54
Chem. Res. Toxicol. 2001, 14, 54-61
Reaction of Trichloroethylene Oxide with Proteins and DNA: Instability of Adducts and Modulation of Functions Hongliang Cai† and F. Peter Guengerich* Department of Biochemistry and Center in Molecular Toxicology, Vanderbilt University School of Medicine, Nashville, Tennessee 37232-0146 Received August 23, 2000
Trichloroethylene (TCE) shows several types of toxicities, some of which may be the result of bioactivation. Oxidation by P450s yields the electrophile TCE oxide. We previously analyzed N6-acyllysine adducts formed from the reaction of TCE oxide with proteins [Cai, H., and Guengerich, F. P. (2000) Chem. Res. Toxicol. 13, 327-335]; however, we had been unable to measure ester adducts under the prolonged conditions of proteolysis and derivatization. Protein amino acid adducts were directly observed by mass spectrometry during the reaction of TCE oxide with the model polypeptides insulin and adrenocorticotropic hormone (ACTH, residues 1-24). The majority (80%) of the protein adducts were unstable under physiological conditions and had a collective t1/2 of ∼1 h, suggesting that they are ester type adducts formed from reactions of Cys, Ser, Tyr, or Thr residues with intermediates formed in TCE oxide hydrolysis. Synthetic O-acetyl-L-Ser and O-acetyl-L-Tyr had half-lives of 1 h and 10 min at pH 8.0, respectively, similar to the stabilities of the protein adducts. The effects of TCE oxide adduct formation on catalytic activities were examined with five model enzymes. No recovery of catalytic activity was observed during the reaction of TCE oxide with two model enzymes for which the literature suggests roles of a Lys, rabbit muscle aldolase and glucose-6-phosphate dehydrogenase. However, in the cases of papain (essential Cys residue in the active site), R-chymotrypsin (critical Ser residue), and D-amino acid oxidase (essential Cys and Tyr residues), time-dependent recoveries of enzyme activity were observed following reaction with TCE oxide or either of two model nucleophiles (dichloroacetyl chloride and acetic formic anhydride), paralleling the kinetics of removal of adducts from insulin and ACTH. Formation of adducts (∼2%) was detected in the direct reaction of TCE oxide with 2′-deoxyguanosine, but not with the other three nucleosides found in DNA. During the reaction of TCE oxide with a synthetic 8-mer oligonucleotide, formation of adducts was observed by mass spectrometry. However, the adducts had a t1/2 of 30 min at pH 8.5. These results indicate the transient nature of the adducts formed from the reaction of TCE oxide with macromolecules and their biological effects.
Introduction 1,1,2-Trichloroethylene (TCE)1 is a volatile organic solvent that is widely used as a degreasing agent, a polymer precursor, and a dry cleaning agent. It has been estimated that approximately 34% of the drinking water supplies tested in the United States have TCE contamination (1). Because of its widespread industrial use and improper disposal, TCE has become a major environmental pollutant and is one of the most abundant organic contaminants found in Superfund sites (1, 2). TCE, like other chlorinated ethylenes, can generate toxic effects and induces liver, lung, and kidney tumors in rodents * To whom correspondence should be addressed: Department of Biochemistry and Center in Molecular Toxicology, Vanderbilt University School of Medicine, Nashville, TN 37232-0146. Telephone: (615) 322-2261. Fax: (615) 322-3141. E-mail: guengerich@ toxicology.mc.vanderbilt.edu. † Current address: Pharmacokinetics, Dynamics, and Metabolism, Pfizer Global R&D, 2800 Plymouth Rd., Ann Arbor, MI 48105. 1 Abbreviations: TCE, trichloroethylene; ACTH, adrenocorticotropic hormone (fragment consisting of residues 1-24); MOPS, 3-(N-morpholino)propanesulfonic acid; Tris, tris(hydroxymethyl)aminomethane; BTEE, benzoyl-L-Tyr ethyl ester; L-BAPNA, N2-benzoyl-L-Arg-p-nitroanilide. The nucleoside abbreviations are standard for the journal.
(3-5). Three major hypotheses have been developed to account for the rodent carcinogenicity: genotoxicity of TCE oxidation products (6), genotoxicity of GSH conjugates (4, 7), and peroxisome proliferation by the oxidation products Cl2CHCO2H and Cl3CCO2H (8, 9). Recently, high TCE exposure (either occupationally or environmentally) has been suggested to be related to the development of immune system dysfunction and autoimmune disorders in humans (10-12). Currently, TCE is undergoing a thorough evaluation and risk assessment by the National Center for Environmental Assessment of the United States Environmental Protection Agency (13). TCE is metabolized via two general pathways, an oxidative pathway in which the first step is catalyzed by P450 and a GSH-dependent pathway in which the initial step is catalyzed by GSH transferase (14). Only a small fraction of TCE is metabolized by the GSH conjugation pathway, primarily by GSH transferase in the liver, followed by further metabolism in the kidney (14). Key oxidative metabolites of the P450 pathway, including Cl3CCHO [chloral (hydrate)], Cl3CCO2H, and Cl2CHCO2H, are thought to be critical for development of toxicity or
10.1021/tx000185n CCC: $20.00 © 2001 American Chemical Society Published on Web 12/07/2000
Trichloroethylene-Protein and -DNA Adducts
carcinogenicity (13, 14). Metabolites of TCE derived from the GSH conjugation pathway that may be responsible for the renal toxic effects of TCE include S-(1,2-dichlorovinyl)glutathione, S-(1,2-dichlorovinyl)-L-Cys, and S-(1,2dichlorovinyl)-L-Cys sulfoxide (13). The metabolism of TCE has been summarized in detail (14), and the toxicity and mutagenicity of TCE and its metabolites have been revisited (13, 15). However, there is much research yet to be done to assess the risk of TCE to human health. Clewell et al. reported a physiologically based pharmacokinetic model for risk assessment of TCE and its metabolites (16-18). This model was based on the view that all Cl3CCHO is generated from rearrangement of an oxygenated P450-TCE intermediate. Previously, physiologically based pharmacokinetic modeling was also based on the concept that all Cl3CCHO is formed from TCE oxide rearrangement (19, 20). In the 1970s, controversy existed over whether an epoxide intermediate is the only route of TCE oxidation (21, 22). Although we found evidence for TCE oxide formation, TCE oxide was found not to be an obligate intermediate in the formation of chloral, and we concluded that Cl3CCHO formation proceeds via chloride migration in an oxygenated P450-TCE intermediate (23). We also suggested that TCE oxide itself may not be the intermediate responsible for irreversible binding to protein and DNA (23). Consistent with our conclusion, Green and Prout (24) concluded that there was evidence in favor of only low levels of TCE oxide formation in the oxidative metabolism of TCE in rat or mouse liver microsomes. Recently, we elucidated the roles of acyl halides derived from TCE oxide hydrolysis in the modification of protein Lys groups (25, 26). Previously, we demonstrated that TCE-protein adducts are much more abundant than any DNA adducts (27). Consistent with our observation, Kautianinen et al. (28) also found that the level of protein binding (1.4 ng/g of protein) was 100 times higher than the level of DNA binding (1.5 pg/g of DNA) in B6C3F1 male mice exposed to TCE at low doses (relevant to human exposure). Pumford and his associates used a polyclonal antibody to detect putative dichloroacetylated proteins in livers of mice exposed to TCE (2). Immunoblots revealed the presence of two major TCE adducts at 50 and 100 kDa in liver microsomal fractions from male B6C3F1 mice. One of the adducts has been reported to be P450 2E1, an enzyme known to oxidize TCE (1, 29-31). However, further characterization of adducts and quantitation has not been reported. TCE oxide has been postulated to be responsible for the covalent binding of TCE with protein and, to a much lesser extent, DNA (32-34). Previously, we had not found a good correlation between conversion of TCE to Cl3CCHO, TCE oxide, and protein adducts in rat, mouse, and human liver microsomes (27). Therefore, a more detailed reinvestigation of TCE oxide and its protein and possible DNA adducts is needed. We have characterized the reaction of TCE oxide with proteins and found that both N6-formyl-Lys and N6(dichloroacetyl)-Lys are formed, with the level of the former exceeding the latter (26). However, our preliminary studies on the stability of O-acyl amino acids indicated that these adducts were unstable under the prolonged conditions needed for protease digestion, derivatization, and HPLC/MS analysis. Because of our concerns about the contributions of unstable adducts to the estimate of the total level, we utilized direct MS
Chem. Res. Toxicol., Vol. 14, No. 1, 2001 55
analysis of modified proteins as a means of characterizing total adduct levels. The methodology was extended to the reaction of TCE oxide with nucleosides and oligonucleotides. We also considered the transient nature of the inhibition of enzymes by TCE oxide in the context of adduct lability.
Experimental Procedures Caution: Care should be exercised in handling TCE oxide, dichloroacetyl chloride, and formic acetic anhydride, all of which are reactive electrophiles. Use adequate skin protection and avoid inhalation. Enzymes. Rabbit muscle aldolase, Leuconostoc mesenteroides glucose-6-phosphate dehydrogenase, R-chymotrypsin, D-amino acid oxidase, papain, catalase, trypsin, proteinase K, super oxidase dismutase, insulin, and adrenocorticotropic hormone (ACTH, residues 1-24) were obtained from Sigma (St. Louis, MO). Spectroscopy. 1H and 13C NMR spectra were recorded on a Bruker AM 400 spectrometer (operating at 400.13 MHz in the 1H mode, Bruker, Billerica, MA) at 25 °C in the Vanderbilt facility. Samples were prepared in either D2O or CD3OD (100.0 at. % D grade, Aldrich Chemical Co., Milwaukee, WI). UV/visible spectroscopy was carried out using a Cary 14/OLIS spectrophotometer equipped with a circulating water bath for temperature control (On-Line Instrument Systems, Bogart, GA). Mass spectrometry was carried out with a Waters 2690 separations module (Alliance) (Waters Corp., Milford, MA) and a Finnigan MAT TSQ/SSQ 7000 series mass spectrometer (Finnigan, Sunnyvale, CA). MS studies were performed using either the positive ion mode (for analysis of insulin, ACTH, and nucleoside adducts) or negative ion mode (for analysis of oligonucleotide adducts) with an electrospray needle voltage of 4.5 kV. N2 was used as the sheath gas (70 psi) to assist nebulization and as the auxiliary gas (10 psi) to assist with desolvation. The stainless steel capillary was heated to 200 °C, and the electrospray interface and mass spectrometer parameters were optimized to obtain maximum sensitivity. The tube lens and the heated capillary were operated at 75 and 20 V, respectively, and the electron multiplier was set at either 1900 or 2000 V. Chemicals. Solvents for HPLC and HPLC/MS were HPLC grade (EM Science, Gibbstown, NJ). An 8-mer oligonucleotide (5′-GCCTAGAT-3′) was purchased from Midland Certified Reagent Co. (The Woodlands, TX). TCE was distilled (atmospheric pressure) to remove inhibitors prior to epoxide synthesis. Reagents for synthesis were ACS grade or better, unless otherwise specified. Syntheses. (1) TCE Oxide. TCE oxide was synthesized by m-chloroperbenzoic acid treatment of TCE and partially distilled (25). Concentrations (in residual TCE) were determined colorimetrically using the 4-(4-nitrobenzyl)pyridine reagent (23, 25, 35). (2) Acetic Formic Anhydride (36). The synthesis has been described previously (26). (3) O-Acetyl-L-Ser. O-Acetyl-L-Ser was synthesized by a published method (37). The general experimental procedure is as follows. The amino acid (10 mmol) was dissolved in 6 N HCl (2 mL). CH3CO2H (2 mL) was added, and the solution was cooled to 0 °C in an ice bath. Acetyl chloride (20 mL) was then added slowly to the beaker. After 30 min, 3 volumes of (CH3CH2)2O was added and the O-acetyl-L-Ser‚HCl precipitated. The compound was recovered by filtration, washed with (CH3CH2)2O, dried in vacuo, and recrystallized from C2H5OH (75% yield): 1H NMR (D2O) δ 2.07 (s, 3H, COCH3), 4.36 (t, 1H, CHCH2), 4.50 (m, 2H, CHCH2); 13C NMR (D2O) δ 20.4 (COCH3), 52.6 (CHCH2), 62.3 (CHCH2), 169.6 (COCH3), 173.6 (COOH). (4) O-Acetyl-L-Tyr. O-Acetyl-L-Tyr was synthesized by published methods (37, 38) with a yield of 45%: 1H NMR (D2O) δ 2.19 (s, 3H, COCH3), 3.12 (m, 2H, CHCH2), 4.12 (t, 1H, CHCH2), 6.75 [d, 2H, CH2C(CHCH)2CO], 7.03 [d, 2H, CH2C(CHCH)2CO]; 13C NMR (D2O) δ 20.7 (COCH3), 35.1 (CHCH2), 54.6
56
Chem. Res. Toxicol., Vol. 14, No. 1, 2001
(CHCH2), 116.3 [CH2C(CHCH)2CO], 122.7 [CH2C(CHCH)2CO], 126.0 [CH2C(CHCH)2CO], 131.2 [CH2C(CHCH)2CO], 169.6 (COCH3), 173.6 (COOH). Reactions. (1) Reaction of TCE Oxide with Insulin or ACTH. The indicated amounts of TCE oxide (10 µL of dry CH3CN) were added to a solution of insulin or ACTH (1 mg/mL in 200 µL of 10 mM NH4CH3CO2, pH 7). After 5 min, 20 µL of the reaction mixture was analyzed directly by HPLC/MS using a syringe pump (flow rate of 0.1 mL/min) with a Vydac C4 guard column (Vydac, Hesperia, CA) (placed between the injection loop and the mass spectrometer). The HPLC/MS analysis procedure is described as follows. Before the 20 µL sample was loaded and injected, the HPLC system was washed thoroughly with 20 mM NH4CH3CO2 for 5 min. The sample was then loaded, injected, and washed with 20 mM NH4CH3CO2 for 2 min. Subsequently, the eluting buffer was changed to H2O/CH3OH (50/50, v/v, containing 0.1% CH3CO2H, v/v), which eluted the insulin or ACTH. In an alternative experiment, the mixture of insulin or ACTH reacted with TCE oxide (100 µL, after reaction for 5 min) was treated with NaOH (adjusted to pH 12) for 5 min. The pH was then adjusted to 5 by addition of 0.1 N HCl. Subsequently, the mixture (20 µL) was subjected to HPLC/MS for analysis of adducts. Control experiments were also carried out with unreacted insulin or ACTH following the same procedure. Quantitation of the adducts was carried out with the assumption that all modified polypeptides yield a response similar to that of the unmodified polypeptide. Peak heights were used, along with knowledge of the number of formyl or dichloroacetyl modifications derived from the presence of multiples of 28 or 110/112 added to the MH+ of the unmodified polypeptide. Normalization was on the basis of the sum of all peak heights, including the unmodified polypeptide. (2) Reaction of TCE Oxide with 8-mer Oligonucleotide (5′-GCCTAGAT-3′). TCE oxide (15 mM) was added to a solution of the 8-mer oligonucleotide (0.78 mM, 100 µL in H2O). After 15 min, a 0.10 µM oligonucleotide solution in H2O/2-propanol (50/50, v/v, containing 2.5 mM imidazole and 2.5 mM piperidine) was made from the reaction mixture (39), and subsequently, 20 µL of the solution was analyzed by HPLC/MS using a syringe pump (flow rate of 20 µL/min) with a Vydac C4 guard column (placed between the injection loop and the mass spectrometer). The solution used for HPLC/MS was H2O/2-propanol (50/50, v/v). A control experiment was carried out with the unmodified 8-mer oligonucleotide. (3) Reaction of TCE Oxide with Nucleosides. TCE oxide was added (to a final concentration of 15 mM) to a solution of either dAdo, dCyd, dGuo, or dThd (1 mM, 100 µL in H2O). After reaction for 15 min, HPLC/MS analysis was performed to identify possible nucleoside adducts. HPLC conditions (Zorbax Rx-C8 column, 2.1 mm × 150 mm, Mac-Mod, Chadds Ford, PA) were as follows: flow rate of 0.2 mL/min; solvent A, 10 mM NH4CH3CO2 in H2O with 2% CH3OH, v/v (pH 4.6); solvent B, 10 mM NH4CH3CO2/CH3OH, pH 7.4, 5/95, v/v; 100% A and 0% B at time zero, 90% A and 10% B after 10 min, 70% A and 30% B after 20 min, 0% A and 100% B after 30 min, 0% A and 100% B after 35 min, 100% A and 0% B after 40 min, and 100% A and 0% B after 50 min. In control experiments, TCE oxide was added to 100 µL of H2O and, after 15 min, nucleosides (1.0 mM, 100 µL) were added to the mixture; the remainder of the analytical procedure was carried out as described for the direct reaction with TCE oxide. Enzyme Assays. (1) Kinetic Assays of O-Acetyl-L-Ser and O-Acetyl-L-Tyr Stability. O-Acetyl-L-Ser or O-acetyl-LTyr (final concentration of 2 mg/mL) was dissolved in 100 mM tris(hydroxymethyl)aminomethane (Tris)-HCl buffer (pH 8.0). At varying times, 1 µL of the solution was spotted onto a TLC plate. The TLC plate was developed in n-butanol/CH3CO2H/H2O (v/v/v, 80/20/20) (40) to separate the acetyl derivatives from the amino acids. The quantitation of O-acetyl-L-Ser and O-acetylL-Tyr was performed after colorimetric reaction with ninhydrin (in acetone, 1%, w/v) (40) and analyzed using a NucleoTech GelExpert system (San Mateo, CA).
Cai and Guengerich (2) Glucose-6-Phosphate Dehydrogenase (41). L. mesenteroides glucose-6-phosphate dehydrogenase (100 µL of a 0.5 mg/ mL solution in 100 mM potassium phosphate buffer, pH 7.8) was reacted with TCE oxide (12 mM final concentration, added in CH3CN,
