214
Chem. Res. Toxicol. 1991,4, 214-217
Hepatotoxicity after 3’-Hydroxyacetanilide Administration to Buthionine Sulfoximine Pretreated Mice Mark A. Tirmenstein and Sidney D. Nelson* Department of Medicinal Chemistry, University of Washington, Seattle, Washington 98195 Received May 21, 1990
The administration of 3’-hydroxyacetanilide, a regioisomer of acetaminophen, to mice failed to produce hepatotoxicity even after the administration of diethyl maleate. In contrast, hepatotoxicity did occur when 3’-hydroxyacetanilide was administered to buthionine sulfoximine pretreated mice. Although the administration of 3’-hydroxyacetanilide in conjunction with either diethyl maleate or buthionine sulfoximine depleted total hepatic glutathione, only the combined buthionine sulfoximine-3’-hydroxyacetanilidetreatment decreased hepatic mitochondrial glutathione concentrations to below 20% of control values. In addition, pretreatment with buthionine sulfoximine increased the amount of 3’-hydroxyacetanilide bound to mitochondrial proteins. These results, in conjunction without previous results on the involvement of mitochondrial damage in the pathogenesis of hepatotoxicity caused by acetaminophen, suggest a probable relationship between mitochondrial damage caused by the buthionine sulfoximine3’-hydroxyacetanilide treatment and hepatotoxicity caused by this treatment.
Introduction 3’-Hydroxyacetanilide (AMAP)’ is a regioisomer of acetaminophen (APAP). Like APAP, AMAP has analgesic and antipyretic activity (1, 2). However, in contrast to APAP, there is no evidence that AMAP is hepatotoxic to mice (3-5) or hamsters (6). Studies indicate that APAP is metabolized by cytochrome P-450 to the reactive metabolite N-acetyl-p-benzoquinone imine (7,8), and that this metabolite can conjugate glutathione (GSH) and arylate protein sulfhydryl groups (9, 10). AMAP is also metabolized by cytochrome P-450 to two diphenolic acetanilides, 2-acetamidohydroquinone and 3‘-hydroxyacetaminophen (5,11,12). These metabolites can be further oxidized to 2-acetamido-p-benzoquinoneand 4-acetamido-o-benzoquinone, respectively, in vivo and in vitro (5, 11). There is also evidence that S-methoxyacetaminophen can be formed from AMAP and that it can be oxidized to the reactive product N-acetyl-3-methoxy-pbenzoquinone imine, and the GSH conjugates of these quinones and the quinone imine have been detected in the bile of mice treated with AMAP (11). In this regard, it appears that the reactive metabolites of APAP and those of AMAP are similar in their reactivity with sulfhydryls and in their ability to conjugate GSH. Recently, we demonstrated that both APAP and AMAP deplete liver GSH levels and produce extensive covalent binding to the proteins of liver homogenates following administration to mice (4). In this study, similar levels of covalent binding to liver homogenates were produced by administering either 1.7 mmol of (250 mg/kg) APAP or 4.0 mmol of (600 mg/ kg) AMAP. Despite the similarity in covalent binding, this dose of APAP was demonstrated to be hepatotoxic while the corresponding AMAP dose was not (4). In addition, our results indicate that APAP yields greater levels of covalent binding to mitochondria and depletes mitochondrialGSH to a much greater extent than AMAP. APAP also reversibly inhibited the ability of isolated hepatic mitochondria to sequester calcium 1 h after administration to mice and was associated with increased mitochondrial calcium levels when mitochondria
* To whom correspondence should be addressed. 0893-228x/91/2704-0214$02.50/0
were isolated at later time points (4). AMAP did not produce these alterations in mitochondrial calcium homeostasis. Studies with a variety of toxic agents have noted a correlation between the depletion of mitochondrial GSH and the development of toxicity (13, 14). The cause and significance of this association is unknown. In the present study, mice were treated with either diethyl maleate (DEM) or buthionine sulfoximine (BSO) to reduce GSH levels prior to AMAP administration. DEM conjugates with GSH (15)while BSO inhibits the synthesis of GSH (16). Administration of GSH-depleting agents may augment the ability of AMAP to deplete mitochondrial GSH and to bind to mitochondrial proteins. Our results suggest that loss of total hepatic GSH did not correlate with 3’hydroxyacetanilide-induced hepatotoxicity. However, there appears to be a relationship among decreased concentrations of mitochondrial GSH, increased covalent binding of AMAP-reactive metabolites to mitochondrial proteins, and the subsequent development of hepatotoxicity in AMAP-treated mice.
Materials and Methods Chemicals. BSO, DEM, and the alanine aminotransferase (ALT)determination kit (ALT 59-10) were purchased from Sigma Chemical Co. (St. Louis, MO). AMAP was obtained from Aldrich Chemical Co. (Milwaukee, WI) and was recrystallized from water prior to use. Sodium phenobarbital was purchased from Spectrum Chemical Manufacturing Co. (Gardena,CA). [ring-UL-l’C]AMAP was purchased from Pathfinder Laboratories Inc. (St. Louis, MO), and Aquasol-:! scintillation cocktail was purchased from New England Research Products (Boston, MA). All other chemicals were acquired from Sigma or Aldrich Chemical Co. Animals. Male Swiss Webster mice from Charles River Breeding Laboratories (Wilmington, MA) weighing 20-25 g were used throughout the study. Mice were given 0.1% sodium phenobarbital in their drinking water for 5 days prior to the experiment. Food was withheld for 16 h immediately prior to and after the administration of the drug. Administration of Drugs. BSO was dissolved in saline at a concentration of 0.18 mmol/mL and administered ip at a dose Abbreviations: AMAP, 3‘-hydroxyacetanilide;APAP, acetaminophen, 4’-hydroxyacetanilide;DEM, diethyl maleate; BSO, bbuthionine sulfoximine; ALT, alanine aminotransferase. 0 1991 American Chemical Society
3'-Hydroxyacetanilide Hepatotoxicity
Chem. Res. Toxicol., Vol. 4, No. 2, 1991 215
Table 1. Effects of Buthionine Sulfoximine, Diethyl Maleate, and 3'-Hydroxyacetanilide on Plasma Alanine Aminotransferase Activity Levels treatmenta units/L control (n = 5 ) 41 f 4b 87 f 11 AMAP (n = 5 ) BSO (n = 5 ) 121 f 39 BSO + AMAP (n = 13) 1880 i 817c DEM (n = 5 ) 83 f 11 DEM + AMAP (n = 3) 82 f 10 ~_____________
~.
...
.
.-y'-r/'---
______
ODrugs were administered ip at the following doses: BSO 3.6 mmol/kg, DEM 3.1 mmol/kg, and AMAP 4.0 mmol/kg body weight. Plasma was collected 24 h after AMAP administration. In experiments in which BSO or DEM was administered in conjunction with AMAP, BSO was administered 3 h prior to AMAP while DEM was injected 1 h prior to AMAP. In animals treated with BSO or DEM only, samples were collected at 27 and 25 h, respectively, after the administration of these drugs. These time points were selected to correspond with the 24-h time point following AMAP administration. All values represent the mean f SEM. 'Statistically significant (p < 0.05) from control, AMAP and BSO treatment groups as determined by two-sample t test for independent samples with unequal variance (Satterthwaite's method). of 3.6 mmol/kg body weight to mice as previously reported (17). In experiments in which both AMAP and BSO were administered to mice, AMAP was administered 3 h after the BSO. DEM was dissolved in com oil and administered a t a dose of 3.1 mmol/kg body weight. The volume of the DEM-com oil solution injected was 8 mL/kg body weight. When AMAP and DEM were given in combination to mice, AMAP was injected 1 h after the DEM. AMAP was dissolved in d i n e at a concentrationof 0.24 mmol/mL and injected ip a t a dose of 4.0 mmol/kg body weight. It was necessary to warm AMAP solutions prior to injections (-40 "C) in order to dissolve the drug and maintain it in solution. Homogenization and Mitochondrial Isolation. Animals were killed by cervical dislocation a t the indicated times. Gall bladders were removed and livers were rinsed and then homogenized in 5 volumes of isolation buffer (250 mM sucrose, 10 mM Tris, mM EDTA, pH 7.4 at 4 "C). Homogenizationwas achieved with 6 strokes of a Potter-Elvehjem tissue grinder. The homogenate was centrifuged a t 96Og for 10 min. The supematant was then centrifuged at 7000g for 15 min to pellet mitochondria. Mitochondria were then washed by resuspending pellets in isolation buffer and centrifuging at 7000g for 15 min. Mitochondrial purity and yield were determined by utilizing the marker enzyme citrate synthase as described and reported previously (4). Mitochondrial pellets were resuspended in appropriate volumes of the isolation buffer prior to the GSH assay. Biochemical Determinations. Proteins were measured according to h w r y et al. (28)as modified by Peterson (19). Plasma ALT levels were assessed with an ALT determination kit (ALT 59-10)and an EMIT clinical pmcesso-pectrophotometer (Syva). Blood was collected by cardiac puncture and then centrifuged at 16000g for 7 min to prepare plamsa. GSH was determined as described by Reed et al. (20) as modified by Fariss et al. (21). However, no phenanthroline was used in the GSH assay, and the EDTA did not interfere as long as it was washed off a t the end of the run. Histopathology Determinations. Liver slices were placed in buffered formalin. Scoring was performed by R. J. McMurtry, M.D., Ph.D., a board-certified pathologist. The following criteria were utilized for scoring: 0, necrosis absent; 1+, necrosis of less than 6% of hepatocytes; 2+, necrosis of 6-25% of hepatocytes; 3+, necrosis of 26-50% of hepatocytes; and 4+, necrosis of more than 50% of hepatocytes. Covalent Binding Determinations. Radiolabeled AMAP was administered a t a dose of 535 pCi/kg body weight. Mitochondria were isolated as before. Following isolation, mitochondrial proteins were precipitated with 10% TCA and washed and counted as previously described (4).
Results The 24-h plasma ALT levels for the various treatment groups are reported in Table I. As expected, AMAP-
Figure 1. Microscopic section of a mouse liver after treatment with BSO-AMAP as described in Table I. The section was stained with hematoxylin-eosin and reveals a solid 2+ centrilobular necrosis (upper right quadrant) with sparing of cells in the portal tract (lower left quadrant). Neither BSO nor AMAP treatments alone caused any evidence of hepatic necrosis, which is consistent with prior reports (3-6,28). After these treatments, cells look uniformly like those in the periportal region.
treated animals exhibited no hepatotoxicity as evidenced by low plasma ALT values. Previously, we found no evidence of hepatotoxicity in mice that received 250,600,800, or 1000 mg/kg AMAP (5). However, a high rate of mor'tality (80%) occurred in mice that received lo00 mg/kg AMAP. These animals died within a few hours after drug administration of respiratory failure. There was no evidence of hepatotoxicity. In addition, BSO and DEM by themselves were not hepatotoxic. The combined administration of DEM and AMAP to mice also did not produce large increases in plasma ALT values. In separate experiments, histological examination of DEM-AMAP-treated animals showed no evidence of liver necrosis. Although there was no evidence of hepatotoxicity, the incidence of mortality in this treatment group was high, about 50% after 24 h, with death occurring as early as 4 h after AMAP administration. The combined administration of BSO and AMAP also resulted in about 20% mortality after 24 h. However, in contrast to the other treatments, the administration of BSO in conjunction with AMAP elevated plasma ALT levels. T o confirm that the BSO-AMAP treatment is indeed hepatotoxic, liver slices were collected and histologically examined 24 h after AMAP administration. Livers were characterized by centrilobular necrosis in affected animals (Figure 1). GSH levels in liver homogenates were measured in animals receiving one of the drug treatments listed in Table 11. After 1h, livers were excised and homogenized. In a previous study (4), we determined that extensive liver GSH depletion and covalent binding occurred by 1h after AMAP administration. Control homogenate GSH values were determined to be 25.4 nmol/mg of protein. All agents depleted homogenate GSH levels below this value. The combined administration of AMAP with either DEM or BSO produced very extensive depletion of homogenate GSH levels. The BSO-AMAP treatment reduced hepatic GSH levels to only 3% of controls at the 1-h time point, and the DEM-AMAP treatment yielded hepatic GSH levels of 6% of controls at this time point. Hepatic oxidized glutathione (GSSG) remained at low or nondetectable levels in all animals. The effects of these drug treatments on mitochondrial GSH levels are shown in Table III. Control mitochondrial GSH values were determined to be 3.49 nmol/mg of pro-
Tirmenstein and Nelson
216 Chem. Res. Toxicol., Vol. 4,No. 2, 1991
Table 11. Effects of Diethyl Maleate, Buthionine Sulfoximine, and 3’-Hydroxyacetanilide on Liver Homogenate Glutathione Levels
treatment’ cc trol A AP
BbO
BSO + AMAP DEM DEM + AMAP
1-h time point, nmol/mg of protein GSH GSSG 0.42 f 0.06 25.4 f 1.14b 5.78 f 0.21‘ 0.49 f 0.06 5.18 f 0.60d ND‘ 0.79 f O . l l c ND 13.8 f 0.46d ND 1.53 f 0.12‘ ND
ODrugs were administered ip a t the following doses: BSO 3.6 mmol/kg, DEM 3.1 mmol/kg, and AMAP 4.0 mmol/kg body weight. Animals were killed and livers were collected 1 h after AMAP administration in all AMAP experiments. In experiments in which BSO or DEM was administered in conjunction with AMAP, BSO was administered 3 h prior to AMAP while DEM was injected 1 h prior to AMAP. In animals treated with BSO or DEM only, animals were killed and livers were collected at 4 and 2 h, respectively, after the administration of these drugs. These time points were selected to correspond with the 1-h time point following AMAP administration. bAll values are the mean of at least 3 determinations f SEM. ‘Statistically significant ( p < 0.001) from control as determined by a two-sample t test for independent samples with unequal variance (Satterthwaite’s method). dStatistically significant (p < 0.001) from control as determined by unpaired Student’s test. ‘Not detectable (
