TOXICOLOGY

AND

APPLIED

104,225-234

PHARMACOLOGY

(1990)

Acetaminophen-induced Alterations and Serum Insulin Concentrations

in Pancreatic ,8 Cells in 66C3Fl Mice

D. V. FmciusoN,*3t D. W. ROBERTS,* H. HAN-SHU,# A. ANDREWS,$ R. W. BENSON,* T. J. BUCCI,$ AND J. A. HINSON* *National

Center for Toxicological Research, Jefferson, Arkansas 72079-9502; t University at Little Rock, Little Rock, Arkansas 72204; and SPathology Associates, Inc., National Center for Toxicological Research, Jefferson, Arkansas 72079-9502

Received

October

16,1989;

accepted

February

ofArkansas

19, I989

Acetaminophen-Induced Alterations in Pancreatic /I Cells and Serum Insulin Concentrations in B6C3Fl Mice. FERGUSON, D. V., ROBERTS, D. W., HAN-SHU, H., ANDREWS, A., BENSON, R. W., Buccr, T. J., AND HINSON, J. A. (1990). Toxicol. Appl. Pharmacol. 104,225-234. Administration of acetaminophen (500 mg/kg) to male B6C3Fl mice resulted in alterations of pancreatic /3 cell ultrastructure. These alterations were characterized by pronounced intercellular spaces, cytoplasmid vacuolization, damaged membranes of cytoplasm, secretory granules, and other organelles, and pyknotic nuclei with disrupted membranes. Concomitant with these changes, acetaminophen also caused increases in serum insulin concentrations from 24 pU/ ml at 0 time to 160 &J/ml at 8 hr and increases in serum alanine aminotransferase (ALT) concentrations from 42 to 13,279 U/liter, which indicated hepatic damage. Quantitation of 3(cystein-9yl)acetaminophen adducts in hepatic 10,OOOgsupematant protein using a particle concentration fluorescence immunochemical assay indicated a positive correlation between binding and the occurrence of the hepatotoxicity consistent with what has been previously reported: however, 3-(cystein-4yl)acetaminophen protein adducts were not detected in pancreatic 10,OOOgsupematant. Immunohistochemical analysis of the liver and pancreas from acetaminophen-treated mice revealed acetaminophen-protein adducts in the centrilobular regions of the liver but not in the pancreatic islets. Doses of 100 and 200 mg/kg produced no evidence of hepatotoxicity and no increase in serum insulin; 300 mg/kg and higher doses produced both hepatotoxicity and increased serum insulin concentrations. A comparison of the time course for the increase in serum levels of ALT and insulin following a toxic dose of acetaminophen indicated that the increase in ALT preceded the increase in insulin. Thus the hepatotoxicity of acetaminophen correlates with the formation of 3-(cystein-4yl)acetaminophen protein adducts in liver, which supports the concept that this toxicity is mediated by the reactive metabolite Nacetyl-p-benzoquinone imine; however, the toxicity of acetaminophen to @cells in the pancreas is apparently not mediated by this mechanism. o 1990 Academic PESS, IIIC.

The analgesic acetaminophen (paracetamol) is generally considered to be safe at therapeutic doses; however, in higher doses it may be toxic (Boyd and Bereczky, 1966). The most commonly reported toxicity is fulminating centrilobular hepatic necrosis, which also occurs in laboratory animals (Mitchell et al., 1973a). The mechanism of the toxicity has 225

been studied extensively (for reviews see Hinson, 1980; Nelson, 1982; Black, 1984). Current evidence suggests that the hepatotoxicity is a result of conversion of the drug to the reactive arylating metabolite, N-acetyl-pbenzoquinone imine, by the cytochrome P450 mixed function oxidase system (Dahlin et al., 1984; Potter and Hinson, 1987). At therapeu004 1-008X/90 $3.00 Copyright 0 I990 by Academic F’res.s, Inc. All rights of reproduction in any form reserved.

226

FERGUSON

tic doses, this metabolite is efficiently detoxified by conjugation with glutathione; however, hepatic glutathione is depleted by larger doses and the metabolite binds to protein sulfhydryl groups (Mitchell et al.. 1973b; Hinson et al., 1982; Potter and Hinson, 1986). Binding correlates with the development of the toxicity (Jollow et al., 1973; Pumford et al., 1989). In addition to the known decrease in hepatic glutathione, it has been previously demonstrated that toxic doses of acetaminophen produce alterations in carbohydrate metabolism. Two hours following a hepatotoxic dose, hepatic glycogen was depleted by as much as 80% and this was accompanied by a marked increase in serum glucose. Serum glucose increased from 75 mg% at 0 time to approximately 250 mg% at 2.5 hr following acetaminophen (Hinson et al., 1983). Subsequently, dramatic increases in serum insulin (INS),’ which peaked at approximately 8 hr and remained elevated at 24 hr, were observed. Serum glucose levels showed an inverse correlation to serum insulin levels; by 4 hr glucose levels returned to control values and hypoglycemia occurred at later times. Twenty-four hours after acetaminophen, serum glucose in treated mice was approximately one-half that observed in control animals (Hinson et al., 1984). In the current study we used electron microscopy to examine the possibility that large doses of acetaminophen in mice may produce alterations in pancreatic p cells. In addition, we examined the possibility that 3-(cystein-s-yl)acetaminophen protein adducts may occur in pancreatic p cells of acetaminophentreated mice. MATERIALS

AND

METHODS

ET AL. was obtained from Spectrum Medical Industries, Inc. (Los Angeles, CA). Fetal calf serum was a product of GIBCO (Grand Island, NY). Protein assaykits were purchased from Bio-Rad Laboratories (Richmond. CA). Universal immunoperoxidase staining kits were supplied by Cambridge Research Laboratory (Cambridge, MA). Animals and acetaminophen administration. Male B6C3Fl mice were obtained from the NCTR pathogenfree breeding colony. Mice were maintained in a conventional animal facility, housed in clear plastic cages with hardwood bedding, and provided with Type 50 10 M laboratory chow (Ralston-Purina, St. Louis, MO). The animals were maintained in a constant temperature and humidity environment on a 12-hr light-dark cycle. Food and water were provided ad libitum throughout the experiments. Acetaminophen was diluted in pyrogen-free saline (Travenol Laboratories, Inc., Deerfield, IL) at a concentration such that the stated dose was administered at 25 &g body wt. The dose, route, and temporal relationship of dosing to termination are stated in the appropriate figure legends and tables. The initial electron microscopy studies were performed at an acetaminophen dose of 500 mgjkg. Because this dose was lethal by the intraperitoneal route used in subsequent experiments it was necessary to use 400 mg/kg as the highest dose in later experiments. For the initial electron microscopy. five groups of three animals each were given 500 mg/kg by gastric intubation and one group was euthanized at 3, 4.5.6.8. and 20 hr after treatment. A sixth group of three (controls) was given 0.5 ml saline by intubation and euthanized after 2 hr. Electron microscopy. Acetaminophen-dosed and saline control animals were anesthetized with ether and the tissues were fixed by aortic perfusion of the animals for 3 min with phosphate buffer (0.2 M, pH 7.0) containing 1% paraformaldehyde and 1.5% ghnaraldehyde, followed by 2 additional min in phosphate buffer containing 4% glutaraldehyde. Islets were teased from the pancreas, immersed overnight in phosphate buffer containing 4% glutaraldehyde, rinsed in phosphate buffer, and postfixed 1 hr in phosphate-buffered 1% osmium tetroxide. The tissues were dehydrated in a graded series of ethanol concentrations, cleared in acetone, infiltrated with acetone/Epon/Araldite, then embedded in 100% Epon/ Araldite mixture. Thin sections (600-800 A) were stained with uranyl acetate and lead and examined on a Phillips EM 20 I electron microscope at 60 kV. Clinical chemistry. Blood samples were obtained from the retroorbital Venus plexus of mice anesthetized with COZ. The blood was allowed to clot at room temperature and serum was separated from the clot with Shure Sep II

Reagents. Acetaminophen was purchased from Sigma Chemical Co. (St. Louis, MO). Spectrapor dialysis tubing

’ Abbreviations used: PBS, phosphate-buffered saline:

COZ, carbon dioxide; ALT, alanine aminotransferase; PCFIA, particle concentration fluorescence immunoassay; INS. serum insulin.

ACETAMINOPHEN-INDUCED (General Diagnostics, Morris Plains, NJ). The serum was stored at -70°C prior to analysis for alanine aminotransferase (ALT) and insulin content. Serum concentrations of ALT were determined as indices of hepatotoxicity with a Baker Encore autoanalyzer and Baker CentrifiChem ALT-optimized reagents (Baker Instrument Co., Allentown, PA) according to the method described by Bergmeyer et al. (1978). Zmmunoassuys. For the immunochemical determination of 3-(cystein-9yl)-acetaminophen protein adducts in tissues, the animals given a maximum of 400 mg/kg ip were sacrificed between 0 and 8 hr later by cervical dislocation and the liver and pancreas were removed. Hepatic tissue was homogenized with a Teflon-coated tissue grinder using a 5: 1 (v/w) ratio of 0.0 1 M sodium phosphate buffer, pH 7.2, containing 8.5% (w/v) NaCl (PBS) prior to immunochemical determination of 3-(cystein-Syl)acetaminophen protein adducts. Pancreatic tissue was homogenized in a 5: 1 (v/w) ratio of PBS using a hand-operated glass tissue grinder. The homogenates were centrifuged at 10,OOOgfor 15 min at 4°C and the supernatants were dialyzed five times against 1 liter of PBS. The dialyzed tissue samples were stored at -7o’C prior to analysis. 3-(cystein-S-yl)acetaminophen protein adducts were quantitated immunochemically using a modification of a competitive immunoassay. This particle concentration thtorescence immunoassay (PCFIA) used a polyclonal antiserum specific for S-(cystein-S-yl)acetaminophen (Roberts et al., 1987; Benson et al., 1989) and a solid-phase antigen prepared by coupling acetaminophen-modified metallothionein to aminosubstituted polystyrene beads using N-succinimidyl 3-(2pyridyldithio)propionate as a coupling reagent. Antibody bound to the solid phase was separated from unbound antibody using specially designed 96-well vacuum manifold plates, and bound antibody was detected using a fluorometer (Pandex Division, Baxter Health Care Products, Mundelein, IL) to quantitate affinity-purified fluorescein-conjugated goat anti-rabbit antisera. The PCFIA had similar limits of detection (20 pmol/mg protein) and recognized the same epitope as did the previously described enzyme-linked immunosorbant assay (Roberts et al., 1987; Potter et al., 1989). Serum insulin was determined using a competitive immunoprecipitation radioimmunoassay kit (Cambridge Medical Technology, Billerica, MA). Insulin levels in unknown samples were calculated by comparison to a standard curve prepared using serial twofold dilutions of the 300 rU/ml insulin standard. A previously described indirect immunoperoxidase technique (Roberts et al., 1989) was used to localize 3(cystein-4yl)acetaminophen adducts in liver and pancreatic tissue. This procedure involved immunohistochemical staining of S-pm-thick sections of microwave-fixed tissue. The sections were stained with the polyclonal antiserum, specific for acetaminophen pro-

PANCREATIC

DAMAGE

227

tein adducts, followed by subsequent staining with a horseradish peroxidase-labeled anti-rabbit antiserum according to the directions outlined in the immunoperoxidase staining kit. Statistical analysis. Student’s t test for two independent means, two-tailed, was used to make comparisons among levels of aminotransferases, insulin, and 3-(cysteinSyl)acetaminophen protein adducts. Differences were considered statistically significant when p d 0.05.

RESULTS Clinical appearance of the mice. Between 3 and 8 hr after acetaminophen treatment, the mice appeared increasingly ruffled, depressed, and inactive. By 20 hr they were recumbent with very shallow respiration. No deaths occurred. The saline-treated controls appeared normal when euthanized at 2 hr. Ultrastructure of pancreatic p cells. The islets from the saline-treated mice contained predominantly p cells (Fig. la) (B), some (Y cells (A), and a few 6 cells (D). The (Y cells were present primarily at the periphery of the islets whereas the /3 cells were centrally located. Clear cells were not observed in this study. The (Ycell granules were uniform in size and shape, with an electron-dense central core and a tight membrane boundary. Granules of p cells were variable in size and in the density of the central core, which was circular, angular, or occasionally crystalloid. The granule membrane was loose and created a halo effect. The 6 cells, usually located adjacent to the (Ycells, were generally larger with more cytoplasmic matrix. Their secretory granules had small, low-density central cores with partial membrane encirclement. In contrast to the islets of the saline-treated control mice, by 3 hr the islets of mice treated with toxic doses of acetaminophen contained areas of prominent intercellular spaces (edema) and a few degenerating p cells. The p cells were swollen and had ruptured cytoplasmic membranes and disrupted internal structure. The secretory granules that remained varied in size and core density. Disruption of the granular membrane reduced

228

FERGUSON

ET

AL.

ACETAMINOPHEN-INDUCED

the halo effect. Membraneous debris was present in the intercellular spaces. Other organelles were not affected at this stage. After 4.5 hr, increased numbers of p cells had disrupted cell membranes and swollen and fragmented rough endoplasmic reticula. Secretory granules had partially or completely lost their granular membranes. Membraneous debris was scattered throughout the cytoplasm while the mitochondria remained undamaged. The (Ycells appeared undamaged. After 6 hr, /3 cells had greatly increased intercellular spaces (Fig. 1b) (S) which were especially pronounced around the capillaries. Damage to the cytoplasmic and granular membranes was increased and loss of cytosol was observed. Cellular membraneous organelles and secretory granules (arrow) were present in the intercellular spaces along with the membraneous debris (m). Macrophages with numerous phagolysosomes were also noted within the islets. After 8 and 20 hr, the lesions increased in severity. The cytoplasmic and granular membrane damage and the cell swelling were greatly increased. Mitochondria in p cells were enlarged at this time. These observations are consistent with the concept that large doses of acetaminophen are toxic to the pancreatic p cells and that the previously observed increases in serum insulin following large doses of acetaminophen to mice (Hinson et al., 1984) were a result of ,f3 cell damage and release of insulin. Examination of target tissuefor acetaminophenprotein adducts. It has previously been shown that hepatotoxicity of acetaminophen correlates with covalent binding of acetaminophen to liver protein (Jollow et al., 1973; Pumford et al., 1989); thus, it seemed reasonable that pancreatic p cell toxicity might similarly correlate with binding of adduct. The

PANCREATIC

DAMAGE

229

possible covalent binding of acetaminophen to pancreatic protein was studied using a PCFIA that specifically recognizes 3-(cysteinS-yl)acetaminophen adducts (Roberts et al., 1987, submitted). As shown in Fig. 2A, 3(cystein&yl)acetaminophen protein adducts were not detected in the 10,OOOg supernatant (S- 10) fractions of pancreatic homogenates at 2.5, 5, or 8 hr after acetaminophen administration; however, serum insulin increased during this period to approximately 10 times that of controls (Fig. 2C). 3-(Cystein-Syl)acetaminophen protein adducts were detected in these same mice in the S-10 supernatant fraction of hepatic protein (Fig. 2B) and peak levels of adducts in hepatic S- 10 fraction were observed immediately before large increases in serum ALT concentrations occurred (Fig. 2D). The loss of these adducts from the hepatic S-10 supernatant fraction correlates with the appearance of 3-(cysteinSyl)acetaminophen protein adducts in the serum, presumably as a result of lysis or leakage of hepatocytes consistent with the previous observation of Pumford et al. (1989). The data in Fig. 2 suggest that the mechanism of toxicity of acetaminophen to pancreatic p cells was not metabolism of acetaminophen to a reactive metabolite and covalent binding to cysteine groups on protein in pancreas, as has been proposed for the hepatic toxicity of acetaminophen. Although the PCFIA used to quantitate 3-(cystein-S-yl)acetaminophen protein adducts in tissue homogenates was quite sensitive (detection limit of 20 pmol of bound acetaminophen/ mg protein), the possibility remained that putative 3-(cystein-S-yl)acetaminophen adducts in pancreatic p cells might not be detected because of the relatively low concentrations of the adducts in tissue homogenates.

FIG. 1. (a) Electron micrograph of pancreatic islet cells from saline control. (b) Electron micrograph of pancreatic islet from acetaminophen-treated mouse, 6 hr after 500 mg/kg orally. The (Ycells are designated A, the /3 cells B, and the 6 cells D. In b, m represents membranous debris and S represents exaggerated intercellular spaces. Note loss of cytosol and disrupted intracellular membranes. The bar represents 2 pm.

230

FERGUSON UI si

0.4-

‘PE o f

0.2-

8 q

0.0 m 0

-0.6

0.6 -

Acelaminophen I)

2

4

6

0

0

2

4

6 -16

I

150-

6

D. ALT

-

Acetaminophen

.5 5 E E’; -3

h f 3-, 0’

0.6-

2oo C. Insulin f-

-10

B. Liver

“O A. Pancreas

: c 5 5 a ;‘E z 5 5

ET AL.

f ‘Isa Fr; 3 =

loo1

.

50 -

-4 I r

k 0 0

I 2

4 Time

SaIllIe I 6

‘, 6

(hours)

0

2

4 Time

6

6

(hours)

FIG. 2. (A) Time course for the formation of immunochemically detectable 3-(cystein-$yl)acetaminophen protein adducts in pancreatic S- 10 of acetaminophen-treated mice. (B) Time course for the formation of 3-(cystein-S-yl)acetminophen protein adducts in hepatic S-IO of acetaminophen-treated mice. (C) Time course of increase in serum insulin of acetaminophen-treated mice. (D) Time course of increase in serum alanine aminotransaminase of acetaminophen-treated mice. Male B6C3FI mice were treated intraperitoneally with 400 mg/kg of acetaminophen. Each point represents the means It standard error of the mean for five animals.

However, calculations indicated that if the adducts were present in the /3 cells at a concentration equivalent to that in the liver (2 nmol/mg protein) and if B cells comprise 1% of the total pancreas, the adducts would have been detectable in the S- 10 supernatant. To rule out more completely the possible presence of 3-(cystein-9yl)acetaminophen adducts in pancreatic p cell proteins, an immunohistochemical assay capable of demonstrating the adducts at the cellular level was used. Acetaminophen protein adducts were observed in the cells of the centrilobular areas of the liver consistent with previous observations (Jollow et al., 1973; Bartolone et al., 1987; Roberts et al., submitted) but could not be demonstrated in the cells of the pancreatic

islets 2.5 hr following acetaminophen istration (data not shown).

admin-

Relationship of hepatotoxicity to increase in serum insulin. Examination of the time course of the changes in serum ALT and serum insulin following a toxic dose of acetaminophen (400 mg/kg) (Figs. 2C and 2D) indicated that the serum ALT increased before the serum insulin did. At 5 hr after acetaminophen, serum ALT was approximately 70% of the amount observed at 8 hr whereas at 5 hr serum insulin was only 38% of the value observed at 8 hr. The relationship between serum insulin and serum ALT was also evaluated as a function of acetaminophen dose. As shown in Table 1, acetaminophen caused no increase in either serum insulin or ALT at doses

ACETAMINOPHEN-INDUCED TABLE 1 DOSE-RESPONSE RELATIONSHIP FOR ACETAMINOPHEN-INDUCEDALTERATIONSIN SERUM INSULIN AND ALT Treatment’

Serum insulin ( dJ/mU

Serum ALT (U/liter)

PANCREATIC

DAMAGE

231

which is not released into the serum following the hepatotoxicity. DISCUSSION

In a previous study, we showed that large doses of acetaminophen produce dramatic alterations in serum glucose concentrations in 13f4 53 f 17 mice. Following acetaminophen administration there was an increase in serum glucose 13-c3 60? 12 from 75 mg% at 0 time to 250 mg% at 2.5 63 + 28 10,497 k 6359 hr concomitant with a 90% decrease in liver glycogen. Subsequently, serum glucose de64 + 9 14,333 + 3443 creased to approximately one-half of control values by 24 hr. Correlated with the decrease n Animals were dosed intraperitoneally with the indicated dose in 0.025 ml/g body wt and euthanized at 8 hr. in serum glucose there was an increase in seb Data represent the means -+ SE for five animals per rum insulin which was maintained at high group (except as noted). levels for the duration of the experiment (24 cThe 400 mg/kg dose resulted in 40% mortality prior hr) (Hinson et al., 1983, 1984). These findto 8 hr and the data for this group are the mean f SE for ings were of interest because similar alterthree animals. ations in serum glucose have been reported to occur in rats treated with the diabetogenic of 100 and 200 mgjkg. However, 300 and 400 agent, streptozotocin. Junod et al. (1969) m& caused elevations in both. found that at 2 hr following streptozotocin Since Grodsky and Forsham (1966) re- administration to the rat, serum glucose inported that liver contains significant levels of creased from approximately 75 mg% to insulin, we investigated the possibility that greater than 200 mg%. The animals were subthe increase in serum insulin that followed sequently hypoglycemic at 10 hr. Correlated toxic doses of acetaminophen might be with this decrease was an approximately caused by lysis of hepatocytes. In saline- eightfold increase in serum insulin, with peak treated mice total hepatic insulin in the S- 10 values at 7 hr. At 24 hr, serum insulin was fraction was determined to be 25,333 f 108 1 decreased and serum glucose increased to U/liver. In acetaminophen-treated mice (400 greater than 300 mg%. This latter increase in mg/kg; 8 hr) this value was 28,897 f 1746 U/ serum glucose was not observed in the acetliver. Acetaminophen significantly decreased aminophen-treated mice; however, the earhepatic ALT levels by 26.3%. The finding lier alterations in serum glucose and serum that acetaminophen did not significantly de- insulin were very similar to those observed crease hepatic insulin levels suggests that lysis following acetaminophen administration to of hepatocytes is not the mechanism of acet- mice (Hinson et al., 1983, 1984). These obaminophen induced increase in serum insu- servations indicate some similarities in the lin. Moreover, the finding that the insulin toxicities of streptozotocin and acetaminolevels were the same in saline- and acetaminphen. We therefore became interested in the ophen-treated hepatic S-10 fractions is con- possibility that acetaminophen may be toxic to the pancreatic p cells. sistent with the previous report that hepatic insulin is localized primarily in the microBy electron microscopy we were able to somal fraction (12~0 et al., 1979) a fraction show that hepatotoxic doses of acetaminoSaline Acetaminophen ( 100 w/kg) Acetaminophen (200w/kg) Acetaminophen (300 w/W Acetaminophen’ (400 mg/kg)

18 + 5’

84 f 30’

232

FERGUSON

phen were associated with injury to pancreatic /3 cells. The predominant initial signs of the toxicity were disruption of cytoplasmic membrane structure, perhaps associated with fluid influx. At later times, loss of granule structure was noted as well as loss of cytoplasm (Fig. 1B). Alterations to other cell types within the islets were not observed. Moreover, because the morphologic changes preceded increases in serum insulin (Fig. 2C) the data suggest that the increase in serum insulin was a result of the fl cell damage. We examined the possibility that the acetaminophen toxicity to the P cells was a result of metabolism of acetaminophen to a covalently bound adduct in the 0 cells. This mechanism seemed reasonable, given the putative mechanism of acetaminophen toxicity to the liver, and because the pancreatic toxicity of streptozotocin, an N-methylnitrosourea, is believed to be mediated by alkylation of macromolecules (Rakieten et al., 1963). Whereas the hepatotoxicity of acetaminophen correlates with the formation of covalently bound acetaminophen protein adducts in hepatic cytoplasm (Jollow et al., 1973) and specifically 3-(cystein-S-yl)acetaminophen protein adducts (Pumford et al., 1989), we saw no evidence that these specific protein adducts were formed in the pancreatic p cells (Fig. 2). Thus, acetaminophen is apparently not metabolized to N-acetyl-p-benzoquinone imine in the p cell, which binds to protein at this site. These results are consistent with previous reports that pancreatic fl cells have only low levels of drug metabolizing capability (Iqbal et al., 1977) and would not be expected to form adducts in appreciable quantities. Also, there is no evidence that the metabolite is formed in the liver and transported to bind to 8 cell protein. Moreover, N-acetyl-p-benzoquinone imine is extremely reactive and would not be expected to be transported from the liver to the pancreatic p cells (Coles et al., 1988). The toxicity of other agents to fl cells has been studied extensively (for review see

ET

AL.

Fischer, 1985). Alloxan, the first diabetogenic agent to be discovered (for review see McLetchie, 1982) is very reactive and easily reduced to dialuric acid and undergoes redox cycling leading to superoxide anion, hydrogen peroxide, and hydroxyl radical formation. Because of its reactivity, the hydroxyl radical has been implicated as a cellular oxidant and has been suggested as the mechanism of alloxan’s toxic action on /3 cells. A similar mechanism may occur in p cells following toxic doses of acetaminophen to mice; however, peroxidation of acetaminophen has not been reported in the @cells and the acetaminophen free radical does not react with molecular oxygen (Fischer et al., 1985; Potter and Hinson, 1986). Thus, it is unclear how acetaminophen could produce toxicity by mechanisms similar to those postulated to occur with alloxan. Although the increase in serum insulin correlated in time (Table 1) with the microscopic development of /3 cell damage, it seemed possible that the increase may be in part a result of the hepatic damage. Significant levels of insulin were present in the S-10 supernatant fraction of control liver homogenates but these levels did not change in livers from hepatotoxic acetaminophen-treated mice. Since total hepatic ALT levels decreased by 26% we concluded that lysis of the hepatocytes is not the source of acetaminophen-induced increase in serum insulin. Moreover, the time course for increase in serum ALT and serum insulin further suggested that this conclusion is correct. The increase in serum ALT versus the increase in serum insulin (Fig. 2) indicated that 5 hr after acetaminophen administration, serum ALT levels were approximately 70% of the levels observed at 8 hr; however, at this time serum insulin levels were 38% of the maximum observed at 8 hr. Thus, the disruption of the plasma membrane of hepatocytes occurred before the substantial increases in serum insulin. Another factor relative to the hepatotoxicity which should be considered in the acetaminophen-

ACETAMINOPHEN-INDUCED

induced increase in serum insulin is the potential decrease in insulin degradation. The liver is the major site for clearance of insulin from plasma (Izzo et al., 1979; Muir et al,, 1986; Duckworth et al., 1987), and major alterations of hepatic function may significantly affect insulin degradation and contribute to increased concentrations of insulin in the blood. The clinical importance of acetaminophen’s toxicity to pancreatic fi cells and the related increase in serum insulin is unclear. Thomson and Prescott (1966) reported on a patient with acetaminophen overdose who developed liver damage and had impaired glucose tolerance following the overdose. Subsequently, the patient recovered and over a period of weeks the glucose tolerance reverted to normal. Unfortunately, since other studies have not examined glucose tolerance in acetaminonhen overdose victims. it is not known if this‘ is a general effect. Additional studies are needed to clarify any clinical importance as well as the mechanism. ACKNOWLEDGMENTS The authors thank Dr. James Crowell, Pathology Associates, Inc., for clinical chemistry analyses, Allan Warb&ton, Pathology Associates, Inc., for immunohistochemical staining of tissues, and Cindy Hartwick for her help in preparing this manuscript.

REFERENCES BARTOLONE, J. B., SPARKS, K., COHEN, S. D., AND KHAIRALLAH, E. A. (1987). lmmunochemical detection of acetaminophen-bound liver proteins. Biochem. Pharmacol. 36,1193-l 197. BENSON, R. W., PLJMFORD,N. R., MCRAE, T. A., HINSON,J. A., AND ROBERTS, D. W. ( 1989). Development of a particle concentration fluorescence immunoassay for 3-(cystein-Syl)acetaminophen adducts. Toxicologist 9,47.

BERGERMEYER, H. U., SCHEIBE, P., AND WAHLEFELD, A. W. (1978). Optimization of methods for aspartic aminotransferase and alanine aminotransferase. Clin. Chem.

4,58-Z.

PANCREATIC

DAMAGE

BLACK, M. (1984). Acetaminophen Annu.

Rev. Med.

233 hepatotoxicity.

35,577-593.

BOYD, E. M., AND BERECZKY, G. M. (1966). Liver necrosis from paraectamol. Brit. J. Pharmacol. 26,606607. COLES, B., WILSON, I., WARDMAN, P., HINSON, J. A., NELSON, S. D., AND KEX~ERER, B. (1988). The spontaneous and enzymatic reaction of N-acetyl-pbenzoquinone imine with glutathione: A stopped-flow study. Arch. Biochem. Biophys. 264,253-260. DAHLIN, D. C., MIWA, G. T., Lu, A. Y. H., AND NELSON, S. D. (1984). N-Acetyl-p-benzoquinone imine: A cytochrome P-450 mediated oxidation product of acetaminophen. Proc. Natl. Acad. Sci. USA 81, 13271331. DUCKWORTH, W. C., HAMEL, F. G., LIEPNIEKS, J. J., PEAVY, D. E., RYAN, M. P., HERMODSON, M. A., AND FRANK, B. H. (1987). Identification of A chain cleavage sites in intact insulin produced by insulin protease and isolated hepatocytes. Biochem. Biophys. Res. Commun. 147,6 15-621. FISCHER,L. J. ( I98 5). Drugs and chemicals that produce diabetes. Trends Pharmacol. Sci. 6,72-75. FISCHER, V., WEST, P. R., NELSON, S. D., HARVISON, P. J., AND MASON, R. P. (1985). Formation of 4-aminophenoxy free radical from the acetaminophen metabolite N-acetyl-pbenzoquinone imine. J. Biol. Chem. 260,11,446-l 1,450. GRODSKY, G. M., ANDFORSHAM, P. M. (1966). Insulin and the pancreas. Annu. Rev. Physiol. 28,347-380. HINSON, J. A. (1980). Biochemical toxicology of acetaminophen. Rev. Biochem. Toxicol. 2,103-129. HINSON, J. A., HAN-HSU, H., MAYS, J. B., HOLT, S. J., MCLEAN, P., AND LETTERER, B. (1984). Acetaminophen-induced alterations in blood glucose and blood insulin levels in mice. Res. Comm. Chem. Path. Pharmacol. 43,38 l-39 1. HINSON, J. A., MAYS, J. B., AND CAMERON, A. M. ( 1983). Acetaminophen-induced hepatic glycogen depletion and hyperglycemia in mice. Biochem. Pharmacol. 32, 1979-1988. HINSON, J. A., MONKS, T. J., HONG, M., HIGHET, R. J., AND POHL, L. R. (1982). 3-(Glutathion-Syl)acetaminophen: A biliary metabolite of acetaminophen. Drug Metab.

Dispos.

10,47-50.

IQBAL, Z. M., VARNES, M. E., YSHIDA, A., AND EPSTEIN, S. S. (1977). Metabolism of benzo(a)pyrene by guinea pig pancreatic microsomes. Cancer Res. 37, 101 l1015. lzzo, J. L., RONCONE, A. M., HELTON, D. L., AND lzzo, M. J. (1979). Subcellular distribution of intraportally injected ‘Z51-labeled insulin in rat liver. Arch. Biochem. Biophys.

198,97-109.

JOLLOW, D. J., MITCHELL, J. R., POTTER, W. Z., DAVIS, D. C., GILLETTE, J. R., AND BRODIE, B. B. (1973). Acetaminophen-induced hepatic necrosis. Il. Role of

234

FERGUSON

covalent binding in vivo. J. Pharmacoi.

E.xp.

Ther.

187,195-202.

JUNOD, A.. LAMBERT, A. E., STAUFFACHER, W., AND RENOLD, A. E. (1969). Diabetogenic action of streptozotocin: Relationship ofdose to metabolic response. J. Clin. Invest. 48,2 129-2 139. MCLETCHIE, N. G. B. (1982). Alloxan diabetes: The sorcerer and his apprentice. Diabetologia 23,12-15. MITCHELL, J. R., JOLLOW, D. J.. POTTER, W. Z., DAVIS, D. C., GILLETTE, J. R., AND BRODIE, B. B. (1973a). Acetaminophen-induced hepatic necrosis. I. Role of drug metabolism. J. Pharmacol. Exp. Ther. 187, 185194. MITCHELL, J. R., JOLLOW, D. J.. POTTER, W. Z., GILLETTE, J. R., AND BRODIE, B. B. (1973b). Acetaminophen-induced hepatic necrosis. IV. Protective role of glutathione. J. Pharmacol. Exp. Ther. 187, 21 I-217. MUIR, A., OFFORD, R. E., AND DAVIES, J. G. (1986). The identification of a major product of the degradation of insulin by ‘insulin proteinase’ (EC 3.4.22.1 I). Biochem.

J. 237,631-637.

NELSON, S. D. (1982). Metabolic activation and drug toxicity. J. Med. Chem. 25,153-165. POTTER, D. W., AND HINSON, J. A. ( 1986). Reactions of N-acetyl-pbenzoquinone imine with reduced glutathione, acetaminophen, and NADPH. Mol. Pharmacol. 30,3 3-4 I.

POTTER, D. W.. AND HINSON, J. A. (1987). Mechanisms of acetaminophen oxidation to N-acetyl-pbenzoqui-

ET AL.

none imine by horseradish peroxidase and cytochrome P-450. J. Biol. Chem. 262,966-973. POTTER, D. W., PUMFORD. N. R., HINSON, J. A., BENSON, R. W., AND ROBERTS, D. W. (1989). Epitope characterization of acetaminophen bound to protein and nonprotein sulfhydryl groups by an enzyme linked immunosorbant assay. J. Pharmacol. Exp. Ther. 248,186-190.

PUMFORD, N. R., HINSON, J. A., POTTER, D. W., RowLAND, K. L.. BENSON, R. W., AND ROBERTS, D. W. (1989). Immunochemical quantitation of 3-(cysteinSyl)-acetaminophen adducts in serum and liver proteins of acetaminophen-treated mice. J. Pharmacol. Exp. Ther. 248, 190-l 96. RAKIETEN, N., RAKIETEN, M. L., AND NADKARNI, M. V. (1963). Studies on the diabetogenic action of streptozotocin. Cancer Chemother. Rep. 29,9 I-98. ROBERTS. D. W., Buccr. T. J., BENSON, R. W.. WARBRITTON,A. R., McRAE,T. A., PUMFORD. N. R., AND HINSON, J. A. Immunohistochemical localization and quantitation of 3-(cystein-9yl)acetaminophen protein adducts in acetaminophen toxicity. Submitted for publication. ROBERTS, D. W., PUMFORD, N. R.. POTTER. D. W., BENSON,R. W., AND HINSON, J. A. (1987). A sensitive immunochemical assay for acetaminophen-protein adducts. J. Pharmacol. Exp. Ther. 241,527-533. THOMSON, J. S., AND PRESCOTT, L. F. (1966). Liver damage and impaired glucose tolerance after paracetamol overdosage. Brit. Med. J. 2,506-507.