Glutamate Dehydrogenase Covalently Binds to a Reactive Metabolite

Division of Toxicology, University of Arkansas for Medical Sciences,. Little Rock, Arkansas 72205-7199, and National Institute of Mental Health,. Nati...
8 downloads 0 Views 318KB Size
Chem. Res. Toxicol. 1996, 9, 541-546

541

Glutamate Dehydrogenase Covalently Binds to a Reactive Metabolite of Acetaminophen N. Christine Halmes,† Jack A. Hinson,† Brian M. Martin,‡ and Neil R. Pumford*,† Division of Toxicology, University of Arkansas for Medical Sciences, Little Rock, Arkansas 72205-7199, and National Institute of Mental Health, National Institutes of Health, Bethesda, Maryland 20892 Received September 12, 1995X

The mechanism of the hepatotoxicity of the analgesic acetaminophen is believed to be mediated by covalent binding to protein; however, critical targets which effect the toxicity are unknown. It has been shown that mitochondrial respiration in vivo is inhibited in mice as early as 1 h following a hepatotoxic dose of acetaminophen, and it is postulated that covalent binding to critical mitochondrial proteins may be important. A time course of mitochondrial proteins stained with anti-acetaminophen in an immunoblot detected two major adducts of 50 and 67 kDa as early as 30 min after a hepatotoxic dose of acetaminophen in mice. To further understand the role of covalent binding to mitochondrial proteins and acetaminophen hepatotoxicity, we have purified and identified a 50 kDa mitochondrial protein which becomes covalently bound to a reactive metabolite of acetaminophen. An N-terminal sequence of the 50 kDa adduct was 100% homologous with the deduced amino acid sequence of glutamate dehydrogenase. In addition, the purified protein was immunochemically reactive with rat liver anti-glutamate dehydrogenase. Enzyme activity of glutamate dehydrogenase was significantly decreased in mice 1 h following hepatotoxic treatment with acetaminophen. These data suggest that acetaminophen hepatotoxicity may in part be mediated by covalent binding to glutamate dehydrogenase.

Introduction Acetaminophen, when used at therapeutic doses, is a safe and effective analgesic and antipyretic; in overdose, it is a hepatotoxin causing centrilobular necrosis (1). Acetaminophen is metabolized by the cytochrome P450 mixed function oxidase system to the reactive intermediate N-acetyl-p-benzoquinone imine (NAPQI)1 (1), which is subsequently detoxified through conjugation to glutathione (2). Following glutathione depletion, NAPQI covalently binds to specific proteins, and this binding correlates with the necrosis (3, 4). It is commonly believed that acetaminophen toxicity is mediated by the covalent binding of NAPQI to a critical protein(s), thus altering function or inducing changes in critical regulatory pathways. Since mitochondrial proteins bind acetaminophen early in the course of toxicity (4), it can be hypothesized that this binding may produce mitochondrial damage and play a critical role in the toxicity of acetaminophen. Previous studies have shown that hepatotoxic doses of acetaminophen cause early morphological changes in mitochondria, including enlargement, fusion, and detached fragments of cristae (5). In addition, acetaminophen inhibits mitochondrial respiration both in vitro (6, 7) and * To whom correspondence should be addressed at the University of Arkansas for Medical Sciences, 4301 W. Markham St., Slot 638, Little Rock, AR 72205. EMAIL: [email protected]. † University of Arkansas for Medical Sciences. ‡ National Institutes of Health. X Abstract published in Advance ACS Abstracts, February 1, 1996. 1 Abbreviations: NAPQI, N-acetyl-p-benzoquinone imine; HEPES, N-(2-hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid); PMSF, phenylmethanesulfonyl fluoride; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate; SDS/PAGE, sodium dodecyl sulfate/ polyacrylamide gel electrophoresis; PVDF, poly(vinylidene fluoride).

0893-228x/96/2709-0541$12.00/0

in vivo (8, 9). This respiratory dysfunction begins as early as 1 h following acetaminophen administration, well before overt toxicity. Hepatotoxic doses of acetaminophen also result in significantly increased mitochondrial calcium levels, although this begins 3 h after treatment (10), suggesting that this perturbation is a late event in the hepatotoxicity. Identification of the mitochondrial proteins to which acetaminophen binds may yield additional insights into further understanding the involvement of mitochondria in the mechanism of acetaminophen toxicity; in this study we report the purification and identification of a 50 kDa mitochondrial protein, glutamate dehydrogenase, to which a reactive metabolite of acetaminophen covalently binds.

Experimental Procedures Materials. Acetaminophen, sucrose, N-(2-hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid) (HEPES), EDTA, phenylmethanesulfonyl fluoride (PMSF), leupeptin, pepstatin, 3-[(3cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS), rat L-glutamic dehydrogenase, and NADH were from Sigma Chemical Co. (St. Louis, MO). Silver staining kit was obtained from Pierce (Rockford, IL). Enhanced Chemiluminescence Western blotting detection reagents were from Amersham Life Science (Buckinghamshire, England). Centricon-10 and centriprep-10 concentrators were from Amicon (Beverly, MA). Poly(vinylidene fluoride) (PVDF) membrane was obtained from Schleicher & Schuell (Keene, NH). Freund’s complete and Freund’s incomplete adjuvant were from Life Technologies, Inc. (Grand Island, NY). Ammonium acetate and 2-ketoglutaric acid were from Aldrich Chemical Co. (Milwaukee, WI). All other chemicals and reagents were of the highest quality and purity available. Time Course. Eight week old male B6C3/F1 mice (26 animals) from Charles River Breeding Labs (Wilmington, VA)

© 1996 American Chemical Society

542 Chem. Res. Toxicol., Vol. 9, No. 2, 1996 were given a hepatotoxic dose of 400 mg/kg of acetaminophen, ip, in 0.9% NaCl or 0.9% NaCl alone and sacrificed at 0.5, 1, 2, or 4 h (2-3 animals/group). Livers were surgically removed, pooled, and homogenized in 3 volumes of 0.25 M sucrose buffer (pH 7.5) containing 10 mM HEPES, 1 mM EDTA, 0.2 mM PMSF, 1 µM leupeptin, and 1 µM pepstatin (homogenization buffer). Mitochondria were prepared similarly to the methods of Fleischer and co-workers (11, 12). Briefly, the homogenate was centrifuged at 3100g for 10 min, and the resulting pellet was centrifuged at 54 000g for 100 min in a discontinuous gradient of 0.25 M/1.6 M sucrose. The interface band was washed twice by centrifugation at 1900g for 10 min, and then the pellet was centrifuged at 92 000g for 60 min in a discontinuous gradient of 0.25 M/1.45 M sucrose. The pellet was washed twice at 39 000g for 10 min, and the resulting pellet was collected and resuspended in 0.25 M sucrose buffer (pH 7.5), containing 10 mM HEPES, and stored at -80 °C. Proteins were separated by sodium dodecyl sulfate/polyacrylamide gel electrophoresis (SDS/PAGE) and immunoblotted as previously reported (13) using anti-acetaminophen previously characterized (14, 15). Protein concentration was determined by the method of Lowry, using bovine serum albumin as the standard (16). Densitometry. Densitometry was performed using a Model GS-670 imaging densitometer (Bio-Rad Laboratories, Hercules, CA) in reflectance mode and analyzed using Bio-Rad Molecular Analyst image analysis software. Peak area was determined using a volume integration of equal band areas. Purification of a 50 kDa Protein. Male B6C3/F1 mice (62 animals) were given 400 mg/kg acetaminophen in 0.9% NaCl and sacrificed 2 h after dosing. Livers were surgically removed, pooled, and homogenized in 5 volumes of homogenization buffer. Mitochondria were prepared according to the method of Fleischer and Kervina using differential gradient ultracentrifugation (11). Briefly, the homogenate was centrifuged at 1000g for 10 min, and the resulting supernatant was centrifuged at 25 000g for 10 min. The pellet was then centrifuged at 25 000g for 10 min, followed by two identical wash steps, at which time the final pellet (mitochondria A) was recovered and homogenized in 0.25 M sucrose buffer (pH 7.5) containing 10 mM HEPES, and stored at -80 °C. To recover mitochondria from the nuclear fraction, the pellet from the 1000g spin was centrifuged at 71 000g for 70 min in a discontinuous gradient of 0.25 M/1.6 M sucrose. The interface band was washed twice at 1900g for 10 min, and the resulting pellet was then centrifuged at 68 400g for 60 min in a discontinuous gradient of 0.25 M/1.45 M sucrose; this pellet was washed twice at 25 000g for 10 min. The resulting pellet (mitochondria B) was collected, homogenized in 0.25 M sucrose buffer (pH 7.5) containing 10 mM HEPES, and stored at -80 °C. Mitochondria A and B were mixed and solubilized in 0.1% CHAPS, and then applied (236 mg) to a BioRad High-Q anion exchange column (20 × 1 cm) which had been equilibrated with 20 mM Tris-HCl buffer (pH 7.5) containing 0.1% CHAPS. Proteins were eluted in a flow rate of 1 mL/min with a linear NaCl gradient to 0.2 M. Fractions were separated by SDS/PAGE and immunoblotted using anti-acetaminophen as described above, except that gels were silver stained and immunoblots were visualized with enhanced chemiluminescence. The fractions which eluted from the column from 0.13 to 0.18 M NaCl were combined and concentrated approximately 12-fold using Centriprep-10 and Centricon-10 concentrators. The fraction which contained the 50 kDa acetaminophen-protein adduct was further purified using a Bio-Rad Model 491 Prep Cell by loading 2.5 mg onto a 7.5% T, 10.5 cm polyacrylamide gel and collecting 1 mL/min fractions. Fractions were analyzed as previously reported (13) and concentrated identically to previous fractions. Amino Acid Sequence Analysis. Five micrograms of the fraction which contained the 50 kDa adduct was separated by SDS/PAGE, transferred to PVDF membrane, and stained with Coomassie blue. An N-terminal amino acid sequence was obtained by microsequencing (17). Briefly, the band containing the 50 kDa adduct was excised and then sequenced on a Beckman LF3600 protein sequencer equipped with a System

Halmes et al. Gold Model 125 on-line PTH analyzer and analyzed with System Gold data analysis. Sequence homology with known proteins was determined by searching the Swiss-Protein database. Antibody Production. Rat liver L-glutamic dehydrogenase from Sigma was further purified using preparative electrophoresis with an 8% T polyacrylamide gel. A female New Zealand White rabbit (2.5 kg) obtained from Myrtle’s Rabbitry (Thompson Station, TN) was immunized with 100 µg of the purified protein in 1.5 mL (1:9 ratio) of Fruend’s complete adjuvant subcutaneously in 15 sites along the back and intramuscularly in the right and left hindquarters. After 6 weeks, the rabbit was boosted in a similar manner with 100 µg of glutamate dehydrogenase in Freund’s incomplete adjuvant, and serum was collected 2 weeks later. Mouse mitochondrial protein (100 µg) and the purified 50 kDa protein (2.5 µg) were then separated by SDS/PAGE, immunoblotted, probed with the antirat glutamate dehydrogenase serum, and visualized with enhanced chemiluminescence. Enzyme Activity. Glutamate dehydrogenase activity was determined according to Rajas (18). Briefly, 8 week old B6C3/ F1 male mice were treated with 400 or 600 mg/kg of acetaminophen, ip, in 0.9% NaCl or 0.9% NaCl alone (6-7 animals per group) and sacrificed at 1 h. Livers were surgically removed and homogenized in 3 volumes of homogenization buffer. The homogenate was centrifuged at 10 000g for 20 min to obtain a crude mitochondrial pellet, which was sonicated for 15 s at power level 5 with a sonicator probe (Branson Sonifier Cell Disruptor 185) and centrifuged at 15 000g for 30 min, at which time the supernatant was collected. Activity was assayed by measurement of NADH oxidation using 2-ketoglutarate as substrate with an absorbance measurement of 340 nm. Total assay volume was 1 mL and contained 10 mM 2-ketoglutarate, 100 mM ammonium acetate, 100 µM NADH, and 0.3 mg/mL protein in 10 mM Tris-HCl buffer (pH 7.3) with 2.5 mM EDTA. Statistical analysis was performed using a Student’s t-test assuming equal variances. The results are expressed as mean ( SE.

Results To characterize the temporal relationship of covalent binding of acetaminophen to mitochondria, we analyzed a time course of the binding in a Western blot using antiserum specific for acetaminophen. Mice were administered a hepatotoxic dose of 400 mg/kg of acetaminophen and sacrificed at 0.5, 1, 2, and 4 h. Mitochondria were prepared by discontinuous gradient ultracentrifugation, separated by SDS/PAGE, transferred to nitrocellulose, and stained with anti-acetaminophen. A 50 kDa and a 67 kDa band were detected at the 0.5 h time point; these bands increased in intensity from 0.5 to 4 h (Figure 1, upper left panel), while total protein content of each lane was approximately equivalent, as shown in Figure 1, lower left panel (gel stained with Coomassie blue). The control group shown was sacrificed at 1 h; controls sacrificed at 0.5, 2, and 4 h showed similar results (data not shown). Numerous other bands appeared at 1 h, similar to previously published results (4). However, we detected a protein fraction which stained with anti-acetaminophen not previously reported, at 120 kDa. To confirm that the 50 kDa band detected in the time course increased in intensity from 0.5 to 4 h, we densitometrically scanned the Western blot and determined volume intensity of the 50 kDa bands using equivalent areas. Binding was determined to be present at 0.5 h, and to increase from 0.5 to 4 h in a linear fashion (Figure 1, right panel). To better understand the relationship between early binding to mitochondrial proteins and ensuing hepatotoxicity, we purified a 50 kDa protein from mouse liver

Glutamate Dehydrogenase Binds Acetaminophen

Chem. Res. Toxicol., Vol. 9, No. 2, 1996 543

Figure 1. Time course of acetaminophen covalently bound to hepatic mitochondrial proteins. (Left panel) Mice were treated with 400 mg/kg acetaminophen and sacrificed at 0.5, 1, 2, and 4 h; controls were sacrificed at 1 h. Time groups consisted of 2 or 3 animals, whose livers were pooled. Mitochondria were prepared using density gradient ultracentrifugation. Hepatic mitochondrial proteins were separated by SDS/PAGE; covalently bound adducts were probed in a Western blot (upper left panel) using anti-acetaminophen (100 µg of protein/lane). Lower left panel, gel stained for total protein with Coomassie brilliant blue (100 µg of protein/lane). (Right panel) Analysis of the 50 kDa adduct. The 50 kDa band was scanned with an imaging densitometer. Peak area was determined using a volume integration of equal band areas.

Figure 2. Separation of mouse liver mitochondrial protein by anion exchange chromatography. Liver mitochondria (236 mg) from mice treated with 400 mg/kg acetaminophen and sacrificed at 2 h were loaded onto a High-Q anion exchange column. (A) High-Q anion exchange profile; (B) SDS/PAGE gels of column fractions silver stained; (C) immunoblots of column fractions stained with anti-acetaminophen.

mitochondria. Mice were treated with 400 mg/kg of acetaminophen and sacrificed 2 h later, as this time point has been shown to have the greatest amount of adducts (19). Mitochondria were prepared by discontinuous gradient ultracentrifugation, and the solubilized mitochondrial proteins were separated using a High-Q anion exchange column (Figure 2). Proteins were eluted with a linear NaCl gradient from 0 to 0.2 M NaCl (Figure 2, panel A). Fractions were then separated by SDS/PAGE,

and gels were either silver stained (Figure 2, panel B) or transferred to nitrocellulose and stained with antiacetaminophen (Figure 2, panel C). The fraction at 130 min contained an adduct at 67 kDa which stained with anti-acetaminophen, which may be the 67 kDa adduct shown in Figure 1 and was previously reported to be formed by 30 min after treatment (4), while the 205 min fraction contained an intensely staining adduct at approximately 47 kDa.

544 Chem. Res. Toxicol., Vol. 9, No. 2, 1996

Figure 3. Mouse liver mitochondrial fractions after preparative electrophoresis separation. The fraction from the High-Q anion exchange column which contained the 50 kDa fraction (2.5 mg) was separated by preparative electrophoresis. Fractions were further separated by SDS/PAGE and immunoblotted. Lanes A represent mitochondria; lanes B represent the fraction obtained from the anion exchange column. (Left panel) Gels stained with Coomassie blue. (Right panel) Immunoblot probed with antiacetaminophen.

Halmes et al.

Figure 4. Western blot of mitochondria (lane A, 100 µg) and purified 50 kDa protein (lane B, 2.5 µg). The blot was probed with anti-glutamate dehydrogenase antibodies obtained from a rabbit immunized with rat glutamate dehydrogenase.

Table 1. Sequence Homology of Peptides from the Acetaminophen 50 kDa Protein and the Deduced Amino Acid Sequence of Glutamate Dehydrogenasea peptide

sequence

N-terminal 50 kDa protein glutamate dehydrogenase

SEAAADREDDPNFFK SEAAADREDDPNFFK

a Amino acid sequence homology of 15 peptides from the N-terminal sequence of the 50 kDa acetaminophen-protein adduct with the deduced amino acid sequence from a cDNA clone encoding for mouse glutamate dehydrogenase (20).

A 50 kDa protein eluted between 0.13 and 0.18 M NaCl (Figure 2, 280-385 min fraction). This fraction was pooled and concentrated, and separated by preparative electrophoresis with a 7.5% T polyacrylamide gel. Fractions from the preparative electrophoresis were separated by SDS/PAGE and stained with Coomassie blue (data not shown). The preparative electrophoresis fractions which contained this 50 kDa protein were then pooled, concentrated, and analyzed by SDS/PAGE; gels were stained for protein with Coomassie blue or immunoblotted, using anti-acetaminophen antiserum (Figure 3). Mitochondrial protein from mice treated with 400 mg/kg acetaminophen and sacrificed 2 h after dosing contains numerous protein adducts (Figure 3, lanes A). The 50 kDa acetaminophen-protein adduct is indicated by lanes B; it appears to be a minor adduct at 2 h following treatment, yet in the time course following a hepatotoxic dose of acetaminophen a 50 kDa adduct stains within 30 min after treatment (Figure 1). The partially purified 50 kDa fraction from the preparative electrophoresis was then separated by SDS/ PAGE and transferred to PVDF membrane, along with unpurified mitochondria, and stained with Coomassie blue to visualize the protein (data not shown). A band at 50 kDa was excised, and an N-terminal amino acid sequence was obtained (Table 1). A search of the SwissProtein database revealed a 100% match between 15 amino acids of this 50 kDa acetaminophen-protein adduct and the deduced amino acid sequence of a mouse cDNA encoding for glutamate dehydrogenase (20) (EC 1.4.1.3).

Figure 5. Glutamate dehydrogenase enzyme activity was determined in mitochondria from mice treated with 400 mg/kg (control activity 1.12 µmol/(min‚mg of protein) or 600 mg/kg (control activity 0.86 µmol/(min‚mg of protein) acetaminophen and sacrificed at 1 h. Activity was measured by NADH oxidation using 2-ketoglutarate as substrate and measuring absorbance at 340 nm.

A rabbit was then immunized with commercially available rat liver glutamate dehydrogenase, and serum was collected and used in a Western blot with crude mitochondria and the purified 50 kDa protein (Figure 4). A single 50 kDa band was detected in the mitochondria, and in addition the antiserum also recognized a single band at 50 kDa in the purified protein, which further confirms that this mitochondrial protein which binds acetaminophen is glutamate dehydrogenase. To determine if the covalent binding by acetaminophen affects the enzyme activity of glutamate dehydrogenase, enzyme activity levels were determined in mice treated with 400 or 600 mg/kg acetaminophen and sacrificed at 1 h. Enzyme activity (µmol/(min‚mg of protein)) was significantly reduced (p