Effects of Hepatotoxic Doses of Acetaminophen and Furosemide on

In contrast with acetaminophen, hepatotoxic doses of furosemide do not ... The gradient was 2 min at 90% A, 0% B, and 10% C, followed by a 20 min line...
0 downloads 0 Views 172KB Size
Chem. Res. Toxicol. 2000, 13, 873-882

873

Effects of Hepatotoxic Doses of Acetaminophen and Furosemide on Tissue Concentrations of CoASH and CoASSG in Vivo Lynette K. Rogers,*,†,‡ Christina J. Valentine,§ Magdalena Szczpyka,§ and Charles V. Smith†,‡ Children’s Research Institute, Children’s Hospital, Columbus, Ohio 43205, Department of Pediatrics, The Ohio State University, Columbus, Ohio 43210, and Department of Pediatrics, Baylor College of Medicine, Houston, Texas 77030 Received April 17, 2000

The effects of hepatotoxic doses of acetaminophen on tissue concentrations of CoA in the thiol form (CoASH) and as the corresponding mixed disulfide with GSH (CoASSG) were determined to test the hypotheses that early oxidant effects of acetaminophen are expressed principally in the mitochondrial compartment and that oxidative shifts in this redox couple could be employed as biomarkers of mitochondrially compartmentalized oxidant stresses. Administration of 400 mg of acetaminophen/kg to male ICR mice did not change CoASSG concentrations at 2, 4, or 6 h, but CoASH levels were lower than in saline-treated control animals at 2 and 4 h (77 ( 8 vs 124 ( 14 nmol/g of liver and 66 ( 9 vs 142 ( 7 nmol/g of liver, respectively). HPLC analyses of acid supernatants from livers of mice treated with acetaminophen in vivo showed a peak that coeluted with an adduct generated in vitro by reaction of CoASH with N-acetyl-p-benzoquinone imine, but extensive efforts to characterize further the putative product formed in vivo have been unsuccessful. Decreases in CoASH levels were not observed in mice given comparably hepatotoxic doses of furosemide, which diminishes the concern that the decreases in CoASH levels observed in the acetaminophen-treated mice were simply secondary to injury. Hepatic CoASSG concentrations were elevated 10-20-fold 2 h after administration of 400 or 500 mg of furosemide/kg, but were not different than in saline-treated control mice at 4 or 6 h. Increases in hepatic concentrations of GSSG were observed after 6 h in both the acetaminophen-treated and the furosemide-treated mice, suggesting that these changes may be more reflective of oxidant responses to hepatic necrosis than of thiol oxidation mechanisms involved in mediating the injury. The results presented here are not consistent with oxidant stress mechanisms in initiation of hepatic necrosis by acetaminophen in vivo, but the data suggest possible roles for mitochondrially compartmentalized oxidant effects of furosemide.

Introduction Excessive doses of acetaminophen can cause hepatic necrosis, and most of the available evidence supports the hypothesis that this cellular damage is mediated by the effects of chemically reactive intermediates generated by metabolism of the parent compound (1). Covalent binding, also called alkylation or arylation, of cellular proteins has always been observed, when examined, with injurious doses of acetaminophen, and it seems logical that covalent modifications would alter the functions of biological molecules in ways that would primarily diminish cellular viability (2, 3). However, studies addressing the contributions of oxidative mechanisms to tissue damage also have been reported (4-8). The central roles played by the glutathione system in reduction of H2O2 and other oxidants, with the concomi* To whom correspondence should be addressed: Children’s Research Institute, Children’s Hospital, 700 Children’s Dr., Columbus, OH 43205. Phone: (614) 722-2781. Fax: (614) 722-2774. E-mail: [email protected]. † Children’s Hospital. ‡ The Ohio State University. § Baylor College of Medicine.

tant production of GSSG, and the availability of analytical methods for measurement of GSSG levels in tissues, cells, and fluids have resulted in GSSG concentrations becoming perhaps the most commonly employed quantitative biomarkers of oxidant stresses (9-11). Changes in the GSH/GSSG ratios also have been employed as criteria for oxidant stresses, in part because of the presumption that GSH/GSSG ratios reflect or even mediate changes in protein thiol/disulfide (PSH/PSSX) ratios (12), which could mediate pathophysiological changes through alterations in protein structure and function. Initial studies showed no elevations of GSSG levels in animals treated with toxic doses of acetaminophen (9, 10). Farber and co-workers reported potentiation of acetaminophen cytotoxicity by pretreatment with 1,3-bis(2chloroethyl)-1-nitrosourea (BCNU),1 which frequently is assumed to alter toxicant-mediated effects primarily 1 Abbreviations: BCNU, 1,3-bis(2-chloroethyl)-1-nitrosourea; NAC, N-acetylcysteine; CoASH, coenzyme A, thiol form; CoASSG, coenzyme A, glutathione-mixed disulfide form; NEM, N-ethylmaleimide; ALT, alanine transaminase; AcCoA, acetyl coenzyme A; NAPQI, N-acetylp-benzoquinone imine; MPT, mitochondrial permeability transition.

10.1021/tx0000926 CCC: $19.00 © 2000 American Chemical Society Published on Web 08/05/2000

874

Chem. Res. Toxicol., Vol. 13, No. 9, 2000

through inhibition of glutathione reductase (5, 13). Although BCNU does inhibit glutathione reductase, the enhanced cytotoxicities of other toxicants in BCNUpretreated cells are more closely correlated with the DNA cross-links that BCNU forms than with inhibition of glutathione reductase (14, 15). Fischer-344 rats treated with doses of BCNU that inhibited glutathione reductase activities by 70% were not more susceptible to hepatotoxicity of acetaminophen administered 18 h later (16). Further, neither the BCNU-pretreated rats nor control rats pretreated with vehicle showed increases in biliary or hepatic concentrations of GSSG in response to administration of hepatotoxic doses of acetaminophen. Later studies in mice similarly revealed no increases in GSSG levels in livers or bile of animals given hepatotoxic doses of acetaminophen (16). The failure of acetaminophen to elevate tissue or bile GSSG concentrations was not consistent with hypotheses regarding oxidant stresses generated during initiation of hepatic injury. Jaeschke (6) and Tirmenstein and Nelson (7, 17) observed increases in hepatic concentrations of GSSG in acetaminophen-treated mice, but only after increases in plasma transaminase activities were observed and later than hepatic damage can be diminished appreciably by administration of N-acetylcysteine (NAC) (18). These data suggest that the observed oxidant effects of hepatotoxic doses of acetaminophen may be more related to consequences of cell damage than to mechanisms involved in the initiation of injury. Oxidative events initiated secondary to cellular damage still might contribute to expansion of a lesion, but the two processes need to be distinguished in studies of mechanisms of injury. Gibson et al. reported modest potentiation of acetaminophen-induced hepatic injury in mice pretreated with ferrous sulfate (8), suggestive of the contributions of Fenton-like reactions to the initiation of injury. However, neither they nor we (19) observed increases in levels of 2,4-dinitrophenylhydrazine-reactive proteins, frequently called “protein carbonyls”, that are characteristic of metal-catalyzed oxidation of proteins. Despite the limited evidence for oxidant stresses early during the metabolism of hepatotoxic doses of acetaminophen, a substantial degree of compartmentalization of oxidative effects could be obscured from detection in whole tissue fractions. Although respiring mitochondria do not metabolize acetaminophen (17), effects of acetaminophen on mitochondria both in vitro and in vivo have been reported (20, 21). Several lines of evidence suggest that mitochondrially compartmentalized oxidant effects might initiate cell death, but direct tests of this hypothesis in vivo are severely limited by the potential for substantial thiol oxidation or reduction or thiol-disulfide exchange reactions during the time required to isolate purified mitochondria from tissues. Mitochondria contain approximately 75% of the CoASH in rat liver (22), and mitochondria contain 90-95% of the CoA in rat cardiac muscle (23). GSSG and GSH equilibrate in vitro with CoASH, forming and reducing, respectively, CoASSG (24-27). Decreases in CoASH and increases in CoASSG contents have been reported in perfused rat livers exposed to hydroperoxides (27). In turn, CoASH and CoASSG levels can be measured in N-ethylmaleimide (NEM)-treated, acid supernatants prepared from freeze-clamped tissue, thus minimizing post vivo artifacts. In these studies, we measured concentrations of CoA in its thiol form (CoASH) and its mixed

Rogers et al.

disulfide with GSH (CoASSG) in freeze-clamped tissues as potential biomarkers for changes in thiol-disulfide status within the mitochondria. As controls for possible effects of cell death on hepatic CoASH, we studied the effects of comparably hepatotoxic doses of furosemide (4-chloro-N-furfuryl-5-sulfamoylanthranilic acid). Furosemide is a diuretic used extensively in clinical medicine (28). Acute tissue damage is not a major problem in clinical use, but single doses of furosemide in excess of 100 mg/kg can produce acute hepatic necrosis in mice. Like acetaminophen, tissue damage by furosemide is accompanied, and possibly mediated, by alkylation of tissue proteins by chemically reactive metabolites generated by hepatic cytochromes P450 (29, 30). In contrast with acetaminophen, hepatotoxic doses of furosemide do not deplete tissue GSH concentrations, depletion of GSH by exogenous agents such as diethyl maleate does not enhance furosemideinduced hepatic damage, and administration of GSH precursors such as NAC does not protect against furosemide-induced necrosis (29-32). For these reasons, furosemide-induced hepatic necrosis in mice is a useful experimental model for testing specific hypotheses, particularly those relating to distinguishing cause from effect, regarding GSH depletion in tissue injury caused by other agents or treatments such as acetaminophen.

Experimental Procedures Animals. Male ICR mice weighing 18-26 g were purchased from Harlan Sprague-Dawley, Inc. (Houston, TX) and were maintained on standard laboratory diet ad libitum for at least 72 h before study. For acetaminophen studies, mice were fasted 16-18 h prior to acetaminophen treatment with 400 mg of acetaminophen/kg, which was dissolved in 0.9% NaCl and administered ip. Furosemide studies were conducted on fed animals given 400 or 500 mg of furosemide/kg as a 10 mg/mL clinical suspension (American Regent Laboratories, Inc., Shirley, NY). Control animals were given equal volumes of vehicle. At 2, 4, or 6 h after treatment, mice were anesthetized with sodium pentobarbital, blood was drawn by cardiac puncture, and plasma was separated by centrifugation at 2000g for 5 min. A portion of liver tissue was freeze-clamped with liquid nitrogencooled tongs and stored at -70 °C. Method I. At the time of analysis, 0.2 g of frozen liver was powdered under liquid N2 and homogenized in 0.8 mL of 3.6% HClO4 with a Dounce homogenizer. Samples were centrifuged at 12700g for 15 min at 4 °C, and the supernatants were analyzed directly by HPLC or were diluted appropriately for GSH and GSSG content measurements. Method II. At the time of analysis, 0.2 g of frozen liver was powdered under liquid N2 and added directly to 0.8 mL of chilled 0.1 M NaHPO4 (pH 7.4) containing 28.5 mM NEM. The mixtures were homogenized immediately in a Dounce homogenizer, added to equal volumes of 4% HClO4, vortexed, and centrifuged at 12700g for 15 min at 4 °C. The supernatants were analyzed directly by HPLC or were diluted appropriately for GSSG content measurements. GSH and GSSG. Total GSH contents were measured in 10% tissue homogenates prepared in 50 mM NaPO4, 50 mM serine borate, and 17.5 mM EDTA (pH 7.4). GSSG levels were measured in NEM-treated acid supernatants prepared as described above. Tissue contents of both GSH and GSSG were measured by the enzyme recycling method described by Adams et al. (9). Hepatic concentrations of GSH, GSSG, and CoA species, which are expressed in moles per wet weight of tissue, were calculated from the concentrations measured in the respective acid supernatants using the simplifying assumptions of tissue densities of 1 g/mL, homogenate final volumes equal

Hepatic CoASH and CoASSG

Chem. Res. Toxicol., Vol. 13, No. 9, 2000 875

Table 1. Hepatic Injury and Glutathione Status in Mice Treated with Acetaminophena acetaminophen (mg/kg) 0 400 0 400 0 400

time (h)

ALTb-d (units/L)

GSHb-d (µmol/g of liver)

GSSGb,d (nmol/g of liver)

GSH/GSSGc

2 2 4 4 6 6

20 ( 2 168 ( 47e,F,G 35 ( 14 1210 ( 404e,F 28 ( 4 653 ( 179e,G

5.73 ( 0.82 ( 0.37e,H 4.01 ( 0.46F 0.65 ( 0.19e,I 3.15 ( 0.46G 5.78 ( 0.55e,G-I

45.2 ( 3.6 21.8 ( 7.4F 33.0 ( 4.1 17.9 ( 8.2G 49.8 ( 14.8 102.2 ( 10.9e,F,G

127 ( 22 32 ( 6e 124 ( 10 54 ( 22e 91 ( 27 61 ( 11

1.08F,G

a Male ICR mice were given 0 or 400 mg of acetaminophen/kg (ip). At 2, 4, or 6 h, animals were anesthetized, blood was obtained by cardiac puncture, and livers were removed and portions immediately freeze-clamped with liquid N2-cooled tongs. Plasma ALT activities and tissue GSH and GSSG concentrations were measured as described in Experimental Procedures. Data are means ( SE from six animals in each group and were assessed statistically by two-way ANOVA, with modified t-tests. Common capital letters indicate groups that are different as determined by modified t-tests (p < 0.05). b Effect of time. c Effect of treatment. d Interaction between time and treatment indicated. e Different than corresponding vehicle-treated controls as determined by modified t-tests (p < 0.05).

to the sums of buffer volumes and tissue weights, and a uniform distribution of analytes in the fractions. Protein. Protein concentrations were measured by the method of Lowry et al. (33) using bovine serum albumin as a standard. ALT. Plasma alanine transaminase (ALT) activities were determined by a Sigma assay kit (procedure 59-UV) from the ALT-mediated conversion of alanine to pyruvate, and reduction of pyruvate with lactate dehydrogenase with detection of the coupled oxidation of NADH. CoASH and CoASSG. Method I. CoASH, CoASSG, and acetyl coenzyme A (AcCoA) were separated by HPLC (34) using an Alltech 7 µm C18 Adsorbosphere column and quantitated by absorbance at 254 nm. Isocratic elution using 0.2 M NaHPO4 (pH 2.8) with 4.5% acetonitrile separated the compounds of interest within 30 min, but to provide adequate separation of CoA species from acetaminophen metabolites in samples obtained 2 h after treatment with acetaminophen, the use of a mobile phase containing 3% acetonitrile was necessary. Standards were prepared in 100 mM NaHPO4 (pH 3.0) for CoASH and AcCoA in the range of 0-40 µM and for CoASSG in the range of 0-4 µM. CoA species were quantitated by comparison of peak areas from samples with peak areas measured in experimentally derived standard curves. Method II. CoASH and CoASSG contents were determined by HPLC using methods derived from the report of Baker and Schooley (35). The CoA species were separated on a Zorbax SBC18 column using the following mobile phases: (A) 25% methanol and 75% water, (B) 65% methanol and 35% water, and (C) 0.1 M tetrabutylammonium hydrogen sulfate (pH 5.0). The gradient was 2 min at 90% A, 0% B, and 10% C, followed by a 20 min linear gradient to 5% A, 85% B, and 10% C, using quantitation by peak area from UV detection at 254 nm. Standards were prepared at the same final concentrations as described in Method I, but dissolved in 0.1 M NaPO4 and 28.5 mM NEM and acidified with equal volumes of 4% HClO4. Preparation and HPLC Analysis of CoA Conjugates. For synthesis of the adducts with acetaminophen and NEM, aqueous solutions of CoASH were treated with N-acetyl-p-benzoquinone imine (NAPQI), which was prepared by oxidation of acetaminophen by silver oxide (36), or with NEM dissolved in acetonitrile. The resulting product mixtures were analyzed by HPLC with a 7 µm C18 Adsorbosphere column and a mobile phase of 20% MeOH and 0.2 M NaH2PO4 (pH 5.0) at a flow rate of 1.0 mL/min, with detection at 254 nm. The concentrations of the products were estimated by comparison of peak areas of the products with areas obtained with known concentrations of acetaminophen and CoASH. Biological samples were analyzed immediately after homogenization and centrifugation. Mass Spectra. Mass spectral characterizations of CoA species and the CoA conjugates were carried out by positive ion FAB-MS on a VG ZAB SEQ multiple-sector quadrupole mass spectrometer. Samples were combined with a matrix of 2-hydroxyethyl disulfide in oxalic acid directly on the probe. Total ion chromatographs were obtained using a 10 V ionization

potential and spectra analyzed for ions of interest. Product ion spectra were acquired from the molecular ion of each compound. Statistics. Data are expressed as means ( SE and were assessed statistically by two-way ANOVA. When the two-way ANOVA indicated an effect of dose or time, or an interaction between these factors, subsequent post hoc testing was performed with modified t-tests to determine differences between groups. All statistical comparisons were performed using SPSS (37).

Results Plasma ALT activities were elevated in the acetaminophen-treated mice as early as 2 h and increased through 4 and 6 h (Table 1), indicating moderate levels of hepatic damage. GSH levels were diminished in the livers of acetaminophen-treated animals at 2 and 4 h to concentrations that were 15% of the levels in salinetreated control animals, but GSH levels in the acetaminophen-treated mice rebounded and exceeded the levels in control animals by 6 h. Hepatic GSSG concentrations tended to be lower in acetaminophen-treated animals than in controls at 2 and 4 h, but statistically significant differences in groups were not observed; however, at 6 h the hepatic GSSG levels in acetaminophen-treated animals were higher than in saline-treated control animals or in the acetaminophen-treated animals at 2 and 4 h. Hepatic GSH/GSSG ratios were lower 2 and 4 h after treatment in acetaminophen-treated animals than in controls, but these ratios were not different after 6 h. In our initial studies of CoASH, CoASSG, and AcCoA, we employed a modification of a method described by King and Reiss (34) in which freeze-clamped tissues were homogenized in HClO4, and the supernatants were examined by HPLC (Method I). In these studies, we measured GSH and GSSG levels in the same HClO4 supernatants and observed hepatic GSSG levels that were 3-4 times higher than we had found in earlier studies using tissue homogenization in neutral buffer containing NEM to minimize post vivo thiol oxidation and thiol-disulfide exchange reactions. Subsequent parallel measurements using the two methods on adjacent portions of freeze-clamped livers revealed GSSG levels in line with our previous studies using the NEM-based method and higher levels with homogenization directly in HClO4 (data not shown). These elevated GSSG levels observed in tissues homogenized in HClO4 are in agreement with the effects observed by Asensi et al. (38) and suggested that similar oxidation and exchange reactions might be affecting our estimates of CoASSG levels. Therefore, we investigated the use of reaction with NEM during tissue homogenization for measurements of

876

Chem. Res. Toxicol., Vol. 13, No. 9, 2000

CoASSG and CoASH, which necessitated measurement of the latter as the CoASH-NEM adduct. Separation of the CoASH-NEM adduct from CoASSG required significant alteration of the HPLC method, which was accomplished with a modification of the elution gradient reported by Baker and Schooley (35). With the modified protocol, the CoASSG concentrations were approximately 1 /2-1/3 of the levels estimated in the same samples homogenized in HClO4. The analysis of the CoA-NEM adduct gave a sharper peak shape than was obtained for CoASH in the previous method, and we observed a lower limit of detection of the CoA-NEM adduct of 2 pmol oncolumn at a signal-to-noise ratio of 3/1. The limit of detection for CoASSG was 6 pmol on-column and was unaffected by the change in method. The day-to-day coefficients of variation were 4.3% for CoASH (assayed as the NEM derivative) and 11.2% for CoASSG with the NEM method. Although AcCoA prepared in a NaHPO4 buffer was well separated by HPLC with the modified procedure, AcCoA prepared in buffer containing NEM gave no peak at the expected retention time, and no additional peaks were detected. All data reported in this paper for CoASH, CoASSG, and GSSG concentrations were obtained using the NEM-based method, as described in Experimental Procedures as Method II. Hepatic CoASH levels were lower in acetaminophentreated mice than in controls at 2 and 4 h, but were no longer different by 6 h (Figure 1A). At the present time, we attribute no mechanistic implications to the relatively small but statistically significant differences between the 4 and 6 h groups that are observed in hepatic CoASH levels of the vehicle-treated control animals. The levels of AcCoA in liver samples analyzed by Method I were 92.6 ( 5.3 vs 62.9 ( 5.1 nmol/g of liver at 2 h and 79.5 ( 5.4 vs 41.7 ( 5.4 nmol/g of liver at 6 h in control and acetaminophen-treated mice, respectively. Hepatic CoASSG concentrations appear to be slightly lower in acetaminophen-treated mice than in control animals, and both groups appear to show lower levels with increasing time post exposure; however, none of these effects were statistically significant (Figure 1B). CoASH/CoASSG ratios were not different in either group at 2 and 4 h or in saline-treated animals after 6 h, but were increased in the acetaminophen-treated animals at 6 h (Figure 1C). The decreases in hepatic CoASH concentrations in acetaminophen-treated animals, the absence of comparable increases in CoASSG concentrations, and the avid alkylation of biological thiols by chemically reactive metabolites of acetaminophen suggested the possibility of alkylation of CoASH. Reaction of CoASH with NAPQI in vitro gave a product that could be detected by HPLC (Figure 2A). Peaks coeluting with this product were not observed in supernatants of livers from saline-treated animals (Figure 2B), but were observed in tissues from acetaminophen-treated animals (Figure 2C). Two hours after administration of acetaminophen, the level of the putative CoASH-acetaminophen adduct was estimated as 25.7 ( 4.5 nmol/g of liver (mean ( SE, n ) 8), using chromatographic peak areas and molar extinction coefficients calculated as the sum of experimentally determined peak areas for acetaminophen and CoASH. To date, extensive efforts to characterize this putative adduct have been unsuccessful. FAB-MS did, however, provide additional evidence regarding the nature of the product of the reaction

Rogers et al.

Figure 1. Effects of acetaminophen on hepatic concentrations of CoASH (A) and CoASSG (B) and CoASH/CoASSG ratios (C). Male ICR mice were treated with 400 mg acetaminophen/kg or equal volumes of saline and sacrificed at 2, 4, or 6 h. Livers were removed, freeze-clamped, and stored at -70 °C. Tissue was ground under liquid N2, homogenized in 0.1 M NaPO4 (pH 7.4) containing 28.5 mM NEM, and then acidified with an equal volume of 4% HClO4. Samples were centrifuged at 12700g, and supernatants were analyzed by HPLC (see Experimental Procedures). Results are expressed as means ( SE (n ) 4), and data were assessed statistically by two-way analysis of variance, with modified t-tests post hoc, with significance indicated at P < 0.05. Asterisks indicate values different from those of the corresponding saline-treated animals as determined by a modified t-test. Common letters indicate groups that are different as determined by modified t-tests.

between CoASH and NAPQI in vitro, with an ion at m/z 917, as expected for the protonated molecular ion of a one-to-one adduct (Figure 3). CoASH, CoASSG, and the product of the reaction of CoASH with NEM also were analyzed for comparison and to facilitate the structural assignment of ions. The primary FAB-MS spectra yielded the corresponding protonated molecular ions [MH+], and product ion spectra from the respective protonated molecular ions (Figure 4A-D) were consistent with the assignments indicated in Figure 5. For the CoASH plus NAPQI derivative, the fragment ion at m/z 410 could

Hepatic CoASH and CoASSG

Chem. Res. Toxicol., Vol. 13, No. 9, 2000 877

Figure 2. HPLC analysis of putative CoASH-acetaminophen adducts. (A) CoASH and NAPQI were reacted in vitro and analyzed by HPLC as described in Experimental Procedures. Liver samples obtained from mice 2 h after treatment with saline (B) or acetaminophen (C) were processed as described for CoA measurements and analyzed by HPLC.

Figure 4. Product ion spectra from the protonated molecular ions of CoASH (A), the CoAS-NEM adduct (B), the CoASNAPQI adduct (C), and CoASSG (D).

Figure 3. FAB-MS spectrum of the product of reaction of NAPQI with CoASH in vitro. NAPQI was mixed with CoASH and the mixture placed immediately on the probe using 2-hydroxyethyl disulfide with oxalic acid as a matrix. The ion at m/z 768 corresponds to the protonated molecular ion of CoASH, and the ion at m/z 917 corresponds to the expected adduct of CoASH and NAPQI.

arise from either of two fragments. The greater intensity of the m/z 410 ion relative to the ion at m/z 428 in the CoASH plus NAPQI spectrum (Figure 4C) than in the spectra in panels A and B of Figure 4 suggests that both modes of fragmentation (2 and b in the proposed scheme) may be contributing to the ion at m/z 410. The data are consistent with the formation of the thioethers 1 and 2 expected from Michael addition of CoASH to NAPQI and NEM, respectively (Figure 5). In contrast with the effects of acetaminophen, hepatic GSH levels were altered only minimally in the furosemide-treated mice (Table 2). Hepatic GSH levels in the mice treated with 500 mg of furosemide/kg were slightly lower than in control animals at 2 and 4 h, but were not different from controls at 6 h. GSH levels were not affected at any time point by 400 mg of furosemide/kg. GSSG concentrations were not affected by either dose of furosemide at 2 or 4 h, but were elevated in both furosemide-treated groups at 6 h. The GSH/GSSG ratios were lower in the animals given 500 mg of furosemide/ kg than in the respective saline-treated controls at all time points and were lower at 4 h in the animals given 400 mg/kg (Table 2). However, the ratios remained above the levels observed in the fasted mice employed in the studies with acetaminophen (Table 1).

Hepatic CoASH contents were not altered significantly by furosemide at any time point, but levels in all treatment groups were greater at 6 h than in the respective groups at 2 and 4 h (Figure 6A). Surprisingly, HPLC analyses indicated that hepatic CoASSG levels were substantially higher in both groups of furosemidetreated animals at 2 h, but CoASSG concentrations returned to control levels at 4 and 6 h (Figure 6B). The increases in CoASSG levels and minimal changes in CoASH levels resulted in dramatic decreases in the CoASH/CoASSG ratios (Figure 6C). The possible coelution of a minor metabolite of furosemide or other substance with CoASSG cannot be excluded completely, but in livers of furosemide-treated mice 2 h but not 4 or 6 h after treatment, we observed greater areas of this peak using a number of different HPLC columns and two distinct mobile phases, plus a number of other mobile phases studied in the process of methods development. At no time have we observed indications of peak shoulders or secondary peaks that might confound the data. We were most concerned about the possible coelution of CoA esters of short chain polar fatty acids and have confirmed chromatographic separation of CoASSG from the acetyl, propionyl, malonyl, methylmalonyl, and succinyl CoA esters (data not shown).

Discussion The decreases in hepatic GSH/GSSG ratios caused by administration of acetaminophen (Table 1) satisfy some definitions of oxidant stresses. However, the early decreases in GSH/GSSG ratios are driven almost entirely by lower GSH concentrations with no increases in GSSG levels to account for the loss of GSH. The consumption

878

Chem. Res. Toxicol., Vol. 13, No. 9, 2000

Rogers et al.

Figure 5. Product ion fragmentation patterns of CoASH derivatives. Common fragmentation patterns are assigned to each of the CoASH-derived species analyzed in Figure 4. Structures 1 and 2 are proposed for the products of the reactions of CoASH with NAPQI and NEM, respectively. Table 2. Hepatic Injury and Glutathione Status in Mice Treated with Furosemidea furosemide (mg/kg) 0 400 500 0 400 500 0 400 500

time (h)

ALTb-d (units/L)

GSHc (µmol/g of liver)

GSSGb,c (nmol/g of liver)

GSH/GSSGb,c

2 2 2 4 4 4 6 6 6

16 ( 3 53 ( 18F 36 ( 5G 29 ( 9 136 ( 41H,I 195 ( 39 30 ( 7 379 ( 90e,F,H,J 1208 ( 297e,G,I,J

11.36 ( 10.17 ( 0.33 9.24 ( 0.57e 11.82 ( 0.60G 10.79 ( 1.15 9.03 ( 0.43e 10.02 ( 0.58F,G 10.65 ( 0.57 9.87 ( 1.16

24.8 ( 37.4 ( 4.8G 51.6 ( 14.1H 17.9 ( 1.0I 33.1 ( 4.0J 57.4 ( 12.1K 41.6 ( 16.4F,I 112.6 ( 29.9e,G,J 102.9 ( 22.5e,H,K

493 ( 84F 283 ( 27G 211 ( 41e,H 669 ( 63I 348 ( 73e,J 183 ( 44e,K 358 ( 113F,I 182 ( 110G,J 136 ( 62e,H,K

0.75F

3.7F

a Male ICR mice were given 0, 400, or 500 mg of furosemide/kg (ip). At 2, 4, or 6 h, animals were anesthetized, blood was obtained by cardiac puncture, and livers were removed and portions immediately freeze-clamped with liquid N2-cooled tongs. Plasma ALT activities and tissue GSH and GSSG concentrations were measured as described in Experimental Procedures. Data are means ( SE from four animals in each group and were assessed statistically by two-way ANOVA, with modified t-tests. Common capital letters indicate groups that are different as determined by modified t-tests (p < 0.05). b Effect of time. c Effect of treatment. d Interaction between time and treatment indicated. e Different than corresponding vehicle-treated controls as determined by modified t-tests (p < 0.05).

of hepatic GSH stores is more readily attributable to formation of the thioether conjugate of an acetaminophen metabolite with GSH (1, 9). Similar changes in hepatic concentrations of GSH and GSSG and in GSH/GSSG ratios can be effected by administration of doses of diethyl maleate that do not cause measurable cellular damage (32), which suggests that changes in GSH/GSSG ratios based entirely on decreases in GSH levels are not necessarily useful biomarkers of oxidative mechanisms of initiation of cellular injury. The absence of an increase in absolute concentrations of GSSG in the acetaminophen-treated animals through 4 h after treatment is not consistent with oxidant stress responses on this pivotal biomarker of antioxidant utilization during this period of maximal drug metabolism (39). The increases in hepatic GSSG concentrations observed 6 h after drug administration are in agreement with the previous reports of Tirmenstein and Nelson (7) and Jaeschke (6) and suggest oxidant activity, but this oxidant response does not precede expression of hepatic injury. Although administration of NAC just before or

within 4 h after acetaminophen is highly effective in attenuating hepatic necrosis in this animal model, NAC given g4 h after administration of acetaminophen shows little amelioration of injury (18, 40). The failure of NAC administration g4 h after acetaminophen to attenuate hepatic necrosis appreciably also indicates that whatever oxidative effects are reflected by the elevation of GSSG levels 6 h after treatment are unlikely to contribute substantially to the extent of hepatic damage. The lack of increases in hepatic CoASSG concentrations and the absence of decreases in CoASH/CoASSG ratios (Figure 1) are not consistent with the most straightforward expression of our working hypothesis regarding mitochondrial compartmentalization of oxidant stress effects of acetaminophen. The data do not prove that oxidant mechanisms, whether mitochondrially compartmentalized or not, do not contribute to acetaminophen-induced necrosis. However, in previous studies using derivatization with monobromobimane and separation of the fluorescent thioether derivatives by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (41,

Hepatic CoASH and CoASSG

Figure 6. Effects of furosemide on hepatic concentrations of CoASH (A) and CoASSG (B) and CoASH/CoASSG ratios (C). Male ICR mice were treated with 400 or 500 mg furosemide/kg or equal volumes of saline and sacrificed at 2, 4, or 6 h. Livers were removed, freeze-clamped, and stored at -70 °C. Tissue was ground under liquid N2, homogenized in 20 mM NaPO4 (pH 7.4) containing 28.5 mM NEM, and then acidified with an equal volume of 4% HClO4. After acidification, samples were centrifuged at 12700g, and supernatants were analyzed by HPLC (see Experimental Procedures). Results are expressed as means ( SE (n ) 4), and data were assessed statistically by two-way analysis of variance, with modified t-tests post hoc, with significance indicated at P < 0.05. Asterisks indicate values different from those of the corresponding saline-treated animals as determined by a modified t-test. Common letters indicate groups that are different as determined by a modified t-test.

42), we observed no diminution of hepatic levels of specific protein thiols in whole tissue homogenates or subcellular fractions of livers of acetaminophen-treated mice studied 2, 4, or 6 h after treatment, with the singular exception of decreases in thiol contents and enzyme activities of carbamoyl phosphate synthetase I in mitochondria of acetaminophen-treated animals 6 h after administration of hepatotoxic doses of acetaminophen. However, neither of these effects was reversed in vitro by disulfide reducing agents such as dithiothreitol. Tirmenstein and Nelson observed decreases in hepatic protein thiol concentrations in acetaminophen-treated mice (7), with data expressed per milligram of protein.

Chem. Res. Toxicol., Vol. 13, No. 9, 2000 879

They noted that the observed decreases in protein thiol contents represented 15 times the amounts of thioether formed by covalent binding of acetaminophen metabolites, thus indicating that formation of protein thioethers with drug metabolites is insufficient to explain the changes in protein thiol status they observed (7). However, the differences they reported could be influenced by dilution of hepatocellular proteins by extravasation of plasma proteins and the entrapment of blood that accompanies acetaminophen-induced hepatic necrosis and swelling in vivo (16, 18). The decreases in hepatic CoASH levels in the acetaminophen-treated mice (Figure 1) also are not readily explained by oxidative mechanisms. CoASSG concentrations are much smaller than are the decreases in CoASH levels, and the CoASSG concentrations are not increased in the treated animals. Huth et al. have reported evidence for formation of mixed disulfides with CoASH and three mitochondrial proteins, with release of CoASH by treatment with dithiothreitol and β-mercaptoethanol (43, 44). However, in the absence of increases in tissue levels of GSSG or CoASSG, S-thiolations of protein thiols by CoASH are less plausible as mechanisms for the decreases in hepatic CoASH levels observed in mice treated with acetaminophen. Substantial protein thiol S-thiolation in the absence of measurable increases in GSSG concentrations would be unusual in that preferential direct oxidation of the protein thiols would be required. Chen et al. recently have reported formation of the corresponding ipso adducts in reactions of GSH or papain with NAPQI in nonbiological reaction systems (45). These authors further demonstrated that the ipso adducts did not react with ascorbate, but did react with GSH to yield acetaminophen and GSSG or PSSG, respectively. They postulated that similar reactions in vivo could explain the late increases in tissue levels of GSSG, through accumulation of ipso adducts formed during the initial phase of acetaminophen metabolism and GSH depletion, with formation of GSSG during GSH resynthesis and recovery. Although additional studies will be necessary to test the possible relevance of related mechanisms in vivo, the ipso adducts characterized by Chen et al. were not stable in the presence of thiols, and protein thiols, which represent the majority of cellular thiol contents and are present throughout the course of acetaminophen-induced hepatic necrosis (41). Further, in the absence of thiols, the GSHNAPQI ipso adduct decomposed in vitro at pH 7.0 with a half-life of 22 min to form acetaminophen (53%) and 2′- and 3′-GS-AP (19 and 28%, respectively). Because of the substantial amounts of 2′-GS-AP Chen et al. observed in the decomposition of the ipso GSH adduct, proposals that ipso adduct mechanisms contribute significantly to acetaminophen metabolism and generation of GSSG in biological systems are difficult to reconcile with the virtual absence of 2′-thioether products observed in vivo (46). Interestingly, the absence of effects by and on ascorbate of acetaminophen metabolism in vivo, despite the rapid reactions of ascorbate with NAPQI or with alkylating metabolites of acetaminophen formed in microsomal oxidation systems, previously led us to suggest the intermediacy of ipso NAPQI adducts as a mechanism that could explain the absence of thiol and ascorbate oxidation with retention of capacity to alkylate nucleophiles (16). These observations, in addition to the absence of in-

880

Chem. Res. Toxicol., Vol. 13, No. 9, 2000

creases in GSSG levels during maximal metabolic generation of the alkylating metabolite, again despite the facile reaction of GSH with chemically synthesized NAPQI to form GSSG and acetaminophen in addition to 3′-GS-AP, indicated to us that the data regarding the effects of acetaminophen metabolism in vivo were not consistent with free NAPQI being a major metabolite of acetaminophen in vivo. The ipso adduct of some nucleophilic moiety with NAPQI offered an explanation for limiting reduction by ascorbate or GSH, while retaining alkylating capacity, if elimination (reversal) or collapse to a stable adduct were avoided. Glutathione S-transferases were our primary candidates for the intermediate nucleophiles (Y in the scheme in ref 16), and the tight limitations on reactions characteristic of free NAPQI suggested to us that the transferases would be closely associated functionally and probably topographically with the activating cytochromes P450. The results of our studies to date are best explained by alkylation of CoASH by reactive metabolites of acetaminophen. The transformations observed with acetaminophen metabolism and conjugation with GSH and covalent binding to protein in vivo are similar to the alkylation reactions observed by exposure of biological thiols to NAPQI in vitro (46). The coelution of an HPLC peak observed in acid supernatants of livers of acetaminophen-treated animals that is not observed in samples from animals not treated with acetaminophen with the product of the reaction of CoASH with NAPQI in vitro (Figure 2) further supports the alkylation hypothesis. The hepatic levels of the postulated conjugate estimated by HPLC analyses (25.7 ( 4.5 nmol/g of liver, n ) 8) do not account for the full 40 nmol/g liver decrement in CoASH levels observed in the acetaminophen-treated animals (Figure 2A), but the further metabolism, nonenzymatic degradation, and export to bile or the vasculature of the postulated intermediate have not been investigated, and could easily account for the net losses of CoA species. Efforts to isolate and characterize the apparent adduct from tissues of animals treated with acetaminophen have not been successful to date. The adduct prepared in vitro exhibits limited chemical stability, and the only conditions we have found that provide effective HPLC separation of this species contain high concentrations of phosphate, which interfere with mass spectral studies. Hepatotoxic doses of furosemide were administered to animals to test the hypothesis that decreases in hepatic levels of CoASH, such as those observed in the acetaminophen-treated mice, are common consequences of acute lethal injury. The potential for tissues to undergo oxidative degradations make determination of the biomarker responses to cell death initiated by distinct mechanisms an essential control in efforts to distinguish potential causes from effects of injury. We measured hepatic CoASH and CoASSG levels in mice given doses of furosemide that were comparably hepatotoxic to the doses of acetaminophen that were studied. The working hypothesis for these studies was that depletion of hepatic CoASH in the acetaminophen-treated animals might represent a result of cell injury, perhaps related to activation of the mitochondrial permeability transition pore (47-49), with extramitochondrial degradation or cellular export of CoA (50). The failure of comparably hepatotoxic doses of furosemide to diminish CoASH levels (Figure 6) indicates that the decreases in CoASH levels observed in the acetaminophen-treated mice are not

Rogers et al.

attributable to common consequences of acute hepatic necrosis. The fact that hepatic CoASH levels in the acetaminophen-treated mice are decreased maximally by 2 h (Figure 1), whereas administration of NAC 2 h after acetaminophen dramatically attenuates the expression of hepatic necrosis in this animal model (18), further strengthens the interpretation of the data that the decreases in hepatic CoASH concentrations in the acetaminophen-treated animals are unlikely to be secondary to cell damage. The mechanisms that are responsible for the decreases in CoASH levels in livers of acetaminophen-treated mice are not defined, but appear to be driven in concert with the metabolism of acetaminophen and occurring well ahead of the irreversible commitment to hepatic necrosis in this experimental model. Furosemide is similar to acetaminophen in that both cause hepatic necrosis in mice by mechanisms that involve generation of chemically reactive metabolites that covalently bind to target tissue proteins (29, 31). In contrast with acetaminophen, the chemically reactive metabolites of furosemide do not react preferentially with cellular GSH, and depletion of GSH does not precede alkylation of protein or initiation of necrosis. In addition, depletion of hepatic GSH by administration of diethyl maleate does not enhance furosemide-induced necrosis, and NAC and other GSH precursors do not ameliorate injury in vivo (18, 32). GSH does attenuate covalent binding to protein of metabolites of furosemide in microsomal activation systems in vitro (31), and cellular concentrations of GSH are diminished by cytotoxic doses of furosemide in isolated mouse hepatocytes (51), but these manifestations of furosemide metabolism and toxicity are not observed in vivo. The absence of an effect of furosemide on hepatic GSSG levels at 2 and 4 h (Table 1) is consistent with the absence of an effect on biliary GSSG efflux in rats (10), although furosemide may be metabolized differently in rats than in mice and is not appreciably hepatotoxic in rats (31). The greater GSSG contents in the livers of furosemidetreated mice than in the saline-treated controls observed after 6 h with both 400 and 500 mg/kg doses are consistent with an oxidant stress response. The increases in GSSG levels 6 h after furosemide treatment, which were similar to the effects observed with acetaminopheninduced liver damage, are more readily attributable to effects secondary to cellular injury than to primary effects of metabolism of the drug. The apparent increases in hepatic concentrations of CoASSG in the furosemide-treated mice 2 h after treatment (Figure 4) indicate an oxidant stress response. The absence of differences in tissue GSSG levels suggests that the responses indicated by the changes in CoASSG levels reflect some unexplained chemical specificity or are directed by cellular compartmentalization. The covalent binding to hepatic proteins of furosemide metabolites is maximal in the first 2 h following intraperitoneal administration to mice of the radiolabeled drug (29, 31), which suggests that the greatest potential for generation of reactive species driven by metabolism of furosemide should occur during that same interval. The apparent increases in CoASSG concentrations in the livers of furosemide-treated mice are observed early in the initiation of injury and, therefore, represent biomarkers of potential mechanisms of the initiation phase of cell killing. Although quite unexpected, our data suggest that the early oxidant stress responses observed in furo-

Hepatic CoASH and CoASSG

semide-treated mice are compartmentalized, most probably in the mitochondria. The data presently available do not, however, establish CoASSG levels as a reliable index of mitochondrial redox status, and substantial work remains to be done to test this hypothesis critically. In addition to serving as a potentially useful biomarker of compartmentalized oxidant stress, CoASSG is a potent vasoconstrictor, active at concentrations as low as 1 pM in vitro and in rats in vivo at doses as low as 0.5 nmol per animal (52). These vasoactive properties could contribute to initiation or expansion of tissue injury in vivo. The data presented here showing no increases in hepatic concentrations of GSSG or CoASSG during the first 4 h of exposure to hepatotoxic doses of acetaminophen do not support hypotheses based on thiol oxidation mechanisms in initiation of acetaminophen-induced hepatic necrosis in vivo. The decreases in hepatic CoASH levels observed in acetaminophen-treated mice are best interpreted as consequences of alkylation of CoASH by chemically reactive metabolites of acetaminophen, but the data available at present are not conclusive. The increases in hepatic CoASSG levels and unchanged GSSG levels observed in mice 2 h after administration of hepatotoxic doses of furosemide suggest oxidant stresses that appear to be compartmentalized, most probably in the mitochondria, but the significance of this surprising observation will require further investigation.

Acknowledgment. We are grateful for the support of Grant GM44263 from the National Institutes of Health and a Student Research Fellowship award (for C.J.V.) from Alpha Omega Alpha.

References (1) Nelson, S. D. (1990) Molecular mechanisms of the hepatotoxicity caused by acetaminophen. Sem. Liver Dis. 10, 267-278. (2) Smith, C. V., Lauterburg, B. H., and Mitchell, J. R. (1985) Covalent binding and acute lethal injury in vivo: How has the hypothesis survived a decade of critical examination? In Drug Metabolism and Disposition: Considerations in Clinical Pharmacology (Wilkinson, G., and Rawlins, M. D., Eds.) pp 161-181, MTP Press, London. (3) Cohen, S. D., Pumford, N. R., Khairallah, E. A., Boekelheide, K., Pohl, L. R., Amouzadeh, H. R., and Hinson, J. A. (1997) Selective protein covalent binding and target organ toxicity. Toxicol. Appl. Pharmacol. 143, 1-12. (4) Wendel, A., Feuerstein, S., and Konz, K.-H. (1979) Acute paracetamol intoxication of starved mice leads to lipid peroxidation in vivo. Biochem. Pharmacol. 28, 2051-2055. (5) Farber, J. L., Leonard, T. B., Kyle, M. E., Nakae, D., Serroni, A., and Rogers, S. A. (1988) Peroxidation-dependent and peroxidation-independent mechanisms by which acetaminophen kills cultured rat hepatocytes. Arch. Biochem. Biophys. 267, 640-650. (6) Jaeschke, H. (1990) Glutathione disulfide formation and oxidant stress during acetaminophen-induced hepatotoxicity in mice in vivo: the protective effect of allopurinol. J. Pharmacol. Exp. Ther. 255, 935-941. (7) Tirmenstein, M. A., and Nelson, S. D. (1990) Acetaminopheninduced oxidation of protein thiols. Contribution of impaired thiolmetabolizing enzymes and the breakdown of adenine nucleotides. J. Biol. Chem. 265, 3059-3065. (8) Gibson, J. D., Pumford, N. R., Samokyszyn, V. M., and Hinson, J. A. (1996) Mechanisms of acetaminophen-induced hepatotoxicity: Covalent binding versus oxidative stress. Chem. Res. Toxicol. 9, 580-585. (9) Adams, J. D., Jr., Lauterburg, B. H., and Mitchell, J. R. (1983) Plasma glutathione and glutathione disulfide in the rat: regulation and response to oxidative stress. J. Pharmacol. Exp. Ther. 227, 749-754. (10) Lauterburg, B. H., Smith, C. V., Hughes, H., and Mitchell, J. R. (1984) Biliary excretion of glutathione and glutathione disulfide in the rat: regulation and response to oxidative stress. J. Clin. Invest. 73, 124-133.

Chem. Res. Toxicol., Vol. 13, No. 9, 2000 881 (11) Smith, C. V. (1991) Correlations and apparent contradictions in assessment of oxidant stress status in vivo. Free Radical Biol. Med. 10, 217-224. (12) Gilbert, H. F. (1995) Thiol/disulfide exchange equilibria and disulfide bond stability. Methods Enzymol. 251, 8-28. (13) Nakae, D., Oakes, J. W., and Farber, J. L. (1988) Potentiation in the intact rat of the hepatotoxicity of acetaminophen by 1,3-bis(2-chloroethyl)-1-nitrosourea. Arch. Biochem. Biophys. 267, 651659. (14) Bodell, W. J., Aida, T., Berger, M. S., and Rosenblum, M. L. (1986) Increased repair of O6-alkylguanine DNA adducts in gliomaderived human cells resistant to the cytotoxic and cytogenetic effects of 1,3-bis(2-chloroethyl)-1-nitrosourea. Carcinogenesis 7, 879-883. (15) Tong, W. P., Dirk, M. C., and Ludlum, D. B. (1982) Formation of the cross-link 1-[N3-deoxycytidyl],2-[N1-deoxyguanosyl]-ethane in DNA treated with N,N′-bis(2-chloroethyl)-N-nitrosourea. Cancer Res. 42, 3102-3105. (16) Smith, C. V., and Mitchell, J. R. (1985) Acetaminophen hepatotoxicity in vivo is not accompanied by oxidant stress. Biochem. Biophys. Res. Commun. 133, 329-336. (17) Tirmenstein, M. A., and Nelson, S. D. (1989) Subcellular binding and effects on calcium homeostasis produced by acetaminophen and a nonhepatotoxic regioisomer, 3′-hydroxyacetanilide, in mouse liver. J. Biol. Chem. 264, 9814-9819. (18) Corcoran, G. B., Racz, W. J., Smith, C. V., and Mitchell, J. R. (1985) Effects of N-acetylcysteine on acetaminophen covalent binding and hepatic necrosis in mice. J. Pharmacol. Exp. Ther. 232, 864-872. (19) Gupta, S., Rogers, L. K., and Smith, C. V. (1994) Biliary excretion of lysosomal enzymes, iron, and oxidized protein in Fischer-344 and Sprague-Dawley rats and the effects of diquat and acetaminophen. Toxicol. Appl. Pharmacol. 125, 42-50. (20) Donnelly, P. J., Walker, R. M., and Racz, W. J. (1994) Inhibition of mitochondrial respiration in vivo is an early event in acetaminophen-induced hepatotoxicity. Arch. Toxicol. 68, 110-118. (21) Vendemiale, G., Grattagliano, I., Altomare, E., Turturro, N., and Guerrieri, F. (1996) Effect of acetaminophen administration on hepatic glutathione compartmentation and mitochondrial energy metabolism in the rat. Biochem. Pharmacol. 52, 1147-1154. (22) Robishaw, J. D., and Neely, J. R. (1985) Coenzyme A metabolism. Am. J. Physiol. 248, E1-E9. (23) Tahiliani, A. G., and Neely, J. R. (1987) A transport system for coenzyme A in isolated rat heart mitochondria. J. Biol. Chem. 262, 11607-11610. (24) Gilbert, H. F. (1982) Biological disulfides: The third messenger? J. Biol. Chem. 257, 12086-12091. (25) Dyar, R. E., and Wilken, D. R. (1972) Rat liver levels of coenzyme A-glutathione and of enzymes in its metabolism. Arch. Biochem. Biophys. 153, 619-626. (26) Chang, S. H., and Wilken, D. R. (1966) Participation of the unsymmetrical disulfide of coenzyme A and glutathione in an enzymatic sulfhydryl-disulfide interchange. J. Biol. Chem. 241, 4251-4260. (27) Crane, D., Haussinger, D., and Sies, H. (1982) Rise of coenzyme A-glutathione mixed disulfide during hydroperoxide metabolism in perfused rat liver. Eur. J. Biochem. 127, 575-578. (28) Ponto, L. L. B., and Schoenwald, R. D. (1990) Furosemide (Frusemide). A pharmacokinetic/pharmacodynamic review (Part I). Clin. Pharmacokinet. 18, 381-408. (29) Mitchell, J. R., Potter, W. Z., Hinson, J. A., and Jollow, D. J. (1974) Hepatic necrosis caused by furosemide. Nature 251, 508-510. (30) Wirth, P. J., Bettis, C. J., and Nelson, W. L. (1976) Microsomal metabolism of furosemide evidence for the nature of the reactive intermediate involved in covalent binding. Mol. Pharmacol. 12, 759-768. (31) Mitchell, J. R., Nelson, W. L., Potter, W. Z., Sasame, H. A., and Jollow, D. J. (1976) Metabolic activation of furosemide to a chemically reactive, hepatotoxic metabolite. J. Pharmacol. Exp. Ther. 199, 41-52. (32) Smith, C. V., Hughes, H., Lauterburg, B. H., and Mitchell, J. R. (1983) Chemical nature of reactive metabolites determines their biological interactions with glutathione. In Functions of Glutathione: Biochemical, Physiological, Toxicological, and Clinical Aspects (Larsson, A., Orrenius, S., Holmgren, A., and Mannervik, B., Eds.) pp 125-138, Raven Press, New York. (33) Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265-275. (34) King, M. T., and Reiss, P. D. (1985) Separation and measurement of short-chain coenzyme-A compounds in rat liver by reversedphase high-performance liquid chromatography. Anal. Biochem. 146, 173-179.

882

Chem. Res. Toxicol., Vol. 13, No. 9, 2000

(35) Baker, F. C., and Schooley, D. A. (1981) Separation of S-acylCoA thioesters and related compounds by reversed-phase ion-pair chromatography. Methods Enzymol. 72, 41-52. (36) Dahlin, D. C., and Nelson, S. D. (1982) Synthesis, decomposition kinetics, and preliminary toxicological studies of pure N-acetylp-benzoquinone imine, a proposed toxic metabolite of acetaminophen. J. Med. Chem. 25, 885-886. (37) Norusis, M. J. (1992) SPSS for Windows. Advanced Statistics, release 5, SPSS Inc., Chicago. (38) Asensi, M., Sastre, J., Pallardo, F. V., de la Asuncion, J. G., Estrela, J. M., and Vina, J. (1994) A high-performance liquid chromatography method for measurement of oxidized glutathione in biological samples. Anal. Biochem. 217, 323-328. (39) Mitchell, J. R., Jollow, D. J., Potter, W. Z., Davis, D. C., Gillette, J. R., and Brodie, B. B. (1973) Acetaminophen-induced hepatic necrosis. I. Role of drug metabolism. J. Pharmacol. Exp. Ther. 187, 185-194. (40) Corcoran, G. B., Todd, E. L., Racz, W. J., Hughes, H., Smith, C. V., and Mitchell, J. R. (1985) Effects of N-acetylcysteine on the disposition and metabolism of acetaminophen in mice. J. Pharmacol. Exp. Ther. 232, 857-863. (41) Gupta, S., Rogers, L. K., Taylor, S. K., and Smith, C. V. (1997) Inhibition of carbamyl phosphate synthetase-I and glutamine synthetase by hepatotoxic doses of acetaminophen in mice. Toxicol. Appl. Pharmacol. 146, 317-327. (42) Taylor, S. K., and Smith, C. V. (1999) Mechanisms of decreases in monobromobimane-derived fluorescence of carbamyl phosphate synthetase-I following hepatotoxic doses of acetaminophen in mice. Toxic Subst. Mech. 18, 67-82. (43) Huth, W., Arvand, M., and Moller, U. (1988) Identification of [1-14C]pantothenic-acid-mediated modified mitochondrial proteins. Eur. J. Biochem. 172, 607-614. (44) Huth, W., Worm-Breitgoff, C., Moller, U., and Wunderlich, I. (1991) Evidence for an in vivo modification of mitochondrial proteins by coenzyme A. Biochim. Biophys. Acta 1077, 1-10.

Rogers et al. (45) Chen, W., Shockcor, J. P., Tonge, R., Hunter, A., Gartner, C., and Nelson, S. D. (1999) Protein and nonprotein cysteinyl thiol modification by N-acetyl-p-benzoquinone imine via a novel ipso adduct. Biochemistry 38, 8159-8166. (46) Hoffmann, K.-J., Streeter, A. J., Axworthy, D. B., and Baillie, T. A. (1985) Identification of the major covalent adduct formed in vitro and in vivo between acetaminophen and mouse liver proteins. Mol. Pharmacol. 27, 566-573. (47) Reed, D. J., and Savage, M. K. (1995) Influence of metabolic inhibitors on mitochondrial permeability transition and glutathione status. Biochim. Biophys. Acta 1271, 43-50. (48) Petronilli, V., Costantini, P., Scorrano, L., Colonna, R., Passamonti, S., and Bernardi, P. (1994) The voltage sensor of the mitochondrial permeability transition pore is tuned by the oxidation-reduction state of vicinal thiols. J. Biol. Chem. 269, 16638-16642. (49) Bremer, J., Wojtczak, A., and Skrede, S. (1972) The leakage and destruction of CoA in isolated mitochondria. Eur. J. Biochem. 25, 190-197. (50) Skrede, S. (1973) The degradation of CoA: Subcellular localization and kinetic properties of CoA and dephospho-CoA pyrophosphatase. Eur. J. Biochem. 38, 401-407. (51) Massey, T. E., Walker, R. M., McElligott, T. F., and Racz, W. J. (1987) Furosemide toxicity in isolated mouse hepatocyte suspensions. Toxicology 43, 149-160. (52) Schulter, H., Meissner, M., van der Geit, M., Tepel, M., Bachmann, J., Groβ, I., Nordhoff, E., Karas, M., Witzel, H., and Zidek, W. (1995) Coenzyme A glutathione disulfide. A potent vasoconstrictor derived from the adrenal gland. Circ. Res. 76, 675-680.

TX0000926