Adenine Adducts with Diepoxybutane: Isolation and Analysis in

Oct 15, 1997 - Adenine Adducts with Diepoxybutane: Isolation and Analysis in Exposed Calf Thymus DNA ... The findings of this study provide a basis fo...
0 downloads 11 Views 323KB Size
Download PDF
Chem. Res. Toxicol. 1997, 10, 1171-1179

1171

Adenine Adducts with Diepoxybutane: Isolation and Analysis in Exposed Calf Thymus DNA Natalia Tretyakova, Ramiah Sangaiah, Ten-Yang Yen, Avram Gold, and James A. Swenberg* Laboratory of Molecular Carcinogenesis and Mutagenesis, Department of Environmental Sciences and Engineering, University of North Carolina, Chapel Hill, North Carolina 27599 Received May 5, 1997X

1,3-Butadiene (BD) is a high-volume industrial chemical and a common environmental pollutant. Although BD is classified as a “probable human carcinogen”, only limited evidence is available for its tumorigenic effects in occupationally exposed populations. Animal studies show a surprisingly high sensitivity of mice to the carcinogenic effects of BD compared to rats (≈103-fold), making interspecies extrapolations difficult. Identification and quantitation of specific BD-induced DNA adducts are important for improving our understanding of the mechanisms of BD biological effects and for explaining the observed species differences. Covalent binding of BD to DNA is probably due to its two epoxy metabolites: 3,4-epoxy-1butene (EB) and 1,2:3,4-diepoxybutane (DEB). Both EB and DEB are direct mutagens producing frameshift and point mutations at both A:T and G:C base pairs. DEB is 100 times more mutagenic than EB and is found in quantity only in tissues of the most sensitive species (mouse). This has led to the suggestion that the higher sensitivity of mice to BD could be due to greater exposure to DEB. The present work was initiated in order to isolate and structurally characterize DEB-induced adenine adducts. The adducts were formed by reacting DEB with free adenine (Ade), 2′-deoxyadenosine (2′-dAdo), and calf thymus DNA followed by HPLC separation and analysis of the products by UV spectrophotometry, electrospray ionization mass spectrometry, and nuclear magnetic resonance. The adenine reaction resulted in three products which were identified as N-3-, N-7-, and N-9-(2′-hydroxy-3′,4′-epoxybut-1′-yl)adenine. These adducts underwent acid-catalyzed hydrolysis to their corresponding (2′,3′,4′-trihydroxybut-1′yl)adenines upon heating or storage. The 2′-dAdo reaction with DEB followed by acid hydrolysis yielded a single adduct, N6-(2′,3′,4′-trihydroxybut-1′-yl)adenine (N6-DEB-Ade). N-3-DEB-Ade and N6-DEB-Ade were also found in hydrolysates of calf thymus DNA exposed to DEB. The amounts of N-3-DEB-Ade (13/103 normal Ade) and N6-DEB-Ade (5/103 normal Ade) were slightly lower than those of the corresponding EB-induced adducts in similar experiments, suggesting comparable reactivity of the two epoxy metabolites of BD toward adenine in DNA. The findings of this study provide a basis for future analyses of BD-induced adenyl DNA adducts in vitro and in vivo.

Introduction 1,3-Butadiene (BD)1 is an important industrial chemical used mainly in the production of synthetic rubber and plastics. Approximately 52 000 workers are potentially exposed to BD in the United States (1). BD is also an environmental pollutant included in the 1990 Clean Air Act. It is found in automobile exhaust, gasoline vapor, cigarette smoke, and stack emission from waste incinerators (2). Epidemiological studies show some association between occupational exposure to BD and lymphatic and hematopoietic cancers in humans (3, 4). BD has been classified by IARC as a “probable human carcinogen” (5). Since a significant portion of the human population is potentially exposed to BD, the toxicity and mutagenicity * Corresponding author. Tel: (919) 966-6139. Fax: (919) 966-6123. X Abstract published in Advance ACS Abstracts, September 15, 1997. (1) Abbreviations: Ade, adenine; 2′-dAdo, 2′-deoxyadenosine; BD, 1,3-butadiene; CT DNA, calf thymus DNA; DAD, diode array detector; DEB, diepoxybutane; EB, 3,4-epoxy-1-butene; FAB MS, fast atom bombardment mass spectrometry; Gua, guanine; ESI MS, electrospray ionization mass spectrometry; IARC, International Agency for Research on Cancer; N-3-DEB-Ade, N-3-(2′-hydroxy-3′,4′-epoxybut-1′-yl)adenine; N-3-DEB-Ade II, N-3-(2′,3′,4′-trihydroxybut-1′-yl)adenine; N-3-DEBAde III, N-3-(2′,3′-dihydroxy-4′-chlorobut-1′-yl)adenine; N-7-DEB-Ade, N-7-(2′-hydroxy-3′,4′-epoxybut-1′-yl)adenine; N-9-DEB-Ade, N-9-(2′hydroxy-3′,4′-epoxybut-1′-yl)adenine.

S0893-228x(97)00068-4 CCC: $14.00

of BD have been extensively investigated in laboratory animals. Rodent inhalation studies showed strikingly high sensitivity of mice to BD-induced tumors (6-8). Rats developed tumors at 1000-8000 ppm BD (6), while an increase in lung tumors in female mice was observed at exposure concentrations as low as 6.25 ppm (7). Furthermore, the target tissues in mice (heart, lung, mammary gland, forestomach, ovary) are different from those in rats (thyroid, testis, uterus, pancreas, mammary gland). The observed dramatic species differences in carcinogenic response to BD are likely to be due to the differences in its metabolism. BD requires metabolic activation for its mutagenicity. It was demonstrated that BD undergoes cytochrome P450-mediated epoxidation to the major metabolite 3,4-epoxy-1-butene (EB) (9). The latter can be hydrolyzed via epoxide hydrolase, conjugated with glutathione via S-transferase, or further oxidized to 1,2: 3,4-diepoxybutane (DEB) (10-12). Studies of BD metabolism in rats and mice showed that mice were making more EB than rats (12). The conversion of EB to DEB was much more significant in mouse than in rat tissues (13). Both EB and DEB were shown to be direct-acting mutagens in the Ames test, induced mutations at hprt © 1997 American Chemical Society

1172 Chem. Res. Toxicol., Vol. 10, No. 10, 1997

and tk loci, and caused genotoxic effects in mice (1417). EB and DEB are therefore thought to be the ultimate carcinogens that react with the nucleophilic sites of DNA causing adduct formation. Interestingly, DEB was 2 orders of magnitude more mutagenic than EB, probably due to its bifunctional nature which makes it capable of forming DNA-DNA and DNA-protein cross-links (18). It has been suggested that the higher sensitivity of mice to BD may be due to the greater exposure to DEB (13). Covalent binding of carcinogens or metabolites to DNA is considered an early critical step in carcinogenesis. To date it is not known which particular DNA adducts are responsible for butadiene mutagenicity and carcinogenicity. EB has been shown to react with calf thymus DNA at N-7 of guanine (Gua) and N-3 of adenine (Ade) in a 10:1 ratio (19, 20). However, the mutational spectra of BD and metabolites show that the number of point mutations at A:T base pairs is equal to or higher than that at the G:C base pairs (17, 21). This suggests that BD-adenine adducts might have a greater mutagenic potential than the guanine adducts. While the major EBadenine adducts have been isolated (20), DEB-induced DNA lesions have not been well characterized. Two adducts were reported from reactions of dAMP with DEB: peak A, UV spectrum with λmax ) 266 nm (major adduct) and peak B, λmax at 260 nm (22). Both adducts were found in DEB-exposed poly(dA-dT)(dA-dT), but only adduct B could be 32P postlabeled and detected in DEBtreated culture of Chinese hamster ovary cells. The identities of the adducts have not been established. Although the N6-DEB-Ade structure was suggested for the adduct B based on FAB MS data, the reported mass spectrum (22) does not give any indication of N6 substitution. From our experience with BD-DNA adducts (20), mass spectrometry data alone do not distinguish between the isomers substituted at different positions of a base and must be accompanied by UV and NMR data to achieve unambiguous identification. The present work was initiated in order to isolate and structurally characterize the DEB-induced adducts with adenine and to quantitate these adducts in DEB-exposed calf thymus DNA (CT DNA). We have previously studied reactions of EB with guanosine, adenine and adenosine (20) and DEB reactions with guanosine (24). On the basis of these earlier findings, we expected to find predominantly N-3 adducts from DEB reactions with free adenine and N-1/N6 adducts from its reactions with 2′deoxyadenosine. The adducts were isolated from reaction mixtures using HPLC and characterized by various spectral methods: nuclear magnetic resonance (NMR), UV spectrophotometry, and electrospray ionization mass spectrometry (ESI MS). The synthesized adducts were used as marker compounds to allow HPLC detection and quantitation of the corresponding species in hydrolysates of calf thymus DNA exposed to DEB.

Experimental Section Caution: DEB is carcinogenic and should be handled with extreme caution. Materials. HPLC-grade water and methanol were purchased from Fisher Scientific (Fair Lawn, NJ) and Malinckrodt (Paris, KY), respectively. DEB (d,l) was obtained from Aldrich (Milwaukee, WI). Calf thymus DNA, nucleobases, and nucleosides were acquired from Sigma Chemical Co. (St. Louis, MO). Solid phase extraction columns (ODS-AQ, 500 mg, 3 mL) were

Tretyakova et al. obtained from YMC (Wilmington, NC). Other chemicals were purchased from Fisher Scientific. Reactions of Adenine (Ade) with DEB. Ade (25 mg in 5 mL of 1:1 solution of methanol/10 mM Tris-HCl buffer, pH 7.2) was incubated with 10× molar excess of DEB for 10 h at 37 °C (environmental shaker). Aliquots of the reaction mixture (0.2 mL) were removed for analysis by HPLC to monitor reaction progress. The reactions were stopped by extracting the remaining DEB with diethyl ether, and the reaction mixtures were separated by HPLC. Reactions of 2′-Deoxyadenosine (2′-dAdo) with DEB. 2′dAdo (25 mg in 5 mL of water) was reacted with 860 mg (0.77 mL) of DEB at 37 °C for 10 h. The unreacted DEB was extracted with 4 × 5 mL of diethyl ether, and the mixture was heated at 80 °C for 1 h to convert the N-1-alkyladenosines into the corresponding N6-alkyladenosines (Dimroth rearrangement, ref 25). In order to release free bases, the solution was subjected to acid hydrolysis (0.2 N HCl at 80 °C for 1 h). The hydrolysate was neutralized with KOH and separated by HPLC. Reactions of DEB with Calf Thymus DNA. Calf thymus DNA (36 mg) was hydrated overnight in 5 mL of distilled water. The resulting solution was diluted to 10 mL with 10 mM TrisHCl buffer (pH 7.2) and sheared using a 10 mL syringe and needles of decreasing size (gauge 18-22); 2 mL aliquots of this solution were incubated with 36 µL of DEB at 37 °C for 18 h (environmental shaker). The mixtures were then extracted with 3 × 2 mL of diethyl ether and subjected to mild acid hydrolysis (0.1 N HCl, 70 °C for 30 min) followed by adjustment of the pH to 7 with KOH and filtration with Centricon 3 filters to remove the DNA backbone. The resulting filtrates were subjected to HPLC analysis. HPLC Analyses. HPLC-UV analyses were performed using C18 reverse-phase columns, a Rheodyne injector (Baxter, Charlotte, NC) equipped with a 2 mL injection loop, two Waters 510 HPLC pumps (Millipore, Milford, MA), and either a Hewlett Packard 1040A diode array detector (DAD) equipped with a HPLC-detection computer system or an Applied Biosystems, Inc. (Ramsey, NJ) 757 absorbance detector. For the primary separation of DEB + Ade reaction mixtures, a Beckman (Fullerton, CA) C18 semipreparative column (250 × 10 mm, 5 µm) was used. The flow rate was 3 mL/min, and the gradient (%A in B) was linear from 15% to 50% A in 30 min, where A was 70% aqueous methanol and B was 20 mM potassium phosphate buffer, pH 5.5. The fractions corresponding to reaction products were collected from multiple HPLC runs, pooled, and further purified with a Beckman octadecyl analytical column (25 × 0.46 cm, 5 µm) at a flow rate of 1.5 mL/min and the same gradient program. A purity index (spectra match value which ranges from 0 to 1000, where 1000 indicates no change in absorption spectra at consecutive time points of a peak) was obtained for the pooled fractions using the DAD data to ensure proper separation. The adducts were freed from phosphate salts using ODS-AQ cartridges (YMC). The cartridges (500 mg solid phase, 3 mL volume) were primed with 6 mL of methanol and 9 mL of water, and the samples were loaded by gravity. The columns were washed with 6 mL of water, and the samples were recovered with 6 mL of methanol. The resulting solutions were evaporated under reduced pressure and redissolved in 50 µL of water; 5-10 µL of this solution was injected in the LC for LC/ESI+ MS analyses. DEB + dAdo reaction mixtures were analyzed using an Altima C18, 250 × 4.6 mm, 5 µm column (Alltech) eluted at 1.5 mL/min with a linear gradient from 15% to 50% A over 30 min, where A was 70% aqueous methanol and B was water. The reaction products were collected and purified as described above. For the analyses of the acid hydrolysates of DEB-exposed CT DNA, a Beckman C18, 250 × 4.6 mm, 5 µm column was eluted at 1.5 mL/min. The gradient program was linear from 5% to 25 % A in 35 min, where A was 70% aqueous methanol and B was 20 mM phosphate buffer, pH 4.7. The peaks were identified from their UV spectra and retention times which were compared to those of the authentic standards. Further evidence for structural assignment was obtained from the ESI+ MS spectra

Adenine Adducts of Diepoxybutane

Chem. Res. Toxicol., Vol. 10, No. 10, 1997 1173

Figure 1. HPLC separation of DEB + adenine reaction mixture: 250 × 10 mm C18 column (Beckman), 3 mL/min, linear gradient from 15% to 50% A in 30 min, A ) 70% methanol/water, B ) 20 mM KH2PO4, pH 5.5; UV detection at 275 nm. of the pooled, desalted fractions. Quantification was performed using a single-wavelength detector set to λ ) 274 nm for N-3DEB-Ade adduct and λ ) 267 nm for N6-DEB-Ade adduct with calibration curves obtained for standard solutions. The calibration was performed as follows: standard solutions containing pure, desalted marker compounds of known molarity (20-50 µM) were prepared in water and stored at -20 °C. The concentrations were confirmed by UV spectrophotometry. Calibration solutions were prepared by a series of dilutions and analyzed by HPLC. The peak areas obtained at 274 nm (N-3 adduct) or 267 nm (N6 adduct) were used to generate standard curves relating concentration to the peak area. These curves were then used to obtain concentrations from peak areas for DNA hydrolysates. UV Spectrophotometry. Amounts of 0.5-1 mg of each adduct were dissolved in 1 mL of distilled water, 0.1 N HCl, and 0.1 N NaOH. The ultraviolet (UV) spectra were recorded on a Shimadzu (Columbia, MD) 160U UV spectrophotometer. Mass Spectrometry. Electrospray ionization mass spectrometry (ESI+ MS) was performed on a Finnigan 4000 quadrupole mass spectrometer (Finnigan-MAT, San Jose, CA) equipped with a pneumatic electrospray source (Analytica of Bradford, Inc., Branford, CT); 5-10 µL of aqueous solution was used for LC/ESI+ MS analysis. LC/ESI+ MS analyses were conducted using a Pharmacia liquid chromatographic system with dual pumps #2248 (Pharmacia LKB Biotechnology, Uppsala, Sweden). Chromatographic separations were accomplished using a capillary C18 column (150 × 0.8 mm, Hypersil, 3 µm particle size; LC Packings, San Francisco, CA) by changing the mobile phase from 3% to 75% aqueous methanol in 18 min, altering the phase to 98% methanol in 4 min, and washing the column with 100% methanol for 3 min. A voltage of 3.1 kV was applied to the electrospray needle, and 70 psi of the nebulizer gas (N2) was applied to stabilize the spray. The voltage difference between the exit of glass capillary and the first skimmer in the differential pumping region (skimmer offset) was optimized at 80 V to obtain (M + H)+ ions of the adducts and was set to 160 V to promote collision-induced dissociation of the (M + H)+ ions. The data were acquired and processed by a Technivent Vector 2 data system (ProLab Resources, Madison,

WI). Full scan mass spectra were obtained by scanning from m/z 50 to 500 every 2 s. NMR Spectrometry. Amounts of 0.5-1 mg of purified, desalted, and desiccated compounds were dissolved in 0.5 mL of DMSO-d6. The 1H NMR spectra were obtained on a Bruker AMX-500 MHz spectrometer, and chemical shifts are reported in ppm relative to the proton resonance of TMS.

Results DEB Reactions with Adenine. Reaction of adenine with DEB at pH 7.2 yielded three products with retention times of 8.0, 9.0, and 13.5 min on a semipreparative column (Figure 1). Preliminary identification of these adducts as N-3-DEB-Ade, N-7-DEB-Ade, and N-9-DEBAde, respectively, was based on comparison of their UV spectra at differing pH values to published spectra of various alkylated adenines (Table 1). The molecular weights of all three products were established as 221 Da since (M + H)+ ions were observed at m/z ) 222 under ESI+ MS conditions (e.g., Figure 2a). When ESI+ MS was performed under conditions promoting collision-induced dissociation (skimmer offset ) 160 V), a fragment peak at m/z ) 136 (protonated adenine) was observed for all three compounds, confirming their structures as adenine adducts (e.g., Figure 2a). The mass spectral data suggested that these adducts were adenines which had captured one molecule of DEB, with the second epoxy ring remaining intact: Ade–CH2–CHOH–CH

CH2,

M = 221

O

Further information on the structures of N-3-, N-7-, and N-9-DEB-Ade was obtained from proton NMR spectra. Proton assignment in the NMR spectrum of N-3DEB-Ade was based on comparison of chemical shifts, multiplicities, coupling constants and integration ratios

1174 Chem. Res. Toxicol., Vol. 10, No. 10, 1997

Tretyakova et al.

Table 1. Retention Times and UV Spectra of the DEB Adducts with Adenine Compared to Reference Compounds tR, min

λmax/λmin

compound

SP columna

AN columnb

pH 7

pH 1

pH 12

MW

N-3-DEB-Ade N-3-DEB-Ade I N-3-DEB-Ade II N-3-DEB-Ade III N-7-DEB-Ade N-9-DEB-Ade N6-DEB-Ade N-3-Me-Ade N-7-Me-Ade N-9-Me-Ade N6-Me-Ade

8.1 4.1 5.8 10.5 9.1 13.0

3.6 2.2 7.3 9.4 4.3 7.2 13.2

274/241 270/236 274/240 274/240 269/235 260/227 267/230 273/244 268/232 261/230 267/230

275/242 270/236 275/242 275/242 272/239 260/231 266/230 273/243 271/237 261/230 267/230

273/245 degraded 273/245 273/245 269/240 262/236 273/240 273/243 270/234 262/228 273/240

221 222 239 257 221 221 239 149 149 149 149

a 25 × 1 cm, 5 µm Beckman column; 15-50% A in 30 min, A ) 70% aqueous methanol, B ) 20 mM phosphate buffer, pH 5.5. b 25 × 0.46 cm, 5 µm Beckman column; the same gradient as above. c Obtained with commercial standards.

Chart 1. Structures of DEB-Ade Adducts

Figure 2. ESI+ MS of N-3-DEB-Ade (a), N-3-DEB-Ade II (b), and N-3-DEB-Ade III (c): Finnigan 4000 mass spectrometer with an Analytica of Bradford pneumatic electrospray source, voltage at the ES needle ) 3.1 kV, N2 pressure ) 70 psi, skimmer offset ) 160 V, scanning from m/z ) 70 to 500 in 2 s.

with the spectra for N-7-(2′,3′,4′-trihydroxybut-1′-yl)guanine (24), adenine, and N-3-methyladenine standards. Both adenine and alkyl chain signals were present in the spectrum (Figure 3 and Table 2). As in the case of N-3EB-Ade adducts, the amino protons were observed as two broad singlets at =8 ppm, probably due to the contribution of the imino form (20). The chemical shifts for H-8 and H-2 protons (7.7 and 8.2 ppm, respectively) were similar to the corresponding signals in the spectra of N-3-

methyladenine and N-3-EB-adenine (20), providing further evidence for N-3 substitution. The diastereotopic 1′- and 4′- methylene protons of the side chain appeared as doublets of doublets, similar to the spectrum of N-7DEB-Gua (24). However, only one hydroxyl proton signal was observed for N-3-DEB-Ade (a doublet at 5.6 ppm), consistent with the presence of a 3′,4′-epoxy group in the molecule. 2′- and 3′-methine protons gave multiplets at 3.0 and 3.9 ppm (Table 2). All spectral evidence supported the structure N-3-(2′-hydroxy-3′,4′-epoxybut-1′-yl)adenine for N-3-DEB-Ade (Chart 1). Early in these studies it was noticed that N-3-DEBAde was unstable and degraded upon storage (even at -20 °C) or during solvent evaporation under vacuum. HPLC analysis of the resulting solution using a 4.6 mm column revealed three major products which were named N-3-DEB-Ade I, II, and III in the order of their retention

Adenine Adducts of Diepoxybutane

Chem. Res. Toxicol., Vol. 10, No. 10, 1997 1175

Figure 3. Proton NMR spectrum of N-3-DEB-Ade in DMSO-d6 (Bruker AMX-500 spectrophotometer, chemical shifts relative to TMS).

Scheme 1. N-3 Adduct Formation in Reactions of DEB with Adenine

times. UV spectra of the products at 3.4 min (N-3-DEBAde II) and 8.4 min (N-3-DEB-Ade III) were identical to that of the original adduct (λmax ) 274 nm, λmin ) 240 nm), suggesting that they were also N-3-substituted adenines. (M + H)+ ions were observed at m/z ) 240 and 258 for N-3-DEB-Ade II and III, respectively (Figure 2b,c). The mass spectrum of N-3-DEB-Ade III had a characteristic isotope pattern of m/z ) 258:260 ) 3:1 suggesting the presence of one chlorine atom in the molecule (Figure 2c). This is consistent with the structures of N-3-(2′,3′,4′-trihydroxybut-1′-yl)adenine (N-3DEB-Ade II) and N-3-(2′,3′-dihydroxy-4′-chlorobut-1′yl)adenine (N-3-DEB-Ade III) (Chart 1). The formation of these products from N-3-(2′-hydroxy3′,4′-epoxybut-1′-yl)adenine (N-3-DEB-Ade) can be explained as follows (Scheme 1). The 3′,4′ epoxy group of the original adduct is highly reactive and can be opened in the presence of nucleophilic species in solution. Simple hydrolysis leads to N-3-DEB-Ade II, and chloride-ion attack at the epoxy group produces N-3-DEB-Ade III. We

have previously reported similar reactions of N-7-(2′hydroxy-3′,4′-epoxybut-1′-yl)guanine, the intermediate formed from reactions of DEB with calf thymus DNA (24). The early peak at 2.1 min (N-3-DEB-Ade I) had an UV spectrum with λmax ) 270 nm, different from the spectrum of the parent compound (Table 1). Structural assignment of this adduct was based on UV, NMR, and mass spectra (Tables 1 and 2) and also on the fact that a similar product was formed from N-9-DEB-Ade (see below). We suggest that N-3-DEB-Ade I is formed from N-3-DEB-Ade through an intramolecular attack of the remaining epoxy ring at N-9 of the same molecule of adenine. This cyclization process results in formation of a 6-membered ring including the N-3 and N-9 positions of the adenine (Scheme 1). The molecular weight of N-3DEB-Ade I was determined to be 222 Da, consistent with the proposed structure. A very short retention time of N-3-DEB-Ade I on a C18 column and its instability at high pH (Table 1) give evidence for the cationic nature of this adduct. Proton NMR spectrum (not shown) suggested substitution at N-3 since the amino hydrogens were seen as two broad singlets at 9.5 and 9.6 ppm (see above). Two hydroxyl proton signals were present in the spectrum (at 4.6 and 4.7 ppm). In general, all signals in the spectrum were shifted downfield compared to other adducts described in this work (Table 2), probably due to the presence of positive charge. Proton NMR analysis of N-7-DEB-Ade supports its identification as an N-7 adenine adduct (Table 2). H-2 and H-8 protons appear as a two-proton singlet at 8.2 ppm. Unlike N-3-DEB-Ade (above) and N6-DEB-Ade (below), the amino protons were observed as a sharp twoproton singlet at 6.88 ppm, suggesting substitution of the imidazole rather than the pyrimidine ring. The pattern of side chain proton signals was similar to that of N-3DEB-Ade (see above), but the 4′-CH2 proton signals were shifted downfield by about 1 ppm (Table 2). The coupling constants were also somewhat larger than those normally seen for epoxides. However, the electrospray ionization mass spectrum shows pseudomolecular ions at m/z ) 222, consistent with the presence of a 3′,4′-epoxy ring. Cyclization to N6 can be ruled out by the sharp twoproton signal of NH2 observed in the NMR spectrum.

1176 Chem. Res. Toxicol., Vol. 10, No. 10, 1997 Table 2. Proton NMR Spectroscopic Data for DEB-Adenine Adducts compound

chemical shift, ppm

coupling constant, Hz assignment

N-3-DEB-Ade 2.61 (dd, 1H) 2.70 (dd, 1H) 3.03 (m, 1H) 3.89(m, 1H) 4.21 (dd, 1H) 4.49 (dd, 1H) 5.56 (d, 1H) 7.73 (s, 1H) 7.83, 7.90 (2 br s, 2H) 8.21 (s, 1H)

J ) 5.1, 2.65 4′-CH2 J ) 5.1, 4.5 4′-CH2 3′-CH 2′-CH J ) 3.9, 14 1′-CH2 J ) 3.9, 14 1′-CH2 2′-OH H-8 NH2 H-2

3.53 (dd, 1H) 3.66 (dd, 1H) 3.70 (m, 1H)) 3.82 (br d, 1H) 4.26 (dd, 2H) 4.51 (dd, 2H) 5.48 (br s, 1H) 6.88 (s, 2H) 8.20 (s, 2H)

J ) 10.5, 7 4′-CH2 J ) 10.5, 5.8 4′-CH2 3′-CH 2′-CH J ) 15, 10 1′-CH2 J ) 15, 3 1′-CH2 2′-CHOH NH2 H-2, H-8

2.48 (dd, 1H) 2.62 (dd, 1H) 2.95 (m, 1H) 3.71 (m, 1H) 4.14 (dd, 1H) 4.26 (dd, 1H) 5.50 (d, 1H) 7.18 (s, 2H) 8.07 (s, 1H) 8.12 (s, 1H)

J ) 5.3, 3.0 J ) 5.3, 4.3

3.39 (br d, 1H) 3.41 (br d, 1H) 3.57 (m, 1H) 3.67 (m, 1H) 4.77 (br m, 1H) 4.60 (br m, 1H) 7.35 (br s, 1H) 8.09 (s, 1H) 8.16 (s, 1H) 12.93 (br s, 1H)

J)6 J)6

N-7-DEB-Ade

N-9-DEB-Ade

J ) 14, 7.9 J ) 4.6, 14 J ) 5.8

4′-CH2 4′-CH2 3′-CH 2′-CH 1′-CH2 1′-CH2 2′-OH NH2 H-8 H-2

N6-DEB-Ade 4′-CH2 4′-CH2 3′-CH 2′-CH 1′-CH2 1′-CH2 N6H H-8 H-2 H-9

N-3-DEB-Ade I 4.08 (m, 1H) 4.21 (m, 1H) 4.60 (m, 2H) 4.78 (d, 1H) 4.80 (2 br dd, 2H) 6.03 (m, 1H) 6.12 (m, 1H) 8.62 (s, 1H) 8.91 (s, 1H) 9.51, 9.62 (2 br s, 2H)

1′-CH2 1′-CH2 4′-CH2 4′-OH 3′-CHOH 4′-CH2 3′-CH 2′-CH H-8 H-2 NH2

Furthermore, the UV and the mass spectra do not support a cyclic structure. UV spectra of this compound at varying pH values completely match the UV spectra of other N-7-substituted adenines (Table 1). In MS experiments, the entire substituent was readily cleaved under CID conditions giving ions of protonated adenine, a fragmentation not favorable for a cyclic adduct. Therefore, we assign the structure N-7-(2′-hydroxy-3′,4′-epoxybut-1′-yl)adenine (N-7-DEB-Ade) to this adduct (Chart 1). Further structural studies are in progress. N-7-DEBAde was also unstable upon storage giving two products: N-7-(2′,3′,4′-trihydroxybut-1′-yl)adenine (tR ) 4.6 min) and an unidentified compound (tR ) 3.5 min) with UV spectrum different from that of the parent compound (λmax ) 280 nm, λmin ) 235 nm).

Tretyakova et al.

Evidence for substitution at N-9 for the adduct identified as N-9-DEB-Ade was provided by the proton NMR spectrum (Table 2). Similar to the N-7 isomer above, NH2 protons were seen as a 2H singlet at 7.2 ppm. H-2 and H-8 protons were resolved, giving two singlets at 8.09 and 8.16 ppm, respectively. The pattern of side chain proton signals was similar to that of N-3- and N-7-DEBAde (Table 2). Spectral data suggested the structure of N-9-(2′-hydroxy-3′,4′-epoxybut-1′-yl)adenine for this adduct (Chart 1). As for N-3- and N-7-DEB-Ade, N-9-DEBAde was unstable when subjected to heating or storage. Interestingly, the major degradation product of N-9-DEBAde (N-9-DEB-Ade I) had similar HPLC retention time and UV and mass spectra to N-3-DEB-Ade I, the product obtained from N-3-DEB-Ade (see above and Table 1). Therefore, we suggest that N-9-DEB-Ade I and N-3-DEBAde I are isomers formed by a similar intramolecular nucleophilic substitution mechanism (Scheme 1). DEB Reactions with 2′-Deoxyadenosine (dAdo). HPLC separation of the DEB + dAdo reaction mixture before hydrolysis showed at least five new products in addition to unreacted dAdo (results not shown). Heating the solution at 80 °C for 1 h in order to induce Dimroth rearrangement of the N-1-dAdo adducts into the corresponding N6-dAdo adducts shifted the retention times of the products from 21-28 min to 12-14 min, but the HPLC chromatogram still contained five peaks which were difficult to resolve. We decided to hydrolyze 2′deoxyribonucleosides to the corresponding nucleobases which should have fewer isomers. Heating the reaction mixture at 80 °C in 0.2 N HCl for 1 h resulted in a hydrolysate which contained only one adduct peak in addition to unreacted adenine (Figure 4, the peak at 12.3 min). UV spectrum of this compound (λmax ) 267 nm, λmin ) 230 nm at pH 7) was consistent with substitution at the N6 of adenine (N6-DEB-Ade, Table 1). ESI+ MS analysis of N6-DEB-Ade resulted in molecular ions (M + H+) at m/z ) 240, corresponding to adenine containing one trihydroxybutyl substituent:

Ade-CH2-CHOH-CHOH-CH2OH, M ) 239 Fragmentation of the (M + H) + ions (m/z ) 240) under CID conditions gave rise to product ion peaks at m/z ) 136, 148, and 119. The molecular ion of N6-DEB-Ade was more stable under CID conditions than those of N-3-, N-7-, or N-9-DEB-Ade and required higher skimmer offset to induce fragmentation. This behavior had previously been observed for other N6-adenine adducts (26). Analysis of the proton NMR spectrum (Table 2) provided further information on the structure of N6-DEBAde. Just as for authentic N6-methyladenine, an H-9 proton resonance was present at 12.93 ppm, and the signals for H-8 and H-2 appeared as singlets at 8.09 and 8.16 ppm, respectively. The N6H signal appeared as a broad singlet at 7.35 ppm (7.5 ppm for the N6-methyladenine standard). This signal and the resonance of the H-9 proton at 12.93 ppm establish substitution at N6 of adenine, since the adenines alkylated at other positions (N-1, N-3, or N-7) lack the H-9 proton (Table 2). The signal for 4′-CH2 of the side chain appeared as two broad doublets at 3.39 and 3.41 ppm, and 1′-CH2 was seen as two multiplets at 4.77 and 4.60 ppm. Based on the spectral evidence, N6-DEB-Ade was identified as N6(2′,3′,4′-trihydroxybut-1′-yl)adenine (Chart 1, N6-DEBAde). No 3′,4′-epoxy intermediate was isolated in this case, since acid hydrolysis was performed prior to HPLC separation.

Adenine Adducts of Diepoxybutane

Chem. Res. Toxicol., Vol. 10, No. 10, 1997 1177

Figure 4. HPLC separation of DEB + dAdo reaction mixture after acid hydrolysis: 250 × 4.6 mm C18 column (Alltech), 1.5 mL/ min, linear gradient from 15% to 50% A in 30 min, A ) 70% aqueous methanol, B ) water; UV detection at 267 nm.

extent of alkylation at N-7 of guanine (24). Comparison of these results to the data previously obtained for EBexposed DNA (20) reveals that DEB appears to induce fewer N-3-adenine adducts than EB does. N-1-DEB-Ade adducts, if formed, readily rearrange into the corresponding N6 species upon heating at 80 °C for 1 h, since no N-1-DEB adducts were observed in reaction mixtures.

Discussion

Figure 5. HPLC separation of an acid hydrolysate of DEBexposed calf thymus DNA: 25 × 0.46 cm C18 column (Beckman), 1.5 mL/min, linear gradient from 5% to 25% A, A ) 70% methanol/water, B ) 20 mM KH2PO4, pH 4.5; UV detection at 275 nm.

Calf Thymus DNA Reactions with DEB. The reaction of calf thymus DNA with diepoxybutane under physiological conditions followed by a mild acid hydrolysis resulted in four HPLC product peaks in addition to unreacted purines (Figure 5). The compounds at 5.0 and 17.2 min were identified as N-3-DEB-Ade II and N6-DEBAde, respectively, based on their retention times, UV spectra, and mass spectra, all of which were compared to those of the corresponding marker compounds. The peak at 11.8 min was the previously described N-7(2′,3′,4′-trihydroxybut-1′-yl)guanine, the major DEBinduced DNA adduct (24). The extent of alkylation of N-3 and N6 positions of Ade in exposed DNA was 13 and 5 adducts/103 normal adenines, respectively. This corresponds to approximately 1/5 (N-3) and 1/10 (N6) the

The goal of the present study was to synthesize various DEB adducts of adenine and to analyze them in DEBexposed calf thymus DNA. DEB reaction with free adenine resulted in alkylation of the most nucleophilic positions of Ade to give N-3-(2′-hydroxy-3′,4′-epoxybut1′-yl)adenine (N-3-DEB-Ade), N-7-(2′-hydroxy-3′,4′-epoxybut-1′-yl)adenine (N-7-DEB-Ade), and N-9-(2′-hydroxy3′,4′-epoxybut-1′-yl)adenine (N-9-DEB-Ade). This suggests an SN2-type nucleophilic substitution which can be expected for DEB and other aliphatic epoxides at physiological conditions (23). N-3-Alkyladenines have been previously reported for reactions of adenine with ethylene oxide (25), propylene oxide (27), and simple methylating agents (28). N-9- and N-7-methyladenines were also previously isolated from reactions of adenine with dimethyl sulfate (28). In the present work, the N-3-adenine adduct was predominant followed by N-9- and N-7-DEBAde. This to some extent correlates with the nucleophilic strengths of the different nitrogens in adenine (29). The presence of deoxyribose at the N-9 position in 2′dAdo dramatically shifted its reactivity compared to free adenine. One product, N6-DEB-Ade, was observed, which probably formed from N-1-DEB-Ade (not isolated) through Dimroth rearrangement (25, 27, 30). We have previously isolated N-1-adenine adducts from reactions of EB with 2′-dAdo and demonstrated their rearrangement to the corresponding N6 adducts upon heating at pH 7 (20). In both cases, adenosine reactions produced much lower yields of the adduct than reactions of free adenine.

1178 Chem. Res. Toxicol., Vol. 10, No. 10, 1997

Tretyakova et al.

3,4-Epoxy-1-butene, the metabolic precursor of DEB, has been reported to form regioisomeric pairs of adenine adducts resulting from attack at C-3 and C-4 of the epoxide (20). Nucleophilic attack at the primary epoxide center can be expected for SN2 subsitution. However, the stereoelectronic effects of the vinyl group apparently lead to accelaration of SN2 attack at the allyl position (C-3) of EB resulting in formation of both regioisomers. The present data suggest that in reactions of DEB with adenine, nucleophilic substitution takes place exclusively at the terminal carbon, resulting in a single regioisomer for each adduct. Similar regiochemistry has been previously reported for reactions of DEB with guanosine (24). Since the internal carbons of DEB are not activated, reaction at the terminal carbon is favored. DEB reaction with a nucleobase leads to opening of one of the two oxirane rings present. The remaining epoxy group is highly reactive and participates in reactions with nucleophilic species present in solution (hydroxide anion, chloride anion, etc). We have previously reported formation of multiple adducts from N-7-(2′hydroxy-3′,4′-epoxybut-1′-yl)guanine, an unstable intermediate from DEB + guanine reaction (24). Similar products were observed for N-3-, N-7-, and N-9-(2′hydroxy-3′,4′-epoxybut-1′-yl)adenines. (2′,3′,4′-Trihydroxybut-1′-yl)adenines were produced through simple hydrolysis (N-3-DEB-Ade II); chlorohydrins were formed through nucleophilic substitution involving chloride anion (N-3-DEB-Ade III). Cyclic products were observed due to reactions of the second epoxy group with the same molecule of adenine (N-3-DEB-Ade I). Alternatively, the remaining epoxy ring may react with another nucleobase in DNA producing cross-links. Brookes and Lawley suggested in 1965 that the powerful cytotoxicity of bifunctional agents, such as bis(2-chloroethyl) sulfide, was due to their ability to cross-link the strands of the DNA molecule preventing its replication (31). No crosslinks were observed in our experiments with adenine and 2′-deoxyadenosine, but a study is underway to identify adenine and guanine cross-links in DEB-exposed oligonucleotides. Since the commercial diepoxybutane used in this study was a racemic mixture mixture of d and l isomers (no meso isomer), the monoepoxy products formed from an SN2-type attack of a nucleobase at the terminal carbon of DEB were expected to be a mixture of enantiomers which can not be separated by conventional HPLC: O Base

O

CH2

CH • CHOH (R) (R)

CH2

or

Base

CH2

CH • CHOH (S) (S)

CH2

Indeed, only one HPLC peak for each adduct was observed under varying HPLC conditions (Figure 1). The stereochemical outcome of hydrolysis of the epoxide (Scheme 1) will depend on the mechanism, since the nucleophilic attack can take place at C-3′ or C-4′, resulting in either retention or inversion of configuration. Inversion at C-3′ should result in diastereomeric pairs of adducts which could in principle be resolved by HPLC. Indeed, there was an indication of their presence under some HPLC conditions (results not shown). The adducts synthesized in the first part of this study were applied as marker compounds to analyze the lesions present in DEB-exposed calf thymus DNA. N-9-DEBAde was not expected to form in DNA since the N-9

position of adenine is blocked by deoxyribose. Both N-3DEB-Ade II and N6-DEB-Ade were found in acid hydrolysates of exposed calf thymus DNA, whereas the N-7 adduct was not detected. Lawley and Brookes previously described alkylation of adenine in deoxyadenylic acid and DNA at N-3, N-7, and N-1 (N6) positions with dimethyl sulfate and propylene oxide (32). The ratio of N-7-Gua to N-3-Ade adducts observed in the present study (5:1) is the same as reported for propylene oxide (32) and is lower than for EB (10:1) (20). N-3-, and N-7-alkylations of purines in DNA result in destabilization of the glycoside bond due to quaternization of the alkylated nitrogen. As a result, both N-3DEB-Ade and N-7-DEB-Gua adducts are subject to spontaneous depurination. N-3-Adenine adducts were reported to be released 6-fold faster than N-7-guanine adducts (33). In addition, both are subject to repair by methylpurine glycosylase. These processes lead to the formation of apurinic sites, which have been shown to cause mutations if not repaired before DNA synthesis takes place (34). On the other hand, the facile release of N-3-adenine and N-7-guanine adducts from DNA can be useful in applying them as noninvasive biomarkers of exposure to BD by monitoring their amounts in urine (35). Unlike the N-3-adenine adducts, N6- and N-1-DEBadenines are not expected to be spontaneously released from DNA. These adducts are also more likely to be promutagenic due to the alkylation in the base-pairing region. The conversion of N-1-alkyladenines to N6 adducts has been reported to be slow but measurable at neutral pH and 37 °C and was slower in DNA than in free nucleotides (33). Although we found N-3 adducts to be more abundant in DEB-exposed DNA than the N6DEB-adenines (13 and 5 adducts/103 normal Ade, respectively), further studies involving measurements of these adducts in tissues of exposed animals are required to estimate the relative importance of various types of adenine adducts for butadiene-induced carcinogenesis.

Acknowledgment. The authors would like to acknowledge Patricia Upton and Gregory Janis for assistance with the manuscript. This work was sponsored in part by a grant from Chemical Manufacturers Association.

References (1) Morrow, N. L. (1990) The industrial production and use of 1,3butadiene. Environ. Health Perspect. 86, 7-8. (2) Pelz, N., Dempster, A. M., and Shore, P. R. (1990) Analysis of low molecular weight hydrocarbons including 1,3-butadiene in engine exhaust gases using an aluminum oxide porous-layer open tubular fused-silica column. J. Chromatogr. Sci. 28, 230-235. (3) Santos-Burgoa, C., Matanoski, G. M., Zeger, S., and Schwartz, L. (1992) Lymphohematopoietic cancer in styrene-butadiene polymerization workers. Am. J. Epidemiol. 136, 834-854. (4) Delzell, E., Sathiakumar, N., Macaluso, M., Hovinga, M., Larson, R., Barbone, F., Beall, C., and Cole, P. A. (1996) A follow-up study of synthetic rubber workers. Toxicology 113, 182-189. (5) International Agency for Research on Cancer (IARC). (1992) 1,3Butadiene. IARC Monogr. Eval. Carcinogen. Risk Chem. Hum. 54, 237-285. (6) Owen, P. E., Glaister, J. R., Gaunt, I. F., and Pullinder, D. H. (1987) Inhalation toxicity studies with 1,3-butadiene. 3. Two year toxicity/carcinogenicity study in rats. Am. Ind. Hyg. Assoc. J. 48, 407-413. (7) Melnick, R. L., Huff, J. E., Roycroft, J. H., Chou, B. J., and Miller, R. H. (1990) Inhalation toxicology and carcinogenicity of 1,3butadiene in B6C3F1 mice following 65 weeks of exposure. Environ. Health Perspect. 86, 27-36.

Adenine Adducts of Diepoxybutane (8) National Toxicology Program (NTP). (1993) NTP Technical report on the toxicology and carcinogenesis studies of 1,3-butadiene, NTP TR No. 434, Research Triangle Park, NC. (9) Malvoisin, E., Lhoest, G., Poncelet, F., Roberfroid, M., and Mercier, M. (1979) Identification and quantitation of 1,2-epoxybutene-3 as the primary metabolite of 1,3-butadiene. J. Chromatogr. 178, 419-425. (10) Malvoisin, E., Mercier, M., and Roberfroid, M. (1982) Enzymatic hydration of butadiene monoxide and its importance in the metabolism of butadiene. Adv. Exp. Med. Biol. 38A, 437-444. (11) Sharer, J. E., Duescher, R. J., and Elfarra, A. A. (1991) Formation, stability and rearrangements of the glutathione conjugates of butadiene monoxide: evidence for the formation of stable sulfurane intermediates. Chem. Res. Toxicol. 4, 430-436. (12) Himmelstein, M., Turner, M. J., Asgharin, B., and Bond, J. A. (1994) Comparison of blood concentrations of 1,3-butadiene and butadiene epoxides in mice and rats exposed to 1,3-butadiene by inhalation. Carcinogenesis 15, 1479-1486. (13) Thronton-Manning, J. R., Dahl, A. R., Bechtold, W. E., Griffith, W. C., Jr., and Henderson, R. F. (1995) Disposition of butadiene monoepoxide and butadiene diepoxide in various tissues of rats and mice following a low-level inhalation exposure to 1,3butadiene. Carcinogenesis 16, 1723-1731. (14) Gervasi, P. G., Citti, L., Del Monte, M., Longo, V., and Benetti, D. (1985) Mutagenicity and chemical reactivity of epoxic intermediates of butadiene, isoprene and structurally related compounds. Mutat. Res. 158, 77-82. (15) Tice, R. R., Boucher, R., Luke, C. A., and Shelby, M. D. (1987) Comparative cytogenetic analysis of bone marrow damage induced in male B6C3F1 mice by multiple exposures to gaseous 1,3butadiene. Environ. Mutagen. 9, 235-250. (16) Cochrane, J. E., and Skopek, T. R. (1994) Mutagenicity of butadiene and its epoxide metabolites: I Mutagenic potential of 1,2-epoxybutene, 1,2,3,4-diepoxybutane and 3,4-epoxybutanediol in cultured human lymphoblasts. Carcinogenesis 15, 713-717. (17) Cochrane, J. E., and Skopek, T. R. (1994) Mutagenicity of butadiene and its epoxide metabolites: II Mutational spectra of butadiene, 1,2-epoxybutene and diepoxybutane at the hprt locus in splenic T cells from exposed B6C3F1 mice. Carcinogenesis 15, 719-723. (18) Lawley, P. D., and Brookes, P. (1967) Interstrand cross-linking of DNA by bifunctional alkylating agents. J. Mol. Biol. 25, 143160. (19) Citti, L., Gervasi, P. G., Turch, G., Bellucci, G., and Bianchini, R. (1984) The reaction of 3,4-epoxy-1-butene with deoxyguanosine and DNA in vitro: synthesis and characterization of the main adducts. Carcinogenesis 5, 47-52. (20) Tretyakova, N. Yu., Lin, Y., Sangaiah, R., Upton, P. B., and Swenberg, J. A. (1997) Identification and quantitation of DNA adducts from calf thymus DNA exposed to 3,4-epoxy-1-butene. Carcinogenesis 18, 137-147. (21) Recio, L., and Meyer, K. G. (1995) Increased frequency of mutations at A:T base pairs in the bone marrow of B6C3F1 lacI

Chem. Res. Toxicol., Vol. 10, No. 10, 1997 1179

(22)

(23) (24)

(25)

(26)

(27)

(28)

(29) (30)

(31)

(32)

(33)

(34)

(35)

transgenic mice exposed to 1,3-butadiene. Environ. Mol. Mutagen. 26, 1-8. Leuratti, C., Jones, N. J., Marafante, E., Kostaiainen, R., Peltonen, K., and Waters, R. (1994) DNA damage induced by the environmental carcinogen butadiene: identification of a diepoxybutane-adenine adduct and its detection by 32P-postlabelling. Carcinogenesis 15, 1903. Singer, B., and Grunberger, D. (1983) Molecular Biology of Mutagens and Carcinogens, Plenum Press, New York. Tretyakova, N. Yu., Sangaiah, R., Yen, T.-Y., and Swenberg, J. A. (1997) Synthesis, characterization, and in vitro quantitation of N-7-guanine adducts of diepoxybutane. Chem. Res. Toxicol. 10, 779-785. Li, F., Segal, A., and Solomon, J. J. (1992) In vitro reaction of ethylene oxide with DNA and characterization of DNA adducts. Chem.-Biol. Interact. 83, 35-54. Esmans, E., Vanhoutte, K., Lemiere, F., and Van Dongen, W. Structural analysis of 2′-deoxynucleot(s)ide adducts with electrospray tandem mass spectrometry. Poster presentation at the 44th ASMS conference, May 1996, Portland, OR. Solomon, J. J., Mukai, F., Fedyk, J., and Segal, A. (1988) Reactions of propylene oxide with 2′-deoxynucleosides and in vitro with calf thymus DNA. Chem.-Biol. Interact. 67, 275-294. Lawley, P. D., and Brookes, P. (1964) Methylation of adenine in deoxyadenylic acid or deoxyribonucleic acid at N-7. Biochem. J. 92, 19c-20c. Pal, B. (1962) Studies on the alkylation of purines and pyrimidines. Biochemistry 1, 558-563. Segal, A., Mate, U., and Solomon, J. J. (1979) In vitro Dimroth rearrangement of 1-(2-carboxyethyl)adenine to N6-(2-carboxyethyl)adenine in single-stranded calf thymus DNA. Chem.-Biol. Interact. 28, 333-344. Lawley, P. D., and Brookes, P. (1965) Molecular mechanism of the cytotoxic action of difunctional alkylating agents and of resistance to its action. Nature 206, 480-485. Lawley, P. D., and Jarman, M. (1972) Alkylation by propylene oxide of deoxyribonucleic acid, adenine, guanine and deoxyguanylic acid. Biochem. J. 126, 893-900. Lawley, P. D., and Brookes, P. (1963) Further studies on the alkylation of nucleic acids and their constituent nucleotides. Biochem. J. 89, 127-138. Gentil, A., Cabral-Neto, J. B., Mariage-Samson, R., Margot, A., Imbach, J. L., Rayner, B., and Sarasin, A. (1992) Mutagenicity of a unique apurinic/apyrimidinic site in mammalian cells. J. Mol. Biol. 227, 981-984. Shuker, D. E. G., and Farmer, P. B. (1992) Relevance of urinary DNA adducts as markers of carcinogen exposure. Chem. Res. Toxicol. 5, 450-460.

TX9700681