The toxicity of acetaminophen and N-acetyl-p-benzoquinone imine in isolated hepatocytes is associated with thiol depletion and increased cytosolic Ca2+

Cytotoxic effects of biphenyl and hydroxybiphenyls on isolated rat hepatocytes

The cytotoxic effects of biphenyl (BP) and its hydroxylated derivatives, o-phenylphenol (OPP), m-phenylphenol (MPP), p-phenylphenol (PPP), 2-biphenylyl glycidyl ether (OPP-epoxide), phenyl-hydroquinone (PHQ), o,o'-biphenol (o,o'-BPol) and p,p'-biphenol (p,p'-BPol), were investigated in freshly isolated rat hepatocytes. OPP, MPP and PPP, at concentration of 0.75 mM, resulted in the loss of intracellular ATP, glutathione (GSH) and protein thiols, causing cell death. OPP-epoxide and BP were less toxic than the OPP isomers. MPP or PPP compared with OPP caused serious impairments in oxidative phosphorylation in mitochondria isolated from rat liver. PHQ (0.75 mM) caused a rapid loss of intracellular ATP which preceded the onset of cell death. PHQ was more toxic than o,o'-BPol or p,p'-BPol. PHQ dissolved in Krebs-Henseleit buffer without hepatocytes was rapidly converted to its corresponding quinone, phenyl-benzoquinone. The cytotoxicity produced by PHQ depends on the rate of formation of reactive intermediates. These results indicate that the addition of a hydroxyl group to the aromatic ring of BP enhances BP-induced cytotoxicity and that the mitochondria are a common target of the OPP isomers and other BP derivatives. In addition, the para- or meta-hydroxyl groups rather than the ortho-hydroxyl group increase the toxicity. The cytotoxicity produced by PHQ depends on the rate of formation of reactive intermediate(s) such as phenyl-benzoquinone.

Combined effects of adrenergic and intravenous anesthetic agents on inositol monophosphate levels in rat liver prisms

Combined effects of adrenergic and intravenous anesthetic agents on phosphatidylinositol (PI) turnover were studied using rat liver prisms incubated with [3H]myo-inositol. Rat liver prisms responded to epinephrine, norepinephrine and phenylephrine dose-dependently with an increase in inositol monophosphate (IP1) formation but they did not respond to ephedrine. Dopamine-induced effects were seen only at concentrations as high as 10(-4) mol.l-1. The enhancement of IP1 formation induced by epinephrine was potentiated by thiamylal at concentrations of 10(-5) mol.l-1 and 10(-4) mol.l-1, remained unaffected by ketamine, fentanyl or midazolam, but was dose-dependently inhibited by droperidol. The present results from in vitro studies of liver cell metabolism suggest that alpha-adrenergic agents in combination with barbiturates may potentiate liver cell damage by activation of PI turnover and interrelated intracellular Ca++ accumulation.

Zonation of acetaminophen metabolism and cytochrome P450 2E1-mediated toxicity studied in isolated periportal and perivenous hepatocytes

To study the mechanism of centrilobular damage developing in the centrilobular region after high doses of acetaminophen (APAP), its metabolism and toxicity were compared in periportal and perivenous hepatocytes isolated by digitonin/collagenase perfusion. Contrary to earlier reports, based on perfusions, no evidence for a periportal dominance of APAP sulfation could be observed. Glucuronidation, the dominant pathway of conjugation at high (5 mM) APAP concentration, was faster in perivenous cells. During primary culture, prolonged exposure (> or = 24 hr) to 5 mM APAP damaged perivenous cells, with a higher P450 2E1 level than periportal cells. When cells were isolated from ethanol-pretreated rats, to induce P450 2E1 levels specifically in the perivenous region, perivenous hepatocytes exhibited enhanced APAP vulnerability and extensive glutathione depletion. In contrast, corresponding periportal cells retained good viability. Isoniazid, an inhibitor of cytochrome P450 2E1, protected cells against APAP toxicity and prevented glutathione depletion. Induction of P450 2E1 also caused a 3-fold increase in the covalent binding of reactive intermediates from [14C]APAP, and this increase was mainly confined to perivenous cells. These results indicate that in rat liver there is only slight perivenous zonation of APAP conjugation and suggest that zone-specific APAP activation, mediated by the regional expression of ethanol-inducible cytochrome P450 2E1, is responsible for the characteristic centrilobular liver damage elicited by APAP.

Silymarin protects against paracetamol-induced lipid peroxidation and liver damage

The effect of silymarin on liver damage induced by acetaminophen (APAP) intoxication was studied. Wistar male rats pretreated (72 h) with 3-methylcholanthrene (3-MC) (20 mg kg-1 body wt. i.p.) were divided into three groups: animals in group 1 were treated with acetaminophen (APAP) (500 mg kg-1 body wt. p.o.), group 2 consisted of animals that received APAP plus silymarin (200 mg kg-1 body wt. p.o.) 24 h before APAP, and rats in group 3 (control) received the equivalent amount of the vehicles. Animals were sacrificed at different times after APAP administration. Reduced glutathione (GSH), lipid peroxidation and glycogen were measured in liver and alkaline phosphatase (AP), gamma-glutamyl transpeptidase (GGTP) and glutamic pyruvic transaminase (GPT) activities were measured in serum. After APAP intoxication, GSH and glycogen decreased very fast (1 h) and remained low for 6 h. Lipid peroxidation increased three times over the control 4 and 6 h after APAP treatment. Enzyme activities increased 18 h after intoxication. In the group receiving APAP plus silymarin, levels of lipid peroxidation and serum enzyme activities remained within the control values at any time studied. The fall in GSH was not prevented by silymarin, but glycogen was restored at 18 h. It was concluded that silymarin can protect against APAP intoxication through its antioxidant properties, possibly acting as a free-radical scavenger.

Comparative effect of nifedipine and S-adenosylmethionine, singly and in combination on experimental rat liver cirrhosis

An experimental rat liver cirrhosis, by means of carbon tetrachloride and ethanol during 8 weeks, was employed to assay the effect of Nifedipine (a calcium antagonist blocker), S-Adenosylmethionine (a precursor of glutathione); singly and in combination on rat liver cirrhosis. A slight decrease of cirrhosis (N.S.) was observed with the pharmacological therapy employed singly. The combination therapy (Nifedipine+S-Adenosylmethionine) significantly inhibited liver cirrhosis (p < 0.01).

Protection from oxidative damage in mouse liver cells

Paracetamol toxicity in mouse hepatocytes involved oxidative stress initiated by the formation of NAPQI. This oxidative component of paracetamol injury is associated with the latter stages of the poisoning process. Ebselen, a drug with GSH-peroxidase activity, was effective in ameliorating these oxidative events.

Paracetamol toxicity and its prevention by cytoprotection with iloprost

In a well-established two phase model of paracetamol toxicity in hamster hepatocytes cell death was accompanied, but not preceded, by a rise in cytosolic free calcium [Ca2+]i. Cell death appears to involve reversible oxidative damage, possibly to the cytoskeleton or mitochondria. In this model low concentrations (10(-8) to 10(-14) M) of iloprost, a stable analogue of prostacyclin, offered protection against the toxic effects of paracetamol. In preliminary studies with a rat liver epithelial cell line transduced with murine P4501A2 the toxicity of paracetamol was attenuated by iloprost. Inhibition of protein synthesis with cycloheximide had no effect on paracetamol toxicity but abolished the cytoprotective effect of iloprost.

Dynamic changes in the distribution of the calcium-activated neutral protease in human red blood cells following cellular insult and altered Ca2+ homeostasis

Mechanistic studies were conducted to examine the relationship between oxidative membrane protein damage, altered Ca2+ homeostasis, and changes in the levels of plasma membrane-bound Ca(2+)-activated neutral protease, microCANP. Alterations in the levels of plasma membrane-bound microCANP in erythrocytes and hemolysate following cumene hydroperoxide (CHP) insult were monitored using SDS-PAGE and immunoblot analyses. Free radical scavengers, antioxidant and EGTA effects on membrane-bound microCANP levels in CHP-treated cells and hemolysate were also examined. CHP (2 mM) addition to red cells caused a significant decrease/loss in intensity of numerous protein bands in the SDS-PAGE pattern, to include bands 1, 2, 2.1, 4.1, 4.2, and an approximately 60-kDa protein. N-acetylcysteine (20 mM), dithiothreitol (50 mM), and dimethylthiourea (50 mM) diminished CHP-mediated membrane protein damage; in contrast, dimethylfuran (50 mM) exacerbated CHP-mediated membrane protein damage. Dimethylsulfoxide (50 mM) was without significant effect. The free radical scavengers and antioxidants differentially affected membrane-bound microCANP levels largely in parallel with their ability to modulate membrane protein damage. Immunoblot analysis of 1 mM CHP-treated red cells revealed a time-dependent loss of membrane-bound microCANP, with a complete loss of microCANP monitored at 8 hr. Treatment of erythrocytes with CHP also resulted in concentration-dependent alterations in the level of membrane-bound microCANP: at 0.5 or 1.0 mM CHP a decreased level of membrane-bound microCANP was detected relative to control, whereas an increase in the level of bound enzyme was monitored from 2 to 4 mM CHP. CHP addition to hemolysate produced a decrease in membrane-bound microCANP levels comparable to that observed with erythrocytes; addition of the Ca2+ chelator EGTA or Calpain Inhibitor I (N-acetyl-leucyl-leucyl-leucyl-nor-leucinal) to hemolysate effectively inhibited this decrease. In contrast, treatment of erythrocytes with Ca2+ in the presence of the Ca2+ ionophore A23187 resulted in change in the SDS-PAGE protein bands and membrane-bound microCANP levels that were comparable to those produced by CHP. Inclusion of EGTA in this system prevented microCANP binding. These data provide evidence for membrane damage and concomitant dynamic alterations in membrane-bound microCANP levels in the red cell or hemolysate following oxidative insult, and show that this process can be modulated by free radical scavengers and antioxidant, simulated by treating cells with Ca2+ in the presence of ionophore, and inhibited by EGTA or Calpain Inhibitor I.

Prooxidant-induced Ca2+ release from liver mitochondria. Specific versus nonspecific pathways

Ca2+ release from mitochondria can be induced by a variety of chemically different prooxidants. Release induced by these compounds is possibly regulated by protein mono(ADP)ribosylation, and leaves mitochondria initially intact. Excessive "cycling" (continuous release and uptake) of Ca2+ by mitochondria leads to their damage, as shown by a decreased membrane potential, fast Ca2+ release, and impairment of ATP synthesis. When cycling is prevented by Ca2+ chelators or by inhibition of the uptake route with ruthenium red, prooxidants still induce Ca2+ release but mitochondria remain intact. It has recently been suggested that formation of a "pore" in the inner mitochondrial membrane participates in the Ca2+ release mechanism. We find that the prooxidant-induced Ca2+ release is not paralleled by sucrose entry into, or K+ release from, or swelling of mitochondria, provided Ca2+ cycling is prevented. Thus, the prooxidant-induced Ca2+ release does not require formation of a "pore." We conclude that the release occurs via a specific pathway.

Cytoskeleton-dependent surface blebbing induced by the polar solvent N-methylformamide

In vivo and in vitro studies performed on the polar solvent N-methylformamide (NMF), as well as on its association with chemotherapeutic agents or X rays, have clearly demonstrated that this compound is capable of inducing changes in biological characteristics of tumor cells, e.g., cell differentiation. However, the mechanism of action of NMF is far from being elucidated. Hence, in order to better clarify such a mechanism an in vitro study was carried out by using mouse fibroblasts in primary culture (MEF) and human melanoma cultured cells (M14). Results obtained by immunocytochemical and ultrastructural methods with doses of NMF ranging from 0.1 to 7% are reported here. As a general rule, a different sensitivity (in terms of cytopathologic changes induced by NMF) was found between the cell types considered. In fact, melanoma cells appeared to be highly susceptible to the action of the drug, undergoing severe morphological modifications represented mainly by a reversible dose and time-dependent cell rounding and surface blebbing. In contrast, NMF-induced injury in MEF cells was characterized mainly by a simple retraction of the cell body. A cytochemical analysis of the expression of certain membrane antigens (e.g., glycoproteins, epidermal growth factor receptor, B2 microglobulin) in NMF-treated M14 cells undergoing blebbing was also carried out. A randomly distributed labeling of such molecules was observed. Accordingly, freeze-fracturing electron microscopic analysis also displayed a random distribution of intramembrane particles over the plasma membrane. When subcellular changes induced by the drug were investigated, a remarkable modification of cytoskeletal components was detected in both cell types. In particular, cross-linked actin microfilament bundles were easily observed in NMF-exposed MEF cells. Finally, when different experimental conditions which perturb calcium ion homeostasis or restore protein thiol group reduced state were analyzed, a noticeable impairment of the blebbing phenomenon was observed. Thus, a target effect of NMF on the microfilament system, probably leading, in turn, to several subcellular changes and cell surface blebbing, can be hypothesized. Such a cytoskeletal element-dependent cytopathology appears to be related to changes of the oxidized state of such molecules as well as to calcium ion perturbations.

Paracetamol toxicity in hamster isolated hepatocytes: the increase in cytosolic calcium accompanies, rather than precedes, loss of viability

Paracetamol is cytotoxic to hamster isolated hepatocytes by a mechanism that does not involve an early increase in [Ca2+]i. Although an increase in [Ca2+]i does occur, it accompanies rather than precedes, loss of viability. Studies with the ionophore, 4-bromo-A23187, suggest that although sustained elevations of [Ca2+]i per se can initiate cell death, this occurs at levels of [Ca2+]i only above 500 nM. This concentration was not achieved on exposure of cells to a cytotoxic concentration of paracetamol for 30 min. The [Ca2+]i-response of hepatocytes to vasopressin stimulation was not altered by exposing the cells to toxic concentrations of paracetamol. This demonstrates that paracetamol does not cause any impairment in the mobilisation or redistribution of Ca2+. The role of elevated levels of [Ca2+]i in mediating chemically-induced cell-killing requires re-evaluation.

Do intracellular Ca2+ activity and hepatic glutathione play a role in the pathogenesis of lithocholic acid-induced cholestasis?

The possible relevance of alterations in intracellular Ca2+ and hepatic glutathione levels (GSH) in the pathogenesis of cholestasis induced by lithocholic acid (LCA) was examined by comparing effects of LCA and acetaminophen on these parameters and bile flow (BF) in rats. Intracellular Ca2+ activity was measured via glycogen phosphorylase a determination in rats given an intravenous bolus injection of either LCA (12 mumol/100 g body wt.), acetaminophen (60 mg/100 g body wt.), or a mixed solution of LCA and acetaminophen. BF was reduced immediately after LCA administration, with a maximum decrease occurring at 60 min followed by an increase to normal values at 210 min. On the other hand, glycogen phosphorylase a activity was elevated during all time periods after LCA treatment. Hepatic glutathione followed the BF curves being markedly depleted at the peak of cholestasis (60 min) and normal in the total recovery period (210 min). In contrast, acetaminophen had no effect on BF but significantly increased glycogen phosphorylase a activity and depleted hepatic glutathione levels. These results suggest that cholestatic effect of LCA is not due to changes in intracellular Ca2+ or hepatic glutathione levels.

The role of the nucleus and other compartments in toxic cell death produced by alkylating hepatotoxicants

Hepatocellular necrosis occurs under a wide range of pathological conditions. In most cases, toxic cell death takes place over a finite span of time, delayed from the point of initial injury and accompanied by homeostatic counterresponses that are varied and complex. The present strategies for discovering critical steps in cell death recognize that (1) different toxins produce similar morphologic changes that precede killing in widely varied cell types, and that (2) lethal events are likely to involve one or more compartmentalized functions that are common to most cells. Investigations of the plasma membrane, endoplasmic reticulum, cytoplasm, mitochondrion, and nucleus have greatly advanced our understanding of acute hepatocellular necrosis. This report examines each compartment but emphasizes molecular changes in the nucleus which may explain cell death caused by alkylating hepatotoxicants. Accumulating knowledge about two distinct modes of cell death, necrosis and apoptosis, indicates that loss of Ca2+ regulation and subsequent damage to DNA may be critical steps in lethal damage to liver cells by toxic chemicals.

N-acetyl-p-benzoquinone imine-induced protein thiol modification in isolated rat hepatocytes

Incubation of isolated rat hepatocytes with N-acetyl-p-benzoquinone imine (NAPQI) or 3,5-dimethyl-N-acetyl-p-benzoquinone imine (3,5-Me2-NAPQI) resulted in a concentration-dependent decrease in the protein thiol content of the mitochondrial, cytosolic and microsomal fractions. On a concentration basis, 3,5-Me2-NAPQI induced a more marked depletion of protein thiols than did NAPQI. Sodium dodecyl sulphate-polyacrylamide gel electrophoretic separation of the proteins of each fraction showed that different proteins had different susceptibilities to modification of their cysteine residues by the quinone imines. A few protein bands showed a decreased protein thiol content following incubation with non-toxic concentrations of quinone imines, whereas other proteins were affected by higher concentrations. Concentrations of quinone imines that were highly cytotoxic induced a general loss of protein thiols. NAPQI-induced protein thiol depletion occurred within 5 min and remained essentially unchanged for at least 30 min. In contrast, protein thiol depletion induced by 3,5-Me2-NAPQI increased over the 30-min time course of the experiment. Toxic concentrations of 3,5-Me2-NAPQI caused the formation of high molecular mass aggregates in all three subcellular fractions after 30 min of incubation. The observed crosslinking was not due to protein disulfide formation. However, no aggregate formation was observed after exposure of hepatocytes to NAPQI. One of the major target proteins of quinone imine-induced protein thiol depletion was a 17 kDa microsomal protein that was identified as the microsomal glutathione S-transferase. Exposure of hepatocytes and isolated liver microsomes to the quinone imines resulted in an up to four-fold increase in the specific activity of the microsomal glutathione S-transferase. In conclusion, our results are consistent with the suggestion of a critical role of protein thiol depletion in quinone imine-induced cytotoxicity.

Relationship between metabolism and cytotoxicity of ortho-phenylphenol in isolated rat hepatocytes

The relationship between the metabolism and the cytotoxicity of ortho-phenylphenol (OPP) was investigated using isolated rat hepatocytes. Addition of OPP (0.5-1.0 mM) to the hepatocytes caused a dose-dependent toxicity; 1.0 mM OPP caused acute cell death. Pretreatment of hepatocytes with SKF-525A (50 microM, a non-toxic level) enhanced the cytotoxicity of OPP (0.5-1.0 mM). This was accompanied by inhibition of OPP metabolism. Conversely, OPP at low concentrations (0.5 or 0.75 mM) was converted sequentially to phenyl-hydroquinol (PHQ) and then to glutathione (GSH) conjugate in the cells. The concentrations of both metabolites, especially PHQ-GSH conjugate, were very low in hepatocytes exposed to 1.0 mM OPP alone as well as with SKF-525A. The cytotoxicity induced by 0.5 mM OPP was enhanced by the addition of diethylmaleate (1.25 mM) which continuously depletes cellular GSH. In contrast, additions to hepatocytes of 5 mM of dithiothreitol, cysteine, N-acetyl-L-cysteine or ascorbic acid significantly inhibited the cytotoxicity induced by 0.5 mM PHQ; GSH, protein thiols and ATP losses were also prevented. Further, these compounds depressed the rate of PHQ loss in hepatocyte suspensions. These results indicate that the acute cytotoxicity caused by the high dose (1.0 mM) of OPP is associated with direct action by the parent compound; at low doses (0.5-0.75 mM) of OPP, the prolonged depletion of GSH in hepatocytes enhances the cytotoxicity induced by PHQ.

Purification, antibody production, and partial amino acid sequence of the 58-kDa acetaminophen-binding liver proteins

Immunochemical analysis of electrophoretically resolved liver proteins from mice administered hepatotoxic doses of acetaminophen has identified two proteins of 44 and 58 kDa as major targets for acetaminophen arylation. In the present study the 58-kDa acetaminophen-binding protein (58-ABP) was purified from mouse liver cytosol by gel permeation chromatography, preparative isoelectric focusing, and polyacrylamide gel electrophoresis. The acetaminophen adducts were visualized on immunoblots using affinity-purified anti-acetaminophen antibodies after each step of the purification. Gel permeation chromatography, under nondenaturing conditions, indicated that the protein is a monomer. Two-dimensional gel electrophoresis demonstrated that the 58-ABP consists of a cluster of four immunochemically reactive isoforms with isoelectric points ranging from 6.2 to 6.6. V-8 protease digestion of the isoforms suggested that they contained similar peptide fragments. The purified 58-ABP was utilized to produce polyclonal antibodies and to determine the amino acid composition and partial sequence of the protein. These antibodies revealed a protein cluster of similar molecular weight and isoelectric points in the cytosol of a human liver specimen. Amino acid analysis of the purified protein indicated that it contains eight cysteine residues (about 1.4% by weight). This low cysteine content raises the possibility that at hepatotoxic doses acetaminophen may also bind to non-thiol sites on the protein. The amino acid sequence of two cyanogen bromide/tryptic peptide fragments revealed that the major immunochemically detectable acetaminophen target in the cytosol is homologous to a selenium-binding protein which has been recently sequenced.

Effect of nifedipine and S-adenosylmethionine in the liver of rats treated with CCl4 and ethanol for one month

An experimental model of toxic liver injury in rats was employed to assay the effect of Nifedipine (a calcium antagonist blocker) and S-Adenosylmethionine (a precursor of glutathione). An important decrease in both perivenular fibrosis and cirrhosis was found. Furthermore, a significant decrease in lactic acid levels was found in the group of animals treated with pharmacologic therapy, although no correlation was seen between lactic acid levels and the different degrees of perivenular fibrosis. No significant variations in ALT and AST enzymes were observed between both groups, as opposed to a significant decrease in LDH enzyme in the Nifedipine+S-Adenosylmethionine group. The results indicate an improvement in the histologic picture of the liver in rats treated by means of pharmacological association, without any change in inflammatory infiltrate and with a slight decrease in necrosis, indicating an action mechanism via creeping fibrosis (instead of a hepatitis pathway).

Acetaminophen-induced cytotoxicity in cultured mouse hepatocytes: effects of Ca(2+)-endonuclease, DNA repair, and glutathione depletion inhibitors on DNA fragmentation and cell death

Hepatotoxic alkylation of mouse liver cells by acetaminophen is characterized by an early loss of ion regulation, accumulation of Ca2+ in the nucleus, and fragmentation of DNA in vitro and in vivo. Acetaminophen-induced DNA cleavage is accompanied by the formation of a "ladder" of DNA fragments characteristic of Ca(2+)-mediated endonuclease activation. These events unfold well in advance of cytotoxicity and the development of necrosis. The present study utilized cultured mouse hepatocytes and mechanistic probes to test whether DNA fragmentation and cell death might be related in a "cause-and-effect" manner. Cells were isolated by collagenase perfusion, cultured in Williams' E medium for 22-26 hr, and exposed to acetaminophen. Aurintricarboxylic acid, a general Ca(2+)-endonuclease inhibitor, and EGTA, a chelator of Ca2+ required for endonuclease activation, significantly decreased DNA fragmentation at 6 and 12 hr and virtually abolished cytotoxicity. N-Acetylcysteine also eliminated DNA fragmentation and cytotoxicity. 3-Aminobenzamide, an inhibitor of poly(ADP-ribose) polymerase-stimulated DNA repair, failed to alter the amount of DNA fragmentation at 6 hr but substantially increased acetaminophen cytotoxicity in hepatocytes at 12 hr. With the exception of when DNA repair was inhibited by 3-aminobenzamide, Ca2+ accumulation in the nucleus, DNA fragmentation, and hepatocyte death varied consistently and predictably with one another. Collectively, these findings suggest that unrepaired damage to DNA contributes to acetaminophen-induced cell death in vivo and may play a role in necrosis in vivo.

A comparison of proteins S-thiolated by glutathione to those arylated by acetaminophen

This study was designed to evaluate whether the same proteins that irreversibly bind reactive electrophiles of drugs also bind glutathione (GSH) under oxidative conditions. Specifically, proteins that can be arylated by acetaminophen were compared to those that form glutathione-protein mixed disulfides (PSSG) after incubation with diamide. Data are presented which suggest that both GSH and acetaminophen bind to a subset of N-ethylmaleimide (NEM)-reactive protein thiols. To evaluate the pattern of proteins that bind GSH, PSSGs were formed in vitro by incubating cytosolic proteins with GSH and diamide. A sensitive procedure was developed in which PSSGs were first reduced with 0.1 mM dithiothreitol (DTT), and the newly exposed protein thiols were labeled with either [3H]NEM (for quantitative analysis) or with fluorescein-5-maleimide (for visual detection). Acetaminophen binding was achieved by incubating cytosolic proteins in vitro with the reactive acetaminophen metabolite, N-acetyl-p-benzoquinoneimine (NAPQI). Proteins from both assays were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose for Western blot analysis. Acetaminophen binding was detected by immunoblotting with an affinity-purified antibody against acetaminophen, and PSSGs were visualized using anti-fluorescein antibodies. In both instances, binding to proteins was observed to be selective. A comparison of the proteins modified by GSH binding with those that bind acetaminophen indicates that the major cytosolic acetaminophen-binding protein of 58 kDa may also be modified by glutathiolation under oxidative conditions.

Glutathione metabolism and its role in hepatotoxicity

Glutathione (GSH) fulfills several essential functions: Detoxification of free radicals and toxic oxygen radicals, thiol-disulfide exchange and storage and transfer of cysteine. GSH is present in all mammalian cells, but may be especially important for organs with intense exposure to exogenous toxins such as the liver, kidney, lung and intestine. Within the cell mitochondrial GSH is the main defense against physiological oxidant stress generated by cellular respiration and may be a critical target for toxic oxygen and electrophilic metabolites. Glutathione homeostasis is a highly complex process, which is predominantly regulated by the liver, lung and kidney.

Acetaminophen-induced cytotoxicity in cultured mouse hepatocytes: correlation of nuclear Ca2+ accumulation and early DNA fragmentation with cell death

Hepatotoxic doses of acetaminophen cause widespread alkylation of liver and early loss of cytosolic Ca2+ regulation. Although the precise location and target of lethal alkylation are not known, Ca2+ accumulation is viewed as a possible link between cell alkylation and cell death. We have recently shown that Ca2+ accumulates in the nucleus and that DNA fragments in vivo before the development of acetaminophen-induced necrosis in mice. The present study examined cultured hepatocytes for nuclear damage and its association with cell death in vitro. Positive results would argue for two key points. (1) Nonparenchymal cell damage does not explain DNA fragmentation induced by acetaminophen in vivo. (2) A chemical that causes necrosis can produce DNA damage considered characteristic of apoptosis. Hepatocytes from NIH Swiss mice were isolated by collagenase perfusion, cultured in Williams' E medium for 24 hr, and exposed to acetaminophen. Cytotoxicity was assessed by lactate dehydrogenase leakage and release of [3H]adenine from a prelabeled nucleotide pool. Genomic DNA fragmentation was assessed quantitatively by colorimetric analysis and qualitatively by agarose gel electrophoresis. Acetaminophen caused DNA damage from 1-4 hr onward and produced significant release of lactate dehydrogenase and [3H]adenine nucleotides at later times. Agarose gel electrophoresis revealed a "ladder" of DNA fragments characteristic of Ca(2+)-mediated endonuclease activation. Cytotoxicity correlated with nuclear Ca2+ accumulation (r greater than 0.895, p less than 0.05) and with percentage DNA fragmentation (r greater than 0.835, p less than 0.05). Nuclear changes in vitro generally reproduced those observed in vivo. Collectively, these findings demonstrate that nuclear Ca2+ accumulation and DNA fragmentation appear as early events that correlate directly with later cytotoxicity. These changes may contribute to acetaminophen-induced injury leading to cell death in vitro and necrosis in vivo.

Use of N-acetylcysteine in clinical toxicology

The major use of N-acetylcysteine in clinical toxicology is in the treatment of acetaminophen (paracetamol) overdosage. The hepatorenal toxicity of acetaminophen is mediated by a reactive metabolite normally detoxified by reduced glutathione. If glutathione is depleted, covalent binding to macromolecules and/or oxidation of thiol enzymes can lead to cell death. Oral or intravenous N-acetylcysteine or oral D,L-methionine mitigates acetaminophen-induced hepatorenal damage if given within 10 hours, but becomes less effective thereafter. In vivo, N-acetylcysteine forms L-cysteine, cystine, L-methionine, glutathione, and mixed disulfides; L-methionine also forms cysteine, thus giving rise to glutathione and other products. Oral therapy with N-acetylcysteine or methionine for acetaminophen poisoning is contraindicated in the presence of coma or vomiting, or if activated charcoal has been given by mouth. Nausea, vomiting, and diarrhea may also occur as a result of oral N-acetylcysteine administration. Anaphylactoid reactions including angioedema, bronchospasm, flushing, hypotension, nausea/vomiting, rash, tachycardia, and respiratory distress may occur 15-60 minutes into N-acetylcysteine infusion (20 hours intravenous regimen) in up to 10% of patients. Following accidental intravenous overdosage, the adverse reactions of N-acetylcysteine are similar but more severe; fatalities have occurred. A reduction in the loading dose of N-acetylcysteine may reduce the risk of adverse reactions while maintaining efficacy. Administration of N-acetylcysteine for a longer period might provide enhanced protection for patients in whom acetaminophen absorption or elimination is delayed. N-acetylcysteine may also have a role in the treatment of toxicity from carbon tetrachloride, chloroform, 1,2-dichloropropane, and other compounds. The possible use of N-acetylcysteine and other agents in the prevention of the neuropsychiatric sequelae of acute carbon monoxide poisoning is an important area for future research.

Chronic dietary iron overload in rats results in impaired calcium sequestration by hepatic mitochondria and microsomes [corrected]

The purpose of these experiments was to determine whether chronic dietary iron overload causes impairment of hepatic mitochondrial and/or microsomal calcium sequestration. Experimental iron overload was produced by feeding three groups of rats a chow diet supplemented with 3.0% (wt/wt) carbonyl iron for up to 8 weeks achieving graded increases in hepatic iron concentrations ranging from 1360 to 3170 micrograms/g. At low levels of iron overload, there were no changes in mitochondrial oxidative metabolism or calcium sequestration, whereas at moderate and high degrees of iron loading, both of these parameters were significantly reduced. In contrast, there were significant decreases in microsomal cytochrome P450 levels and microsomal calcium sequestration at all three levels of iron loading. These abnormalities occurred at hepatic iron concentrations at which the authors have previously found evidence of hepatic organelle lipid peroxidation. These alterations in organelle calcium sequestration may impair intracellular calcium homeostasis in the liver and contribute to subsequent cellular injury.

Comparative cytotoxic effects of acetaminophen (N-acetyl-p-aminophenol), a non-hepatotoxic regioisomer acetyl-m-aminophenol and their postulated reactive hydroquinone and quinone metabolites in monolayer cultures of mouse hepatocytes

Toxic effects of acetaminophen (paracetamol, N-acetyl-p-aminophenol, APAP) in monolayer cultures of mouse hepatocytes developed over a period of 18 hr. N-Acetyl-m-aminophenol (AMAP) was approximately 10-fold less toxic than APAP, despite the fact that it bound covalently to a greater extent to hepatocyte macromolecules. AMAP did not deplete glutathione to as great an extent as APAP, indicating that their reactive metabolites may bind to different proteins or that oxidative damage in addition to arylation of proteins may be involved in the development of cell death. The toxicity of 3-methoxy-acetyl-p-aminophenol was similar to that of APAP, whereas the other hydroquinone and quinone metabolites were 8-10 times more cytotoxic than APAP. The potencies of these analogs were in the order: acetyl-m-aminophenol-p-benzoquinoneimine greater than or equal to 2,5-dihydroxyacetanilide greater than or equal to 3-methoxy-p-benzoquinone greater than or equal to N-acetyl-p-benzoquinone imine (NAPQI) greater than or equal to acetyl-m-aminophenol-o-benzoquinone greater than or equal to 3-hydroxy-acetyl-p-aminophenol. The relative toxic potencies of the hydroquinone and quinone metabolites of AMAP were comparable to that of NAPQI, and do not readily explain the marked difference between the cytotoxic effects of AMAP and APAP.

Mechanisms of cell death

Two distinct morphological patterns of cell death have been recognized, termed necrosis and apoptosis. Apoptosis, or programmed cell death, occurs in both physiological and pathological conditions. It arises due to an elevation of cytosolic free calcium concentration resulting in activation of a nuclear endonuclease. Activated endonuclease produces oligonucleosome-length DNA fragments. This DNA cleavage can directly precipitate cell death. Both glucocorticoids and TCDD may induce apoptosis by production of a heat labile factor that mediates calcium influx whereas tributyltin causes the opening of calcium channels. Evidence that perturbation in calcium homeostasis is an important event in cell necrosis is becoming increasingly persuasive, but the events that propagate the lesion are still unclear. Despite evidence for cytoskeletal disruption, activation of degradative enzymes such as proteases and phospholipase A2 and stimulation of other enzymes such as poly (ADP-ribose) polymerase, the exact role that these play in cell killing is not resolved. Indeed, recently the radical dichotomy between apoptosis and necrotic cell death has come into question. It is clear that further work is required to determine the role played by some elements of the apoptotic process in chemically induced cell death.

Prevention of paracetamol-induced liver injury by fructose

Hepatic cell injury was studied in an in vitro system using rat liver slices incubated in two stages. During the first 2 hr slices were exposed to 10 mM paracetamol, this was absent during the subsequent 4 hr of incubation. Cell damage was quantified at the end by measuring leakage of lactic dehydrogenase, increase in water content and potassium loss. Treatment of slices with 20 mM fructose in the second period of incubation prevented paracetamol-induced damage. The effect of fructose was not modified by the continued presence of paracetamol in the second incubation period. The inhibition of glycolysis either with 1 mM NaF or 10 microM iodoacetate blocked the effect of fructose. The protective effect afforded by fructose was not duplicated by the addition of lactate. All these findings strongly suggest an increase in intracellular ATP levels as the most probable explanation for the protective effect of fructose, and point to fructose as a potentially useful therapeutic tool for protection of the liver late in paracetamol intoxication.

The killing of cultured hepatocytes by N-acetyl-p-benzoquinone imine (NAPQI) as a model of the cytotoxicity of acetaminophen

The killing of isolated hepatocytes by N-acetyl-p-benzoquinone imine (NAPQI), the major metabolite of the oxidation of the hepatotoxin acetaminophen, has been studied previously as a model of liver cell injury by the parent compound. Such studies assume that the toxicity of acetaminophen is mediated by NAPQI and that treatment with exogenous NAPQI reproduces the action of the endogenously produced product. The present study tested these assumptions by comparing under identical conditions the toxicity of acetaminophen and NAPQI. The killing of hepatocytes by acetaminophen was mediated by oxidative injury. Thus, it depended on a cellular source of ferric iron; was potentiated by 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU), an inhibitor of glutathione reductase; and was sensitive to antioxidants. By contrast, the cytotoxicity of NAPQI was not prevented by chelation of ferric iron; was unaffected by BCNU; and was insensitive to antioxidants. Thus, the killing of cultured hepatocytes by NAPQI occurs by a mechanism different from that of acetaminophen. The killing by NAPQI was preceded by a collapse of the mitochondrial membrane potential and a depletion of ATP. Monensin potentiated the cell killing, and extracellular acidosis prevented it. These manipulations are characteristic of the toxicity of mitochondrial poisons, and are without effect on the depletion of ATP and the loss of mitochondrial energization. Thus, mitochondrial de-energization by a mechanism unrelated to oxidative stress is a likely basis of the cell killing by NAPQI. It is concluded that treatment of cultured hepatocytes with NAPQI does not model the cytotoxicity of acetaminophen in these cells.

Delayed calcium channel blockade with diltiazem reduces paracetamol hepatotoxicity in mice

1. Diltiazem (30 mg kg-1 body weight, intraperitoneally) given to mice 9 h after paracetamol (450 mg kg-1, orally) reduced liver damage, as judged by plasma aspartate aminotransferase activity (median 186, range 6-602 IU 1(-1), n = 18 vs 466, range 23-3872 IU 1(-1) in 18 saline-treated controls; P less than 0.05) with comparable reductions in mortality (14% vs 33%, respectively; NS). 2. Regenerative activity, as judged by mitotic figures in tissue removed at 30 h after paracetamol, was significantly higher in mice treated at 9 h with diltiazem (median 0.83 per high power field vs 0.1 in saline-treated controls; P less than 0.05). 3. Diltiazem administered earlier or later than 9 h showed reduced efficacy and in some cases potentiated toxicity, as did nifedipine (40 mg kg-1 in divided doses up to 9 h).

Cytoprotective effects of calcium channel blockers. Mechanisms and potential applications in hepatocellular injury

Deregulation of calcium homeostasis is strongly implicated in the development of cellular injury, including in hepatocytes, and is thought to be the limiting step in transition to an irreversible stage. Calcium channel blockers appear to exert their cytoprotective effects through several mechanisms. These may involve blockade of L-(long-lasting)-type calcium channels, reduction of oxidative stress, antagonism at inflammatory mediator receptor sites and interaction at other intracellular sites. Studies relating to the liver are few but suggest that calcium channel blockers may have a role to play in limiting hepatocellular damage, especially those arising from exposure to a variety of toxic agents.

Effects of endurance training and exercise on tissue antioxidative capacity and acetaminophen detoxification

Both acute acetaminophen toxicity and physical exercise are accompanied by structural and functional damage to tissues. For acute acetaminophen toxicity, this damage occurs mainly in the liver. This damage, which is believed to be initially caused by oxidation and/or arylation, occurs only after depletion of liver glutathione (GSH). GSH normally protects against oxidation and/or arylation. Prolonged physical exercise also depletes GSH in the body. We hypothesized that with endurance training (repeated oxidant stress) tissues will develop mechanisms to prevent GSH depletion. Our results show that, for the same amount of submaximal exercise, trained rats are able to maintain their levels of GSH or their GSH redox status (in the liver, heart, skeletal muscle and plasma) in contrast to their untrained counterparts. Also, upon administration of acetaminophen, trained rats show a less pronounced depletion in liver GSH than untrained rats. We also hypothesized that training may lead to improved maintenance of tissue GSH homeostasis because of induction in the enzyme pathways of protection. We observe that training significantly increases (50-70%) glutathione peroxidase and reductase, glucose-6-phosphate dehydrogenase, and catalase activity in heart and skeletal muscle. Since GSH, in addition to providing cellular protection, also functions in other physiological processes including transport and metabolism, the training-induced benefits seen here may have more far-reaching consequences than ever before realized.

Cytoskeleton as a target in menadione-induced oxidative stress in cultured mammalian cells: alterations underlying surface bleb formation

Several in vitro and in vivo studies have suggested that surface bleb formation during oxidative cell injury is related to alteration in cytoskeleton organization. Various cell lines different in origin and growth characteristics were exposed to 2-methyl-1,4-naphthoquinone (menadione) which is known to induce bleb formation and cytotoxicity by generating considerable amounts of oxygen-reactive species. Treated cells were analyzed by means of immunocytochemistry and electron microscopy in order to investigate the morphological and molecular features underlying bleb generation. The results obtained indicate that menadione-induced bleb formation is a widely observed phenomenon present mainly in round or mitotic cells. Surface blebs appear free of organelles and contain only few ribosomes and amorphous material. Occasionally, they undergo detachment from the cell surface as large cytoplasmic vesicles. Bleb surfaces with protein clusters as well as bald blisters with an almost exclusive localization of intramembrane particles on their narrow base were detected using freeze-fracture techniques. Immunocytochemical investigations performed on menadione-exposed cells revealed that some surface proteins (collagen IV, sialo-proteins, beta 2 microglobulin and fibronectin) and adhesion molecules (vinculin) underwent changes in their expression over the bleb surface. Moreover, different behavioural characteristics of actin microfilaments, vimentin and keratin intermediate filaments and microtubules was observed. Alpha-actinin, vimentin and microtubular proteins (tubulin, MAPs and tau) were detected within the blebs. On the other hand, actin and keratin filaments appeared to be absent. The results presented here demonstrate that cytoskeletal structures and the microfilament system in particular, represent important targets in menadione-induced morphological changes in cultured cells. These changes appear to lead to the redistribution of several cytoskeletal and membrane proteins as well as dissociation of the cytoskeleton network from its anchoring domains in the plasma membrane thus generating sites of structural weakness where blebs would arise and progressively grow. Experimental evidence supporting a crucial role of thiol oxidation and elevation of cytoplasmic calcium concentration in bleb formation is also provided.

Extensive alteration of genomic DNA and rise in nuclear Ca2+ in vivo early after hepatotoxic acetaminophen overdose in mice

Hepatotoxic doses of acetaminophen cause early impairment of Ca2+ homeostasis. In this in vivo study, 600 mg/kg acetaminophen caused total nuclear Ca2+ and % fragmented DNA to rise in parallel from 2-6 hr, followed by large later increases mirroring frank liver injury. Agarose gel electrophoresis revealed substantial loss of large genomic DNA from 2 hours onward, with accumulation of DNA fragments in a ladder-like pattern resembling apoptosis. Extensive late cleavage of DNA probably resulted from cell death, whereas degradative loss of large genomic DNA at 2 hours arose at an early enough point to contribute to acetaminophen-induced liver necrosis in mice.

Calcium channel blocking agents protect against acetaminophen-induced cytotoxicity in rat hepatocytes

The effect on acetaminophen-induced cytotoxicity of three calcium channel blocking agents--diltiazem, verapamil and gallopamil--was studied in primary cultures of rat hepatocytes and compared with the chelating agent EGTA. Using the measurement of cytosolic lactate dehydrogenase (LDH) as an index of cytotoxicity, it was demonstrated that a 1-hr pretreatment with calcium channel blocking agents protected cells against acetaminophen cytotoxicity, but were less effective than EGTA. These data suggest that influx of extracellular Ca2+ into the cells could have a role in the genesis of hepatocyte injury by acetaminophen.

Oxidative stress in mitochondria: its relationship to cellular Ca2+ homeostasis, cell death, proliferation, and differentiation

A variety of chemically different prooxidants causes Ca2+ release from mitochondria. This prooxidant-induced Ca2+ release occurs from intact mitochondria via a route which is physiologically relevant and may be regulated by protein monoADP-ribosylation. When the released Ca2+ is excessively 'cycled' by mitochondria (continuously taken up and released) the inner membrane is damaged. This leads to a decreased ability of mitochondria to retain Ca2+, uncoupling of mitochondria, and an impairment of ATP synthesis, which in turn deprives the cell of the energy necessary for the proper functioning of the Ca2+ ATPases of the endoplasmic (sarcoplasmic) reticulum, the nucleus and the plasma membrane. The ensuing rise of the cytosolic Ca2+ level cannot be counterbalanced by the damaged mitochondria which, under normoxic conditions, act as a safety device against an increase of the cytosolic Ca2+ concentration. The impaired ability of mitochondria to retain Ca2+ may lead to cell death. However, there is also evidence emerging that release of Ca2+ from mitochondria may be physiologically important for cell proliferation and differentiation.

The uptake of the cyanobacterial hepatotoxin microcystin by isolated rat hepatocytes

Microcystin-YM a cyclic heptapeptide hepatotoxin isolated from the cyanobacterium Microcystis aeruginosa was radiolabeled with 125I, and used to investigate the uptake of the toxin by freshly isolated rat hepatocytes. The uptake was temperature dependent with apparent activation energy of 18 kcal/mole (77 kJ/mole) for the initial rate of uptake. Uptake of non-toxic (10-20 nM) doses of microcystin by hepatocytes continued with time, the intracellular to extracellular distribution ratio for the toxin was 70 at 60 min for 10(6) cells/ml. Uptake of higher doses of microcystin (100 nM and more) stopped when the cells blebbed: a toxic response of hepatocytes to microcystin. Uptake of microcystin by hepatocytes was inhibited 70-80% by the addition of 10 microM sodium deoxycholate or bromsulphthlein, compounds that protect hepatocytes from the toxic effects of microcystin.

Interactions between N-acetyl-p-benzoquinone imine and fluorescent calcium probes: implications for mechanistic toxicology

Intracellular free calcium ([Ca2+]i) homeostasis has been implicated as an early target in both cellular necrosis and apoptosis. In this study, we have used peripheral blood mononuclear cells (PBMC) as target cells to investigate the effects of several reactive metabolites associated with drug toxicity on [Ca2+]i in order to delineate further early events in cytotoxicity. Compounds implicated in both drug-induced necrosis (N-acetyl-p-benzoquinone imine; NAPQI) and drug hypersensitivity (sulfamethoxazole hydroxylamine; SMX-HA) were examined and their effects on [Ca2+]i compared with those of the T cell mitogen phytohemagglutinin (PHA; 1.5 micrograms/ml) and the calcium ionophore ionomycin (2.5 microM). PHA and ionomycin produced characteristic elevations in [Ca2+]i as monitored by an increase in the fluorescence of fluo-3-loaded cells. SMX-HA did not significantly affect [Ca2+]i at concentrations previously shown to be cytotoxic to PBMC (100 and 500 microM), suggesting that Ca2+ homeostasis is not an early target for SMX-HA toxicity. Addition of NAPQI (250 microM) to fluo-3-loaded cells produced a marked decrease in fluorescence which was not reversed by ionomycin. Conversely, addition of NAPQI to cells loaded with indo-1 resulted in a rapid increase in fluorescence. This effect, however, was found to be attributable to NAPQI addition per se rather than to an increase in [Ca2+]i. HPLC and fluorescence analysis of samples generated from the decomposition of NAPQI revealed the presence of several products which fluoresced intensely at the excitation/emission wavelength pairs of a number of fluorescent probes commonly used to monitor [Ca2+]i.(ABSTRACT TRUNCATED AT 250 WORDS)

Differential effects of arylating and oxidizing analogs of N-acetyl-p-benzoquinoneimine on red blood cell membrane proteins

Incubation of human red blood cell membranes (white ghosts) with N-acetyl-p-benzoquinone imine (NAPQI), a toxic metabolite of acetaminophen, or with either an arylating or an oxidizing analog of NAPQI, resulted in the inhibition of membrane ion transporting systems and the modification of cytoskeletal proteins. NAPQI and 2,6-dimethyl-NAPQI, which primarily arylates protein thiols, inhibited the calmodulin-activated Ca pump ATPase activity, the basal (calmodulin-independent) Ca pump ATPase activity and the Na,K pump ATPase activity. In contrast, 3,5-dimethyl-NAPQI, which primarily oxidizes protein thiols, caused selective inhibition of the calmodulin-activated Ca pump ATPase activity. Sodium dodecyl sulfate gel electrophoresis of red blood cell (RBC) membrane proteins revealed that NAPQI and 2,6-dimethyl-NAPQI, but not 3,5-dimethyl-NAPQI, decreased the intensity of band 3 corresponding to the anion transporter, whereas NAPQI as well as 2,6-dimethyl-NAPQI, and to a lesser extent 3,5-dimethyl-NAPQI, caused a decrease of cytoskeletal protein bands, including spectrin, actin, and bands 4.1 and 4.2. These modifications were associated with increased formation of high molecular weight protein aggregates that did not enter the gel. Treatment of 3,5-dimethyl-NAPQI-exposed ghosts with the reducing agent dithiothreitol (DTT), resulted in the recovery of the affected cytoskeletal protein bands. Conversely, the modifications caused by NAPQI and 2,6-dimethyl-NAPQI were only partially reversed by DTT treatment. Taken together our results suggest that NAPQI and its two analogs modified ion transporting systems and cytoskeletal proteins by reacting with protein thiols. Both oxidation and arylation of protein thiols can alter the functional properties of important RBC membrane proteins. Of the two reactions, arylation appeared to be the less specific and more damaging event.

Metabolism of acetaminophen by cultured rat hepatocytes. Depletion of protein thiol groups without any loss of viability

Over the course of 4 hr, the metabolism of acetaminophen (APAP) by cultured rat hepatocytes resulted in a depletion of protein thiols and an accumulation of oxidized glutathione (GSSG) in the medium. With 20 mM APAP, arylation and the formation of glutathione mixed disulfides accounted for a loss of 22% of the total protein thiols in the absence of any loss of viability. With 20 mM APAP and an inhibition of glutathione reductase by 1.3-(2-chloroethyl)-1-nitrosourea (BCNU), protein thiols were depleted by 40% by arylation and the formation of glutathione mixed disulfides, again without a loss of viability. With 20 mM APAP and BCNU in the presence of 20 mM deferoxamine, there was still little or no cell killing after 8 hr despite a loss now of almost 60% of the total protein thiols. These data do not support the hypothesis that a depletion of protein thiols is related to the toxicity of APAP. One millimolar APAP and BCNU killed 60% of the hepatocytes within 4 hr. In this circumstance, the loss of protein thiols was not attributable to either arylation by APAP metabolites or the formation of glutathione mixed disulfides. The antioxidant N,N'-diphenyl-phenylenediamine prevented the cell killing and the loss of protein thiols, a result implicating a role for lipid peroxidation in the depletion of protein-bound thiols. However, protein thiol depletion under these circumstances is not necessarily related to the lethal cell injury and most likely represents an epiphenomenon of the peroxidation of cellular lipids.

Importance of protein thiols during N-methyl-N'-nitro-N-nitrosoguanidine toxicity in primary rat hepatocytes

N-Methyl-N'-nitro-N-nitrosoguanidine (MNNG), a potent toxicant in isolated rat hepatocytes, was evaluated for its mechanism of cytotoxicity. This direct acting toxicant generates an alkylating carbonium ion that covalently binds to cell macromolecules, depletes nonprotein thiols (NPT), and subsequently kills cells. In this study MNNG depleted protein thiols (PT) in a two-phase process. The first phase (about 30% depletion) occurred rapidly, in parallel with the depletion of NPT. After a plateau, a second phase of PT depletion occurred 5-8 min prior to cell death. Indole-3-carbinol (I-3-C), added prior to MNNG, did not alter the depletion of NPT nor the first phase of PT depletion. However, cell killing was substantially retarded and was still immediately preceded by the second phase of PT depletion. The addition of o-phenanthroline or 5,10-dihydroindeno[1,2-b]indole (DHII) prior to MNNG did not alter the first phase of PT depletion, but partially protected (about 30%) against the depletion of NPT. However, o-phenanthroline or DHII completely protected against the MNNG-induced loss of cell viability and the second-phase depletion of PT. When DHII was added after MNNG and prior to the expected second phase of PT depletion, that depletion was markedly depressed, as was the subsequent loss of cell viability. We conclude that MNNG covalent binding, depletion of NPT, and first-phase depletion of PT may be necessary, but insufficient to kill cells. We propose that rapid depletion of cellular antioxidants predisposes the cell to oxidative stress and that oxygen toxicity is responsible for the second-phase depletion of PT and the final cytotoxic events. The fact that the second-phase depletion of PT is required for and immediately precedes cell death suggests the importance of critical but as yet unidentified target thiol proteins in MNNG hepatotoxicity.