TOXICOLOGY

AND

APPLIED

PHARMACOLOGY

113, 19-29

( 1992)

Purification, Antibody Production, and Partial Amino Acid Sequence of the 58-kDa Acetaminophen-Binding Liver Proteins JOHN B. BARTOLoNE,*,tY’ RAYMOND B. BIRGE,*~t’2 STEVEN J. BULERA,* MARY K. BRUNO,* ERVANT V. NISHANIAN,* STEVEN D. COHEN,? AND EDWARD A. KH~IRALLAH*~~ *Department of Molecular and Cell Biology and tDepartment of Pharmacology and Toxicology, The University of Connecticut, Storrs. Connecticut 06269-312.5

ReceivedAugust 1, 199I ; acceptedNovember 7, 199 1 While considered safe at therapeutic doses, the widely used analgesic acetaminophen (4-hydroxyacetanilide, N-acetyl-p aminophenol, paracetamol, APAP) can in overdose result in severe centrilobular liver necrosis (Boyd and Bereczky, 1966; Mitchell et al., 1973). The hepatotoxicity is thought to result from the cytochrome P-450-mediated oxidation of Immunochemical analysis of electrophoretically resolved liver APAP to an electrophilic intermediate, N-acetyl-pbenzoproteins from mice administered hepatotoxic doses of acet- quinonimine (NAPQI) (Dahlin et al., 1984). NAPQI is iniaminophen has identified two proteins of 44 and 58 kDa as tially detoxified by conjugation to reduced glutatbione (Potter major targets for acetaminophen arylation. In the present study et al., 1974). However, following an acute overdose, cellular the 58kDa acetaminophen-binding protein (58-ABP) was pu- reserves of glutathione become sufficiently depleted (Jollow rified from mouse liver cytosol by gel permeation chromatography, preparative isoelectric focusing, and polyacrylamide gel et al., 1973; Mitchell et al., 1974) to allow covalent binding electrophoresis. The acetaminophen adducts were visualized on of NAPQI to hepatic proteins with resulting toxicity. Despite considerable research, the exact events responsible immunoblots using affinity-purified anti-acetaminophen antihepatotoxicity remain unknown. Albodies after each step of the purification. Gel permeation chro- for APAP-induced matography, under nondenaturing conditions, indicated that the though lipid peroxidation (Wendel et al., 1979), protein sulfprotein is a monomer. Two-dimensional gel electrophoresis hydryl oxidation (Moore et al., 1985), and alterations in celdemonstrated that the 58-ABP consists of a cluster of four im- lular calcium levels (Corcoran et al., 1988; Orrenius et al., munochemically reactive isoforms with isoelectric points ranging 1989; Tsokos-Kuhn, 1989; Boobis et al,, 1990) have been from 6.2 to 6.6. V-8 protease digestion of the isoforms suggested postulated as mechanisms leading to the hepatotoxicity, the that they contained similar peptide fragments. The purified 58ABP was utilized to produce polyclonal antibodies and to de- most widely accepted hypothesis supports the contention that termine the amino acid composition and partial sequence of the the covalent binding of APAP metabolites to critical protein targets is associated with the initiation of deleterious processes protein. These antibodies revealed a protein cluster of similar molecular weight and isoelectric points in the cytosol of a human which culminate in cell death (Mitchell et al., 1973; Jollow liver specimen. Amino acid analysis of the purified protein in- et al., 1973; Hinson, 1980). Although no clear mechanistic dicated that it contains eight cysteine residues (about 1.4% by linkages between the covalent binding of APAP to the reweight). This low cysteine content raises the possibility that at sulting liver toxicity have been established, there have been hepatotoxic doses acetaminophen may also bind to nonthiol sites no reports of APAP-induced cell damage in the absence of on the protein. The amino acid sequence of two cyanogen bro- adduct formation. Hence an important objective in submide/tryptic peptide fragments revealed that the major immustantiating whether covalent binding may mediate the cynochemically detectable aceteminophen target in the cytosol is totoxic effects of APAP requires the identification and charhomologous to a selenium-binding protein which has been reacterization of specific protein adducts which are generated cently sequenced. 0 1992 Academic PW, IIIC. following an acute overdose (Nelson and Pearson, 1990). Immunochemical analysis with affinity-purified antiAPAP antibodies has demonstrated that the covalent binding ’ Present address: Richardson-Vicks, Inc., 1 Far Mill Crossing, Shelton, of APAP to hepatic proteins in mice was highly selective to CTO6484. proteins of 44 and 58 kDa with lesser binding to proteins of ’ Present address: Department of Molecular Oncology, The Rockefeller 130 and 33 kDa (Bartolone et al,, 1987, 1988; Birge et al., University, 1230 York Ave., New York, NY. 10021-6399. 3 To whom all correspondence should be addressed. 1989). Good correlations between covalent binding to the

Purification, Antibody Production, and Partial Amino Acid Sequence of the 58kDa Acetaminophen-Binding Proteins. BARTOLONE,J.B., BIRGE,R.B., BULERA,~. J., BRUNO,M.K., NISHANIAN, E. V., COHEN, S. D., AND KHAIRALLAH, E. A. (1992). Tuxicol. Appl. Pharmacol. 113, 19-29.

19

0041-008X/92 $3.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

20

BARTOLONE

58-kDa acetaminophen-binding protein (58-ABP) and the resulting cellular damage have been reported in mice (Birge et al., 1989; Bartolone et al., 1989a; Beierschmitt et al., 1989; Brady et al., 1990) and in humans (Birge et al., 1990). Others have confirmed that the majority of the immunochemically detectable 3-(cystein-S-yl)acetaminophen adducts present in mouse liver were associated with a cytosolic protein of similar molecular weight (Pumford et al., 1990). As an initial step toward a better understanding of the cellular consequences of the covalent binding of APAP this paper describes the purification, partial amino acid sequence, and production of antibodies against this major APAP target. MATERIALS

AND METHODS

Reagents Ultrapure electrophoretic grade Tris, glycine, acrylamide, and NY-methylene-bis-acrylamide were obtained from ICN BiomedicaIs, Inc. (Costa Mesa, CA). Sodium dodecyl sulfate and trypsin were obtained from BoehringerMannheim Biochemicals (Indianapolis, IN). Cyanogen bromide was purchased from Aldrich Chemical (Milwaukee, WI). ‘2SI-conjugated goat antirabbit IgG was purchased from DuPont New England Nuclear (Boston, MA). Nitrocellulose membranes (0.2 pm) were obtained from Schleicher and Schuell (Keene, NH). Silver staining kits were purchased from Bio-Rad Laboratories (Richmond, CA). Purified 3-(cystein-S-yl)acetaminophen was kindly provided by Dr. Sidney Nelson, University of Washington. All other biochemicals and reagents were purchased from Sigma Chemical Co. (St. Louis, MO). Animals and Treatment Protocol Three-month-old male Crl:CD-1 mice (Charles River Laboratories, Wilmington, MA) were maintained in stainless steel cages in temperature-controlled quarters with a 12-hr light/dark cycle. Food (Purina Rodent Chow 5001, Ralston Purina Company, St Louis, MO) and water were provided ad libitum. After an overnight fast the mice were dosed with 600 mg acetaminophen/kg, p.o., in 50% propylene glycol/water (injection volume, 10 ml/kg) and killed by cervical dislocation 4 hr later. This drug treatment has previously been reported to result in significant protein binding by 4 hr with subsequent elevation of plasma sorbitol dehydrogenase and progression to hepatic centrilobular necrosis (Ginsberg et al., 1982; Placke et al., 1987). Control animals were given vehicle only. Tissue Samples To ensure that not all the 58-ABP to be isolated was arylated by APAP, livers from 3 control mice were added to a pool from 12 APAP-treated mice. Livers were perfused via the portal vein with ice-cold 10 mM phosphate buffer, pH 7.4, containing 0.9% NaCl (PBS), and homogenized (1:4) in a buffer containing 0.25 M sucrose, 10 mM Tris-HCl, and 1 mM MgC12 (pH 7.4; STM buffer). Since we have previously shown that the 58-ABP is primarily a cytosolic protein, liver homogenates were centrifuged for 60 min at 105,000 g and the supematants were utilized for subsequent purification of the protein. To determine whether a similar protein exists in humans, a liver sample from a 17-year-old Caucasian male was obtained from the National Disease Research Interchange (Philadelphia, PA). Sections of liver had been removed and frozen at -80°C within 1 hr after death. Approximately 1 g of tissue was homogenized I:9 (w/v) in STM buffer and the cytosolic fraction was utilized for two-dimensional electrophoretic resolution and immunochemical analysis as described below.

ET AL. Pol.vacrylamide Gel Electrophoresis For one-dimensional SDS-PAGE the tissue samples were diluted 1:1 with sample buffer containing 2% SDS, 5% mercaptoethanol, 20% glycerol, and 0.025% bromophenol blue in 100 mM Tris-HCl at pH 7.0, heated at 100°C for 5 min, and run at 30 pg of protein per lane. Protein content was determined by the method of Lowry et al. ( 195 1) using bovine serum albumin (BSA) as a standard. Fractions obtained after gel permeation chromatography or isoelectric focusing were diluted 3: 1 (v/v) in 4X sample buffer and heated and 40-~1 aliquots were analyzed per lane. The proteins were resolved according to molecular weight on discontinuous 10% SDS-PAGE (Laemmli, 1970). Two-dimensional gels were run according to O’Farrell(l975) except that 0.1% SDS was added to the isoelectric focusing sample buffer to enhance resolution in the first dimension. Duplicate gels were routinely run to assess protein purification and immunochemical specificity. One-dimensional gels were stained with 0.1% Coomassie brillant blue in 50% methanol and 3% acetic acid and destained with 25% methanol and 10% acetic acid; twodimensional gels were silver stained (Monissey, 198 1). Western Blotting and Immunostaining Proteins resolved by one- or two-dimensional polyacrylamide gels were transblotted (Bio-Rad Trans-Blot Cell) onto 0.2-pm nitrocellulose membranes (Towbin et al., 1979) at 80 V for 6 hr in 25 mM Tris/ 192 mM glycine buffer (pH 8.3) and 20% methanol. After blotting, the nitrocellulose membranes were rinsed in 15 mM T&buffered saline (TBS) at pH 7.4 containing 0.05% Tween 20 (TBS-Tween) for 1 hr to remove excess SDS and then blocked overnight at 4°C in TBS containing 3% BSA. For detection of APAPbound proteins, membranes were incubated with affinity-purified anti-APAP or anti-58-ABP antibodies for 2-3 hr at room temperature (Bartolone et al., 1988) and, after three 5-min washes in TBS-Tween, incubated for 60 min with ‘251-conjugated goat anti-rabbit IgG (10 &I in 100 ml TBS-BSA). Following immunostaining, the nitrocellulose membranes were again washed extensively and air dried and the immunoreactive proteins were visualized after exposure to Kodak XAR-5 film for 8-24 hr at -70°C. Anal.vtical Gel Permeation Chromatography of APAP-Bound Proteins Molecular weights of the major cytosolic APAP-binding proteins were estimated under nondenaturing conditions using Sephadex G-200 superfine gel permeation chromatography. The column (1.5 X 30 cm) was maintained at 4°C and equilibrated with PBS containing 1 mM EDTA. P-Amylase (200 kDa), alcohol dehydrogenase (150 kDa), bovine serum albumin (67 kDa), and ovalbumin (43 kDa) were used as molecular weight standards. Approximately 2.25 mg of cytosolic protein was applied to the column and eluted at a flow rate of approximately 1.0 ml/hr. Protein elution was monitored at 280 nm using a Pharrnacia UV-2 flow spectrophotometer (Pharmacia. Piscataway, NJ) and fractions collected every 15 min were analyzed immunochemically for the presence of APAP-bound proteins. The native molecular weights of the 58- and 44-kDa APAP-bound proteins were estimated from a plot of their average elution volume (K,,) versus the logarithm of the molecular weights of the protein standards. Purification of the 58-ABP (I) Preparative gel permeation clrromatograplry. For preparative scale purification of the 58-ABP, a Sephadex G-150 supertine column (5 X 40 cm) was utilized. Approximately 2 g of cytosolic protein was loaded onto a column previously equilibrated with 50 mM ammonium bicarbonate (pH 7.4) 0.1 mM dithiothreitol, and 1 mM EDTA at 4°C. The proteins were eluted at a flow rate of approximately 0.4 ml/min and the outflow was monitored at 280 nm. Fractions were collected at 12-min intervals and analyzed for APAP-bound proteins by one-dimensional SDS-PAGE electrophoresis followed by Western blotting using anti-APAP antibodies. Fractions enriched in the 58-ABP were pooled and dialyzed for 24 hr at 4°C against three l-liter changes of 10 mM Tris-HCl (pH 7.0).

ISOLATION

OF THE 58-kDa ACETAMINOPHEN-BINDING

(2) Preparative isoelectric foeusing Following dialysis, the pooled Sephadex G- 150 fractions containing the 58-ABP were mixed with 2% ampholytes (80% pH 5-7 and 20% pH 3-10) and urea to yield a 4 M final concentration and then subjected to preparative isoelectric focusing (IEF) using a Rotofor apparatus (Bio-Rad Laboratories). This protein mixture (60 ml) was electrofocused at 4°C at a constant 12 W until 2000 V was reached (usually within 4 hr). All IEF fractions were kept at -70°C until analyzed immunochemically. Those IEF fractions containing the 58-ABP were then further resolved using two-dimensional PAGE and either immunostained with anti-APAP antibodies or silver stained to determine the purity of each fraction. (3) Excision of the 58-ABP from polyacrylamide gels The IEF fraction enriched with the most 58-ABP was concentrated approximately fivefold using a Centricon microconcentrator with a IO-kDa cutoff filter and then resolved on 10% SDS-polyacrylamide gels. After electrophoresis the gels were briefly stained for 5 min with 0.1% Coomassie brillant blue in 50% methanol and 1% acetic acid, destained in 50% methanol for 30 min, and then placed in HPLC-grade water for 10 min. Each lane contained approximately 0.5- 1.Opg of the 58-ABP as visually estimated from the Coomassie blue staining of standards containing known amounts of proteins. The 58kDa stained bands were then excised with a razor blade from each slab gel and either utilized directly to produce polyclonal antibodies or electroeluted to provide samples for amino acid analysis and peptide sequencing. (4) Electroelution of the 58-ABP from polyac~lamide gel slices Gel slices containing the 58-ABP were placed into sterile 0.5-ml microfuge tubes and a hole was punched into the top and bottom of each tube using a 22gauge needle. The tubes were inserted into Centricon microconcentrators with a IO-kDa cutoff filter and placed into a Centrilutor (Amicon, Danvers, MA). The reservoir was filled with a buffer (pH 7.8) consisting of 25 mM Tris base, 0.2 M glycine, 0.45 mM EDTA, and 0.05% SDS. Electroelution at 100-400 V was continued until all visible Coomassie blue staining was removed from the acrylamide gel slices (about 5 hr). The eluted proteins were assayed for purity after SDS-PAGE by silver staining and for APAP binding by reactivity with anti-APAP antibodies.

.4nti-5%ABP Antibody Production Female New Zealand white rabbits weighing about 3 kg were injected subcutaneously with emulsified gel slices containing approximately 50 pg of the purified 58-ABP in 3 ml PBS and complete Freund’s adjuvant (SO: 50, v/v). Monthly booster injections consisted of a suspension of the 58ABP emulsified in an equal volume of Freund’s incomplete adjuvant. Blood was collected from the central artery of the ear and stored in l-ml aliquots at -70°C. After the second booster, the sera tested positive by ELISA against the purified antigen. The specificity of the antibodies against the 58-ABP was tested against cytosolic proteins from control and APAP-treated mice at serum dilutions up to 1:20,000. A dilution of 1:10,000 was found to be optimal for most applications. To assesswhether the polyclonal antibodies could detect both native and APAP-conjugated 58-ABP, cytosolic proteins were resolved by two-dimensional SDS-PAGE, blotted, and detected with ‘zsIconjugated goat anti-rabbit IgG. Negative controls consisted of incubating the nitrocellulose membranes with the same dilution of preimmune rabbit serum. Characterization of the Purified 58-kDa Acetaminophen-Binding

Protein

(I) V-8 protease digestion of the 58-ABP isoforms To determine if the four isoforms of the 58-ABP were structurally related, each protein was excised from the two-dimensional gels of the Rotofor fraction most highly enriched with the protein (fraction 3) and digested with V-8 protease as described by Cleveland et al. (1977). Pools of each of the isoforms from four separate gels were placed over a 5% stacking and a 14% resoiving SDSPAGE gel and then overlayed with sample buffer containing 125 mM TrisHCl (pH 6.8). 0.1% SDS, and 1 mM EDTA. After 15 min, the reservoirs

PROTEIN

21

were filled with running buffer and the equilibrated slices were sequentially overlayed with 20 ~1 of sample buffer containing 20% glycerol and 0.002% bromophenol blue and then with 10 ~1 of sample buffer containing 10% glycerol and 0.8 PLgof V-8 protease. The gels were run at 10 mA until the bromophenol dye front reached the interface between the stacking and resolving gel and the power was shut off for 15 min to allow time for the protease digestion to occur. Electrophoresis was then resumed until the dye front was 1 cm from the bottom of the gel. After electrophoresis, the peptides were transblotted onto nitrocellulose (0.2 pm) for 8 hr at 10 V and then for an additional 30 min at 40 V. Gels were either silver stained to detect the peptides resolved from each protein or blotted to nitrocellulose and probed with the anti-58-ABP antibody to determine which peptide contained recognizable 58-ABP epitopes. (2) Amino acid analysis Following electroelution, the purified 58-ABP was concentrated approximately fivefold using a Centricon microconcentrator with a IO-kDa M, cutoff filter and then dialyzed against three l-liter changes of 0.05% SDS containing 2.0 mrvrammonium bicarbonate to remove the Tris-glycine salts. The dialyzed 58-ABP was transferred to a 1.5-ml Eppendorf microfuge tube, frozen using an ethanol-dry ice bath, and lyophilized to complete dryness in a Savant Speedvac (New York, NY). Residual SDS in the dried pellet was removed by adding 50 ~1 of HPLC-grade water and 450 ~1 of ice-cold acetone for 16 hr at -20 C. After centrifugation at 12,OOOg,the acetone-containing supematant was discarded and the washed pellet was dissolved in 50 ~1 of 70% formic acid. Five microliters of the sample was utilized for determination of protein content and amino acid analysis while the remainder was stored at -70°C for peptide sequence analysis. Analysis of the amino acid composition was conducted using postcolumn derivatization with ninhydrin on a Beckman 7300 Amino Acid Analyzer. The purified protein was hydrolyzed at 115°C for 16 hr in 100 ~1 of 6 N HCl containing 0.2% phenol with 4 nmol norleucine added as an internal standard. The sample was dried and redissolved in sample diluent buffer containing 4 nmol homoserine as an internal standard for the loading procedure. To determine the cysteine content of the 58-ABP, 7 pg of purified protein was treated with 3% performic acid and incubated at 4°C for 16 hr. The sample was then dried and rehydrated in water prior to acid hydrolysis and amino acid analysis. In an effort to determine the extent of APAP binding to the cysteinyl residues of the purified protein, an aliquot of a chemically synthesized 3-(cystein&yl)acetaminophen standard was also treated with performic acid and analyzed as described above. (3) Peptide isolation and partial sequencing of the 58 ABP Approximately 13.3 rg of the purified 58-ABP was digested with cyanogen bromide and trypsin (Stone et al., 1989). Since the amino acid analysis had indicated that the purified protein contained approximately 10 methionine residues, 4.2 ~1 of 0.5 M cyanogen bromide in acetonitrile (about a IOOO-fold excess) was added to the protein sample dissolved in 70% formic acid. After incubation at room temperature for 24 hr in the dark, the sample was dried in a Speedvac, redissolved in 50 ~1 of 8 M urea and 0.4 M ammonium bicarbonate, and incubated with 5 ~1 of 45 mM DTT for 15 min at 50°C to reduce cystine residues. The resulting cysteine residues were carboxymethylated by incubation with 5 ~1 of 100 mM iodoacetamide in the dark for 20 min. The protein sample was then diluted with 147 pl of HPLC-grade water and further digested with 2.7 ~1 of a 0.1 mg/ml solution of trypsin for 24 hr at 37°C. The cyanogen bromide/tryptic digest was then injected onto a narrow bore (2.1 mm X 25 cm) Vydac Cl8 reverse-phase column (Separations Group, Hysperia, CA) with a 5-pm support bed and analyzed on a Waters HPLC system (Model 72 1). The peptide fragments were eluted using a discontinuous gradient consisting of a mixture of 0.06% trifluoroacetic acid (TFA) in water (solvent A) and 0.052% TFA in 80% acetonitrile (solvent B) at a flow rate of 0.7 ml/min. The gradient consisted of solvent A containing (1) 2-37.5% B for 60 min. (2) 37.5-75% B for the next 30 min, and (3) 75-

22

BARTOLONE (native

Mr)

VOID

200

kd

I

2

4

6

6

ET AL.

150

kd

67

I

10

12

14

16

16

20

22

kd

43

I

24

26

fraction

26

30

32

34

36

kd I

36

40

42

44

-558

kd

-44

kd

46

number

FIG. 1. Western blot analyses of liver cytosol containing APAP-bound proteins eluted from Sephadex G-200. Livers were collected from mice killed 4 hr after APAP administration and cytosolic proteins (2.25 mg) were subjected to Sephadex G-200 gel permeation chromatography. Aliquots of the fractions were resolved by 1D SDS-PAGE and transblotted to nitrocellulose for immunochemical detection with anti-APAP antibodies. The position of the chromatoeraohic - - migration of known protein standards is indicated above the Western blot; the electrophoretic migrations of the 44- and 5%kDa APAP-binding proteins are indicated on the right side of the figure.

98% B for the next 15 min. Elution was monitored at 210 nm and the peptides were collected by peak using an Isco Model 2 150 Fraction Collector. Two well-resolved peptide fragments (5 1 and 60) were sequenced on an Applied Biosystems Model 471 Amino Acid Sequencer with each amino acid identified as the phenylthiohydantoin (PTH) derivative using an online Applied Biosystems Model 120 A HPLC system (Stone et al.. 1989).

RESULTS

period of time (Fig. 3). Immunochemical analysis with anti-APAP antibodies indicated that the 5%ABP is primarily detectable in fractions 146- 162 (Fig. 3, center). Figure 3 (bottom) suggests that the faint APAP adducts of similar molecular weight that were detected in fractions 102- 118 are a different protein since they do not crossreact with the anti-58-ABP antibodies. Isoelectric focusing of the Sephadex fractions enriched with the 58-ABP re-

Molecular Weight Estimates of the Native APAP-Bound Proteins Cytosolic proteins from APAP-treated mice were fractionated under nondenaturing conditions by gel permeation chromatography using Sephadex G-200. APAP-bound proteins were detected immunochemically with anti-APAP antibodies. Consistent with our previous findings two protein bands of 44 and 58 kDa accounted for the majority of APAP binding with minor protein adducts also detected (Fig. 1). The relationship between the average elution volume (&J and the logarithm of the native molecular weights of the protein standards was established (Fig. 2). From this relationship the native 58-ABP is estimated to be a monomer with a molecular weight similar to its apparent molecular mass on SDS gels. By contrast, the 44-kDa APAP protein adduct appears to be a component of a much larger protein complex. By extrapolation, one can estimate that the 44kDa adduct is a component of a multimeric protein complex with a molecular mass of about 350 kDa.

Purification of the 5%ABPfrom Mouse Liver A preparative Sephadex G- 150 column permitted good resolution of large amounts of cytosolic proteins in a short

0.7

Kav

0.6 0.5 0.4 alcohol

dehydrogenase

0.3 ‘\ 44-AEP

0

I

10,000

I

‘\

‘\ ‘1

J

II/l4l

100,000 Molecular

Weight

(Daltons)

FIG. 2. Sephadex G-200 calibration curve. Molecular weight protein standards included P-amylase (200 kDa), alcohol dehydrogenase ( 150 kDa), bovine serum albumin (67 kDa), and ovalbumin (43 kDa). The molecular weights of the 5%ABP and the 44-ABP were estimated from the plot of their average elution volume (K,,) versus the logarithm of the molecular weights (MJ of the protein standards.

23

58

ABP

29 24

58 ABP

44 ABP

58

ABP

FIG. 3. Elution of APAP-bound proteins on preparative Sephadex G-l 50. Two grams of cytosolic protein was subjected to preparative Sephadex G150 gel permeation chromatography. Aliquots of the fractions were resolved by 1D SDS-PAGE and transblotted to nitrocellulose for immunochemical detection of APAP-bound proteins. (Top) The Coomassie protein staining profile. (Center and bottom) Identical Western blots probed with anti-APAP and anti-58-ABP antibodies, respectively. The electrophoretic migration of the 5%ABP is indicated on the right.

vealed that the protein was primarily concentrated in the three IEF fractions (Fig. 4A). Western blotting with the anti-APAP antibody confirmed that the major 58-kDa protein in these fractions contained bound APAP (Fig. 4B). When the best-resolved IEF fraction (3) was electroeluted and subjected to one-dimensional electrophoresis, a single band of 58 kDa predominated when the gels were either silver stained or immunostained with antiAPAP antibodies (Fig. 4C). After two-dimensional PAGE, fraction 3 was seen to contain four clearly resolvable 5% kDa proteins with isoelectric points ranging between 6.2 and 6.6 (Fig. 5). The relative abundance of the two more basic proteins (Fig. 5B) appeared to correlate with the extent of immunoreactivity with the anti-APAP antibody

(Fig. 5C). Conversely, the protein with the most acidic isoelectric point was least abundant and was least detectable when probed with the anti-APAP antibody. However, the peptide maps of the V-8 protease digests of all four 58-kDa proteins were virtually identical (Fig. 6). Since these 5%kDa proteins appeared to contain very similar peptide fragments, all four 58-ABP isoforms were pooled and utilized for polyclonal antibody production, amino acid analysis, and peptide sequencing. Anti-58ABP

Antibodies

Antibodies that recognize the 58-ABP by ELISA were obtained within 2 months after the first immunization. A 1: 10,000 dilution of the immune sera was found to be optimal

24

BARTOLONE

ET AL.

Amino Acid Composition Mr

IEF 4BP

.,

w

,5a

I-L

Western B

Electroe C

ttion

FIG. 4. Preparative isoelectric focusing of the 5%ABP. Sephadex fractions ( 146 162) were further purified by isoelectric focusing in a ROTOFOR preparative isoelectrofocusing device. The figure depicts IEF fractions 2, 3, and 4, which were stained with Coomassie blue (A) or immunostained with anti-APAP antibodies (B). Protein and immunostaining of IEF fraction 3 following electroelution are depicted in (C).

The results of the amino acid analysis indicated that the 58-ABP contained approximately 550 amino acids (Table 1). Compared to the average amino acid content of 207 unrelated proteins &lapper, 1977), the amino acid composition of the 58-kDa protein suggested a high content of glycine (16.5%) and proline (6.5%) and a relatively low content of alanine (4.5%) and cysteine (1.4%). It was not possible to determine the number of APAP molecules bound to the hydrolyzed 58-ABP since the cysteinyl adducts released upon acid hydrolysis were below the limits of detection of the Amino Acid Analyzer. However, it was not likely that the cysteine content of the protein was underestimated due to APAP binding since per-formic acid treatment of chemically -

IEF

for detection of the protein on Western blots. Immunoblots of the Sephadex G-150 fractions probed with the anti-58 ABP antibodies confirmed that the protein adduct was monomeric; no other Sephadex fraction containing higher molecular weight proteins contained 58-kDa subunits that were immunoreactive with the anti-58ABP antibody (Fig. 3, bottom). Immunochemical analysis of hepatic cytosol from APAP-treated as well as control mice revealed that a 58-kDa protein was the only one immunostained in either sample (Fig. 7). Arylation did not appear to shift the migration of any of the isoforms. By contrast, when probed with antiAPAP antibodies, several protein targets in addition to the 58-ABP were detected in APAP-treated but not in control cytosol. Two-dimensional PAGE analysis indicated that all four 58-ABP isoforms could also be detected with the anti58-ABP antibody in liver cytosol from control mice (Fig. 8A). Preimmune rabbit sera served as a control and did not possess any immunoreactivity with cytosol from either control or APAP treated mice.

Immunochemical Evidencefor the Presenceof the 58-ABP in Human Liver To determine if similar 58-kDa proteins exist in human liver, a cytosolic extract from a human liver specimen was resolved by two-dimensional PAGE, electroblotted to nitrocellulose, and then probed with the anti-58-ABP antiserum. The antiserum detected four protein spots with electrophoretie migration similar to that of the isoforms observed in mouse liver cytosol (compare Figs. 8A and 8B).

FIG. 5. Two-dimensional SDS-PAGE analysis of the 5%ABP. Unfractionated APAP-treated liver cytosol (A) and IEF fraction 3 (B) were silver stained and compared to a Western blot of IEF fraction 3 immunostained with anti-APAP antibodies. Samples were resolved in the fust dimension by isoelectric focusing and then according to molecular weight using SDSPAGE. Arrowheads point to the positions of the four 58-ABP isofonns.

ISOLATION

OF THE 58-kDa ACETAMINOPHEN-BINDING

25

PROTEIN

A comparison of the sequences of these two peptides (Table 2) with those in Genbank and the Swiss-Protein databases revealed complete identity with a 56-kDa mouse liver selenium-binding protein that has been recently cloned from a mouse liver cDNA library by Bansal et al. (1990). The sequences deduced from peptides 51 and 60 of the 58-ABP aligned perfectly with residues 187-195 and 236-255, respectively, of the selenium-binding protein. In addition the amino acid composition of the 58-ABP also compared favorably with that of the 56-kDa selenium-binding protein (Table 1). The overestimation of the glycine content of the 58-ABP may have resulted from residual glycine contained in the buffer used for protein electroelution. DISCUSSION

FIG. 6. V-8 protease digestion of the 58-ABP isoforms. The four isoforms of the 58-ABP were cut out individually from 2D SDS-PAGE of ROTOFOR fraction 3 and digested with V-8 protease. All detectable peptides appeared to be recognized immunochemicaBy by the anti-58-ABP antibodies. Protease digests of two other nonrelated proteins extracted from the 2D gels showed distinctly different patterns and none of their peptides were recognized by the antibody (data not shown).

It has long been postulated that tissue injury can result as a consequence of the covalent binding of chemically reactive metabolites of xenobiotics to cellular macromolecules (Miller and Miller, 1952). Although the correlation between the extent of total covalent binding and the incidence of cell death has been well documented (Brodie et al., 197 1; Mitchell et al., 1973, 1974; Jollow et al., 1973, 1974), some compounds have been reported to decrease cytotoxic effects without decreasing covalent binding (Labadorios et al., 1977; DeValia et al., 1982;Albano et al., 1985; Tee et al., 1986), while other

APAP

CONTROL

synthesized 3-(cystein-Syl)-APAP prior to protein hydrolysis yielded virtually quantitative recovery of the cysteine residues (data not shown).

SequenceAnalysis of Two Peptides of the 5%ABP An initial attempt to sequence the amino terminus of the purified 58-ABP directly from proteins bound to polyvinylidene difluoride membranes (Immobilon-P) was unsuccessful possibly as a result of a blocked N-terminus. As an alternative, 13.3 pg of the purified 58-ABP was sequentially cleaved, first with cyanogen bromide and then with trypsin, and the resulting peptides were separated by a reverse-phase HPLC column (Fig. 9). The amino acid sequences of two well-resolved peptide fragments (peaks 5 1 and 60) are shown in Table 2. Nine amino acids were sequenced from the first peak (5 1) and 19 amino acids were sequenced from the second (peak 60). The presence of the methionine residue in the second sequence was unexpected but may be attributed to the incomplete cyanogen bromide digestion of the peptide resulting from the neighboring threonine residue (Kathy Stone, personal communication, Protein Sequencing Facility, Yale University).

FIG. 7. Detection of specific antibodies against the 58-ABP. Western blots of liver cytosolic proteins from control and APAP-treated mice were immunostained with either anti-APAP or anti-58-ABP sera. Preimmune serum were used for controls. The relative migrations of the proteins used as molecular weight standards are indicated.

26

BARTOLONE -

PI

IEF

+

FIG. 8. Immunochemical detection of the 5%ABP in human liver. Proteins of the 105,000g liver supematant obtained from (A) untreated mice or (B) humans were resolved on 2D SDS-PAGE, transblotted, and immunochemically analyzed with anti-58-ABP antibody. Arrowheads indicate the 58-ABP proteins.

compounds bind to proteins without causing toxicity (Wiley et al., 1979; Monks et al., 1982; O’Brien et al., 1985; Roberts et al., 1990). Until target proteins are better characterized, the contribution of adduct formation to toxicity will remain unknown and covalent binding data can only be interpreted as an index of the exposure of a tissue to reactive metabolites (Gillette, 1974). Recent studies have shown that the arylation of proteins by APAP is not random and that adducts of approximately 44 and 58 kDa account for most of the immunochemically detectable covalent binding (Bartolone et al., 1987, 1988; Beierschmitt et al., 1989; Birge et al., 1989). At least six different experimental paradigms have demonstrated an association between the arylation of the 58-kDa protein and APAP toxicity. First, the binding to the 58-ABP in vivo increased when subthreshold doses of APAP were exceeded (Bartolone et al., 1987). Second, in hepatocytes from phenobarbital pretreated mice, the increased hepatotoxicity after an APAP overdose correlated with the increased binding to the 58-ABP (Birge et al., 1989). Third, in 3-month-old CD1 mice that were more susceptible to APAP than 2-monthold animals, the increased hepatotoxicity was associated with greater 58-ABP arylation even though total radiolabeled APAP binding in the two groups of mice was not different (Beierschmitt et al., 1989). Fourth, arylation of a 58-kDa protein was associated with APAP damage not only in liver but also in other extrahepatic target organs such as kidney

ET AL.

and lung (Bartolone et al., 1989a). Fifth, a similar 58-kDa protein was found to be targeted in the liver of a human APAP fatality (Birge et al., 1990). Sixth, when piperonyl butoxide was administered 2 hr after APAP, the protective effect of this mixed-function oxidase inhibitor was associated with a diminished accumulation of the 58-kDa adduct even though total macromolecular binding of [3H]APAP was not altered (Brady et al., 1990). These correlations are suggestive of a potential mechanistic link between the hepatotoxicity and the arylation of the 58-ABP. Recently, Pumford et al. ( 1990) have independently contirmed that a cytosolic protein of similar molecular weight is the major APAP adduct detectable immunochemically. Clearly, characterization and eventual identification of this protein will provide significant insight into the role of its arylation by APAP in the ensuing toxicity. This report describes the purification and partial characterization of the 58-ABP. The protein behaved as a monomer under native conditions on gel permeation chromatography; however, on two-dimensional electrophoresis the 58-ABP appeared to consist of four proteins with isoelectric points

TABLE 1 The Amino Acid Composition of the 58-kDa Protein: Comparison with the Amino Acid Composition of the 56-kDa Selenium-Binding Protein Amino acid

Mole %

No. of residues in 58-ABP

No. of residues in 56-SBP

Alanine Arginine Aspartic acid and asparagine Cysteine Glutamic acid and glutamine Glycine Histidine Isoleucine Leucine Lysine Methionine Phenylalanine Proline Serine Threonine Tyrosine Valine Tryptophan

4.48 4.40

24.1 24.2

24 19

10.40 1.43

57.3 1.9

43 9-10

10.09 16.50 1.91 4.55 9.12 4.66 1.83

55.6 90.9 10.5 25.1 50.3 25.7

45 44 12 21 45 26

3.17

10.1 17.5

11 16

6.50 6.88 4.86 3.33 5.88

35.8 31.9 26.8 18.3 32.4

33 36 22 17 32

ND

-

11

Note. Amino acid analysis of approximately 1.5 pg of the 58-kDa acetaminophen-binding protein (58-ABP) was conducted as described under Materials and Methods. Each value is the average of triplicate analyses. The tryptophan content was not determined (ND). The values for the amino acid composition of the 56-kDa selenium-binding protein (56-SBP) were calculated from the data provided by Bansal et al. ( 1990).

ISOLATION

CONTROL CNBR/TRYPTIC

OF THE 5%kDa ACETAMINOPHEN-BINDING

BLRNK DIGEST

CNBRlTRYPTIC

DIGEST

60.00

80.00

100.00

FIG. 9. Reversed-phase HPLC peptide mapping of CNBr- and trypsindigested fragments of the 58ABP. The acetone-washed 58-ABP pellet was dissolved in 70% aqueous formic acid, digested first with CNBr and then trypsin, and injected onto a narrow-bore Vydac C 18 reverse-phase column (25 cm X 2.1 mm). The peptide fragments of the 5%ABP were separated using a gradient elution as described under Materials and Methods (bottom). Peaks 5 1 and 60 were utilized for peptide sequencing. (Top) A control gradient elution. The x-axis is “time in minutes” and the y-axis is Azlo.

ranging from 6.2 to 6.6. The fact that all four proteins appeared to bind APAP in proportion to their relative protein concentration and that they all contained very similar peptide fragments after V-8 protease digestion makes it likely that the four 58kDa proteins are structurally related isoforms. A chemical basis for the different isoelectric points of the four isoforms has not been ascertained. An antibody raised against the purified protein recognized all four 58kDa isovariants irrespective of whether the protein was derived from control or from APAP-treated mice and all four isovariants were also detectable in human liver. These observations suggest that (1) the presence of bound APAP is not required for antibody recognition of the 58-ABP, nor does it interfere with such reaction, and (2) APAP binding did not appear to alter the electrophoretic mobility of the isoforms. Hence the differences in the isoelectric points of the 58-ABPs cannot be attributed to differential amounts of APAP arylation of a single protein. Rather it is possible that the isoforms may be a result of posttranslational modifications such as phosphorylation or acetylation. This remains to be established. Amino acid analysis of the purified protein indicated that the 58-ABP contained eight cysteine residues (about 1.4% by weight). It is unlikely that the covalent binding of APAP resulted in an underestimation of the cysteine content of the protein since (1) 20% of the original liver homogenate utilized for purification of the protein was obtained from control liver and (2) immunohistochemical analysis has revealed that, even though the 58-ABP was uniformly distributed throughout the hepatic lobules, its arylation 4 hr after APAP administration was detectable primarily in the centrilobular

PROTEIN

27

zones of the liver (Bartolone et al., 1989b; Emeigh-Hart et al., 1990). Thus, the initial preparation utilized to purify the 58-ABP contained a very significant proportion of nonarylated protein. In addition pet-formic acid treatment of chemically synthesized 3-(cystein-4yl)-APAP prior to amino acid analysis yielded virtually quantitative recovery of cysteine residues. Cysteine residues on proteins are considered the major sites of APAP arylation in vivo (Streeter et al., 1984; Hoffman et al., 1985). Based on our previous observations, the native 58-ABP isoforms are very reactive with N-ethylmaleimide at pH 7.4 (Bartolone et al., 1988) and can undergo reversible glutathiolation under oxidative conditions (e.g., upon incubation with 0.1 IIIM diamide in vitro; Birge et al., 199 la). These data suggest that at physiological pH the 58-ABP is a very nucleophilic protein and, by virtue of its reactive thiol content, may be well suited to bind NAPQI. The consequences of such binding remain unclear. It might directly mediate toxicity through alterations of some critical, as yet undefined function. Alternatively, the protein may decrease further NAPQI availability for interaction with other critical targets and thereby serve a protective function. The concept that the 58-ABP may be cytoprotective would be analogous to the role of metallothionein in binding and detoxifying heavy metals (Karin, 1985). However, the relatively low cysteine content of the 58-ABP (8 residues; 1.4%) compared to that of metallothionein (20 cysteine per 61 residues; 33%) (Fowler, 1989) may not support the latter hypothesis unless at hepatotoxic doses APAP also binds to nonthiol sites on the protein. A search of GenBank and the Swiss-Protein databases indicated that the sequence of two of the peptide fragments obtained from the purified 58-ABP possessed complete identity with a 56-kDa selenium-binding protein whose function has not been clearly defined (Bansal et al., 1990). In addition, the subcellular localization, amino acid composition, and isoelectric point of the 58-ABP are also apparently similar to those reported for the selenium binding TABLE 2 The Amino Acid Sequenceof Peptides 51 and 60 of the 58-ABP Peptide 5 1 -Gly-Tyr-Asp-Phe-Try-Tyr-Ghr-Pro-ArgPeptide 60 -His-Glu-Ile-Ile-Gln-Thr-Leu-Gln-Met-Thr-Asp-Gly-~u-I~ePro-Leu-Glu-Ile-ArgNote. Peptide sequence data were obtained from the cyanogen bromide/ tryptic digests of the purified 58-ABP described in the legend to Fig. 9. The amino acid sequence of peptides 51 and 60 were found to be identical to those in positions 187-195 and 236-254 of the 56-kDa selenium-binding protein recently reported by Bansal et al. (1990).

28

BARTOLON IE ET AL.

protein (Bansal et al., 1989). Despite these strong correlations, actual confirmation of the identity of the 58-ABP with the selenium-binding protein will require a complete comparison of the cDNA sequences. These experiments are currently in progress. ACKNOWLEDGMENTS This study was supported by grants from the University of Connecticut Research Foundation, NIH (Grant GM-31460) N.I.E.H.S. (Grant ES07 163), and the Center for Biochemical Toxicology. The authors thank the University of Connecticut Biotechnology Center and Dr. Kenneth Williams and Ms. Kathy Stone of the Protein Sequencing Facility at Yale University for their excellent technical support.

REFERENCES Albano, E., Rundgren, M., Harvison, P., Nelson, S., and Moldeus, P. (1985). Mechanisms of N-acetyl-pbenzoquinoneimine cytotoxicity. Mol. Pharmacol. 28,306-3 Il. Bansal, M. P., Mukhopadhyay, T., Scott, J., Cook, R. G., Mukhopadhyay, R., and Medina, D. (1990). DNA sequencing of a mouse liver protein that binds selenium: Implications for selenium’s mechanism of action in cancer prevention. Carcinogenesis 11, 207 l-2073. Bansal, M. P., Obom, C. J., Danielson, K. G., and Medina, D. (1989). Evidence for two selenium binding proteins distinct from glutathione peroxidase in mouse liver. Carcinogenesis 10, 54 I-546. Bartolone, J. B., Beierschmitt, W. P., Birge, R. B., Emeigh Hart, S. G., Wyand, S., Cohen, S. D., and Khairallah, E. A. (1989a). Selective acetaminophen metabolite binding to extrahepatic proteins: An in viva and in vitro analysis. Toxicol. Appl. Pharmacol. 99, 240-249. Bartolone, J. B., Birge, R. B., Sparks, K., Cohen, S. D., and Khairallah, E. A. (1988). Immunochemical analysis of acetaminophen covalent binding to proteins: Partial characterization of the major acetaminophenbinding liver proteins. Biochem. Pharmacol. 37,4763-4774. Bartolone, J. B., Cohen, S. D., and Khairallah, E. A. (1989b). Immunohistochemical localization of acetaminophen-bound liver proteins. Fundam. Appl.

Toxicol.

13, 859-862.

Bartolone, J. B., Sparks, K., Cohen, S. D., and Khairallah, E. A. (1987). Immunochemical detection of acetaminophen-bound liver proteins. Biochem. Pharmacol. 36, 1193-l 196. Beierschmitt, W. P., Brady, J. T., Bartolone, J. B., Wyand, D. S., Khairallah, E. A., and Cohen, S. D. (1989). Selective protein arylation and the age dependency of acetaminophen hepatotoxicity in mice. Toxicol. Appl. Pharmacol.

98,5 17-529.

Birge. R. B., Bartolone, J. B., Cohen, S. D., Khairallah, E. A., and Smolin, L. A. (1991a). A comparison of proteins S-thiolated by glutathione to those arylated by acetaminophen. Biochem. Pharmacof. 42 (Supplement), S197-S207. Birge, R. B., Bartolone, J. B., Emeigh Hart, S., Nishanian, E., Tyson, C. A., Khairallah, E. A., and Cohen, S. D. (1990). Acetaminophen hepatotoxicity: Correspondence of selective protein arylation in human and mouse liver in vitro, in culture and in vivo. Toxicol. Appl. Pharmacol. 105, 472-482. Birge, R. B., Bartolone, J. B., McCann, D. J., Mangold, J. B., Cohen, S. D., and Khairallah, E. A. (1989). Selective protein arylation by acetaminophen and 2,6-dimethylacetaminophen in cultured mouse hepatocytes from phenobarbital-induced mice and uninduced mice. Biochem. Pharmacol. 38,4429-4439. Birge, R. B., Bulera, S. J., Bartolone, J. B., Ginsburg, G. L., Cohen, S. D., and Khairallah, E. A. (1991b). The arylation of mouse liver microsomal

membrane proteins by acetaminophen are associated with the release of a 44-kDa acetaminophen-binding protein complex into the cytosol. Toxicol. Appl. Pharmacol.

109,443-482.

Boobis, A. R., Seddon, C. E., Nasseri-Sina, P., and Davies, D. S. (1990). Evidence for a direct role for intracellular calcium in paracetamol toxicity. Biochem. Pharmacol. 39, 1277- 128 I. Boyd, E. M., and Bereczky, G. M. (1966). Liver necrosis from paracetamol. Br. J. Pharmacol. 26,606-6 14. Brady, J. T., Birge, R. B., Khaimllah, E. A., and Cohen, S. D. (1990). Posttreatment protection with piperonyl butoxide against acetaminophen hepatotoxicity is associated with changes in selective but not total covalent binding. In Biological Reactive Intermediates IV (C. M. Witmer, R. R. Snyder, D. J. Jollow, G. F. Kalf, J. J. Kocsis. and I. G. Sipes, Eds.), pp. 685-688. Plenum Press, New York. Brodie, B. B., Reid, W. D., Cho, A. K., Sipes, I. G., and Gillette, J. R. (197 1). Possible mechanism of liver necrosis caused by aromatic organic compounds. Proc. Natl. Acad. Sci. USA 68, 160-164. Cleveland, D. W., Fisher, S. G., Kirschner, M. W., and Laemmli, U. K. (1977). Peptide mapping by limited proteolysis in sodium dodecyl-sulfate and analysis by gel electrofocusing. J. Biol. Chem. 252, 1102-I 106. Corcoran, G. B., Bauer, J. A., and Lau, T. W. D. (1988). Immediate rise in intracellular calcium and glycogen phosphorylase a activity upon acetaminophen covalent binding leading to hepatotoxicity in mice. Toxicology 50, 157-167. Dahlin, D. C., Miwa, G. T., Lu, A., and Nelson, S. D. (1984). N-acetyl-pbenzoquinoneimine: A cytochrome P-450 mediated oxidation product of acetaminophen. Proc. Natl. Acad. Sci. 81, 1327- 133 1. DeValia, J. L., Oglivie, R. C., and McLean, A. E. M. (1982). Dissociation of cell death from covalent binding of paracetamol by flavones in a hepatocyte system. Biochem. Pharmacol. 31,3745-3749. Emeigh-Hart, S. G., Cartun, R. W., Wyand, D. S., Khairallah, E. A., and Cohen, S. D. (1990). Selective protein arylation by acetaminophen: Immunochemical localization in mouse liver, kidney and lung. Toxicologist 112, 229.

Fowler, B. A. (1989). Biological roles of high atlinity metal-binding proteins in mediating cell injury. Comments Toxicol. 3, 27-46. Gillette. J. R. (1974). A perspective on the role of chemically reactive metabolites of foreign compounds in toxicity. I. Correlation of changes in covalent binding of reactive metabolites with changes in the incidence and severity of toxicity. Biochem. Pharmacol. 23, 2927-2938. Ginsberg, G. L., Placke, M. E., Wyand, D. S., and Cohen, S. D. (1982). Protection against acetaminophen-induced hepatotoxicity by prior treatment with Fenitrothion. Toxicol. Appl. Pharmacol. 66, 383-399. Hinson. J. A. ( 1980). Biochemical toxicology of acetaminophen. In Reviews of Biochemical Toxicology (E. Hodgson, J. R. Bends, and R. M. Philpot, Eds.), Vol. 2, pp. 103-130. Elsevier Press, New York. 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. Jollow, D. J., Mitchell, J. R., Potter, W. Z.. Davis, D. C., Gillette, J. R.. and Brodie, B. B. (1973). Acetaminophen-induced hepatic necrosis. II. Role of covalent binding in vivo. J. Pharmacol. Exp. Ther. 187, 185-194. Jollow, D. J., Mitchell, J. R., Zampaghone, N., and Gillette, J. R. (1974). Bromobenzene-induced liver necrosis. Protective role of glutathione and evidence for 3,4 bromobenzene oxide as the hepatotoxic metabolite. Pharmacology 11, 15 1- 169. Karin, M. (1985). Metallothioneins: Proteins in search of a function. Cell, 41, 9-10.

ISOLATION

OF THE 58-kDa ACETAMINOPHEN-BINDING

Klapper, M. H. (1977). The independent distribution of amino acid near neighbor pairs into polypeptides. Biochem. Biophys. Res. Commun. 78, 1018-1024. Labadorios, D., Davis, M., Portman, B., and Williams, R. (1977). Paracetamol-induced hepatic necrosis in the mouse: Relationship between covalent binding, hepatic glutathione depletion and the protective effect of alpha-mercaptopropionyl-glycine. Biochem. Pharmacol. 26, 31-35. Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head bacteriophage T4. Nature (London) 227,680-685. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (195 1). Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265-275. Miller, E. C., and Miller, J. A. (1952) In vivo combinations between carcinogens and tissue constituents and their possible role in carcinogenesis. Cancer

Res. 12, 547-556.

Mitchell, J. R., Jollow, D. J., Potter, W. Z., Gillette, J. R., and Brodie, B. B. (1973). Acetaminophen-induced hepatic necrosis. IV. Protective role of glutathione. J. Pharmacol. Exp. Ther. 187, 2 11-2 17. Mitchell, J. R., Thorgeirsson, S. S., Potter, W. Z., Jollow, D. J., and Keiser, H. (1974). Acetaminophen-induced hepatic injury: Protective role of glutathione in man and rational therapy. Clin. Pharmacol. Ther. 16, 676684. Monks, T. J., Hinson, J. A., and Gillette, J. R. (1982). Bromobenzene and pbromophenol toxicity and covalent binding in vivo. Life Sci. 30, 841848. Moore, M., Thor, H., Moore, G., Nelson, S., Moldeus, P., and Orrenius, S. (1985). The toxicity of acetaminophen and N-acetyl-pbenzoquinoneimine in isolated hepatocytes is associated with thiol depletion and increased cytosolic Ca’+. J. Biol. Chem. 260, 13,035-13,040. Morrissey, J. H. (198 I). Silver stain for proteins in polyacrylamide gels: A modified procedure with enhanced uniform sensitivity. Anal. Biochem. 117,307-310.

Nelson, S. D., and Pearson, P. G. (1990). Covalent and non-covalent interactions in acute lethal cell injury caused by chemicals. Annu. Rev. Pharmacol.

Toxicol.

30, 169-195.

O’Brien, K. A. F., Smith, L., and Cohen, G. M. (1985). Differences in nap thalene-induced toxicity in the mouSe and rat. Chem. Biol. Interact. 55, 109-122. O’Farrell, P. H. (1975). High resolution two-dimensional electrophoresis of proteins. J. Biol. Chem. 250, 4007-402 1, Orrenius, S., McConkey, D. J., Bellomo, G., and Nicotera, P., (1989). Role of calcium in cell killing. Trends Pharmacof. lo,28 l-285.

29

PROTEIN

Placke, M. E., Wyand, D. S., and Cohen, S. D. (1987). Extrahepatic lesions induced by acetaminophen in the mouse. Toxicol. Pathol. 15, 381-387. Potter, D. W., Pumford, N. R., Hinson, R. W., Benson, R. W., and Roberts, D. W. (1989). Epitope characterization of acetaminophen bound to protein and non-protein suhhydryl groups by an enzyme-linked immunosorbent assay.J. Pharmacol. Exp. Ther. 248, 182-189. Potter, W. Z., Thorgeirsson, S. S., Jollow, D. J., and Mitchell, J. R. (1974). Acetaminophen-induced hepatic necrosis. V. Correlation of hepatic necrosis, covalent binding and glutathione depletion in hamsters. Pharmacology 12, 129-143. Pumford, N. R., Hinson, J. A., Benson, R. W., and Roberts, D. W. (1990). Immunoblot analysis of protein containing 3-(cystein-9yl)acetaminophen adducts in serum and subcellular liver fractions from acetaminophentreated mice. Toxicol. Appl. Pharmaol. 104, 52 l-532. Roberts, S. A., Price, V. F., and Jollow, D. J. (1990). Acetaminophen structure-toxicity studies: In vivo covalent binding of a nonhepatotoxic analog, 3-hydroxyacetanilide. Toxicol. Appl. Pharmacol. 105, 195-208. Stone, K. L., LoPresti, M. B., Crawford, J. M., DeAngelis, R., and Williams, K. R. (1989). Enzymatic digestion of proteins and HPLC peptide isolation. In A Practical Guide to Protein and Peptide Purification for Microsequencing, (P. T. Matsudaira, Ed.), pp. 3 l-47. Academic Press, New York. Streeter, A. J., Dahlin, D. C., Nelson, S. D., and Baillie, T. A. (1984). The covalent binding of acetaminophen to protein. Evidence for cysteine residues as major sites of arylation in vitro. Chem. Biol. Interact. 48, 349366. Tee, L., Boobis, A., Huggett, A., and Davies, D. (1986). Reversal of acetaminophen toxicity in isolated hamster hepatocytes by dithiothreitol. Toxicol.

Appl. Pharmacol.

83, 294-3

14.

Towbin, H., Staehelin, T., and Gordon, J. (I 979). Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Nat/. Acad. Sci. USA 76,4350-4354. Tsokos-Kuhn, J. 0. (1989). Evidence in vivo for elevation of intracellular free calcium in the liver after diquat, acetaminophen and CCL. Biochem. Pharmacol.

38,3061-3065.

Wendel, A., Feuerstein, S., and Konz, K. (1979). Acute paracetamol intoxication of starved mice leads to lipid peroxidation in vivo. Biochem. Pharmacol. 28,205 l-2055. Wiley, R. A., Hanzlik, R. P., and Gillesse, T. (1979). Effect of substituents on in vitro metabolism and covalent binding of substituted bromobenzenes. Toxicol.

Appk

Pharmacol.

49,249-255.