in a depletion of GSH. ESR study shows that the depletion of GSH by BSO caused an increase in the cellular level of Cr(V) intermediate without affecting the level of Cr(III) complex, whereas the levels of Cr(V) and (III) were both suppressed by the depletion of GSH by DEM. GSH depletion by DEM decreased cellular uptake of chromate more than that by BSO. Glutathione reductase activity in cells, which is capable of reducing Cr(VI), was unaffected by BSO, however a strong inhibition was observed in cells treated with DEM, indicating that DEM affects not only GSH levels but also Cr(VI)-reducing enzyme activity. Thus, the depletion of GSH in cells increases Cr(V) formation, based upon the effect of BSO. The role of intracellular GSH in Cr(VI) reduction is discussed.

Role of cellular glutathione and glutathione S-transferase in the expression of alkylating agent cytotoxicity in human breast cancer cells

Glutathione (GSH) and glutathione S-transferases (GSTs) play an important role in the protection of cells against toxic effects of many electrophilic drugs and chemicals. Modulation of cellular GSH and/or GST activity levels provides a potentially useful approach to sensitizing tumor cells to electrophilic anti-cancer drugs. In this study, we describe the interactions of four representative alkylating agents (AAs), melphalan, 4-hydroperoxy-cyclophosphamide (4HC), an an activated form of cyclophosphamide, 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU), and cisplatin, with GSH and GST in the human breast cancer cell line MCF-7. Depletion of cellular GSH pools by approximately 80% by treatment of the cells with the GSH synthesis inhibitor buthionine sulfoximine (BSO) sensitized the tumor cells to each AA to a different extent, with dose-modifying factors of 2.39, 2.21, 1.64, and 1.27 observed for melphalan, 4HC, cisplatin, and BCNU, respectively. Treatment of the cells with the GST inhibitor ethacrynic acid (EA) failed to show any significant effects on the cytotoxicity of these AAs. However, EA did potentiate the cytotoxicity of melphalan when given in combination with BSO, an effect that may be due to a more complete depletion of cellular GSH levels by the combined modulator treatment. Following a 1-hr exposure to cytotoxic-equivalent concentrations of these AAs, GSH levels decreased substantially in the case of 4HC and BCNU, but increased by 30-50% in the case of cisplatin and melphalan. BSO pretreatment largely blocked this effect of cisplatin and melphalan on cellular GSH, while it further enhanced the GSH-depleting activity of both 4HC and BCNU. On the basis of these results, it is concluded that (a) GSH affects the cytotoxicity of different AAs to different extents, (b) basal GST expression in MCF-7 cells does not play a major role in AA metabolism, (c) EA can potentiate the enhancing effect of BSO on melphalan cytotoxicity in MCF-7 cells, and (d) depletion of cellular GSH by pretreatment with BCNU or cyclophosphamide may correspond to a useful strategy for enhancing the anti-tumor activity of other AAs given in a sequential combination.

Exogenous glutathione attenuates the antiproliferative effect of buthionine sulfoximine

Buthionine sulfoximine (BSO) inhibits proliferation of human lung carcinoma A549 cells in a manner that does not correlate with intracellular glutathione (GSH) depletion, nor does it reflect overt toxic effects of BSO. However, BSO inhibits uptake by A549 cells of cystine, which is an essential amino acid for cell growth in culture. Thus, it is hypothesized that inhibition of cellular cystine uptake is, or is partially, responsible for the antiproliferative effect of BSO. It has been shown that the gamma-glutamyl amino acid transport system plays a role in cystine transport across cell membranes. This transport system requires extracellular GSH for its operation. BSO, by inhibiting intracellular GSH synthesis, would reduce GSH export and decrease extracellular GSH levels. Therefore, the present study was undertaken to examine the effect of exogenously added GSH on BSO inhibition of cellular cystine uptake and its relationship to the antagonistic effect of GSH on BSO antiproliferation. A549 cells were treated with 10 mM BSO and exogenous GSH was added to these BSO-treated cultures. Effects of exogenous GSH on BSO antiproliferation and cellular GSH depletion were determined simultaneously as a function of time. The effect of GSH on BSO inhibition of cystine accumulation was measured using [35S]cystine. The results obtained demonstrate that exogenously added GSH partially overcame BSO antiproliferation. The GSH antagonistic effect did not correlate with repletion of intracellular GSH, but it did correlate with recovery of BSO-inhibited cystine accumulation. Exogenous GSH also enhanced proliferation of non-BSO treated cells at concentrations below 1.0 mM. The results of this study suggest that BSO inhibition of cystine uptake may represent one mechanism by which BSO exerts its antiproliferative effect. The antagonistic effect of exogenous GSH on BSO antiproliferation may result from recovery of BSO-inhibited cystine uptake, although other mechanisms responsible for the GSH antagonistic effect may also exist.

Significance of the proportion of binucleate cells in the micronucleus assay: a methodological study

Using treatment with cytochalasin-B (Cyt-B) for the induction of a cytokinetic block, the significance of the proportion of binucleate cells (BNC) in the micronucleus (MN) assay was investigated in a methodological study. A Chinese hamster cell line V79 was used in which MN were induced by radiation. In complementary tests the radiation effect in inducing MN was enhanced by depletion of the cellular glutathione content with buthionine sulfoximine (BSO). The data indicated that the concentration of Cyt-B is the major factor which determines the proportion of BNC. This proportion was shown to be independent of radiation dose and of BSO. Furthermore, the MN frequency was not related to the percentage of BNC. Therefore, a high proportion of BNC may be practical for the MN assay, but may not make the technique more accurate.

Biochemical regulation of the activity of gamma-glutamylcysteine synthetase from rat liver and kidney by glutathione

Rat liver and kidney gamma-glutamylcysteine synthetase (gamma GCS) had similar catalytic properties and consisted of heavy and light subunits, but the molecular structure of the two enzymes was not the same as evidenced by the results of SDS-PAGE and disc gel electrophoresis. Unlike kidney enzyme, most of liver gamma GCS was in a reduced enzyme form which did not have disulfide linkage between heavy and light subunits. Although the oxidized form of the two enzymes which subunits were linked with disulfide bond(s) could be dissociated to a similar extent by GSH, liver gamma GCS was inhibited by GSH to a much greater extent. These results suggest that the relative sensitivity of the gamma GCS enzymes to inhibition by GSH might be related to the inherent dissociability of heavy and light subunit of gamma GCS.

Modifications of cellular thiols during growth and squamous differentiation of cultured human bronchial epithelial cells

Thiol modifications during growth and differentiation of cultured normal human bronchial epithelial cells was studied by analysis of their content and redox state of low-molecular-weight thiols and protein thiols. Subculture of the cells with trypsin decreased the cellular content of the major low-molecular-weight thiol, i.e., reduced glutathione, although the glutathione content had returned to levels comparable to those before subculture already after 4 h in conjunction with cell attachment. During subsequent culture, increases in the cellular contents of glutathione, total cysteine equivalents, and total protein thiols occurred. These modifications in the amounts and redox balance of thiols were transient and preceded the major growth phase. Exposure of cells at clonal density to either diethylmaleate, a thiol-depleting agent, or buthionine sulfoximine, an inhibitor of glutathione synthesis, decreased the proliferative ability of the cells as demonstrated by a markedly decreased colony forming efficiency. Moreover, in mass cultures exposed to buthionine sulfoximine, a marked depletion of the glutathione content was again accompanied by inhibition of growth. Exposure of the cells to agents known to induce growth arrest and terminal squamous differentiation, i.e., fetal bovine serum, Ca2+, or transforming growth factor-beta 1, resulted in increased levels of reduced glutathione. No consistent alteration in the contents of the other thiols was noted. Overall, the results demonstrate consistent variations in the amounts and redox state of cellular thiols, particularly reduced glutathione, supporting a role of thiols in regulation of growth and squamous differentiation of human bronchial epithelial cells.

Toxicity of precocene II in rat hepatocyte cultures: effects of serum and culture time

Omission of serum from culture medium markedly increased the toxicity of the cytochrome P450-activated toxin precocene II in 24-h cultures of rat hepatocytes. In addition, the depletion of glutathione and inhibition of cytochrome P450 in 24-h cultures increased and decreased, respectively, the toxicity of precocene II, as seen in vivo. Precocene II was also toxic to 72-h cultures, although the effect of the serum-free medium was attenuated. However, depletion of glutathione and inhibition of P450 had little or no effect on the toxicity in 72-h cultures. Therefore, the mechanism of precocene II-induced cell death appears to change with time in culture.

Evidence for embryonic peroxidase-catalyzed bioactivation and glutathione-dependent cytoprotection in phenytoin teratogenicity: modulation by eicosatetraynoic acid and buthionine sulfoximine in murine embryo culture

Phenytoin teratogenicity may result from embryonic, peroxidase-catalyzed bioactivation of phenytoin to a toxic reactive free radical intermediate for which embryonic glutathione (GSH) is cytoprotective. This hypothesis was tested in embryo culture using 5,8,11,14-eicosatetraynoic acid (ETYA), a dual inhibitor of two peroxidase systems, prostaglandin synthetase, and lipoxygenases. Embryos from CD-1 mice were explanted on Gestational Day 9.5 (vaginal plug, Day 1) and incubated for 24 hr at 37 degrees C in culture medium (35% male rat serum, 15% fetal bovine serum, and 50% Waymouth's medium) saturated with 5% CO2 in air. Initially, a nonembryotoxic concentration of ETYA (0,40,80, or 100 microM) was established within its peroxidase inhibitory range (Ki = 4-8 microM). Subsequently, embryos were incubated with vehicle alone, a therapeutic concentration of phenytoin alone (20 micrograms/ml or 80 microM), ETYA alone (40 microM), or phenytoin and ETYA combined. ETYA alone below 100 microM had no effect on yolk sac diameter (YSD), crown-rump length (CRL), somite development (SD), anterior neuropore closure (ANPC), or turning, but at 100 microM reduced CRL, YSD, and SD (p < or = 0.05). Phenytoin alone was embryotoxic, causing reduced CRL, YSD, and SD (p < or = 0.0001). Phenytoin and ETYA (40 microM) together resulted in an increase in YSD, SD, and CRL relative to those with phenytoin alone (p < or = 0.01), indicating that inhibition by ETYA of embryonic, peroxidase-catalyzed bioactivation of phenytoin is cytoprotective. GSH may play a critical role in detoxifying a phenytoin free radical or subsequent activated oxygen species, thereby reducing covalent binding, lipid peroxidation, and oxidative stress that may initiate embryotoxicity or death. To test this hypothesis, embryos were cultured in the presence or absence of 1 mM buthionine sulfoximine (BSO), an inhibitor of GSH synthesis, for 3 hr, at which time BSO was washed out and the embryos were incubated for 24 hr in fresh culture medium containing 80 microM phenytoin or its vehicle. Soluble thiols, including GSH, and disulfides, including oxidized GSH (GSSG), were measured using high-performance liquid chromatography. Immediately after BSO treatment, there were no differences in the concentrations of GSH or GSSG between BSO-exposed embryos and controls. However, at 24 hr, GSH concentrations in untreated embryos increased almost 17-fold over those at 3 hr concentrations, while GSH in BSO-exposed embryos were reduced to 15% of control values (p = 0.0008).(ABSTRACT TRUNCATED AT 400 WORDS)

Influence of intracellular glutathione concentration of lipofuscin accumulation in cultured neonatal rat cardiac myocytes

The objective of this study was to investigate the relationship between intracellular glutathione (GSH) and lipofuscinogenesis in an established model system of cultured postmitotic neonatal rat cardiac myocytes exposed to moderate oxidative stress with respect to culture conditions (cells grown at 21% oxygen pressure). Intracellular glutathione was depleted by exposing cell cultures to buthionine sulfoximine (BSO), a specific inhibitor of gamma-glutamylcysteine synthetase; this caused an increase in lipofuscin-specific autofluorescence, reflecting lipofuscin accumulation. Cell cultures exposed to 100 microM BSO exhibited the following reactions, as compared to control cells: the intracellular glutathione level decreased 78%, 86%, and 89%, and lipofuscin-specific autofluorescence increased 26%, 28%, and 77% after 5, 8, and 14 days of treatment, respectively. Compared to the glutathione levels found in newly excised heart tissue, concentrations in cultured cardiac myocytes were lower during the first few days after culture establishment, probably a result of the cell preparation technique used. Due to this effect, differences between BSO-treated cells and control cells in regard to glutathione concentration and lipofuscin accumulation were more pronounced during the remainder of the 2-week culture period. Lipid peroxidation measured as thiobarbituric acid reactive substances was not increased in BSO-treated cells. These results are in agreement with the oxidative-stress theory of lipofuscinogenesis previously proposed from the authors' laboratory. The authors show that reduced GSH level leads to a simultaneous increase in accumulation of lipofuscin in cardiac myocytes, possibly by increasing the level of cytosolic hydrogen peroxide.

Responses of bovine lymphocytes to heat shock as modified by breed and antioxidant status

We tested whether resistance of lymphocytes to heat stress is modified by breed, intracellular glutathione content, and extracellular antioxidants. In the first experiment, lymphocytes from Angus (Bos taurus, non-heat-tolerant), Brahman (B. indicus, heat-tolerant), and Senepol (B. taurus, heat-tolerant) heifers (12 heifers per breed) were cultured at 45 degrees C for 3 h to evaluate thermal killing, at 42 degrees C for 12 h in a 60-h phytohemagglutinin-induced proliferation test, and at 42 degrees C for 1 h to measure induction of heat shock protein 70 (HSP70). Killing at 45 degrees C was affected by breed x temperature (P < .01); the decrease in viability caused by a temperature of 45 degrees C was greater for Angus than for Brahman or Senepol. For phytohemagglutinin-stimulated lymphocytes, heating to 42 degrees C reduced [3H]thymidine incorporation equally for all breeds. Viability at the end of culture was affected (P < .001) by a breed x temperature interaction because the decrease in viability caused by culture at 42 degrees C was greatest for lymphocytes from Angus heifers. Heat shock for 1 h at 42 degrees C caused a two- to threefold increase in intracellular concentrations of HSP70, but there was no interaction of temperature with breed. In another experiment (with lymphocytes harvested from three Holstein cows), buthionine sulfoximine, a glutathione synthesis inhibitor, inhibited (P < .01) proliferation of phytohemagglutinin-stimulated lymphocytes at 38.5 and 42 degrees C. Addition of the antioxidants glutathione or thioredoxin to culture did not reduce the effects of heating to 42 degrees C on proliferation.(ABSTRACT TRUNCATED AT 250 WORDS)

Overcoming tumor necrosis factor-alpha resistance of human renal and ovarian carcinoma cells by combination treatment with buthionine sulfoximine and tumor necrosis factor-alpha. Role of tumor necrosis factor-alpha mRNA down-regulation in tumor cell sensitization

BACKGROUND

Previous studies have reported the glutathione plays a central role in a wide range of cellular functions, including protection, detoxification, transport, and metabolism. Buthionine sulfoximine (BSO), a specific inhibitor of gamma-glutamyl-cysteine synthetase, depletes intracellular glutathione. The study investigates the cytotoxic effect of BSO and tumor necrosis factor-alpha (TNF-alpha) used in combination on TNF-alpha-resistant human renal and ovarian cancer cells.

METHODS

Cytotoxicity was determined by a 1-day microculture tetrazolium dye assay. TNF-alpha mRNA was examined by Northern blot analysis.

RESULTS

Combination treatment of TNF-alpha-resistant R4 and R11 human renal cell carcinoma cells with BSO and TNF-alpha overcame their resistance to TNF-alpha. In addition, the combination of BSO and TNF-alpha resulted in a synergistic cytotoxic effect on TNF-alpha-resistant OVC-8 and C30 human ovarian cancer cells. Treatment of R4, R11, and OVC-8 cells with TNF-alpha in combination with glutathione or N-acetyl-cysteine (NAC) showed an antagonistic cytotoxic effect. A possible mechanism of resistance to TNF-alpha in tumor cells is the expression of TNF-alpha mRNA or protein. R4 cells and OVC-8 cells constitutively expressed mRNA for TNF-alpha. Treatment of R4 cells or OVC-8 cells with BSO down-regulated the expression of TNF-alpha mRNA; however, treatment with TNF-alpha up-regulated the expression of TNF-alpha mRNA. When BSO was used in combination with TNF-alpha, the level of TNF-alpha mRNA enhanced by TNF-alpha was markedly reduced. Incubation of R4 cells with glutathione or NAC also down-regulated the expression of TNF-alpha mRNA. R11 and C30 cells did not constitutively express mRNA for TNF-alpha, and the BSO treatment had no effect on the TNF-alpha mRNA level.

CONCLUSIONS

This study demonstrates that the combination of BSO and TNF-alpha can overcome the TNF-alpha resistance of tumor cells and that depletion of intracellular glutathione and down-regulation of TNF-alpha mRNA by BSO may play a role in the enhanced cytotoxicity seen with the combination of BSO and TNF-alpha. There may not be always a correlation between the expression of TNF-alpha mRNA in tumor cells and their resistance to TNF-alpha. The synergistic effect obtained with established renal cell carcinoma cells and ovarian cancer cells suggests that combination treatment with TNF-alpha and BSO could have clinical application in the therapy of TNF-alpha-resistant tumors.

Lipid hydroperoxide induced mitochondrial dysfunction following acute ethanol intoxication in rats. The critical role for mitochondrial reduced glutathione

It has been found that acute ethanol (EtOH) intoxication of rats caused depletion of mitochondrial reduced glutathione (GSH) of approximately 40%. A GSH reduction of similar extent was also observed after the administration to rats of buthionine sulphoximine (BSO), a specific inhibitor of GSH synthesis. Combined treatment with BSO plus EtOH further decreased mitochondrial GSH up to 70% in comparison to control. Normal functional efficiency was encountered in BSO-treated mitochondria, as evaluated by membrane potential measurements during a complete cycle of phosphorylation. In contrast a partial loss of coupled functions occurred in mitochondria from EtOH- and BSO plus EtOH-treated rats. The presence in the incubation system of either GSH methyl monoester (GSH-EE), which normalizes GSH levels, or of EGTA, which chelates the available Ca2+, partially restores the mitochondrial phosphorylative efficiency. Following EtOH and BSO plus EtOH intoxication, the presence of fatty-acid-conjugated diene hydroperoxides, such as octadecadienoic acid hydroperoxide (HPODE), was detected in the mitochondrial membrane. Exogenous HPODE, when added to BSO-treated mitochondria, induced, in a concentration-dependent system, membrane potential derangement. The presence of either GSH-EE or EGTA fully prevented a drop in membrane potential. The results obtained suggest that fatty acid hydroperoxides, endogenously formed during EtOH metabolism, brought about non-specific permeability changes in the mitochondrial inner membrane whose extent was strictly dependent on the level of mitochondrial GSH.

Effect of glutathione levels on aflatoxin B1-DNA binding in livers and kidneys of male rats and hamsters pretreated with buthionine sulfoximine and dimethylmaleate

The effects of pretreatment of buthionine sulfoximine (BSO) alone or in combination with diethylmaleate (DEM) on glutathione (GSH) levels and aflatoxin B1 (AFB1)-DNA binding have been examined in livers and kidneys of young male Fischer rats and Syrian golden hamsters 2 h after an intraperitoneal injection of [3H]AFB1 (0.4 mg/kg body wt.). Animals were treated with BSO (4 mmol/kg body wt.) alone at 4 h and 2 h or with DEM (3 mmol/kg body wt.) at 4 h and BSO at 2 h before AFB1 injection. Hepatic AFB1-DNA binding was about 29.0 and 6.0 pmol/mg DNA in rats and hamsters, respectively. In rats, BSO increased AFB1-DNA binding by about 40% with a drop in GSH by 70%. Treatment with DEM-BSO increased AFB1-DNA binding by about 80% with a concomitant drop in GSH in both species. In hamsters, BSO increased AFB1-DNA binding by only 10% with a 50% drop in GSH. The kidneys of both species have lower GSH levels and AFB1-DNA binding than their respective liver tissues. The effect of BSO alone or in combination with DEM on both GSH levels and AFB1-DNA binding are comparable even though BSO alone is less effective in both species. The role of modulation of GSH levels on AFB1-DNA binding and hence biological effects of AFB1 in these two species is discussed.

Characterization of resistance mechanisms to cis-diamminedichloroplatinum(II) in three sublines of the CC531 colon adenocarcinoma cell line in vitro

Cisplatin resistance was developed in sublines of the CC531 rat colon adenocarcinoma cell line by continued low level drug exposure. Two relatively stable lines were obtained (RL2 and RL4) which were 6- and 20-fold more resistant to cisplatin. In addition, a subline more sensitive than the parent line by a factor of 2 (RLS) was obtained by subculture from a treated tumor. Mechanisms of resistance to cisplatin were investigated in these four lines, with the aim of determining the relative contributions of different resistance mechanisms at various resistance levels. Drug accumulation linearly decreased with increasing drug resistance. A 20-fold resistance was associated with only a 5-fold decrease in accumulation, suggesting that other resistance mechanisms may be involved in the total degree of resistance. Intracellular glutathione, measured fluorometrically, also increased with increasing resistance, varying by a factor of 4 between the most and least resistant lines. Reduction of glutathione levels by buthionine sulfoximine to parent line levels increased sensitivity but the cells remained considerably more resistant than parent cells. Resistant lines cultured in the absence of drug became progressively more sensitive, without accompanying changes in total glutathione levels. DNA-drug adducts, the presumed toxic lesion, were measured immunocytochemically. Initial levels decreased with increasing platinum resistance, although not proportional to resistance (factor of 5 decrease for 20-fold resistance). Drug dose ratios for equal initial adducts were similar to dose ratios for equal drug accumulation, implying that intracellular concentrations solely determine DNA adduction and that differences in glutathione level had little influence on the proportion of drug which eventually formed adducts. After 48 h, a better correlation between remaining adducts and resistance was found (factor 12 less adducts for 20-fold resistance). This implies that repair of adducts was important in determining survival. These data indicate that decreased drug accumulation played a proportionally greater role in the moderately resistant cell line and that adduct repair played a progressively greater role in the highly resistant cell line.

Reactive naphthalene metabolite binding to hemoglobin and albumin

Earlier work has shown that the murine Clara cell cytotoxicant, naphthalene, is metabolized to reactive metabolites which deplete glutathione or, in the absence of sufficient glutathione, become bound covalently to tissue macromolecules. Correlations between bound metabolite levels in the lung with injury suggests an association between reactive metabolite binding and toxicity. In this study we examine the formation of covalent naphthalene adducts with hemoglobin and albumin in mice to determine whether these serve as useful indices of exposure and metabolism for a chemical which shows a glutathione threshold. Covalent binding of radioactivity from [3H]naphthalene to both albumin and hemoglobin was dose dependent and a glutathione threshold was observed. At early times after naphthalene administration, the formation of albumin adducts was 10- to 30-fold higher than that of hemoglobin adducts. Hemoglobin and albumin adduct levels decreased by apparent first-order processes with half-lives of 11.5 and 1.8 days, respectively. These half-lives are consistent with the turnover of these blood proteins in the mouse. Pretreatment with buthionine sulfoximine resulted in higher levels of albumin adduct but in no alteration of hemoglobin adduct levels in comparison with control. In contrast, diethylmaleate pretreatment increased the level of hemoglobin adduct but not albumin adduct. The antibody to naphthalene mercapturates recognized the hemoglobin adduct(s) but not the albumin adduct(s). Comparison of the data from ELISA (standardized using hydroxymercaptodihydronaphthalene) and radiochemical analysis yielded curves with identical slopes; the absolute levels of adduct found by ELISA were approximately half those measured with radiochemical techniques.(ABSTRACT TRUNCATED AT 250 WORDS)

Glutathione related enzymes in cis-diamminedichloroplatinum (II)-sensitive and-resistant human ovarian carcinoma cells

A cis-diamminedichloroplatinum (II) (CDDP)-resistant cell line (NOS2CR) demonstrated 7.4-fold greater resistance to CDDP compared with the parental cell line (NOS2) established from a patient with serous cystadenocarcinoma of the ovary. We investigated the role of enzyme systems associated with glutathione (GSH) in these cell lines. The GSH content was almost identical in both cell lines. Preincubation with 50 microM DL-buthionine-S, R-sulfoximine (BSO), an inhibitor of gamma-glutamyl cysteine synthetase, for 24 hr reduced the IC50 in both NOS2 and NOS2CR cells. Glutathione-S-transferase pi (GST-pi) activity and mRNA level in NOS2CR cells were higher than in NOS2 cells. However, gamma-glutamyltranspeptidase (GGT) activity in NOS2CR cells was 2.4-fold less than in NOS2 cells. The GST activity and mRNA level in both cell lines were constant when the cells were exposed to CDDP. Exposure to CDDP for 48 hr increased the GGT mRNA level 4.4 and 1.8 times in NOS2 and NOS2CR cells, respectively, compared with no exposure. By exposure to CDDP for 48 hr, the GGT activities in NOS2 and NOS2CR cells were increased 1.6-and 2.5-fold, respectively, compared with no exposure. The above data provide the first evidence that GGT activity and GGT mRNA are induced by CDDP in human carcinoma cell lines.

Interactions of nitroglycerin and sulfhydryl-donating compounds in coronary microvessels

Previous studies have shown the effect of nitroglycerin on coronary microvessels < 100 microns in diameter is markedly enhanced by L-cysteine. These studies were performed to examine the mechanisms responsible for this effect. Under control conditions, nitroglycerin caused potent dilations of large (> 200 microns diam) coronary microvessels while having minimal effects on small (< 100 microns diam) coronary microvessels [peak relaxations 85 +/- 4 vs. 23 +/- 3% (mean +/- SE) of endothelin-1-constricted vessels, respectively]. L-Cysteine (100 microM) and N-acetylcysteine (100 microM) markedly enhanced nitroglycerin-induced relaxations of small coronary microvessels (peak relaxation 84 +/- 6 and 87 +/- 12%, respectively) while having no effect on relaxations of vessels > 100 microns. In contrast, neither L-methionine (100 microM) nor glutathione (100 microM) enhanced nitroglycerin's vasodilation of small coronary microvessels. The effects of L-cysteine and N-acetylcysteine on the augmentation of nitroglycerin vasodilatation in smaller coronary microvessels was abolished in the presence of buthionine sulfoximine (100 microM), a potent inhibitor of intracellular glutathione synthesis. Buthionine sulfoximine had no effect on the vasodilatation produced by nitroprusside. These data demonstrate that, in smaller coronary microvessels, L-cysteine and N-acetylcysteine enhance nitroglycerin-induced vasodilatation by increasing intracellular glutathione concentrations. Intracellular glutathione, formed from either L-cysteine or N-acetylcysteine, may participate in the formation of an intermediate of nitroglycerin biotransformation or may maintain a redox potential within coronary microvessels that favors enzymatic bioconversion of nitroglycerin.

Mitomycin C sensitivity in human bladder cancer cells: possible role of glutathione and glutathione transferase in resistance

In this study, we have examined the relationship between sensitivity to mitomycin C (MMC) and glutathione (GSH) and glutathione transferase (GST) levels using a panel of three unrelated human bladder cancer cell lines. J82, HT-1197, and SCaBER. Cell lines HT-1197 and SCaBER were about 2- and 4.5-fold more resistant to MMC as compared to J82. Although the GSH level did not differ significantly in these cell lines, GST activity in HT-1197 and SCaBER cells were higher by about 2.3- and 6.0-fold, respectively, as compared to J82. Similar to GST activity, GST pi content was highest in the most insensitive cell line and lowest in J82 cells. The cytotoxicity of MMC was increased significantly in these cells by a 1-h pretreatment with a nontoxic concentration of ethacrynic acid (EA), an inhibitor of GST activity. EA pretreatment resulted in a marked GSH depletion as well as GST activity inhibition in both of these cells. Although pretreatment of J82 and SCaBER cells with a nontoxic concentration of D,L-buthionine-S,R-sulfoximine (BSO) caused similar GSH depletion, the cytotoxicity of MMC was enhanced only in SCaBER cells. The differential effect of BSO on MMC cytotoxicity in these cell lines appeared to be due to the differences in the extent of GSH regeneration after removal of BSO. While a marked GSH regeneration occurred in J82 cells within 1 h after BSO removal, such an effect was not observed in SCaBER cells. Combined treatment of these cells with BSO and EA produced a greater potentiation of MMC cytotoxicity in both the cell lines when compared to BSO or EA treatment alone. We conclude that GSH/GST levels may affect the sensitivity of human bladder cancer cells to MMC.

Characterization of quinone reductase, glutathione and glutathione S-transferase in human myeloid cell lines: induction by 1,2-dithiole-3-thione and effects on hydroquinone-induced cytotoxicity

In this study, we have characterized quinone reductase (QR), glutathione (GSH), glutathione S-transferase (GST) and their induction by a chemoprotector, 1,2-dithiole-3-thione (D3T), in the human myeloid cell lines ML-1 and HL-60. In addition, we also examined the toxicity of hydroquinone (HQ), a benzene metabolite, to these two cell lines. Both of the cell lines contain a basal level of cellular GSH, which is similar in the two cell lines. Although ML-1 cells contain much higher QR specific activity than HL-60 cells, which are relatively QR deficient, the GST specific activity of ML-1 cells is 1.8 times less than that of HL-60 cells. Immunoblot experiments showed that the GST in these two cell lines is GST pi. In addition, HL-60 cells exhibit 4.5 times more myeloperoxidase specific activity than ML-1 cells. Inclusion of D3T in the cultures could induce significant increases in cellular GSH content and QR activity, but not GST activity in either cell line. Treatment with HQ caused both inhibition of cell proliferation and loss of cell viability in these two myeloid cell lines. HQ treatment also resulted in a significant depletion of cellular GSH, which preceded the loss of cell viability. Pretreatment of both cell lines with buthionine sulfoximine, an inhibitor of GSH biosynthesis, markedly increased HQ-induced toxicity. In contrast, the presence of dicumarol, a QR inhibitor, failed to potentiate HQ-induced toxicity in ML-1 cells. On the other hand, pretreatment of these two myeloid cell lines with D3T significantly protected against HQ-induced inhibition of cell proliferation and cell death. Therefore, the above results suggest that GSH but not QR is an important factor involved in the toxicodynamics of HQ in these myeloid cells.

Effect of cigarette smoke and its condensates on alveolar epithelial cell injury in vitro

The oxidant-antioxidant balance in the airspaces of the lungs may be critical in protecting the lungs from the effects of cigarette smoke. We studied the effect of cigarette smoke and its condensates on the detachment, attachment, and proliferation of the A549 human alveolar epithelial cell line, in an in vitro model of cell injury and regeneration and the protective effects of antioxidants. Whole and vapor phase cigarette smoke decreased 51Cr-labeled A549 cell attachment, increased cell detachment, and decreased cell proliferation, as assessed by [3H]thymidine uptake. Freshly isolated rat type II alveolar epithelial cells showed an enhanced susceptibility to smoke-induced cell lysis when compared with the A549 cell line. Reduced glutathione (GSH) (400 microM) protected against the effects of cigarette smoke exposure on cell attachment, proliferation, and detachment. Depletion of intracellular GSH with buthionine sulfoxamine enhanced the epithelial cell detachment injury produced by smoke condensates. We conclude that cigarette smoke and its condensates cause an oxidant-induced injury to A549 human type II alveolar epithelial cells. Both intra- and extracellular GSH have important roles in protecting epithelial cells from the injurious effects of cigarette smoke.

Intracellular glutathione levels regulate Fos/Jun induction and activation of glutathione S-transferase gene expression

Induction of glutathione S-transferase Ya and NAD(P)H:quinone reductase gene expression by a variety of chemical agents is mediated by regulatory elements, EpRE and ARE, composed of two adjacent AP-1-like binding sites and activated by Fos/Jun heterodimeric complex (AP-1). Recent studies show that chemical induction of glutathione S transferase Ya and quinone reductase gene expression is associated with an induction of c-fos and c-jun gene expression and AP-1 binding activity. In this report we present evidence that the AP-1 binding activity and the expression of chloramphenicol acetyltransferase activity from an EpRE Ya-cat gene construct are induced by an increase in intracellular oxidant levels. We observe that lowering the glutathione levels with buthionine sulfoximine, an inhibitor of gamma-glutamylcysteine synthetase, or diamide, a thiol-oxidizing agent, stimulates both basal and chemical-inducible expression of chloramphenicol acetyltransferase activity from EpRE Ya-cat and the AP-1 binding activity. Furthermore, we observe that the induction of these activities by a variety of chemical agents is inhibited by thiol compounds N-acetylcysteine and glutathione. These findings suggest that diverse chemicals that induce the AP-1 complex, leading to the AP-1-mediated transcriptional activation of glutathione S-transferase Ya gene expression, may act through a common mechanism involving the production of reactive oxygen species and depletion of reduced glutathione.

S-thiolation and irreversible oxidation of sulfhydryls on carbonic anhydrase III during oxidative stress: a method for studying protein modification in intact cells and tissues

S-thiolation of carbonic anhydrase III (CA III) in cultured rat hepatocytes under oxidative stress was studied by immunodetection on nitrocellulose blots of isoelectrofocusing gels. In cells treated with menadione, three S-thiolated forms of CA III were detected, whereas only two forms were observed in hepatocytes treated with t-butyl hydroperoxide. Two "nonreducible" oxidized forms of CA III were also detected on nitrocellulose blots. These forms increased with the amount of stress and were the only modified forms of CA III in buthionine sulfoxide-treated hepatocytes containing 10-fold less glutathione than control hepatocytes. These experiments support the concept that S-thiolation protects CA III from irreversible oxidation during oxidative stress. Partly and fully S-thiolated forms of CA III were easily detected in both male and female hepatocytes by the immunoblotting method, although female cells contained 15-fold less CA III than did male liver. S-thiolated forms of CA III were also detected in rat skeletal muscle and heart showing the utility of this method for determining the effect of oxidative stress on specific S-thiolatable protein in several tissues in vivo.

Phase I clinical trial of intravenous L-buthionine sulfoximine and melphalan: an attempt at modulation of glutathione

PURPOSE

A phase I dose-escalation trial of intravenous (IV) L-buthionine-SR-sulfoximine (BSO) with melphalan (L-PAM) was performed to determine the toxicity and biologic activity of BSO, administered as a short infusion every 12 hours, and the combination of BSO plus L-PAM.

PATIENTS AND METHODS

Twenty-eight patients with refractory malignancies received 30-minute infusions of BSO every 12 hours for 6 to 10 doses in week 1 followed in week 2 by either IV L-PAM (15 mg/m2) alone or BSO as in week 1 with L-PAM. Patients received the combination in week 5 (course no. 2) if they received L-PAM alone during week 2 and vice versa. BSO doses ranged from 1.5 g/m2 to 13.104 g/m2.

RESULTS

The only toxicity observed with BSO infusions was occasional nausea/vomiting. Evaluation of 23 paired courses (L-PAM plus BSO v L-PAM) showed significantly (P < .001) greater leukopenia and thrombocytopenia with L-PAM plus BSO. No other significant toxicity was noted. Measurement of intracellular glutathione (GSH) levels in peripheral mononuclear cells (PBLs) of all patients receiving BSO showed a consistent, non-dose-dependent, linear decrease in GSH with repeated BSO doses. Maximal GSH depletion (40% of baseline values, absolute values 200 to 250 ng/10(6) PBLs) was noted after the sixth BSO dose; extended BSO dosing schedules beyond six total BSO doses did not further deplete GSH. Evaluation of gamma-glutamylcysteine synthetase (GCS) activity showed marked inhibition near the end of each infusion with near complete recovery of GCS activity before the next BSO dose. The pattern of GCS inhibition mirrored the plasma BSO concentrations with peak values (level 6, 4 to 8 mmol/L L,R+L,S BSO) observed at the end of the infusion with a rapid decrease in plasma concentrations with an estimated half-life (t1/2) of less than 2 hours. Differential elimination of the R+S stereoisomers was observed. Analysis of L-PAM pharmacokinetics showed marked interpatient variability and a significant decrease in total-body clearance (P = .01) and volume of distribution (P = .03) in courses with L-PAM plus BSO as compared with L-PAM alone.

CONCLUSION

This study shows that BSO alone and in combination with L-PAM can be safely given to patients, but that a schedule of short infusions every 12 hours does not result in GSH depletion less than 30% of baseline values.

Comparative cytotoxicity of angiotensin-converting enzyme inhibitors in cultured rat hepatocytes

Captopril and enalapril, angiotensin-converting enzyme inhibitors (ACEIs), have been associated with idiosyncratic hepatotoxicity. Such drug reactions may be caused by the formation of reactive metabolites by cytochrome P450 isozymes, which can then cause direct or immune-mediated toxicity. Previously, we have demonstrated that enalapril cytotoxicity in primary cultures of rat hepatocytes was due, at least in part, to cytochrome P450-dependent metabolism, and that glutathione was involved in the detoxification process. In the present study, we extended our investigations into mechanisms of cytotoxicity, using rat hepatocyte cultures, to captopril and three recently marketed ACEIs: fosinopril, lisinopril and quinapril. After 24 hr of exposure to lisinopril or enalaprilat (the deesterified metabolite of enalapril), hepatocytes did not show any evidence of cytotoxicity, measured by lactate dehydrogenase leakage, even at 10 mM drug concentrations. The other ACEIs were toxic to the liver cells, with the rank order of toxicity as quinapril (LC50 = 0.28 mM) > fosinopril (LC50 = 0.4 mM) > enalapril (LC50 = 2.0 mM) > captopril (LC50 = 20 mM). In vivo pretreatment of rats with pregnenolone-16 alpha-carbonitrile to induce isozymes of the P450 3A subfamily significantly enhanced the cytotoxicities of quinapril, fosinopril and enalapril but did not affect captopril cytotoxicity. Pretreatment with P450 inducers selective for other isozyme subfamilies (ethanol, beta-naphthoflavone and phenobarbital) did not alter the in vitro toxicity of any of the ACEIs. Co-incubation with SKF525-A (15 microM) or troleandomycin (0.1 mM) reduced the hepatocidal toxicities of quinapril, fosinopril and enalapril. Preincubation with buthionine sulfoximine (2 mM) enhanced the cytotoxicities of quinapril, fosinopril, enalapril and captopril. The results of this study indicate that like enalapril, quinapril and fosinopril can also undergo P450 3A-dependent bioactivation and require maintenance of glutathione status for detoxification, and that captopril causes cytotoxicity independent of cytochrome P450 metabolism.

Protection of retinal pigment epithelium from oxidative injury by glutathione and precursors

PURPOSE

This study was performed to examine the effect of exogenous glutathione (GSH) or its precursor amino acids on oxidative injury in cultured human retinal pigment epithelium (RPE).

METHODS

Cultured human RPE cells were suspended in Krebs-Henseleit medium, and 150 microM t-butylhydroperoxide was added. Cell viability was assessed by 0.2% trypan blue exclusion 30, 60, and 120 minutes after the addition of GSH or its amino acid precursors.

RESULTS

Added GSH provided protection at concentrations of 0.01 mM and higher. The amino acid precursors for GSH, glutamate, cysteine, and glycine also protected against injury, but this required at least 0.1 mM of each amino acid. Inhibition of intracellular GSH synthesis by inclusion of 1 mM buthionine sulfoximine eliminated the protection by added amino acids but did not alter the protection by added GSH.

CONCLUSIONS

These results indicate that protection by the amino acid precursors is mediated through synthesis of GSH, and they also show that exogenous GSH can provide protection against oxidative injury.

In vivo therapeutic potential of combination thiol depletion and alkylating chemotherapy

The effect of administering the thiol modulating agent buthionine sulfoximine (BSO) in conjunction with alkylating chemotherapy was investigated in vivo in the mouse KHT sarcomas and bone marrow stem cells. Tumour response to treatment was assessed by an in vivo to in vitro excision assay and bone marrow survival was determined in vitro by CFU-GM. Glutathione (GSH) depletion and recovery kinetics were determined at various times after treatment using high performance liquid chromatography (HPLC) techniques. Following a single 2.5 mmol kg-1 dose of BSO, tumour GSH reached a nadir of approximately 40% of control 12-16 h after treatment. Bone marrow GSH was depleted to approximately 45% of control 4-8 h after treatment but recovered to normal by 16 h. When a range of doses of CCNU, mitomycin C, cyclophosphamide or melphalan (MEL) were given 16 h after mice were exposed to a 2.5 mmol kg-1 dose of BSO, only the antitumour efficacy of MEL was effectively enhanced (by a factor of approximately 1.4). This BSO-MEL combination appeared to be selective for the tumour as the bone marrow toxicity was not increased beyond that seen for MEL alone. Since increasing the administered dose of BSO neither increased the extent of thiol depletion in the tumour nor enhanced the antitumour efficacy of MEL, three other protocols for delivering the thiol depletor were explored. BSO was given either as multiple 2.5 mmol kg-1 doses administered at 6 or 16 h intervals or continuously at a concentration of 30 mM supplied in the animals' drinking water. Both multi-dose BSO pretreatments were found to increase both the antitumour efficacy and normal tissue toxicity of MEL such that no advantage compared to the single dose combination was achieved. In contrast, maintaining the thiol depletor in the drinking water led to an approximately 1.7-fold increase in the antitumour efficacy of MEL without any corresponding increase in bone marrow stem cell toxicity. For the various pretreatment strategies it was possible, in all cases, to account for the presence or absence of a net therapeutic benefit on the basis of the tumour and bone marrow GSH depletion and recovery kinetics.

Free radical scavenger depletion in post-ischemic reperfusion brain damage

In the present study the influence of pretreatment with various GSH depletors such as buthionine sulfoximine (BSO) and diethylmaleate (DEM) was investigated in rats following cerebral post-ischemic reperfusion. Moreover, the effect of diethyldithiocarbamic acid (DDC), inhibitor of endogenous Cu,Zn-SOD, was evaluated. A significant depletion (40% of control value) of GSH levels was observed 24 h after DEM administration; after 48 h the value reached control levels. BSO showed maximal GSH depletion (59%) 24 h after administration and it was constant for almost 48 h. DDC administration caused a marked decrease (60%) of Cu,Zn-SOD activity 4 h after the injection and induced a marked decrease in percentage of survival with respect to control (untreated, ischemic) rats, when administered 4 h before ischemia. BSO and DEM prolonged the survival time of animals when administered 24 h before ischemia. This last paradoxical effect is unclear at present, but it might be due to an influence on glutamate cascade.

Effect of depletion of glutathione by buthionine sulfoximine on lipid peroxidation in rats acutely treated with ethanol

1. The effect of depletion of glutathione (GSH) by DL-buthionine-S,R-sulfoximine (BSO) on lipid peroxidation in rats acutely treated with ethanol was investigated. 2. BSO pretreatment has not been found to potentiate an increase in liver, brain and erythrocyte lipid peroxide levels.

Differential effect of cyanohydroxybutene on glutathione synthesis in liver and pancreas of male rats

1-Cyano-2-hydroxy-3-butene (CHB), an aliphatic nitrile found in cruciferous vegetables, causes a two- and sevenfold elevation in reduced glutathione (GSH) in rat liver and pancreas, respectively, after oral administration of 200 mg/kg. While this dose is also associated with pancreatotoxicity, a single 100 mg/kg dose or multiple lesser doses show the same effect, although somewhat reduced in magnitude, with no concomitant toxicity. In an attempt to identify the mechanism of this increase, we investigated the effect of CHB on GSH synthesis by examining the effect of buthionine sulfoximine (BSO), an inhibitor of GSH synthesis, on CHB-induced GSH elevation. Male Fischer 344 rats received 3 mmol BSO/kg ip 24 and 34 hr following CHB or corn oil. The CHB-mediated elevation in hepatic and pancreatic GSH was eradicated by BSO, suggesting that increased synthesis was responsible. The rate-limiting step in synthesis is gamma-glutamyl cysteine synthetase (GCS); the limiting substrate is cysteine. Therefore, CHB effects on GCS activity and hepatic and pancreatic cysteine equivalents were investigated. When rats were treated by gavage with CHB (100 mg/kg), hepatic GCS mRNA concentrations were increased 24 hr after treatment and hepatic cysteine equivalents were significantly elevated 4 hr following CHB. No significant elevation in hepatic GCS activity was observed, however, even 24 hr following CHB. Pancreatic cysteine equivalents were elevated at both 4 and 8 hr after CHB treatment. However, there was no detectable GCS mRNA or activity in pancreas, in either control or treated animals. Furthermore, CHB had no direct effect on the activity of GCS purified from kidney, regardless of whether GSH was present or absent. These results suggest that the mechanism of CHB-mediated induction of GSH may involve early increases in GSH precursors as well as a later increase in GCS mRNA. The mechanism of GSH elevation identified in these studies may hold therapeutic or prophylactic implications.

Enhanced rate of H2O2 release from bovine pulmonary artery endothelial cells induced by TGF-beta 1

We have previously reported that transforming growth factor-beta 1 (TGF-beta 1) produces a "prooxidant" effect on cultured bovine pulmonary artery endothelial cells (BPAEC) [S. K. Das and B. L. Fanburg. Am. J. Physiol. 261 (Lung Cell. Mol. Physiol. 5): L249-L254, 1991]. This effect was found to be associated with a lowering of total cellular GSH (A.C. White, S. K. Das, and B.L. Fanburg. Am. J. Respir. Cell Mol. Biol. 6: 364-368, 1992). In this study, we demonstrate a twofold increase in the rate of extracellular H2O2 release from BPAEC after a 72-h exposure to TGF-beta 1 (2 ng/ml, added at times 0 and 48 h). Increasing and decreasing the levels of cellular GSH with diethylmaleate (DEM, 0.05 mM) and buthionine sulfoximine (BSO, 0.01 mM), respectively, did not affect the rate of TGF-beta 1-induced increase in H2O2 release when compared with the individual effects of these reagents on control cells. The addition of BSO (0.01 mM) to control cells failed to demonstrate an increase in the rate of H2O2 release, despite a more profound decrease in cellular GSH by these cells than detected in cells treated with TGF-beta 1 alone. Moreover, a single dose of TGF-beta 1 (2 ng/ml) induced a 63-85% increase in the rate of H2O2 release within 16 h of exposure, well before the previously demonstrated lowering of cellular GSH. These results indicate that the increase in H2O2 production by TGF-beta 1-stimulated BPAEC is associated with, but does not appear to be the result of, a lowering of cellular GSH. This study further suggests that the TGF-beta 1-induced H2O2 production occurs at a site inaccessible to detoxification by GSH.

Glutathione, cell proliferation, and 1,3-bis-(2-chloroethyl)-1-nitrosourea in K562 leukemia

We have pursued our findings of glutathione reductase (GSSG-R) deficiency and disturbed glutathione in cancer patients treated with 1,3-bis-(2-chloroethyl)-1-nitrosourea (BCNU), by investigating how thiol metabolism, cell proliferation, and the nitrosourea interact in human K562 leukemia. Fasting cells arrested in G greatly increased their reduced glutathione (GSH) in response to growth factors. The rise in thiol began after several hours, peaked before DNA synthesis, and resulted from increased production. BCNU inactivated GSSG-R rapidly, and later retarded, doubled, and greatly prolonged GSH formation before stopping DNA synthesis. Pretreatment unlike post treatment with buthionine-S-R-sulfoximine (BSO) diminished BCNU's ability to block GSSG-R. Enzyme inhibition decreased with falling cellular GSH. In the leukemia system as in vivo, sequential BCNU-induced thiol alterations heralded delayed antiproliferative effects. Drug timing markedly affected both thiol and DNA syntheses. By destroying GSSG-R and delaying the upregulation of thiol synthesis while escalating GSH utilization and requirements, the nitrosourea created a striking and previously unrecognized window of vulnerability for GSH-dependent processes. During this period, altered GSH metabolism could contribute indirectly to BCNU's pleiotropic effects by interfering with DNA alkylation repair, glucose decarboxylation, deoxyribose formation, and possibly by influencing other aspects of proliferation. Acquired GSSG-R deficiency was also an early and sensitive marker for prodrug breakdown and activation.

Glutathione depletion increases tyrosinase activity in human melanoma cells

The aim of the present work was to estimate the effect of intracellular glutathione depletion on melanogenesis in human melanoma cells. We determined tyrosine hydroxylation activity, the rate-limiting step of the pathway, and 14C-melanin formation, an assay reflecting the global eumelanogenic pathway. Intracellular glutathione was depleted by treatment with buthionine-S-sulfoximine, a well-known inhibitor of gamma-glutamylcysteine synthetase. The intracellular depletion of glutathione was substantial after 20 h of incubation with 50 microM buthionine-S-sulfoximine, although a significant effect could be observed after 6 h. Tyrosine hydroxylase activity increased in parallel with glutathione depletion, to reach 160% with respect to the control values during 24 h of buthionine-S-sulfoximine treatment. We have found the response to buthionine-S-sulfoximine to be dose dependent and the two different human cell lines HBL and LND1 to have similar, if not identical, responses. 14C-melanin formation assay revealed even greater activation, up to 400% of the control values. This indicates that glutathione depletion may have two distinct effects: first, a direct one on tyrosinase activity and, second, an effect on the promotion of eumelanogenesis. The stimulation of tyrosine hydroxylase can be explained by a possible inactivation of the enzyme by endogenous thiol compounds rather than by a direct effect of buthionine-S-sulfoximine itself on tyrosinase. The data suggest that thiol compounds may play a role for stimulation of melanogenesis by ultraviolet radiation.

Bcl-2 inhibition of neural death: decreased generation of reactive oxygen species

The proto-oncogene bcl-2 inhibits apoptotic and necrotic neural cell death. Expression of Bcl-2 in the GT1-7 neural cell line prevented death as a result of glutathione depletion. Intracellular reactive oxygen species and lipid peroxides rose rapidly in control cells depleted of glutathione, whereas cells expressing Bcl-2 displayed a blunted increase and complete survival. Modulation of the increase in reactive oxygen species influenced the degree of cell death. Yeast mutants null for superoxide dismutase were partially rescued by expression of Bcl-2. Thus, Bcl-2 prevents cell death by decreasing the net cellular generation of reactive oxygen species.