Formation and Mass Spectrometric Analysis of DNA and Nucleoside

Vanderbilt University School of Medicine, Nashville, Tennessee 37232, and ... of Chemistry and Physics, Union University, Jackson, Tennessee 38305...
0 downloads 0 Views 166KB Size
Download PDF
Chem. Res. Toxicol. 2004, 17, 45-54

45

Formation and Mass Spectrometric Analysis of DNA and Nucleoside Adducts by S-(1-Acetoxymethyl)glutathione and by Glutathione S-Transferase-Mediated Activation of Dihalomethanes Glenn A. Marsch,†,‡ Sisir Botta,† Martha V. Martin,† W. Andrew McCormick,† and F. Peter Guengerich*,† Department of Biochemistry and Center in Molecular Toxicology, Vanderbilt University School of Medicine, Nashville, Tennessee 37232, and Department of Chemistry and Physics, Union University, Jackson, Tennessee 38305 Received July 17, 2003

The dihalomethane CH2Cl2 is an industrial solvent of potential concern to humans because of its potential genotoxicity and carcinogenicity. To characterize DNA damage by dihalomethanes, a rapid DNA digestion under acidic conditions was developed to identify alkali labile DNA-dihalomethane nucleoside adducts using HPLC-electrospray mass spectrometry. DNA digestion worked best using pH 5.0 sodium acetate buffer, a 30 min incubation with DNase II and phosphodiesterase II, and a 2 h acid phosphatase digest. DNA was modified with S-(1acetoxymethyl)glutathione (GSCH2OAc), a reagent modeling activated dihalomethanes. Adducts to G, A, and T were detected at high ratios of GSCH2OAc/DNA following digestion of the DNA with the procedure used here. The relative efficacy of adduct formation was G > T > A . C. The four DNA nucleosides were also reacted with the dihalomethanes CH2Cl2 and CH2Br2 in the presence of glutathione (GSH) and GSH S-transferases from bacteria (DM11), rat (GST 5-5), and human (GST T1-1) under conditions that produce mutations in bacteria. All enzymes formed adducts to all four nucleosides, with dGuo being the most readily modified nucleoside. Thus, the pattern paralleled the results obtained with the model compounds GSCH2OAc and DNA. CH2Cl2 and CH2Br2 yielded similar amounts of adducts under these conditions. The relative efficiency of adduct formation by GSH transferases was rat 5-5 > human T1-1 > bacterial DM11, showing that human GSH transferase T1-1 can form dihalomethane adducts under the conditions used. Although the lability of DNA adducts has precluded more sophisticated experiments and in vivo studies have not yet been possible, the work collectively demonstrates the ability of several GSH transferases to generate DNA adducts from dihalomethanes, with G being the preferred site of adduction in both this and the GSCH2OAc model system.

Introduction Halogenated hydrocarbons are of toxicological concern because of their extensive use in industry and their potential to be toxic and carcinogenic. Dihalomethanes are of interest, particularly CH2Cl2. The annual U.S. production of CH2Cl2 is currently about 3 × 108 kg (1, 2). Under certain conditions, high doses of CH2Cl2 can produce liver and lung tumors in mice (3-6). These findings have led to regulation of the limits of human exposure to CH2Cl2 (7, 8). However, the extrapolation between rodents and humans has not been fully defined, and the scientific basis of the risk assessment is still not well-established (9-12). We have been interested in CH2Cl2, its metabolism, and the relationship of the products to genotoxicity for several years (13-18). Ahmed et al. (19) and Kubic and Anders (20) first demonstrated that dihalomethanes are oxidized by P450s to CO, and subsequently, P450 2E1 * To whom correspondence should be addressed. Tel: (615)322-2261. Fax: (615)322-3141. E-mail: [email protected]. † Vanderbilt University School of Medicine. ‡ Union University.

Scheme 1. Dual Pathway for CH2Cl2 Metabolism

was found to be the dominant enzyme in this process (21). Another route of metabolism was shown to be GSH conjugation by GSTs1 (22) (Scheme 1). Subsequently, some of the θ-class GSTs were shown to be the most active in conjugating CH2Cl2 (23). Enhanced bacterial 1 Abbreviations: GSCH CH Cl, S-(2-chloroethyl)GSH; GSCH OAc, 2 2 2 S-(1-acetoxymethyl)GSH; GST, glutathione S-transferase; DNase, deoxyribonuclease; PDE, phosphodiesterase; NTA, nitrilotriacetate agarose; CID, collision-induced dissociation; ESI, electrospray ionization; MS, mass spectrometry; SRM, selected reaction monitoring. The abbreviations for the DNA bases and nucleosides are standard for this journal.

10.1021/tx034156z CCC: $27.50 © 2004 American Chemical Society Published on Web 12/04/2003

46

Chem. Res. Toxicol., Vol. 17, No. 1, 2004 Scheme 2. Adduct Structures of GSCH2OAc-Nucleoside Products

genotoxicity of dihalomethanes due to GSH conjugation has also been reported (24-27). An inherent problem in these assays was the dihalomethane mutagenicity observed in the absence of added GST and GSH. In our own work, we expressed mammalian GSTs in a Salmonella typhimurium tester strain and could clearly demonstrate the GST-induced mutagenicity of CH2Br2, CHBrCl, and CH2Cl2 (14) and, subsequently, other dihalomethanes (17). HCHO, the final breakdown product of the GST reaction (Scheme 1) (19), was shown not to be responsible for the mutagenicity (14, 17, 28, 29). The nature of the DNA adducts resulting from GSH conjugation of dihalomethanes has been a subject of interest. An inherent problem is the chemical stability of the DNA adducts. We found that GSCH2OAc could serve as a model of the putative S-(1-halomethyl)GSH products, and reaction of this electrophile with dGuo yielded the N2-guanyl adduct (14). A similar approach was used to prepare adducts of the other three nucleosides (Thd, dAdo, and dCyd) (18). Other work in this laboratory with GSTs showed that GST DM11, from a Methanobacterium species that can live on CH2Cl2 as a carbon source (30), was also capable of activating methylene dihalides to genotoxins (17). Kayser and Vuilleumier have shown that radiolabel from both [14C]CH2Cl2 and [35S]GSH was bound to DNA when incubations were done with GSH, CH2Cl2, and GST DM11 (31). Work with single-stranded homooligonucleotides (20-mers) yielded a pattern of G > C > A > T for GST DM11. In our own preliminary analysis of DNA modification by GSCH2OAc, we could only identify the Thd adduct following digestion (18). We have now extended our DNA analysis and have observed G, T, and A adducts in the apparent order G > T > A. Also, we used a different in vitro system with either bacterial GST DM11, rat GST 5-5, or human GST T1-1 with CH2Cl2 or CH2Br2 and were able to detect DNA adducts formed with all four bases catalyzed by these GST enzymes.

Experimental Procedures Caution: GSCH2OAc and GSCH2CH2Cl are potential mutagens and carcinogens and should be handled carefully according to appropriate environmental safety and health protocols. Chemicals. GSH, nucleosides, PDE II, DNase II, and acid phosphatases I, II, IV, and VII were purchased from Sigma Chemical Co. (St. Louis, MO) and used without further purification. CH2Cl2 and CH2Br2 were purchased from Aldrich Chemical Co. (Milwaukee, WI). The half-mustard GSCH2CH2Cl (32) and the dihaloalkane model compound GSCH2OAc (14) were prepared in this laboratory as described in detail earlier.

Marsch et al. GST Expression and Purification. 1. Rat GST 5-5. Rat GST 5-5 was expressed in Escherichia coli BL21 (DE3)pLys (Stratagene, La Jolla, CA) using a modification of a previously described method (33). Cells were transformed with the plasmid pKK233-2 containing the GST 5-5 (His-tagged) cDNA (14, 33). The cells were grown to an OD600 of 0.4-0.6 in Terrific Broth (Becton, Dickinson and Co., Sparks, MD) containing kanamycin (100 µg L-1) and induced with 0.1 mM D-β-isopropylthiogalactoside. The culture was incubated for 18 h at 28 °C, and cells were harvested by centrifuging at 3 × 103g for 15 min. Bacterial cells were resuspended in sonication buffer (20 mM potassium phosphate, pH 7.5, containing 25 µM phenylmethylsulfonyl fluoride, 1 µg leupeptin mL-1, 20% glycerol (v/v), and 2 mM 2-mercaptoethanol). Sonication was performed for 3 × 10 min intervals, with 10 min intermittent periods. The material was centrifuged at 105g for 45 min, and the supernatant was collected for purification. The cytosol (containing 5 mM imidazole, 2 mM 2-mercaptoethanol, and 0.5 M KCl) was loaded onto a 1.5 × 10 cm NiNTA column (Qiagen, Valencia, CA). The Ni-NTA column was previously equilibrated with 20 mM potassium phosphate buffer, pH 8.0, containing 20% glycerol (v/v), 0.5 M KCl, 2 mM 2-mercaptoethanol, and 5 mM imidazole. The NTA column was washed with 10 column volumes of equilibration buffer, followed by washing with 10 column volumes of the same buffer containing 10 mM imidazole instead of 5 mM imidazole, and eluted with a 500 mL linear gradient of 10 mM imidazole to 200 mM imidazole in 20 mM potassium phosphate buffer (pH 8.0) containing 20% glycerol (v/v), 0.5 M KCl, and 2 mM 2-mercaptoethanol. Eluted fractions were assayed using SDS-PAGE (12% gel, w/v). The fractions containing the highest concentrations of GST 5-5 were pooled and concentrated 4-fold, using Millipore Diaflow PM10 ultrafiltration membranes (Millipore, Bedford, MA). Concentrated samples were dialyzed twice (4 °C, 8 h) against 100 volumes of 30 mM potassium phosphate buffer (pH 8.0) containing 10% glycerol (v/v) and 2 mM 2-mercaptoethanol. 2. Human GST T1-1. Human GST T1-1 was expressed in E. coli Bl21 (DE3)pLys cells (33). The cells were transformed with the expression plasmid pET 24d containing the GST T1h (His-tagged) cDNA. The expression and purification conditions for GST T1-1 were the same as those described for GST 5-5 (vide supra). 3. Bacterial GST DM11. GST DM11 was expressed in E. coli and purified as described previously (33). Preparation of Duplex DNA Solutions. Double-stranded calf thymus DNA (Sigma) was dissolved in H2O and sonicated for 20 min at 58% amplitude in a Fisher FS9H sonicator water bath (Fisher Scientific, Pittsburgh, PA). The DNA was repurified by a phenol:CHCl3:isoamyl alcohol extraction (25:24:1, v/v/v) followed by three C2H5OH precipitations. The resulting DNA pellets were resuspended in H2O, and the purity and doublestranded character of the DNA were monitored by UV spectroscopy. Formation of Duplex DNA Modified By the Model Compound GSCH2OAc. GSCH2OAc (0, 42, 85, or 170 mg) was quickly dissolved in 1.5 mL of ice-cold 83 mM sodium acetate buffer (pH 5.0) and immediately mixed with 5.0 mg of calf thymus DNA (dissolved in 1.0 mL of sodium acetate buffer). The reaction progressed for 30 min at room temperature. The DNA was precipitated with two volumes of cold C2H5OH, washed twice with 70% C2H5OH, and dried in vacuo. The first step in the DNA digest used porcine spleen DNase II and bovine spleen PDE II, which hydrolyzed DNA to 3′-phosphate nucleotides during incubation for 30 min at 37 °C. The concentrations used were the following (18): DNA, 1 mM (measured in nucleotides); sodium acetate buffer (pH 5.0), 83 mM; MgSO4, 8.3 mM; NaCl, 20 mM; DNase II, 2 U (µg DNA)-1; and PDE II, 0.74 U (mg DNA)-1. Type II acid phosphatase from white potato [3.0 U (mg DNA)-1] was then added to the solution, and the incubation continued for 2 h. More details of the DNA digest optimization and a

Glutathione-CH2-DNA Adducts scheme summary are presented in the Supporting Information section. Enzymes were removed with Amicon Centricon-30 membranes (Mr cutoff ) 30 kDa), and the filtrate was stored at -80 °C until further analysis. The lability of the dihalomethane adducts necessitated a rapid, efficient digest with all aspects of the experiment performed in 1 day (28). Formation of Nucleoside Adducts in the Presence of Dihaloalkanes and GSTs. 1. Bacterial GST DM11 Reactions. In one set of reactions, aliquots of CH2Cl2-saturated H2O [203 mM at 25 °C (34)] (corresponding to 0, 2.6, 5.1, 10, 20, or 30 µmol), 500 µg of nucleosides (125 µg each of dGuo, dAdo, Thd, and dCyd, for a total of 2.0 µmol), and GSH (310 µg, 1.0 µmol) were added to 0.10 M potassium N-(2-hydroxyethyl)piperazineN′-(2-ethanesulfonate) buffer (pH 7.0) in a total sample volume of 0.50 mL. The GSH and nucleotides were added as 20 mM and 4 mg mL-1 solutions, respectively. GST DM11 (13 nmol) was added last, and the samples were incubated for 30 min at 37 °C. The pH was then lowered to 5.0 with CH3CO2H to attenuate the rate of adduct degradation (18). GST DM11 protein was removed with Amicon Centricon-10 membranes (Mr cutoff ) 10 kDa). Reactions were stored at -80 °C or in dry ice and immediately assayed by HPLC-MS after thawing. In another set of reactions, aliquots of CH2Br2-saturated H2O [67 mM at 25 °C (34)] corresponding to 0, 2.6, 5.1, and 10 µmol were used instead of CH2Cl2. Otherwise, all reagents and procedures were the same as with the CH2Cl2 reactions (vide supra). 2. Rat GST 5-5 Reactions. These reactions were performed as in the CH2Cl2 + GST DM11 set, with the following exceptions. For rat GST 5-5 + CH2Cl2 reactions, aliquots of CH2Cl2 (from a 203 mM CH2Cl2-saturated H2O solution) corresponding to 0, 2.0, 4.1, 8.1, 16, or 24 µmol were added to a set of six samples. Rat GST 5-5 (7.5 nmol) was added to each sample. In the other set of rat GST 5-5 reactions, aliquots of CH2Br2 (from a 67 mM CH2Br2-saturated H2O solution) corresponding to 0, 2.0, 4.0, or 8.0 µmol were used instead of CH2Cl2. Rat GST 5-5 (7.5 nmol) was added to each CH2Br2 sample. Otherwise, all of the reagents and protocols were the same as with the GST DM11 + CH2Cl2 reactions (vide supra). 3. Human GST T1-1 Reactions. These reactions were done as in the CH2Cl2 + GST DM11 set, with the following exceptions. For human GST T1-1 + CH2Cl2 reactions, CH2Cl2 (from a 203 mM CH2Cl2-saturated H2O solution) corresponding to 0, 2.0, 4.1, 8.1, 16, or 24 µmol) was added to each of a set of six samples. Human GST T1-1 (12 nmol) was added to each sample. In the other set of human GST T1-1 reactions, CH2Br2 (from a 67 mM CH2Br2-saturated H2O solution) corresponding to 0, 2.0, 4.0, or 8.0 µmol was used instead of CH2Cl2. Human GST T1-1 (12 nmol) was added to each sample. Reagents and protocols were the same as for the GST DM11 + CH2Cl2 reactions (vide supra). HPLC-MS of Adducts. 1. HPLC System. Dihaloalkanenucleoside adduct samples were analyzed using HPLC-MS. The autosampler and HPLC system consisted of an Alliance 2690 Separations Module from Waters (Milford, MA), and the ThermoElectron TSQ Quantum HPLC-ESI-MS instrument used a Surveyor autosampler and HPLC pump (ThermoElectron, San Jose, CA). A H2O/CH3OH gradient was used to separate dihaloalkane adducts. Buffer A contained CH3OH and 10 mM NH4CH3CO2 (pH 4.5), 2-98, v/v. Buffer B contained CH3OH and 10 mM NH4CH3CO2 (pH 4.5), 95-5, v/v. When samples were introduced by direct infusion, a Harvard Apparatus (Holliston, MA) model 22 syringe pump was used at a flow rate of 10-20 µL min-1. A small bore Precision ODS octadecylsilane column (5 µm, 2.1 mm × 100 mm, Mac-Mod, Chadds Ford, PA) was used in the separations. The flow rate was initially 400 µL min-1 and decreased linearly to 200 µL min-1 after 4 min, during which time the pump valve diverted column eluent to waste as a desalting measure. Thereafter, the flow rate was 200 µL min-1 during the rest of the chromatography. The gradient was as follows: 0-4 min, B ) 0%; 4-7 min, B ) 0-5%; 7-18 min, B

Chem. Res. Toxicol., Vol. 17, No. 1, 2004 47 ) 5-25%; 18-23 min, B ) 25-100%; 23-26 min, B ) 100% (all v/v). An on-line UV monitor (Waters Corporation) was used to record separated analytes absorbing at 260 nm in order to correlate ions with molecules bearing DNA base chromophores. 2. MS of Adducts. MS was performed using TSQ-7000 and ThermoElectron Quantum (ThermoElectron) triple-stage quadrupole mass spectrometers, with a standard API-1 ESI source outfitted with a 150 µm ID (TSQ 7000) or a 100 µm ID (TS Quantum) deactivated fused silica capillary. The mass spectrometers were operated in the positive ion mode. MS detection parameters were optimized using dGuo as a standard and depended on the type of experiment conducted and the flow rate of introduced sample. Typical TSQ 7000 parameters for a flow rate of 200 µL min-1 from an HPLC pump were as follows: N2 sheath gas, 78 psi; N2 auxiliary gas, 12 psi; spray voltage, 5.7 kV; tube lens offset, 110 V; capillary temperature, 200 °C. Typical TSQuantum parameters were as follows: N2 sheath gas, 21 psi; N2 auxiliary gas, 25 psi; spray voltage, 4.1 kV; tube lens offset, 100 V; capillary (ion transfer tube) temperature, 280 °C. Data acquisition and spectral analysis for the TSQ 7000 were conducted using Finnigan ICIS software, version 8.3.2, on a Digital Equipment Co. Alpha workstation. TS Quantum data were processed with Xcalibur Software, version 1.3, from ThermoElectron on a Dell Optiplex GX240 2.4 GHz Pentium IV computer using a Microsoft Windows 2000 operating system to control all instruments and process data. For qualitative analysis, different scan modes were applied. Simple full scans covered the m/z range from 100 to 2000 with scan times up to 2 ms/amu (the range was extended to check condensation products). For the TSQ 7000, tandem MS experiments were conducted with Ar at 2.5 mTorr as the collision gas and a collision voltage of -22 V. For the TSQuantum, tandem MS experiments were conducted with Ar at 1.5 mTorr as the collision gas and a collision voltage of -20 to -26 V. In MS/MS experiments, daughter scans were performed to screen the fragmentation pattern of the parent ions of interest. High sensitivity, highly specific SRM experiments were performed to identify the source molecules of characteristic daughter ions. Analytes were assigned as dGuo, dAdo, and dCyd adducts if they released GSH ions (m/z ) 308, collision voltage -26 V) and lost deoxyribose (loss of m/z 116, collision voltage -20 V). An analyte was assigned as a Thd adduct if it released both m/z ) 273 and m/z ) 290 fragments (collision voltage -22 V).

Results Optimization of DNA Digestion Conditions. Because dihalomethane adducts are known to be alkali labile, a relatively rapid DNA digest using acid conditions was developed previously (18) and further optimized here (Scheme 1, Supporting Information). The first step consists of a simultaneous digest with DNase II and PDE II, resulting in 3′-monophosphate nucleotides, and the second step is a digest to nucleosides by a nonspecific acid phosphatase. Three buffer systems were tested at pH 4.5, 5.0, 5.5, and 6.0 with citrate, acetate, and succinate buffers (results not shown). Using calf thymus DNA unmodified by carcinogen, the enzymes catalyzed the release of most nucleosides and resulted in the cleanest chromatography baseline for a given time of digest if pH 5.0 acetate buffer was used. The dihalomethane-dGuo adduct has reasonable stability at pH 5.0, with an estimated half-life of 38 h (18). To optimize enzyme incubation times, DNA was reacted with the half-mustard GSCH2CH2Cl and the resulting adducts were released as nucleosides after the DNA digest. In summary, the DNase II + PDE II digestion was efficient at hydrolyzing DNA to nucleotides, but the acid phosphatase digestion did not completely

48

Chem. Res. Toxicol., Vol. 17, No. 1, 2004

Marsch et al.

Figure 2. Tandem MS of dGuoCH2SG and ThdCH2SG adducts. (A) Fragmentation of the m/z 587.3 dGuo adduct ion (tR ) 18.3 min) by HPLC-ESI-MS/MS. (B) Fragmentation of the m/z 562.2 Thd adduct ion (tR ) 20.5 min) by HPLC-ESI-MS/MS. The collision voltage in both cases was -22 V.

Figure 1. HPLC-MS-SRM analysis of DNA damaged by GSCH2OAc, showing adducts to A, T, and G. (A) UV trace of DNA + GSCH2OAc reaction digested to nucleosides. (B) LCESI-SRM of GSCH2OAc in DNA digest buffer but with no DNA. The SRM experiment is designed to detect ions with the same m/z as the dAdo adduct and that lose 116 Da after CID. (C) HPLC-ESI-SRM of DNA digested to nucleosides (with no GSCH2OAc added). The SRM experiment is designed to detect ions in the DNA digest that have the same characteristics as dAdoCH2SG. (D) GSCH2OAc + DNA reaction, with detection of the dAdo adduct. (E) GSCH2OAc + DNA reaction, with detection of the dGuo adduct. (F) GSCH2OAc + DNA reaction, with detection of the Thd adduct. SRM controls for detection of dGuo and Thd adducts are not shown here.

hydrolyze all nucleotide adducts to nucleoside adducts. The conditions required for efficient DNA digest are detailed in the Experimental Procedures section (vide supra); specifically, the data showing optimal digestion times are supplied in the Supporting Information section. DNA Adducts Formed by the Model Compound GSCH2OAc. High concentrations of GSCH2OAc added to calf thymus DNA solutions formed high levels of adduct to Gua and less to Ade and Thy bases. MS analysis of nucleoside adducts generated with GSCH2OAc was undertaken previously, and the MS signatures of these adducts are well-characterized (18). HPLC-MSSRM experiments (Figure 1) readily detected an m/z 587 analyte (Figure 1E) that eluted at 18.3 min and lost both an 116 Da fragment (deoxyribose) and resulted in an m/z 308 daughter [GSH + H]+. The elution time and fragmentation of the analyte strongly suggest a dGuoCH2SG adduct. Also detected was an m/z 571 analyte (Figure 1D) that eluted at 21.6 min, lost a 116 Da deoxyribose, and released singly ionized m/z 308 [GSH + H]+. The elution time and fragmentation of this analyte strongly suggest a dAdoCH2SG adduct (18). Also detected was an m/z 562 analyte (Figure 1F) that eluted at 20.5 min and upon fragmentation resulted in the m/z 273 ion [GSH -

SH]+ and the 290 m/z ion [dThdCH2SH + H]+, indicative of a ThdCH2SG adduct. Controls for dGuo and Thd adducts are not shown, but no peaks corresponding to the tR ) 18.3 min peak (dGuoCH2SG) or the tR ) 20.5 min peak (ThdCH2SG) were observed. These adduct designations were confirmed by MS/MS experiments (Figure 2). CID of the m/z 587.2 ion released the signature fragments of dGuoCH2SG, and CID of the m/z 562.2 ion resulted in a typical MS/MS spectrum for ThdCH2SG. Other background peaks are present in the ThdCH2SG MS/MS spectrum, which was acquired online from the low-intensity ThdCH2SG peak (Figure 1, chromatogram F). The level of dAdo adduct present in the samples was too small for a tandem MS experiment, however. Nevertheless, the 21.6 min peak (Figure 1D) was not observed in either control (Figure 1B,C). The adduct yield was a very sensitive function of the concentration of GSCH2OAc. While 153 mM GSCH2OAc yielded abundant dGuo adduct (0.43% of dGuo nucleoside ion signal), at 39 mM model compound, dGuo adduct was barely detectable (data not shown). The same high sensitivity of Thd and dAdo adduct yield to GSCH2OAc concentration was observed. The explanation appeared not to be rapid hydrolysis of the unstable GSCH2OAc prior to reaction (18). For example, in one experiment, GSCH2OAc was added to the DNA solution as a solid; in other experiments, it was dissolved in water first and then added to the DNA solution. The same sensitivity of adduct formation to GSCH2OAc concentrations was seen. It is possible that adduct recovery is inefficient if the adduct concentration is low. The actual ratio of dGuo adduct to dGuo could be calculated in the cases where chromatogram peaks corresponding to dGuo adduct and dGuo were both detected using UV measurements (260 nm). At the highest concentrations of GSCH2OAc (153 mM), 1-2% of the dGuo nucleoside pool was modified by GSCH2OAc.

Glutathione-CH2-DNA Adducts

Chem. Res. Toxicol., Vol. 17, No. 1, 2004 49

Figure 4. Tandem MS of adducts produced by enzymemediated conjugation of GSH and dihaloalkanes. (A) Fragmentation of the m/z 587 dGuo adduct ion (tR ) 17.1 min) by HPLCESI-MS/MS. (B) Fragmentation of the m/z 571 dAdo adduct ion (tR ) 20.7 min) by HPLC-ESI-MS/MS. (C) Fragmentation of the m/z 562 Thd adduct ion (tR ) 17.6 min) by HPLC-ESI-MS/MS. The collision voltage was -22 V.

Figure 3. HPLC-ESI-SRM analysis of nucleosides damaged by GST-catalyzed dihaloalkane-GSH conjugates. (A) Top to bottom: dGuo nucleoside, dGuo adduct SRM analysis of reaction with rat 5-5 + GSH but no CH2Cl2, dGuo adduct SRM analysis of rat GST 5-5 + GSH + 50 mM CH2Cl2 reaction, Thd adduct SRM analysis of reaction of rat GST 5-5 + GSH but no CH2Cl2, and Thd adduct SRM analysis of rat GST 5-5 + GSH + 50 mM CH2Cl2 reactions. (B) Top to bottom: dGuo nucleoside, dCyd adduct SRM analysis of reaction of rat GST 5-5 + GSH but no CH2Cl2, dCyd adduct SRM analysis of rat GST 5-5 + GSH + 50 mM CH2Cl2 reaction, dAdo adduct SRM analysis of reaction of rat GST 5-5 + GSH but no CH2Cl2, and dAdo adduct SRM analysis of rat GST 5-5 + GSH + 50 mM CH2Cl2 reaction.

Synthesis of Nucleoside Adducts by GSTs. Nucleoside incubations with GST enzymes, dihalomethanes, and GSH were undertaken to determine whether dihalomethane-GSH adducts to nucleosides could be detected and, if so, to compare the efficiency of adduct formation by three GSTs. HPLC-MS-SRM experiments demonstrated that bacterial GST DM11, rat GST 5-5, and human GST T1-1 enzymes all formed GSH-dihalo-

methane adducts with all four nucleosides (Figure 3; chromatograms only shown for rat 5-5 + CH2Cl2). For the dGuo, dAdo, and dCyd adducts, SRM experiments monitored the loss of a deoxyribose or the release of singly ionized GSH. After CID, the Thd adduct generated by GST enzymes released the m/z 273 ion [GSH - SH]+ and the m/z 290 ion [dThdCH2SH + H]+. These are all of the same specific decompositions used to detect nucleoside adducts obtained from GSCH2OAc model compound (18), suggesting that adducts formed by enzyme activation are the same structures as those formed by the model electrophile. Further augmenting this conclusion for dAdo and dGuo adducts is the observation that the retention times of the GST-generated nucleoside adducts were very similar to tR values for DNA adducts generated by GSCH2OAc (e.g., compare retention times in Figure 1 with those in Figure 3). The previously reported tR for dCyd adduct formed to nucleosides by GSCH2OAc (18) was also very similar to the tR of dCyd adduct that formed in nucleosides from the reaction of dihalomethanes with GSTs. However, the tR for the enzyme-generated Thd adduct (tR ) 17.6 min) was shorter than the tR for Thd adduct generated by GSCH2OAc (tR ) 20.5 min). Interestingly, the reactions of CH2Cl2 + GSH + GSTs showed two additional species (tR ) 12.8 and 21.5 min) corresponding to the m/z 587 dGuoCH2SG adduct (Figure 3A, middle chromatogram). Neither peak was observed in the control SRM chromatogram showing the loss of deoxyribose (Figure 3A, second from top), and the two m/z 587 species do not release a GSH ion (data not shown). These analytes were not identified but may be other condensations of dGuo with activated dihalomethane. The assignment of adduct structures resulting from GST activation of GSH and dihalomethane was verified by MS/MS after CID of singly ionized adduct species (Figure 4). MS/MS experiments with dGuoCH2SG (Figure 4A) and dAdoCH2SG (Figure 4B) both indicated a loss of

50

Chem. Res. Toxicol., Vol. 17, No. 1, 2004

Marsch et al.

Figure 5. Adduct formation by GSH-dihaloalkane conjugates as a function of substrate concentration. (A) GST DM11 + CH2Br2, (B) GST DM11 + CH2Cl2, (C) rat GST 5-5 + CH2Br2, (D) rat GST 5-5 + CH2Cl2, (E) human GST T1-1 + CH2Br2, or (F) human GST T1-1 + CH2Cl2. The ordinate is the ratio of dGuo (9), Thd (2), dCyd (0), or dAdo (b) adduct ion detected by HPLC-ESI-SRM relative to the dGuo nucleoside ion detected under identical MS conditions.

deoxyribose (116 Da) and a resulting GSH ion (m/z 308), whereas CID of ThdCH2SG (Figure 4C) released the m/z 290 ion characteristic of this adduct (18). The amount of adduct present in the dCyd adduct peak was not sufficient to yield a good fragmentation spectrum of dCyd adduct. Using the ratio of adduct ion detected to dGuo ion detected for all samples, relative levels of adduct could be assessed by plotting adduct ion levels vs dihalomethane concentrations (Figure 5). DM11 (Figure 5A,B) and human GST T1-1 (Figure 5E,F) data were fit to hyperbolic plots, as were the data for adduct production by rat GST 5-5 and CH2Cl2 (Figure 5D). Using rat GST 5-5, the adduct yield vs [CH2Br2] data were curved but a closer analysis showed that a linear fit was better (Figure 5C).2

For all GSTs and DNA bases, CH2Cl2 incubations yielded as much adduct as CH2Br2 under these conditions (which are not necessarily optimized for steady state condiditons), as indicated by the adduct level produced by both dihalomethanes at the same concentrations (compare Figure 5A with 5B, 5C with 5D, and 5E with 5F). With GST DM11 and CH2Br2 or CH2Cl2, the dGuo adduct ion was detected at a level 2-3 times greater than 2 Many parameters measured with the mammalian GST enzymes 5-5 and T1-1 show relatively nonhyperbolic fits of product formation vs concentration of dihaloalkanes (13, 17, 33). However, we emphasize that these fits are crude because of the very limited number of points and should not be overemphasized. An important issue limiting the size of data sets was the instability of the nucleoside adducts and the need to analyze the reactions quickly. In addition, because it was necessary to compare several different sets of experiments within the same time frame, the number of reactions that we could assess by HPLC-MS was limited.

Glutathione-CH2-DNA Adducts

Chem. Res. Toxicol., Vol. 17, No. 1, 2004 51

adduct, especially to dGuo, although not apparently to the extent of the rat enzyme. Second, the GSTs exhibit modest differences in relative efficiencies of dGuo adduct formation and smaller differences in efficiencies of adduct formation to the other three nucleosides (Figure 6). If GST concentrations are normalized for the reactions, rat GST 5-5 formed 3.7× the dGuo adduct yield as GST DM11. Rat GST 5-5 formed only 2× as much Thd adduct as human GST T1-1 and 1.8× as much Thd adduct as GST DM11. For dAdo, Thd, and dCyd, human GST T1-1 and bacterial DM11 catalyzed the same yields of adducts under these conditions (Figure 6).

Discussion Figure 6. Adduct yields for each base as a function of GST used. The adduct yield is shown for each nucleoside with a CH2Cl2 concentration of 50 mM. In the experiments, the rat GST 5-5 concentration was 15 µΜ, the human T1-1 concentration was 24 µM, and the bacterial DM11 concentration was 26 µM. Adduct levels were normalized to the GST DM11 concentration. Adduct yields formed from GST human T1-1 (Figure 5F) were multiplied by 1.1, and adduct yields formed by rat 5-5 (D) were multiplied by 1.7.3

Thd, dCyd, or dAdo adduct ion, respectively (Figures 5A,B and 6). With rat GST 5-5, the dGuo adduct ion was detected at a level 4-5 times greater than the Thd, dCyd, or dAdo adduct ions, respectively (Figures 5C,D and 6). With human GST T1-1, the dGuo adduct ion was also detected at a level 4-5 times greater than the Thd, dCyd, or dAdo adduct ion, respectively (Figures 5E,F and 6). As compared to rat GST 5-5 and human GST T1-1, GST DM11 formed more adducts to dAdo, Thd, and dCyd relative to dGuo adduct formation, but rat GST 5-5 formed the most adduct to all bases. Human GST T1-1 formed more dGuo adduct than did DM11. The rat GST 5-5 was clearly superior to the other GSTs in its ability to catalyze the formation of GSH adducts under these conditions. The most adduct was produced by the rat GST 5-5 + 50 mM CH2Cl2 reaction, which yielded a dGuo adduct ion/dGuo ion ratio of 3 × 10-4 (Figure 5D). However, because the concentrations of the original GST enzyme preparations were different, it was not possible to use the same concentrations of bacterial, rat, and human GSTs in the experiments: the rat GST 5-5 concentration was 15 µΜ, the human T1-1 concentration was 24 µM, and the bacterial DM11 concentration was 26 µM. Adduct levels were thus normalized to the GST DM11 concentration.3 If this dGuo value is denoted as unity, then the relative efficiency of adduct formation to other bases or by other enzymes can be approximated from Figure 6. The relative efficiencies of dGuo adduct formation by human GST T1-1 and GST DM11 were 0.49 and 0.27, respectively. As compared to the dGuo adduct yield by rat GST 5-5, the yields of dCyd, dAdo, and Thd adducts by rat GST 5-5 were 0.25, 0.25, and 0.20, respectively. The relative yields of dCyd, dAdo, and Thd adduct formation by GST DM11 were 0.11, 0.09, and 0.11, respectively, and the relative yields of dCyd, dAdo, and Thd adduct formation by human T1-1 were 0.11, 0.11, and 0.10, respectively. Two points can be made concerning these results. The human enzyme is clearly capable of generating nucleoside 3 Adduct yields formed from GST human T1-1 (Figure 5F) were multiplied by 1.1, and adduct yields formed by rat 5-5 (Figure 5D) were multiplied by 1.7. Thus, the adjusted dGuo adduct ion/dGuo ion ratio was 5 × 10-4 (Figure 6).

The genotoxicity of dihalomethane-GSH conjugates is probably a major factor in dihalomethane mutagenicity (14, 17, 28, 29). However, the electrophilic metabolites in the series GSCH2X (X ) halogen atom) are extremely reactive (35) and their stability presents a major hindrance in the study of mutagenic behavior of dihalomethanes. The analogue GSCH2OAc is a well-characterized compound with better stability (t1/2 ) 11 s at pH 8.0) (18) and readily forms adducts to nucleosides in aqueous solutions. Thus, GSCH2OAc is an excellent model compound for the CH2Cl2 conjugate GSCH2Cl. GSCH2OAc can also be synthesized as a solid and stored for months at -80 °C under anhydrous conditions. GSCH2OAc formed adducts with A, T, and G bases of calf thymus DNA at high concentrations of model compound. dCydCH2SG probably was not detected because it decomposes rapidly under all conditions present in these experiments (14, 18) and did not survive the enzymatic decomposition to nucleosides. When GST reactions with nucleosides were assayed by MS immediately, the dCyd adduct could be detected. However, GST-mediated adduct formation in calf thymus DNA has not been successful to date under the conditions attempted in our laboratory. In contrast with DNA adduct formation, the GST-catalyzed synthesis of nucleoside adducts was rapid, requiring no digestion of DNA to nucleosides. We elected to present data as adduct ion peak area divided by dGuo adduct ion peak area. Using GST enzymes and nucleotides as a target, the ratio of modified nucleotides to dGuo was ∼1/104. Our experience to date with GSCH2OAc indicates that the level of modification of bases in DNA is probably 1/10-1/102 that seen with free nucleosides. Much of the difference may be due to poor recovery following DNA digestion, especially at low concentrations of adduct, but another aspect may be the difficulty of forming the four adducts (18) in double-stranded DNA. In the context of duplex DNA, the bases possess atoms involved in hydrogen bonding at sites of attack and are therefore probably less accessible. Because HCHO is the end product of GST activation of dihalomethanes, we attempted to detect potential HCHO adducts resulting from addition of HCHO to calf thymus DNA. In these experiments, the HCHO concentration was 50 mM and the DNA base concentration was 10 mM bases; the sample was incubated for 1 h at 37 °C. Within the sensitivity of our experiments, no HCHO adducts were detected, consistent with a lack of HCHO mutagenicity in our S. typhimurium systems (14) and other work (17, 28, 29). Quantifying GSH nucleoside adducts by MS is not trivial. Standards for Thd, dAdo, and dCyd adducts have

52

Chem. Res. Toxicol., Vol. 17, No. 1, 2004

been synthesized previously (18) but are unstable to storage, and the pyrimidine adducts in particular are highly unstable even under acidic pH conditions. Previous work (18) on these adducts established their fragmentation patterns by CID, but their lability obviated using these as standards in our experiments. The major issue is that the ionization efficiencies of nucleoside adducts released in the DNA digest was usually unknown. We reported the Thd, dCyd, and dAdo adduct yields from HPLC-ESI-SRM experiments as adduct ion peak areas divided by the dGuo ion peak area. To maintain consistency in reporting these adduct ion yields relative to dGuo ion, we monitored the loss of deoxyribose (m/z 116) in the mass spectrometer for dGuo and all adducts except ThdCH2SG. As the Thd adduct does not easily lose the deoxyribose moiety under the MS conditions used in this study, we monitored the loss of the m/z 272 fragment (GSH minus SH) for this adduct. Without stable isotope standards, we have not been able to establish ionization efficiencies for all nondGuo adducts and we cannot relate ion peak areas in an MS experiment to an accurate yield of adduct as a ratio of the modified to unmodified nucleoside bases. However, a good semiquantitative yield of guanosine adducts was calculated, because several of the GSCH2OAc + DNA samples yielded enough dGuoCH2SG that the peak corresponding to the adduct was visible in the chromatogram showing UV detection at 260 nm. Because the GSCH2 moiety of the dGuo adduct is covalently attached to the N2 exocyclic amine of guanine base, the π-electron structure of Gua is not significantly perturbed. Thus, the absorption characteristics of dGuo adduct were taken to be identical to unmodified dGuo, and the UV peak area of dGuoCH2SG divided by the UV peak area of dGuo yielded a good value of true ratio of dGuo modified by model compound GSCH2OAc. This true ratio was compared to the ion peak ratio, and it was found for the TSQ 7000 mass spectrometer that the ionization efficiency of dGuo was 2.7 times that of dGuoCH2SG; the ratios of adduct ion/dGuo ion peak areas can be multiplied by 2.7 to obtain a better estimate of the yield. We therefore estimate that the highest concentrations of GSCH2OAc (153 mM) used in this study, when added to 6.1 mM DNA bases, resulted in 2% of dGuo modified to form dGuoCH2SG, and we estimated ∼0.04% of dAdo and Thd bases damaged by GSCH2OAc. In some previous studies in which GSCH2OAc was incubated with dGuo nucleoside, yields of >10% were obtained (18). In general, the yield of adducts from DNA reactions was 10 times less than for nucleosides, even at very high concentrations of model compound. Another factor limiting the quantification of adducts is that the concentration of dGuo in a DNA digest sample must be the same as the concentration of dAdo, Thd, and Cyd in the sample. This relation holds in calf thymus DNA, but we are also assuming that the nucleases and acid phosphatase II liberate all nucleoside adducts equally well, with minimal incomplete digestion of highly polymerized DNA to nucleosides. In fact, while the DNase II and PDE II steps were efficient at releasing nucleotides from DNA, the acid phosphatases did not completely remove the 3′-phosphates, and some GSCH2CH2-GMP adduct was detected in the optimization experiments with GSCH2CH2Cl (Supporting Information). Dihalomethane adduct formation in calf thymus DNA using bacterial GST DM11 and radiolabeled CH2Cl2 and

Marsch et al.

GSH has been observed by Kayser and Vuilleumier (30). Using GST DM11 and CH2Cl2 to form adducts in 20-mer single-stranded DNA polymers, they showed that the level of adduct formation by GST DM11 was polydG > polydC > polydA > polydT, with polydGC forming about three times the CH2Cl2 adduct as polydAT (31). We found that the levels of adduction to nucleosides catalyzed by GST DM11 were dGuo > Thd g dCyd g dAdo (Figure 5A,B). In our studies, Thd formed relatively more adduct, perhaps because Thd is a better target nucleophile than is the T base in the context of duplex DNA. However, in our study, dGuo + dCyd formed 3.5× more adduct together than did dAdo + Thd, assuming similar ionization efficiencies for the four nucleosides. This result is very similar to that obtained by Kayser and Vuilleumier (31). Comparisons of DNA adduction by GSCH2OAc to nucleoside adduction by GSTs and dihalomethanes show that the relative adduct yield was dGuo > Thd > dAdo for the former case and dGuo > Thd g dAdo g dCyd for the latter case. Except for the lack of dCyd adduct found from the incubations of DNA with GSCH2OAc, these relative adduct yields are consistent. dCyd adducts could not be isolated from the GSCH2OAc + DNA reaction because a 2.5 h DNA digest at 37 °C was required, by which time all dCyd adduct had degraded. In the case of the GST-mediated adduct, the enzyme was added to a nucleoside mix and immediately after the incubation, this mixture was either stored at -80 °C or was injected onto the HPLC system. Under these conditions, dCyd adduct was found and at levels comparable to Thd and dAdo adduct. It is also possible that there is a pH dependence on adduct formation: DNA reactions with model compounds were performed at pH 5.0, but the GST + GSH + dihalomethane reactions were accomplished at pH 7.0. In addition, adduct stabilities may be functions of pH but not in the same manner for each adduct. (The dGuo adduct is known to be alkaline labile (18).) If dAdo adduct forms more readily at pH 7.0 than at pH 5.0, then a greater yield of dAdo adduct will be identified from the GST + dihalomethane incubations. Finally, adduct formation to each base in duplex DNA may not occur with the same efficiencies as adduction to free nucleosides. Support for our protocol of reporting adduct ion yields relative to dGuo ion yields comes from previous work in this laboratory (32), in which DNA adducts by haloethylGSH compounds were quantified using the 4-(p-nitrobenzyl)pyridine assay. HaloethylGSH compounds generate adducts that are chemically very similar to those produced by dihalomethanes; the conjugation of GSH to the nucleoside is via an ethylene bridge vs a methylene bridge. When 0.5 mg of DNA was incubated with 35 mM GSCH2CH2Cl, 11 nmol adduct/mg DNA was formed (32). This 1.4% yield is close to the 0.5-0.9% yields resulting when 31 mM GSCH2CH2Cl was added to 2 mg of DNA in our DNA digest optimization studies. We employed slightly lower concentrations of chloroethylGSH and four times the DNA mass in our reactions. As Humphreys et al. (32) found that adduct formation vs chloroethyl GSH concentration was roughly linear, our slightly lower values of half-mustard DNA adduct reported in this manuscript by MS are in good agreement with previous work in this laboratory.

Glutathione-CH2-DNA Adducts

Chem. Res. Toxicol., Vol. 17, No. 1, 2004 53

Conclusions Dihalomethanes are important chemicals with regard to potential toxicity. CH2Cl2 and other dihalomethanes are genotoxic after activation by GSTs, but defining and quantifying DNA adducts have been difficult due to the documented instability at neutral pH (14, 18, 31). We utilized two approaches for characterization of the DNA adducts. Following further optimization of DNA digestion conditions developed earlier (18), we treated calf thymus DNA with GSCH2OAc and identified the G, T, and A adducts (Figures 1 and 2). Previously, only the T adduct could be identified in DNA digests (18). Another system was used in which GST enzymes were incubated with CH2Cl2 or CH2Br2 (plus GSH) in the presence of deoxyribonucleosides (Figure 5). As in the experiments with DNA and GSCH2OAc, the G adduct predominated; the other three nucleosides also yielded adducts. This method was more sensitive than the one involving DNA digestion, although the present results should not be overinterpreted regarding the enzymatic specificity of individual reactions. The increased sensitivity of the MS equipment used in this work as compared to the previous study (18) provided a considerable advantage and was an important factor in the refinement of the analyses. Further technical advances should also provide better assays. In summary, our work provides further evidence that dihalomethanes can be activated to DNA adducts by human GST (T1-1), although apparently not as well as rat GST (5-5), and that the dGuo adduct is the dominant one, consistent with limited data available from HPRT mutation spectra produced with CH2Cl2 in Chinese hasmster ovary cells (39).

Acknowledgment. This work was supported in part by U.S. Public Health Service Grants R01 ES10546 and P30 ES00267. We thank D.L. Hachey and M.L. Manier of the Vanderbilt Mass Spectrometry Resource Facility for their assistance, R.N. Armstrong and N. V. Stourman for the GST DM11 plasmid (originally kindly provided by S. Vuilleumier), and D. Akridge for assistance in the preparation of the manuscript.

(6) (7) (8) (9)

(10) (11) (12)

(13)

(14)

(15)

(16)

(17)

(18)

(19)

(20)

(21)

Supporting Information Available: Protocols used in optimizing the DNA digest under acidic conditions, using the half-mustard GSCH2CH2Cl as a model. HPLC-ESI-SRM experiments showing half-mustard DNA adducts and plots showing the effect of incubation time on the release of half-mustard nucleoside adducts. This material is available free of charge via the Internet at http://pubs.acs.org.

References (1) Agency for Toxic Substances & Disease Registry (1993) Toxicological Profile for Methylene Chloride, U.S. Department Health & Human Services, Atlanta. (2) Kari, F. W., Foley, J. F., Seilkop, S. K., Maronpot, R. R., and Anderson, M. W. (1993) Effect of varying exposure regimens on methylene choride-induced lung and liver tumors in female B6C3F1 mice. Carcinogenesis 14, 819-826. (3) Burek, J. D., Nitschke, K. D., Bell, T. J., Wackerle, D. L., Childs, R. C., Beyer, J. D., Dittenber, D. A., Rampy, L. W., and McKenna, M. J. (1984) Methylene chloride: A 2 year inhalation toxicity and oncogenicity study in rats and hamsters. Fundam. Appl. Toxicol. 4, 30-47. (4) Andersen, M. E., Clewell, H. J., III, Gargas, M. L., Smith, F. A., and Reitz, R. H. (1987) Physiologically based pharmacokinetics and the risk assessment process for methylene chloride. Toxicol. Appl. Pharmacol. 87, 185-205. (5) Nitschke, K. D., Burek, J. D., Bell, T. J., Kociba, R. J., Rampy, L. W., and McKenna, M. J. (1988) Methylene chloride: a 2-year

(22)

(23)

(24)

(25)

(26)

(27)

(28)

(29)

inhalation toxicity and oncogenicity study in rats. Fundam. Appl. Toxicol. 11, 48-59. Anderson, M. W., and Maronpot, R. R. (1993) Methylene chlorideinduced tumorigenesis. Carcinogenesis 14, 787-788. Hanson, D. (1991) OSHA to cut methylene chloride exposure limits. Chem. Eng. News 69, 8. Hanson, D. (1997) Methylene chloride exposure sharply limited. Chem. Eng. News 75, 7. Clewell, H. J., III (1995) Incorporating biological information in quantitative risk assessment: an example with methylene chloride. Toxicology 102, 83-94. Rhomberg, L. (1995) Use of quantitative modelling in methylene chloride risk assessment. Toxicology 102, 95-114. Huff, J., Bucher, J., and Barrett, J. C. (1996) Methylene chloride. Science 272, 1083-1084. Green, T. (1997) Methylene chloride induced mouse liver and lung tumours: an overview of the role of mechanistic studies in human safety assessment. Hum. Exp. Toxicol. 16, 3-13. Reitz, R. H., Mendrala, A., and Guengerich, F. P. (1989) In vitro metabolism of methylene chloride in human and animal tissues: use in physiologically based pharmacokinetic models. Toxicol. Appl. Pharmacol. 97, 230-246. Thier, R., Pemble, S. E., Taylor, J. B., Humphreys, W. G., Persmark, M., Ketterer, B., and Guengerich, F. P. (1993) Expression of mammalian glutathione S-transferase 5-5 in Salmonella typhimurium TA1535 leads to base-pair mutations upon exposure to dihalomethanes. Proc. Natl. Acad. Sci. U.S.A. 90, 8576-8580. Thier, R., Mu¨ller, M., Taylor, J. B., Pemble, S. E., Ketterer, B., and Guengerich, F. P. (1995) Enhancement of bacterial mutagenicity of bifunctional alkylating agents by expression of mammalian glutathione S-transferase. Chem. Res. Toxicol. 8, 465-472. Thier, R., Pemble, S., Kramer, H., Taylor, J. B., Guengerich, F. P., and Ketterer, B. (1996) Human glutathione S-transferase T1-1 enhances mutagenicity of 1,2-dibromoethane, dibromomethane, and 1,2,3,4-diepoxybutane in Salmonella typhimurium. Carcinogenesis 17, 163-166. Wheeler, J. B., Stourman, N. V., Thier, R., Dommermuth, A., Vuilleumier, S., Rose, J. A., Armstrong, R. N., and Guengerich, F. P. (2001) Conjugation of haloalkanes by bacterial and mammalian glutathione transferases: mono- and dihalomethanes. Chem. Res. Toxicol. 14, 1118-1127. Marsch, G. A., Mundkowski, R. G., Morris, B. J., Manier, M. L., Hartman, M. K., and Guengerich, F. P. (2001) Characterization of nucleoside and DNA adducts formed by S-(1-acetoxymethyl)glutathione and implications for dihalomethane-glutathione conjugates. Chem. Res. Toxicol. 14, 600-608. Ahmed, A. E., Kubic, V. L., and Anders, M. W. (1977) Metabolism of haloforms to carbon monoxide. I. In vitro studies. Drug Metab. Dispos. 5, 198-204. Kubic, V. L., and Anders, M. W. (1978) Metabolism of dihalomethanes to carbon monoxide. III. Studies on the mechanism of the reaction. Biochemical Pharmacol. 27, 2349-2355. Guengerich, F. P., Kim, D.-H., and Iwasaki, M. (1991) Role of human cytochrome P-450 IIE1 in the oxidation of many low molecular weight cancer suspects. Chem. Res. Toxicol. 4, 168-179. Ahmed, A. E., and Anders, M. W. (1976) Metabolism of dihalomethanes to formaldehyde and inorganic halide. I. In vitro studies. Drug Metab. Dispos. 4, 357-361. Meyer, D. J., Coles, B., Pemble, S. E., Gilmore, K. S., Fraser, G. M., and Ketterer, B. (1991) Theta, a new class of glutathione transferases purified from rat and man. Biochem. J. 274, 409414. Jongen, W. M. F., Alink, G. M., and Koeman, J. H. (1978) Mutagenic effect of dichloromethane on Salmonella typhimurium. Mutat. Res. 56, 245-248. van Bladeren, P. J., Breimer, D. D., Rotteveel-Smijs, G. M. T., and Mohn, G. R. (1980) Mutagenic activation of dibromomethane and diiodomethane by mammalian microsomes and glutathione S-transferases. Mutat. Res. 74, 341-346. Green, T. (1983) The metabolic activation of dichloromethane and chlorofluoromethane in a bacterial mutation assay using Salmonella typhimurium. Mutat. Res. 118, 277-288. Osterman-Golkar, S., Hussain, S., Walles, S., Anderstam, B., and Sigvardsson, K. (1983) Chemical reactivity and mutagenicity of some dihalomethanes. Chem.-Biol. Interact. 46, 121-130. Graves, R. J., Callander, R. D., and Green, T. (1994) The role of formaldehyde and S-chloromethylglutathione in the bacterial mutgenicity of methylene chloride. Mutat. Res. 320, 235-243. Gisi, D., Leisinger, T., and Vuilleumier, S. (1999) Enzymemediated dichloromethane toxicity and mutagenicity of bacterial and mammalian dichloromethane-active glutathione S-transferases. Arch. Toxicol. 73, 71-79.

54

Chem. Res. Toxicol., Vol. 17, No. 1, 2004

(30) La Roche, S. D., and Leisinger, T. (1990) Sequence analysis and expression of the bacterial dichloromethane dehalogenase structural gene, a member of the glutathione S-transferase supergene family. J. Bacteriol. 172, 164-171. (31) Kayser, M. F., and Vuilleumier, S. (2001) Dehalogenation of dichloromethane by dichloromethane dehalogenase/glutathione S-transferase leads to formation of DNA adducts. J. Bacteriol. 183, 5209-5212. (32) Humphreys, W. G., Kim, D.-H., Cmarik, J. L., Shimada, T., and Guengerich, F. P. (1990) Comparison of the DNA alkylating properties and mutagenic responses caused by a series of S-(2haloethyl)-substituted cysteine and glutathione derivatives. Biochemistry 29, 10342-10350. (33) Wheeler, J. B., Stourman, N. V., Armstrong, R. N., and Guengerich, F. P. (2001) Conjugation of haloalkanes by bacterial and mammalian glutathione transferases: mono- and vicinal dehaloethanes. Chem. Res. Toxicol. 14, 1107-1117. (34) Maczynska, Z. (1995) Halogenated methanes with water. In Solubility Data Series 60 (Horvath, A. L., and Getzen, F. W., Eds.) pp 146-175, Oxford University Press, New York. (35) Bo¨hme, V. H., Fischer, H., and Frank, R. (1949) Darstellung und Eigenschaften der R-halogenierten Thioa¨ther. Ann. Chem. 563, 54-72.

Marsch et al. (36) Koga, N., Inskeep, P. B., Harris, T. M., and Guengerich, F. P. (1986) S-[2-(N7-Guanyl)ethyl]glutathione, the major DNA adduct formed from 1,2-dibromoethane. Biochemistry 25, 21922198. (37) Kim, D.-H., Humphreys, W. G., and Guengerich, F. P. (1990) Characterization of S-[2-(N1-adenyl)ethyl]glutathione formed in DNA and RNA from 1,2-dibromoethane. Chem. Res. Toxicol. 3, 587-594. (38) Cmarik, J. L., Humphreys, W. G., Bruner, K. L., Lloyd, R. S., Tibbetts, C., and Guengerich, F. P. (1992) Mutation spectrum and sequence alkylation selectivity resulting from modification of bacteriophage M13mp18 with S-(2-chloroethyl)glutathione. Evidence for a role of S-[2-(N7-guanyl)ethyl]glutathione as a mutagenic lesion formed from ethylene dibromide. J. Biol. Chem. 267, 6672-6679. (39) Graves, R. J., Trueman, P., Jones, S., and Green, T. (1996) DNA sequence analysis of methylene chloride-induced HPRT mutations in Chinese hamster ovary cells: comparison with the mutation spectrum obtained for 1-2-dibromoethane and formaldehyde. Mutagenesis 11, 229-233.

TX034156Z