Glutathione Transferase Zeta-Catalyzed Bioactivation of

Department of Pharmacology and Physiology, University of Rochester Medical Center, 601 Elmwood Avenue, Box 711, Rochester, New York 14642, and Molecul...
6 downloads 5 Views 237KB Size
Download PDF
650

Chem. Res. Toxicol. 2004, 17, 650-662

Glutathione Transferase Zeta-Catalyzed Bioactivation of Dichloroacetic Acid: Reaction of Glyoxylate with Amino Acid Nucleophiles Wayne B. Anderson,† Philip G. Board,‡ and M. W. Anders*,† Department of Pharmacology and Physiology, University of Rochester Medical Center, 601 Elmwood Avenue, Box 711, Rochester, New York 14642, and Molecular Genetics Group, Division of Molecular Medicine, John Curtin School of Medical Research, Australian National University, Canberra ACT 2601, Australia Received May 22, 2003

Dichloroacetic acid (DCA) is a drinking water contaminant, a therapeutic agent, and a rodent carcinogen. Glutathione transferase zeta (GSTZ1-1) catalyzes the biotransformation of a range of R-haloalkanoates and the cis-trans isomerization of maleylacetoacetate. GSTZ1-1 catalyzes the bioactivation of fluorine-lacking dihaloacetates to S-(R-halocarboxymethyl)glutathione, a reactive intermediate that covalently modifies and inactivates the enzyme or is hydrolyzed to glyoxylate. The purpose of this study was to examine the GSTZ1-1-catalyzed bioactivation of DCA, including the reaction of DCA-derived glyoxylate with amino acid nucleophiles and the characterization of the structures and kinetics of adduct formation by LC/MS. The binding of [1-14C]DCA-derived label to bovine serum albumin required both GSTZ1-1 and GSH, whereas the binding to dialyzed rat liver cytosolic protein was increased in the presence of GSH. Studies with model peptides (antiflammin-2 and IL-8 inhibitor) indicated that glyoxylate, rather than S-(R-chlorocarboxymethyl)glutathione, was the reactive species that modified amino acid nucleophiles. Both addition (+74 Da) and addition-elimination (+56 Da) adducts of glyoxylic acid were observed. Addition adducts (+74 Da) could not be characterized completely by mass spectrometry, whereas addition-elimination adducts (+56 Da) were characterized as 2-carboxy4-imidazolidinones. 2-Carboxy-4-imidazolidinones were formed by the rapid equilibrium reaction of glyoxylate with the N-terminal amino group of antiflammin-2 to give an intermediate carbinolamine (Keq ) 0.63 mM-1), which slowly eliminated water to give an intermediate imine (k2 ) 0.067 hour-1), which rapidly cyclized to give the 2-carboxy-4-imidazolidinone. Glucose 6-phosphate dehydrogenase was inactivated partially by glyoxylate when reactants were reduced with sodium borodeuteride, which may indicate that glyoxylate reacts with selective lysine -amino groups. The results of the present study demonstrate that GSTZ1-1 catalyzes the bioactivation of DCA to the reactive metabolite glyoxylate. The reaction of glyoxylate with cellular macromolecules may be associated with the multiorgan toxicity of DCA.

Introduction DCA1

and other haloacetic acids are formed from humic and fulvic acids during the chlorination of raw water supplies (4). Concentrations of up to 160 µg DCA/L of finished drinking water have been reported (5). DCA is also a metabolite of tetrachloroethylene, trichloroethylene, trichloroacetic acid, and chloral hydrate (6-8). DCA is used for the management of congenital lactic acidosis at doses typically between 25 and 50 mg/kg/day (9) and has been investigated for the therapeutic treatment of a range of disorders including diabetes, hyperlipoproteinemia, and myocardial and cerebral ischemia (10). The therapeutic effects of DCA are associated with * To whom correspondence should be addressed. Tel: 585-275-1678. Fax: 585-273-2652. E-mail: [email protected]. † University of Rochester Medical Center. ‡ Australian National University. 1 Abbreviations: DCA, dichloroacetic acid; hGSTZ1c-1c, human GSH transferase zeta-variant c; BSA, bovine serum albumin; G6PDH, glucose 6-phosphate dehydrogenase; AF, antiflammin-2 (His-Asp-MetAsn-Lys-Val-Leu-Asp-Leu); IL8, IL-8 inhibitor (NR-acetyl-Arg-Arg-TrpTrp-Cys-Arg-amide); CAM-IL8, carbamidomethyl-IL8; TCEP, tris(2carboxyethyl)phosphine hydrochloride; MS3, MS/MS/MS.

its inhibition of pyruvate dehydrogenase kinase and the resulting activation of pyruvate dehydrogenase (11). A range of toxic and carcinogenic effects is associated with DCA. DCA is toxic to the liver, kidney, testes, and peripheral nervous system of rats (12). In humans, reversible neuropathies have been reported (12). Other toxicities associated with DCA include ocular (13) and developmental toxicities (14). DCA is hepatocarcinogenic in rats (15) and mice (16). The mechanisms of DCAinduced toxicity and carcinogenicity are incompletely understood (12). GSH transferase zeta (GSTZ1-1) is an evolutionarily conserved class of cytosolic GSH transferases that was first identified in 1997 (17). Four allelic variants of hGSTZ1-1 (-1a-1a, -1b-1b, -1c-1c, and -1d-1d) have been identified (18, 19). GSTZ1-1 is identical with maleylacetoacetate isomerase, which catalyzes the cis-trans isomerization of maleylacetoacetate to fumarylacetoacetate (20), the penultimate step in the tyrosine degradation pathway. GSTZ1-1 also catalyzes the GSH-dependent biotransformation of DCA and other R-haloalkanoic acids (21, 22); R-monohaloalkanoic acids are biotrans-

10.1021/tx034099+ CCC: $27.50 © 2004 American Chemical Society Published on Web 04/03/2004

GSTZ1-1-Catalyzed Bioactivation of DCA

formed to stable GSH conjugates, and dihaloacetic acids are biotransformed to glyoxylate (22). Glyoxylate may be oxidized to oxalate, reduced to glycolate, transaminated to glycine, or metabolized to carbon dioxide (23). GSTZ1-1 also catalyzes the bioactivation of fluorinelacking R,R-dihaloalkanoic acids. S-(R-Halocarboxymethyl)GSH, a reactive intermediate formed during the GSTZ1-1-catalyzed biotransformation of dihaloacetates, covalently modifies and inactivates GSTZ1-1 (24-26) or is hydrolyzed to glyoxylate. The electrophilic aldehyde moiety of glyoxylate may also react with protein-bound nucleophiles (27, 28). The biotransformation of xenobiotics to reactive intermediates that covalently modify cellular macromolecules is associated with their toxic and other deleterious effects (29, 30). The purpose of the present study was to examine the reaction of DCAderived metabolites with amino acid nucleophiles, to characterize the structures of DCA-derived adducts, and to characterize the kinetics of adduct formation. The results presented herein demonstrate that GSTZ1-1 catalyzes the bioactivation of DCA and that glyoxylate is the primary reactive species responsible for the modification of amino acid nucleophiles. The amino acid targets, structures, and reaction kinetics for some glyoxylate-peptide adducts are presented.

Experimental Procedures Materials. DCA (>99% pure), sodium glyoxylate (>99% pure), GSH, BSA, AF, sodium borodeuteride, G6PDH (Torula yeast), D-glucose 6-phosphate, and β-NADP+ were purchased from Sigma-Aldrich-Fluka (Milwaukee, WI). [1-14C]DCA (57 mCi/mmol) was obtained from American Radiolabeled Chemicals (St. Louis, MO) and was diluted with DCA to a specific activity of 28.5 mCi/mmol for some experiments. [35S]GSH (942 Ci/mmol) was purchased from New England Nuclear (Boston, MA) and was diluted with GSH to a specific activity of 50 mCi/ mmol. [14C]Methylated molecular weight markers were purchased from Amersham Biosciences (Piscataway, NJ). Recombinant 6X N-terminal His-tagged hGSTZ1c-1c was prepared as previously described (26). IL8 was purchased from Bachem (King of Prussia, PA). TCEP was purchased from Pierce (Rockford, IL). HPLC grade water and acetonitrile were purchased from EM Science (Gibbstown, NJ) and J. T. Baker (Phillipsburg, NJ), respectively. Determination of Protein Concentrations. Protein concentrations were determined by the method of Bradford (31) with the Bio-Rad (Hercules, CA) dye-binding reagent and with BSA as the standard. Reactions with BSA. Reaction mixtures (200 µL) contained hGSTZ1c-1c (150 µg/mL), [1-14C]DCA (1 mM, 28.5 mCi/mmol), GSH (1 mM), and BSA (75 mg/mL) in 0.1 M potassium phosphate (pH 7.4). In some experiments, reaction mixtures contained DCA (1 mM) and [35S]GSH (1 mM, 50 mCi/mmol). Control reaction mixtures lacked the unlabeled substrate or hGSTZ1c-1c or both. Reaction mixtures were prepared on ice, initiated with the addition of the radiolabeled substrate, and incubated for 1 h at 37 °C in a shaking water bath. Unbound substrates were removed immediately after incubation by ultrafiltration at 4 °C with Microcon YM-10 ultrafiltration devices (10 000 molecular weight cutoff; Millipore, Bedford, MA) with three dilution cycles. Dilutions were carried out with 200 µL of ice-cold 0.1 M potassium phosphate (pH 7.4). The ultrafiltration retentates were diluted with 350 µL of 0.1 M potassium phosphate (pH 7.4) after the final concentration cycle. Reactions with Rat Liver Cytosol. The liver from a male, F344 rat (∼500 g, Taconic Farms, Germantown, NY) was homogenized in 3 vol of ice-cold 0.1 M potassium phosphate buffer (pH 7.4) that contained a protease inhibitor cocktail (1 mL/20 g liver; P8340, Sigma). Cytosol was prepared by dif-

Chem. Res. Toxicol., Vol. 17, No. 5, 2004 651 ferential centrifugation as previously described (32). The cytosol was dialyzed at 4 °C against 2 L of 0.1 M potassium phosphate (pH 7.4) with three buffer changes. Reaction mixtures (100 µL) contained rat liver cytosol (10 µg protein/µL), [1-14C]DCA (1 mM, 57 mCi/mmol), and GSH (1 mM) in 0.1 M potassium phosphate (pH 7.4). Control reaction mixtures lacked GSH. Reaction mixtures were prepared on ice, initiated with the addition of [1-14C]DCA, and incubated for 2 h at 37 °C in a shaking water bath. The reaction mixtures were stored at -80 °C until analyzed. SDS-PAGE and Phosphorimaging Analysis. Reaction mixtures were heated in the presence of β-mercaptoethanol (3.5%, v/v) for 4 min at 100 °C. For some experiments, DTT (50 mM) or TCEP (10 mM) was used in place of β-mercaptoethanol. The proteins were resolved by SDS-PAGE on 12% polyacrylamide gels with a Mini-Protean III apparatus (Bio-Rad). Gels loaded with reaction mixtures that contained BSA were stained with Coomassie R-250 and destained. The gels were dried between a layer of cellophane and plastic wrap with the DryEase Mini-gel Drying System (Novex, San Diego, CA) and placed in direct contact with a phosphor screen for 10 days. To improve sensitivity, the SDS-PAGE-resolved cytosolic proteins were electrophoretically transferred to 0.45 µm nitrocellulose membranes (Bio-Rad). The membranes were stained with Ponceau S (Sigma), air-dried, and placed in direct contact with a phosphor screen for 7 days. Gel images were obtained with the Molecular Dynamics Phosphorimager SI scanner (Amersham Biosciences, Sunnyvale, CA) and analyzed with ImageQuant v. 5.2 software. Bioactivation and Structure Elucidation Studies with AF, IL8, and CAM-IL8. Reaction mixtures (1 mL) contained AF or IL8 (50 µM), DCA (1 mM), GSH (1 mM), and hGSTZ1c1c (20 µg/mL) in 0.1 M potassium phosphate (pH 7.4). Control reaction mixtures lacked GSH. Positive control reaction mixtures contained peptide (as above) and glyoxylate [5 mM (AF) or 30 mM (IL8)] in 0.1 M potassium phosphate (pH 7.4). To minimize the formation of peptide disulfide, positive control reaction mixtures that contained IL8 also contained TCEP (0.5 mM). Other reaction mixtures (0.5 mL) contained CAM-IL8 (50 µM), TCEP (0.5 mM), and glyoxylate (50 mM) in 0.1 M potassium phosphate (pH 7.4). CAM-IL8 was prepared by incubating IL8 (1.25 mM) with TCEP (12.5 mM) and iodoacetamide (25 mM) in 0.1 M potassium phosphate (pH 7.4) for 2 h at 20-22 °C in the dark; 20 µL of this solution was used in reaction mixtures (0.5 mL) that contained glyoxylate. Reaction mixtures were incubated for 16-24 h at 37 °C in a shaking water bath and stored at -80 °C until analyzed. Kinetic Studies. Glyoxylate concentrations were g10-fold greater than the peptide concentration to ensure pseudo firstorder kinetics. Reaction mixtures (1.5 mL) contained AF (50 µM) and 0.5, 1, 2, 4, 8, 16, or 32 mM glyoxylate in 0.1 M potassium phosphate (pH 7.4). Other reaction mixtures (1.0 mL) contained IL8 (50 µM), TCEP (0.5 mM), and 10, 20, 30, 40, 50, or 60 mM glyoxylate in 0.1 M potassium phosphate (pH 7.4). A different glyoxylate stock solution was used to prepare each reaction mixture series. The reaction mixtures were prepared on ice, initiated with the addition of glyoxylate, and immediately placed in a shaking water bath at 37 °C. Samples of the reaction mixtures were removed at the indicated times and were immediately frozen by immersion in liquid nitrogen. Samples at the 0 h time point were frozen e1 min after the addition of glyoxylate. The samples were stored at -80 °C until analyzed. Reduction of AF-Containing Reaction Mixtures with Sodium Borodeuteride. Samples (100 µL) of the reaction mixtures were mixed with 25 µL of freshly prepared, ice-cold sodium borodeuteride (400 mM, 80 mM final concentration, or as indicated). The reaction mixtures were incubated at 20-22 °C for 30 min and then placed on ice until analyzed by LC/MS. Preliminary studies with reaction mixtures that contained the highest glyoxylate concentration (32 mM) indicated that the amount of the reduced product was maximal with 80 mM sodium borodeuteride.

652

Chem. Res. Toxicol., Vol. 17, No. 5, 2004

LC/MS Analysis. Samples (100 µL) of the reaction mixtures that contained IL8, DCA, GSH, and hGSTZ1c-1c (and the associated GSH-lacking controls) were incubated with TCEP (2 µL, 100 mM, ∼2 mM final concentration) for 5 min at 20-22 °C in order to reduce IL8 disulfide and IL8-GSH mixed disulfide. Otherwise, samples were thawed and immediately injected (75 µL, ∼3750 or ∼3000 pmol for sodium borodeuteridereduced samples) on a quadrupole ion trap mass spectrometer (LC/MSD Trap, Agilent Technologies, Palo Alto, CA) operated in positive ion mode and equipped with an 1100 series LC system and a diode-array detector. Samples were kept at room temperature for less than 3 min before injection. A Vydac 218TP52 C18 column (2.1 mm × 250 mm; 5 µm particle size; 300 Å pore; Grace-Vydac, Hesperia, CA) and Waters µBondapak C18 precolumn (Waters Corp., Milford, MA) were employed. Solvent A contained 0.2% formic acid in water, and solvent B contained 0.2% formic acid in acetonitrile. For the DCA bioactivation studies with AF, the peptide mixtures were eluted with a linear gradient from 0 to 40% solvent B over 20-45 min at a flow rate of 0.3-0.5 mL/min. For the DCA bioactivation studies with IL8 and for all kinetic studies, the peptide mixtures were eluted with a linear gradient from 15 to 35% solvent B over 12 min at a flow rate of 0.5 mL/min. The mobile phase was increased to 95% solvent B between injections to ensure complete elution of all peptide-related components; none were detected. The UV absorbance of the column effluent was measured at 215 nm. For the kinetic studies, the mass spectrometer was operated in total ion mode, and all ions between 100 and 1200 m/z were recorded. Nebulizer pressure, dry gas flow rate, and dry gas temperature were 40-50 psi, 9-10 L/min, and 350 °C, respectively. The mass spectrometer operational parameters were optimized for the ion corresponding to the [M + 2H]2+ ion of the unmodified peptide. The peptide (80-100 µM) was infused at a flow rate of 200-300 µL/h into a stream of 20% solvent B at 0.3-0.5 mL/min. MS/MS and MS3 analyses were carried out on ions of interest. The fragmentation amplitude (V) was adjusted to obtain optimal fragmentation. G6PDH Inactivation Studies. Reaction mixtures (90 µL) contained G6PDH (1.25 units, according to the manufacturer’s reported activity) and glyoxylate (20 mM) in 0.1 M potassium phosphate (pH 7.4). Positive control reaction mixtures contained acetaldehyde (20 mM) in place of glyoxylate. Three different stock solutions of glyoxylate and acetaldehyde were employed. Control reaction mixtures lacked glyoxylate. Reaction mixtures were prepared on ice, initiated with the addition of G6PDH, and incubated for 1 or 4 h at 37 °C in a shaking water bath. Ten microliters of freshly prepared, ice-cold sodium borodeuteride (50 mM, 5 mM final concentration) was added, and the reaction mixtures were incubated for an additional 20 min at 20-22 °C. Preliminary studies indicated that maximal inactivation was obtained with 5 mM sodium borodeuteride. Ten microliters of water was added to control reaction mixtures. Reaction mixtures were placed on ice until assayed for G6PDH activity. G6PDH assay mixtures (1.0 mL) contained freshly prepared NADP+ (2 mM), glucose 6-phosphate (4 mM), and 5 µL of the original reaction mixture (0.0625 units of G6PDH) in 0.1 M potassium phosphate (pH 7.4). Preliminary studies indicated that these concentrations of NADP+ and glucose 6-phosphate gave near maximal activity. Assay mixtures were prepared at room temperature (20-22 °C) in a quartz cuvette and initiated with the addition of enzyme. The change in absorbance (e0.55 au) at 340 nm from 0.5 to 3.5 min was used as the measure of activity. The rate of absorbance change was approximately linear over this time period. Kinetic Analysis. The concentrations of the peptides and adducted peptides were estimated by integration of the UV (215 nm) recording. Peak area as a percentage of the sum of all peak areas was multiplied by the starting peptide concentration (50 µM) to estimate individual concentrations. The sum of all peak areas remained constant with time, which indicated that the peptides and glyoxylate-adducted peptides shared similar molar absorptivities. Peptide concentrations at infinite time (C∞) were

Anderson et al. estimated by fitting the data to a one-phase exponential decay curve [y ) (y0 - plateau) × e-kx + plateau] where the plateau equals C∞. The pseudo first-order rate constants (kobs) for peptide loss were determined from the slope of a plot of ln[(C0 - C∞)/(Ct - C∞)] vs time. The instantaneous rates of adduct formation were determined by fitting the data to a one-phase exponential association curve [y ) ymax × (1 - e-kx)] and calculating the tangent to the curve at each time point. The kobs for adduct formation were determined from slopes of plots of the instantaneous rates vs the concentration of unmodified peptide at each time point. The 72 h time point was excluded from the data for rate determinations since it was close to C∞ and introduced a large error in the Ct - C∞ value. For the kinetic analysis of 2-carboxy-4-imidazolidinone formation, k2 and Keq were approximated by plotting the pseudo first-order rate constants vs glyoxylate concentrations and fitting the data to the MichaelisMenten equation, where k2 corresponded to the turnover number (kcat) and Keq corresponded to 1/Km. For the kinetic analysis of +74 Da addition adduct formation, kf, kr, and Keq were determined by plotting the pseudo first-order rate constants vs glyoxylate concentrations and fitting the data to the equation for a line, where kf corresponded to the slope, kr corresponded to the y-intercept, and Keq corresponded to the negative reciprocal of the x-intercept. Data Analysis. Microsoft Excel 97 and GraphPad Prism version 3.03 were used for calculations and graph generation. Theoretical peptide and fragment ion masses were generated with General Protein Mass Analysis for Windows version 5.0 (Lighthouse Data, Odense, Denmark) and Protein Prospector (http://prospector.ucsf.edu/). Data are expressed as means ( SD.

Results DCA Bioactivation Studies with BSA and Rat Liver Cytosol. GSTZ1-1 catalyzes the bioactivation of DCA to S-(R-chlorocarboxymethyl)GSH, a reactive intermediate that covalently modifies and inactivates the enzyme or is hydrolyzed to glyoxylate (24-26). To determine whether DCA-derived reactive intermediates react with protein-bound nucleophiles, BSA was included in reaction mixtures that contained [1-14C]DCA, GSH, and hGSTZ1c-1c. BSA was not modified with 14C when incubated with [1-14C]DCA alone, with [1-14C]DCA and GSH, or with [1-14C]DCA and hGSTZ1c-1c (Figure 1) but was, however, modified with 14C when incubated with [1-14C]DCA, GSH, and hGSTZ1c-1c (Figure 1). Hence, GSTZ1-1 catalyzes the bioactivation of DCA, and DCAderived reactive intermediates react with both GSTZ1-1 and non-GSTZ1-1 amino acid nucleophiles. To determine whether BSA was modified by S-(R-chlorocarboxymethyl)GSH, reaction mixtures were prepared with [35S]GSH. BSA was modified by [35S]GSH in the presence and absence of DCA and hGSTZ1c-1c, and similar levels of modification were determined by spot densitometry for all reaction mixtures (data not shown). Reduction with β-mercaptoethanol, DTT, or TCEP prior to SDS-PAGE analysis had no effect on the relative extents of modification. To determine whether other proteins are targets of DCA-derived reactive intermediates, dialyzed rat liver cytosol was incubated with [1-14C]DCA in the presence or absence of GSH. Rat liver cytosolic protein was modified by [1-14C]DCA, and the extent of the modification was increased in the presence of GSH (Figure 2). The binding of [1-14C]DCA was particularly increased for proteins with estimated masses of 25.1, 26.5, 39.9, 41.5, and 58.2 kDa. The binding of [1-14C]DCA may have been slightly decreased for a protein with an estimated mass of 16.0 kDa.

GSTZ1-1-Catalyzed Bioactivation of DCA

Figure 1. GSH- and GSTZ1-1-dependent binding of [1-14C]DCA to BSA. BSA was incubated with [1-14C]DCA (lane a); [1-14C]DCA and GSH (lane b); [1-14C]DCA and hGSTZ1c-1c (lane e); or [1-14C]DCA, GSH, and hGSTZ1c-1c (lane f), as described in the Experimental Procedures. Alternatively, hGSTZ1c-1c was incubated with [1-14C]DCA (lane c) or [1-14C]DCA and GSH (lane d). Unbound radioactivity was removed by ultrafiltration with three dilution cycles. The proteins were resolved by SDS-PAGE on 12% acrylamide gels. The gels were loaded with 100 µg of BSA and 0.2 µg of hGSTZ1c-1c, and the resolved proteins were stained with Coomassie R-250 (A) and analyzed by phosphorimaging (B). M, molecular weight markers; M*, 14C-methylated molecular weight markers.

DCA Bioactivation Studies with Model Peptides. Studies with model peptides were undertaken to identify the amino acid targets and structures of DCA-derived adducts. AF was chosen because it contains the nucleophilic amino acid Lys. IL8 was chosen because it contains the nucleophilic amino acid Cys and because it contains an acetylated N-terminus. Reaction mixtures were analyzed by LC/MS. Studies with AF. Except for a small amount of a species (AF + O) whose mass (m/z 551.0, [M + 2H]2+) corresponded to the mass of oxidized AF, no adducted peptides were observed in reaction mixtures that contained AF, DCA, and hGSTZ1c-1c (Figure 3A). Two adducted peptides (AF + 56 no. 1 and 2) were, however, observed in reaction mixtures that contained AF, DCA, GSH, and hGSTZ1c-1c (Figure 3B). The mass of both adducted peptides (m/z 571.0, [M + 2H]2+) corresponded to the mass of the unmodified peptide (m/z 543.0, [M + 2H]2+) plus glyoxylic acid (74 Da) minus water. No peaks were observed in the extracted ion chromatogram of the ion that corresponds to the mass of AF adducted by S-(Rchlorocarboxymethyl)GSH (m/z 724.5, [M + 2H]2+; data not shown). Adducted peptides with masses and tR values identical to those previously observed were also observed in reaction mixtures that contained AF and glyoxylate (Figure 3C). Reaction mixtures that contained AF and glyoxylate also contained a small amount of a third adducted peptide (AF + 112). The mass of the peptide (m/z 599.0, [M + 2H]2+) corresponded to the mass of the unmodified peptide plus two molecules of glyoxylic acid minus two molecules of water.

Chem. Res. Toxicol., Vol. 17, No. 5, 2004 653

Figure 2. Potentiation of [1-14C]DCA binding to rat liver cytosolic protein by GSH. Dialyzed rat liver cytosol was incubated with [1-14C]DCA in the absence (lane a) or presence (lane b) of GSH. The proteins were resolved by SDS-PAGE (150 µg/ lane) on 12% acrylamide gels. The resolved proteins were transferred to nitrocellulose membranes and stained with Ponceau S (A) and analyzed by phosphorimaging (B). M*, 14Cmethylated molecular weight markers.

MS/MS analysis of unmodified AF (m/z 543.0, [M + 2H]2+) gave several b, b - NH3, b - H2O, b + H2O, y, and miscellaneous fragment ions (Table 1). MS/MS analysis of the first adducted peptide (AF + 56 no. 1, m/z 571.0, [M + 2H]2+) gave a few b and y series fragment ions, an abundant y6 + H2O ion, and a highly abundant ion (m/z 549.0, [M + 2H]2+) that resulted from the neutral loss (44 Da) of CO2 (data not shown). b5 and b7 ions were shifted by +56 m/z, whereas y6, y6 + H2O, y7, and y8 ions were not shifted in mass. The y8 ion, although not shifted in mass, was reduced to very low abundance as compared with its abundance in the MS/MS spectrum of unmodified AF. MS3 analysis of the ion (m/z 549.0, [M + 2H]2+) that corresponded to the neutral loss of CO2 gave a more complete set of fragment ions (Table 1). All singly charged N-terminal fragment ions that were observed [b2 - (b8 + H2O)] were shifted by +12 m/z, and all doubly charged N-terminal fragment ions that were observed [b62+ - (b8 + H2O2+)] were shifted by +6 m/z. None of the y series ions [(y2 - H2O) - y8] were shifted in mass, although the y8 ion, like in the MS/MS spectrum, was reduced to very low abundance as compared with its abundance in the MS/MS spectrum of unmodified AF. The MS/MS and MS3 spectra of both adducted peptides (AF + 56 no. 1 and AF + 56 no. 2) were identical except that the relative abundances of some fragment ions were slightly different (data not shown). The data indicated that the adducts were located on the N-terminal His and not on Lys-5 and that the adduction resulted in the formation of diastereomers since the spectra of both adducted peptides were identical. The mass and location of the modification were consistent with a reaction between glyoxylate and the

654

Chem. Res. Toxicol., Vol. 17, No. 5, 2004

Figure 3. LC/MS analysis of the reaction of AF with glyoxylate. AF was incubated with DCA and hGSTZ1c-1c (A); with DCA, GSH, and hGSTZ1c-1c (B); or with glyoxylate (C) for 16-24 h at 37 °C in 0.1 M potassium phosphate (pH 7.4). The reaction mixture that contained AF and glyoxylate was incubated for an additional 30 min at 20-22 °C with sodium borodeuteride (4 mM) (D). The reaction mixtures were analyzed by LC/MS in total ion mode, and all ions between 100 and 1200 m/z were recorded. Peak 1, AF + O (m/z 551.0, [M + 2H]2+); 2, AF (m/z 543.0, [M + 2H]2+); 3, AF + 56 no. 1 (m/z 571.0, [M + 2H]2+); 4, AF + 56 no. 2 (m/z 571.0, [M + 2H]2+); 5, AF + 112 (m/z 599.0, [M + 2H]2+); and 6, AF + 59 (m/z 572.5, [M + 2H]2+).

N-terminal amino group to form an imine (Scheme 1) and that the diastereomers may be attributed to syn and anti conformations of the imine. To test this hypothesis, reaction mixtures that contained the adducted peptides were reduced with sodium borodeuteride. Sodium borodeuteride provided greater mass resolution between the reduced and the nonreduced adducted peptides than sodium borohydride (i.e., 3 vs 2 Da), which was important since MS analysis was carried out on [M + 2H]2+ ions. Reduction with sodium borodeuteride led to the formation of a new chromatographically resolved species (AF + 59, Figure 3D). The mass of the new species (m/z 572.5, [M + 2H]2+) was 3 Da greater than the mass of the adducted peptides and, therefore, agreed with the mass expected from sodium borodeuteride reduction of an imine. MS/ MS analysis of the m/z 572.5 ion also supported the proposed NR-carboxymethyl structure (Scheme 1). All singly charged N-terminal fragment ions that were observed [(b2 - H2O) - (b8 + H2O)] were shifted by +59 m/z, and all doubly charged N-terminal fragment ions that were observed [(b6 - NH32+) - (b8 + H2O2+)] were shifted by +29.5 m/z (Table 1). An ion corresponding to the mass of a His immonium ion that had been shifted by +59 m/z was also observed. None of the y series ions that were observed [(y2 - H2O) - y8] were shifted in mass, and the relative abundance of the y8 ion was

Anderson et al.

similar to its abundance in the MS/MS spectrum of unmodified AF. Interestingly, the formation of the NRcarboxymethylated peptide was not accompanied by a reduction in the peak area of the adducted peptides but was, rather, accompanied by a reduction in the peak area of the unmodified peptide. The data indicated that the adducted peptides were not imines since they were not reduced by sodium borodeuteride. The observed decrease in the abundance of the y8 ion in the mass spectra of the adducted peptides indicated a bridge structure between His-1 and Asp-2. A range of aldehydes and ketones, including acetaldehyde, react with peptide N-termini to form 4-imidazolidinones (33-35). The structures of the adducted peptides were, therefore, characterized as 2-carboxy-4-imidazolidinones (Scheme 1). The introduction of a new chiral center at the 2-position of the ring would explain the formation of diastereomers. The data also indicated that the reaction mixtures contained a population of the intermediate carbinolamine (Scheme 1) or imine since the NR-carboxymethylated peptide was produced upon reduction with sodium borodeuteride but that these species reverted back to the unmodified peptide on LC/MS analysis. Both imines and carbinolamines are susceptible to reduction with sodium borodeuteride; carbinolamines and imines are reduced to secondary amines (36). Studies with IL8 and CAM-IL8. No adducted peptides were observed in reaction mixtures that contained IL8, DCA, and hGSTZ1c-1c (Figure 4A). A small quantity of an adducted peptide (IL8 + 74) was, however, observed in reaction mixtures that contained IL8, DCA, GSH, and hGSTZ1c-1c (Figure 4B). The mass of the adducted peptide (m/z 539.3, [M + 2H]2+) corresponded to the mass of the unmodified peptide (m/z 502.4, [M + 2H]2+) plus glyoxylic acid (74 Da). No peaks were observed in the extracted ion chromatogram of the ion that corresponds to the mass of IL8 adducted by S-(R-chlorocarboxymethyl)GSH (m/z 683.9, [M + 2H]2+; data not shown). Incubation of IL8 with high concentrations of glyoxylate for 24 h gave five partially resolved and adducted peptides (Figure 4C). Four of the adducted peptides (IL8 + 74 no. 1-4) had identical masses (m/z 539.3, [M + 2H]2+) and similar tR values to the adducted peptide (IL8 + 74) that was previously observed. The mass (m/z 576.2, [M + 2H]2+) of the fifth adducted peptide (IL8 + 148), which eluted as a broad peak, corresponded to the mass of the unmodified peptide plus two molecules of glyoxylic acid. The peak corresponding to unmodified IL8 was replaced by two partially resolved peptides with masses (m/z 502.4, [M + 2H]2+) that also corresponded to unmodified IL8. These partially resolved and apparently unmodified peptides were observed immediately after addition of glyoxylate (Figure 4D); no incubation was required. Mass spectra of IL8 and glyoxylate-adducted IL8 also contained ions that corresponded to the neutral loss of water and the adduction of potassium (i.e., [M + K]1+, [M + 2K]2+, and [M + H + K]2+) (data not shown). Furthermore, the mass spectra of the m/z 539.3 adducted peptides (IL8 + 74 no. 1-4) contained significant amounts of the ion (m/z 502.4, [M + 2H]2+) that corresponded to unmodified IL8, and the mass spectrum of the m/z 576.2 adducted peptide (IL8 + 148) contained a significant amount of the ions that corresponded to unmodified IL8 and singly adducted IL8 (i.e., m/z 502.4 and m/z 539.3). MS/MS analysis of the m/z 539.3 and m/z 576.2 ions gave only fragment ions that corresponded to the unmodified

GSTZ1-1-Catalyzed Bioactivation of DCA

Chem. Res. Toxicol., Vol. 17, No. 5, 2004 655

Table 1. Theoretical and Observed Ion Masses for AF and Glyoxylic Acid-Adducted AF

ion

typea

His immonium y2 - H2O b2 - H2O y2 b2 b5 - NH32+ b6 - NH32+ y3 b62+ b3 - H2O a72+ b72+ NKVL b8 - NH32+ b82+ b8 + H2O2+ b4 c4 -NH32+ DMNKV - H2O y5 DMNKV b5 - NH3 b5 y6 b6 - H2O b6 DMNKVLD y7 b7 y8 b8 b8 + H2O

AF

AF/decarboxylated imidazolidinone-1 (AF + 56 no. 1)

NR-carboxymethyl AF (AF + 59)

MS/MS of (m/z 543.0, [M + 2H]2+)

MS3 of (m/z 549.0, [M + 2H]2+)

MS/MS of (m/z 572.5, [M + 2H]2+)

theoretical mass (monoisotopic)

observed mass

abundance

mass shift

observed mass

abundance

mass shift

observed mass

abundance

mass shift

110.07 229.12 235.08 247.13 253.09 305.13 354.66 360.21 363.17 366.12 405.72 419.72 455.30 468.72 477.23 486.23 498.18 515.20 534.26 570.27 587.38 588.28 609.25 626.27 701.42 707.33 725.34 816.39 832.46 838.42 947.49 953.45 971.46