Synthesis and Biochemical Characterization of N1-, N2-, and N7

Ewan D Booth , Joanne D Kilgour , Stephen A Robinson , William P Watson ... Minko , M. Todd Washington , Louise Prakash , Satya Prakash , R. Stephen L...
0 downloads 0 Views 447KB Size
Download PDF
126

Chem. Res. Toxicol. 1996, 9, 126-132

Synthesis and Biochemical Characterization of N1-, N2-, and N7-Guanosine Adducts of Butadiene Monoxide Rebecca R. Selzer and Adnan A. Elfarra* Department of Comparative Biosciences and Environmental Toxicology Center, University of Wisconsin, Madison, Wisconsin 53706 Received June 13, 1995X

1,3-Butadiene is a known rodent carcinogen, but the molecular mechanisms of its carcinogenicity are poorly understood. Butadiene monoxide (BM), a known mutagenic metabolite of 1,3-butadiene, was previously shown to react with guanosine to yield two N7-guanine adducts. In the present study, eight guanosine adducts of BM were purified and characterized as diastereomeric pairs of N7-(2-hydroxy-3-buten-1-yl)guanosine (G-1 and G-3), N7-(1-hydroxy3-buten-2-yl)guanosine (G-2 and G-5), N2-(1-hydroxy-3-buten-2-yl)guanosine (G-4 and G-7), and N1-(1-hydroxy-3-buten-2-yl)guanosine (G-6 and G-8) on the basis of stability studies and analyses by UV, 1H NMR, and fast atom bombardment mass spectrometry. While the N7adducts exhibited half-lives of approximately 50 (G-1 and G-3) and 90 h (G-2 and G-5) upon incubation for 192 h in 100 mM phosphate buffer (pH 7.4) at 37 °C, the N1- and N2-adducts remained stable. When guanosine was reacted with excess BM in phosphate buffer (pH 7.4) at 37 °C, adduct formation exhibited pseudo-first-order kinetics, with the N7-adducts being formed approximately 10-fold more favorably than the N1- and N2-adducts. When incubations were carried out at lower BM concentrations, the N7-adducts remained the major detectable adducts, but the N2-adducts were also detectable at equimolar BM and guanosine concentrations, and the N1-adducts were detectable at a 5-fold molar excess of BM. These results, which provide clear evidence that guanosine can be alkylated at multiple sites following 1,3-butadiene exposure, may aid in the development of useful biomarkers for exposure to 1,3-butadiene. The results may also contribute to a better understanding of the molecular mechanisms of 1,3butadiene-induced carcinogenicity.

Introduction Exposure of rats and mice to 1,3-butadiene, a chemical used extensively in the industrial production of synthetic rubber and plastics and detected in automobile exhaust and cigarette smoke, resulted in the formation of tumors at multiple sites (1, 2). Mice were considerably more susceptible to 1,3-butadiene-induced carcinogenicity than rats; however, the biochemical basis for this species difference has not been determined. The International Agency for Research on Cancer has classified 1,3-butadiene as a “possible” human carcinogen (3), but a recent epidemiological study suggested a definitive increase in the incidence of various lymphohematopoietic cancers in industrial workers exposed to 1,3-butadiene (4). In addition, an increase in hprt mutant frequency in the peripheral lymphocytes of nonsmoking workers exposed to butadiene (3.5 ( 7.5 ppm) over controls was reported (5). While these results clearly show that current levels of occupational exposure to 1,3-butadiene may not be adequate to protect workers from the mutagenic/carcinogenic effects of 1,3-butadiene, the molecular mechanisms of 1,3-butadiene-induced carcinogenicity remain unclear. Butadiene monoxide (BM,1 3,4-epoxy-1-butene) is a major metabolite of 1,3-butadiene both in vivo and in vitro by microsomal monooxygenases in both mice and

rats (6-8) and with multiple human cytochrome P450 enzymes and human myeloperoxidase (9, 10). Both enantiomers of BM were formed in rat liver microsomal incubations (11). BM, a mutagen (12) and weak carcinogen in mouse skin-painting studies (13), was postulated to be responsible for the mutagenicity of 1,3-butadiene in the S9 activated Salmonella typhimurium mutagenicity assay (14). Mutants isolated in mice exposed to 1,3butadiene or BM were found to have both transversion and transition base pair substitutions, as well as both +1 and -1 frame shift mutations (15). In addition, activated K-ras genes were reported in a large number of tumors induced by 1,3-butadiene exposure (16). Citti et al. studied the reaction of racemic BM with guanosine in vitro, characterizing two products after cleavage of the sugar moieties of the adducts (17). These adducts were identified as regioisomeric N7-(2-hydroxy3-buten-1-yl)guanine and N7-(1-hydroxy-3-buten-2-yl)guanine on the basis of UV and MS data. The NMR spectra of the isolated products were, however, not conclusive with regard to the regiochemistry of the two products (17). Furthermore, because products were only characterized after cleavage of the sugar moiety, the chemical stability of the modified nucleosides was not characterized. In addition, because diastereomeric products were not resolved, the relative chemical reactivities of the two BM enantiomers were not determined.

* Corresponding author: Dr. Adnan A. Elfarra, Department of Comparative Biosciences, University of Wisconsin School of Veterinary Medicine, 2015 Linden Dr. W., Madison, WI 53706; telephone, (608) 262-6518; fax, (608) 263-3926; internet, [email protected]. X Abstract published in Advance ACS Abstracts, December 1, 1995.

1 Abbreviations: BM, butadiene monoxide; ACN, acetonitrile; FAB/ MS, fast atom bombardment mass spectrometry; G-1 and G-3, diastereomeric N7-(2-hydroxy-3-buten-1-yl)guanosine; G-2 and G-5, diastereomeric N7-(1-hydroxy-3-buten-2-yl)guanosine; G-4 and G-7, diastereomeric N2-(1-hydroxy-3-buten-2-yl)guanosine; G-6 and G-8, diastereomeric N1-(1-hydroxy-3-buten-2-yl)guanosine.

0893-228x/96/2709-0126$12.00/0

© 1996 American Chemical Society

Butadiene Monoxide-Guanosine Adducts

In this study, eight BM-guanosine adducts were synthesized, purified, and characterized as diastereomeric pairs of N7-(2-hydroxy-3-buten-1-yl)guanosine (G-1 and G-3), N7-(1-hydroxy-3-buten-2-yl)guanosine (G-2 and G-5), N2-(1-hydroxy-3-buten-2-yl)guanosine (G-4 and G-7), and N1-(1-hydroxy-3-buten-2-yl)guanosine (G-6 and G-8). The stability of these adducts in phosphate buffer (pH 7.4) at 37 °C was characterized. In addition, the effect of BM concentration on adduct formation and the pseudofirst-order kinetic reaction rates were determined in phosphate buffer (pH 7.4) at 37 °C.

Experimental Procedures Materials. Racemic BM, trifluoroacetic acid, and deuterium oxide were obtained from Aldrich Chemical Co. (Milwaukee, WI). Guanosine was obtained from Sigma Chemical Co. (St. Louis, MO). HPLC grade acetonitrile (ACN) was obtained from EM Science (Gibbstown, NJ). NMR supplies were obtained from Wilmad Glass Co. (Buena, NJ). All other chemicals were of the highest grade commercially available. Caution: BM is a known mutagen and carcinogen in laboratory animals and must be handled using proper safety measures. Synthesis of BM-Guanosine Adducts. Syntheses of N7adducts were carried out as described by Citti et al. (17), with some modifications. Guanosine (56.6 mg, 0.2 mmol) was dissolved in 3.5 mL of glacial acetic acid and heated at 60 °C, in a 12 mL vial with a screw-top Teflon-coated cap, until the guanosine dissolved. A molar excess of BM (0.161 mL, 2 mmol) was added, and the solution was incubated in a Dubnoff shaking water bath at 50 °C for 5 h. At 5 h the reaction mixture was diluted with an equal volume of acetone and four volumes of ethyl ether. The precipitate, separated by centrifugation and dissolved in 20 mM phosphate buffer (pH 5.5), was analyzed by HPLC. Four peaks, in addition to the starting material, were resolved. During the time course of BM-guanosine adduct formation under physiological conditions (see the following), four additional peaks were also found to increase with time and became the sole products as pH was increased from 7.4 to 10.0. These N1and N2-adducts were synthesized by dissolving guanosine (30 mg, 0.1 mmol) in 6 mL of 100 mM potassium phosphate buffer, adjusted to pH 10 with KOH and containing 100 mM potassium chloride with heat (60 °C), and adding an excess of BM (0.3 mL, 3.7 mmol). The reaction was incubated for 5 h at 37 °C. At 5 h the reaction mixture was extracted twice with four volumes of ethyl ether to remove excess BM. The product was lyophilized to dryness, dissolved in water, and analyzed by HPLC. Four peaks, in addition to guanosine, were resolved. HPLC Purification of BM-Guanosine Adducts. HPLC separations of the crude reaction mixtures were performed with a 20 µL injection volume on a Beckman Ultrasphere 5 µm ODS reverse-phase analytical column (250 × 4.6 mm i.d.), using a Beckman gradient-controlled HPLC system (Irvine, CA) equipped with a Beckman diode array detector (Model 168) and dualchannel UV detection at 260 and 280 nm. Use of a linear gradient program starting at 10 min from 0% to 100% pump B over 8 min [pump A, 1% ACN (pH 2.5); pump B, 10% ACN (pH 2.5)], at a rate of 1 mL/min, resolved the four major peaks from each reaction mix, whose retention times on a typical chromatogram were 22.26, 24.38, 24.77, and 26.53 min for the N7-adducts and 25.53, 27.99, 28.54, and 29.28 min for the N1- and N2adducts (Figure 1). These peaks were assigned as G-1 through G-8 in the order of their retention times, respectively. The purified alkylated guanosines (see the following) were used to prepare standard curves. The limit of detection attained by standard curves (0.1-1000 µg/mL) by using this HPLC method was 0.25 µg/mL (r values were greater than 0.999) for all adducts. Bulk separation of crude material was accomplished with a 0.5 mL injection volume on a Beckman Ultrasphere 5 µm ODS reverse-phase semipreparative column (250 × 10 mm i.d.) with the mobile phases described earlier and a linear gradient starting at 10 min and going from 0% to 75% for pump B over

Chem. Res. Toxicol., Vol. 9, No. 1, 1996 127 15 min at a flow rate of 3 mL/min. Adducts were collected from the two separate reaction mixes by using a Gilson 203 fraction collector (Middleton, WI). Peak solutions were fractionated twice to obtain greater than 95% purity, as determined by HPLC, and lyophilized to dryness. Identification of BM-Guanosine Adducts. UV spectra of each peak were obtained with a Beckman diode array detector at pH 2.5 in the mobile phase described earlier. Positive ion fast atom bombardment mass spectra (FAB/MS) were obtained by using a Kratos MS-50 ultrahigh-resolution mass spectrometer fitted with a Kratos DS-55 data system (Manchester, United Kingdom). An Ion Tech FAB gun utilizing xenon as the FAB gas was used with a direct insertion FAB probe. Approximately 50-100 µg of each adduct was analyzed with a spectrum ranging from m/z 50 to 400, using a 3-nitrobenzyl alcohol or glycerol matrix. All spectra are matrix subtracted. Proton NMR spectra of all eight adducts, dissolved in D2O at concentrations ranging from 3 to 15 mg/mL, were obtained on a Bruker spectrometer (Karlsruhe, Germany) at 400 or 500 MHz. N1 and N2-adduct spectra were performed at 5 °C in order to shift the water peak downfield, away from several sugar peaks. Decoupling experiments were performed on G-7 and G-8 to confirm proton assignments. All four N1- and N2-adduct NMR spectra were also obtained in Me2SO-d6 under the same conditions. Chemical shifts are reported in ppm with the H2O peaks as an internal standard. BM-Guanosine Adduct Formation at Physiological Conditions. The reactions of BM and guanosine were carried out in 4 mL vials capped with Teflon septa. Guanosine (10 mM) and BM (10, 25, 50, and 500 mM, added to the vial via microsyringe injection through the septum) were reacted in 3 mL of 100 mM phosphate reaction buffer (pH 7.4) containing 100 mM potassium chloride in a shaking water bath at 37 °C. Reactions were stopped at 60 min by four consecutive extractions with two volumes of ethyl ether each to remove unreacted BM. Samples were then analyzed for the presence of adducts by HPLC, as described earlier. For the pseudo-first-order kinetic experiment, guanosine (3.1 or 9.6 mM) and BM (750 mM) were reacted as described earlier. Lower concentrations of BM (188 mM) did not produce a linear response, possibly due to extensive voltilization and decomposition of BM (8). At the 750 mM BM concentration used in the kinetic study, a large excess of BM should still exist at the end of the reaction, on the basis of the expected loss of BM due to volatilization and decomposition. Samples (0.25 mL), withdrawn at timed intervals by syringe to minimize the loss of BM, were extracted three times with four volumes of ethyl ether (1 mL) to remove unreacted BM before analysis by HPLC was carried out as described earlier. Controls were performed to determine whether the ether extraction had an effect on adduct recovery. Purified adduct was dissolved 1 mg/mL in 100 mM phosphate reaction buffer and diluted with buffer to concentrations ranging from 5 to 1000 µg/mL. Six replicate samples were prepared at each concentration: three were extracted with ether as described earlier and the other three remained unextracted. Samples were analyzed by HPLC to determine adduct recovery. All samples showed greater than 90% recovery. Guanosine Adduct Stability Studies. All eight purified adducts were dissolved individually (100 ppm) in 100 mM phosphate buffer (pH 7.4) containing 100 mM potassium chloride. These solutions were placed in a shaking water bath at 37 °C; samples were withdrawn at 24 h intervals and analyzed by HPLC, under the conditions described earlier, to determine the disappearance of the parent adduct peaks. In other experiments, 3 mL of synthetic N7-adducts (2 mM) were either boiled for 30 min or incubated at 80 °C with 0.6 mL of 0.1, 1, or 2 N HCl. Aliquots of the acid-incubated solutions were removed at 5, 90, or 300 min, neutralized with NaOH, and analyzed by HPLC, using the conditions described earlier, for both the disappearance of the parent guanosine adduct peaks and the formation of the corresponding guanine adduct peaks (G-1 or G-3 hydrolysis resulted in the detection of a new peak at 18.97 min, and G-2 or G-5 hydrolysis resulted in the detection

128 Chem. Res. Toxicol., Vol. 9, No. 1, 1996

Figure 1. HPLC chromatogram of the eight products formed by reaction of racemic BM with guanosine. Peaks G-1 and G-3 were identified as diastereomers of N7-(2-hydroxy-3-buten-1-yl)guanosine, G-2 and G-5 as diastereomers of N7-(1-hydroxy-3buten-2-yl)guanosine, G-4 and G-7 as diastereomers of N2-(1hydroxy-3-buten-2-yl)guanosine, and G-6 and G-8 as diastereomers of N1-(1-hydroxy-3-buten-2-yl)guanosine. of a new peak at 20.76 min). The guanine adduct peaks exhibited UV spectra similar to those reported by Citti et al. (17). Synthetic N1- and N2-adducts were dissolved in 1 M KOH and incubated at 95 °C for 2 h. Aliquots were withdrawn every 30 min, and the pH was adjusted to 2.5 with HCl and immediately analyzed by HPLC for the disappearance of the parent adduct peaks.

Results Reaction of racemic BM with guanosine under physiological conditions yields eight products, G-1, G-2, G-3, G-4, G-5, G-6, G-7, and G-8 (Figure 1), which were separated by HPLC. With absorbance maxima of 260 nm for G-1 and G-3, 258/260 nm for G-2 and G-5, and 258 nm for G-4, G-6, G-7, and G-8, the UV maxima and spectral shapes (data not shown) suggested that all eight products were nitrogen-substituted guanosine adducts (18). Identification of the isomers was achieved by examination of their proton NMR and FAB/MS spectra. Proton NMR spectra of the eight products show the presence of both the guanosine and BM moieties. The assignment of protons (Table 1, Figures 2 and 3) was achieved by comparing the chemical shifts, multiplicities, and integration ratios of the protons in the products with the published spectra of guanosine (19) and the spectra of 3-butene-1,2-diol, as well as by several decoupling experiments (data not shown). The differentiating assignments between regioisomers of the N7-adducts (Figure 3) were based on the expected greater downfield chemical shifts of the hydrogens adjacent to the positively charged N7 of guanosine compared to the hydrogens adjacent to the hydroxyl group. The proton NMR spectra of G-1 and G-3 appeared nearly identical, as did the spectra of G-2 and G-5. One representative spectrum for each regioisomer is given in Figure 2, while the chemical shifts and J values for all compounds are given in Table 1. The relative downfield shifts of the two protons, assigned to H9, in G-1 (4.36, 4.50 ppm) and G-3 (4.29, 4.55 ppm) led to the conclusion that the attached carbon was adjacent to the N7 of guanosine, identifying these products as the diastereomers of N7-(2-hydroxy-3-buten-1-yl)guanosine (Figure 3). The chemical shifts of the H9 protons of G-2 (3.92, 3.99 ppm) and G-5 (3.96, 4.01 ppm) were smaller than those of the H9 protons of G-1 and G-3, suggesting that this

Selzer and Elfarra

carbon was adjacent to the hydroxyl group. These two adducts also contain a single proton doublet of triplets, H3, shifted downfield (5.43 and 5.48 ppm, respectively), suggesting that these products have a single proton on the carbon attached to the N7 of guanosine and thus were identified as the diastereomers of N7-(1-hydroxy-3-buten2-yl)guanosine (Figure 3). Further evidence for these structural assignments was provided by the relative downfield shifts of the assigned H1 protons of G-2 (6.02 ppm) and G-5 (6.05 ppm) compared to those of G-1 (5.82 ppm) and G-3 (5.81 ppm) as the H1 protons of G-2 and G-5 are in closer proximity to the deshielding N7 of guanosine. Assignment of protons from the sugar moieties was confirmed by comparison to the decoupling experiments of G-7 and G-8 to follow. The absence of a signal for the C8 proton in all N7-adduct spectra is consistent with the previous observation that the C8 protons of N7-alkylguanosines rapidly exchange with the deuterium of the solvent (20). G-4 and G-7 were determined to be diastereomeric pairs of a single regioisomer by nearly identified proton NMR spectra (Figure 2C, Table 1). The adducts were determined to be N2-substituted guanosine adducts on the basis of the presence of only one of the two N2 protons at 6.6 ppm when the NMR was performed in Me2SO-d6 (data not shown). The alkylation with BM was determined to have taken place at the carbon adjacent to the double bond of the BM moiety, as evidenced by the splitting of the remaining N2 proton into a doublet at 6.6 ppm when the spectra of G-4 and G-7 were obtained in Me2SO-d6 (data not shown). This demonstrates coupling to only a single proton on the adjacent carbon, and thus G-4 and G-7 were identified as the diastereomers of N2(1-hydroxy-3-buten-2-yl)guanosine. Further evidence for the proton assignments in D2O was obtained by homonuclear decoupling of the G-7 sample. Saturation of H1 at 5.87 ppm was used to identify H3 at 4.62 ppm. Saturation of H3, in turn, identified the H9 protons at 3.67 and 3.75 ppm. Saturation of H8 (4.13 ppm) of the sugar moiety helped to determine the signals of the H10 protons, while other sugar protons were identified by comparison with decoupling experiments performed on G-8. Further evidence for the identity of G-4 and G-7 as N2adducts was their stability under alkaline conditions. If alkylation occurred at the N1, N3, O6, or N7 positions of guanosine, the adducts would be degradable by incubation in 1 M KOH (100 °C, 2 h); however, adducts attached at the exocyclic amino groups of purines are stable under these conditions (21). In this study, the stability of adducts G-4 and G-7 for 2 h under these conditions (data not shown) provides further evidence for adducts with substitutions at the exocyclic amino group. As expected, adducts G-6 and G-8 were found to be unstable under these conditions and were fully degraded after 30 min. G-6 and G-8 were identified as N1-substituted guanosine adducts on the basis of NMR and UV data. Again, diastereomeric pairs of a single regioisomer are evident due to the near identity of the two spectra (Figure 2D, Table 1). Proton assignments were confirmed by several homonuclear decoupling experiments on G-8. Saturation of H3 of G-8 was used to identify H9 and confirm H1. Saturation of various signals originating from the sugar moiety were also confirmatory. Saturation at 3.85 ppm (H10) affected the signals at 4.24 ppm (H8), saturation at 4.76 ppm (H6) affected signals at 4.42 (H7) and 5.91 ppm (H2), and saturation at 4.42 ppm (H7) affected signals at 4.24 (H8) and 4.76 ppm (H6).

Butadiene Monoxide-Guanosine Adducts

Chem. Res. Toxicol., Vol. 9, No. 1, 1996 129

Table 1. 500 MHz 1H NMR Data for Guanosine Adducts of BM in D2O (J Values Reported in Hertz) H1 a G-1 5.82 J ) 16.85, 10.87, 5.80 G-3 5.81 J ) 16.91, 10.80, 5.89 G-2 6.02 J ) 17.19, 10.56, 6.50 G-5 6.05 J ) 17.24, 10.61, 6.4 G-4 5.88 J ) 16.0, 11.0, 5.0 G-7 5.87 J ) 17.2, 11.0, 6.0 G-6 5.91 J ) 17.3, 10.8, 6.3 G-8 5.97 J ) 17.2, 10.4, 6.4

H2

H3

5.95 4.50 J ) 3.57 NAb

H4

H5

H6

5.19 5.26 4.55 J ) 10.68 J ) 17.27 J ) 4.30, 4.12 5.18 5.22 4.55 J ) 10.58 J ) 17.35 NA

H7

4.26 J ) 5.65, 5.52 5.93 4.48 4.24 J ) 2.22 NA J ) 5.00, 5.49 5.94 5.43 5.40 5.32 4.53 4.25 J ) 2.18 J ) 5.09, NA J ) 10.54 J ) 17.29 J ) 3.53, J ) 5.63, 3.21 5.56 5.98 5.48 5.44 5.35 4.56 4.31 J ) 2.78 J ) 5.68, J ) 10.60 J ) 17.20 J ) 3.65, J ) 5.91, 4.95 3.38 5.61 5.91 4.66 5.23 5.28 4.84 4.40 J ) 4.5 J ) 3.5, NA J ) 10.5 J ) 17.5 J ) 5.0, 4.5 J ) 5.0, 5.0

H8 4.16 J ) 5.43, 5.39 4.13 NA 4.13 J ) 3.01, 2.83 4.17 J ) 5.66, 2.75 4.13 J ) 3.5, NA

5.93 J ) 5.0

4.62 J ) 5.0, NA

5.23 J ) 10.5

5.28 J ) 17.5

4.81 4.39 4.13 J ) 5.5, 5.0 J ) 5.0, 5.0 J ) 4.5, 3.5

5.88 J ) 5.5

4.52 J ) 6.0, 6.0

5.20 J ) 10.5

5.30 J ) 17.0

4.69 4.36 4.19 J ) 5.5, 5.5 J ) 4.5, 4.5 J ) 3.5, 3.0

5.91 J ) 6.0

4.58 J ) 6.4, 6.0

5.26 J ) 10.4

5.35 J ) 17.2

4.76 4.42 4.24 J ) 5.6, 5.6 J ) 4.8, 4.0 J ) 3.6, 3.6

H9

H10

4.36/4.50c 3.73/3.86c J ) 6.69/NAd J ) 12.98, 2.84 3.71/3.84c 4.29/4.55c J ) 13.8, J ) 12.91, 7.5/3.1d 3.21 3.92/3.99c 3.72/3.87c J ) 12.19, J ) 13.12, 4.19/6.25d 5.67/2.05d 3.96/4.01c 3.76/3.89c J ) 12.35, J ) 13.04, 4.28/6.26d 3.14/2.45d 3.67/3.77c 3.73/3.83c J ) 12.8, J ) 11.6, 4.8/NAd 6.3/4.3d 3.67/3.75c 3.76/3.85c J ) 12.5, J ) 11.5, 5.0/3.0d 6.0/4.5d 4.07/4.14c 3.78/3.85c J ) 12.9, J ) 14.7, 3.8/2.5d 8.0/3.3d 4.13/4.21c 3.82/3.89c J ) 12.8, J ) 15.2, 4.0/2.8d 8.4/4.4d

H0

8.13

8.12

8.21

8.00

a Proton numbering refers to assignments given in Figure 3. Multiplicities can be found on the spectra in Figure 2. b NA: J values not available from spectra. c Two chemical shifts are given corresponding to the two protons of the methylene group. d Two J values are given corresponding to the two protons of the methylene group.

The loss of the N1 proton present at 10.5 ppm in the spectra in Me2SO-d6 suggested alkylation at the N1 position (data not shown). In addition, UV spectra match well with other alkylated N1-guanosine adducts, but not with O6-guanosine adducts (18). When G-6 and G-8 are incubated with 0.1 N HCl at 80 °C, the resulting hydrolysis products do not have the UV spectral shape and λmax characteristic of N3-alkylated guanines (λmax ) 263 nm), but rather match those of other N1-alkylated guanines (λmax ) 252 nm). While the alkylation site on the BM moiety cannot be ascertained from the NMR spectra, the structures of G-6 and G-8 are tentatively assigned as the diastereomers of N1-(1-hydroxy-3-buten2-yl)guanosine on the basis of the following criteria. First, the N2-regioisomers formed under the same conditions reacted at the carbon adjacent to the double bond of the BM moiety as evidenced by the NMR results. Second, styrene oxide, an analogue of BM, was reported to react at the N1 and N2 positions of guanosine to yield adducts only at the carbon adjacent to the benzene ring (22-24). Mass spectra of all adducts (G-1 to G-8) provided further evidence for their identity. Mass spectra for each of the N7-adducts did not yield the expected molecular ion (m/z 354 for M, as the N7-adduct already carries a positive charge); however, a fragment at m/z 222, consistent with the loss of the sugar moiety (Figure 4), was seen with all four N7-adducted compounds. This is a common fragmentation pattern of N7-guanosine adducts due to the quaternary nitrogen at N7, which allows the adduct to fragment into an ion at BH+ (where B is the base) rather than the BH2+ fragment normally observed with DNA adducts. G-2 and G-5, as opposed to G-1 and G-3, exhibited a prominent guanine fragment (m/z 152; Figure 4C,D), possibly because of facile cleavage of the linkage between the allylic carbon of the BM moiety of G-2 and G-5 and the N7 of the guanine moiety. The G-2 diastereomer shows an m/z 192 fragment corresponding to the loss of a CH2dO moiety from the molecule. The fragments at m/z 110 and 124, which were observed with G-1, G-2, and G-3, correspond to the pyrimidine and the pyrimidine plus one nitrogen atom N2-adduct

from the adjacent imidazole ring, respectively. The G-5 diastereomer exhibited an m/z 136 fragment corresponding to the loss of the exocyclic amino group from the guanine fragment (m/z 152). G-4, G-6, G-7, and G-8 adducts did yield the expected molecular ion (m/z 354 for M + 1) in the mass spectra (Figure 5). The m/z 222 fragment corresponding to the loss of the sugar plus one proton appeared in all four spectra. The N2-isomers contained an m/z 185 peak not evident in the N1 spectra, which may correspond to the loss of two molecules of water, in addition to the loss of the sugar moiety. All four adducts again showed the m/z 152 peak representing the guanine moiety. Peak G-8 also contained m/z 337, which constitutes the loss of a hydroxyl group from the M + 1 peak, and m/z 376, corresponding to the M + Na peak. When guanosine (10 mM) was reacted with varying amounts of BM (10, 25, 50, and 500 mM) for 1 h in 100 mM phosphate buffer (pH 7.4) at 37 °C, the N7- and N2adducts were detected at all BM concentrations (Figure 6). However, the levels of the N7-adducts detected were significantly higher than those of the N2-adducts. The N1-adducts were detected only at 50 (5-fold molar excess) and 500 (50-fold molar excess) mM BM. Concentrations of BM lower than 10 mM did not lead to the detection of any of the N1- and N2-guanosine adducts by the HPLC method, as the amounts of adducts formed were below the limits of detection. At an unlimiting BM concentration (750 mM), product formation exhibited linear time dependency for all adducts, from 0 to 90 min, whether reactions were carried out at 3.1 or 9.6 mM guanosine (Figure 7). Indeed, the observed linear increase in rate with increasing guanosine concentration provided evidence for the reaction being carried out under pseudofirst-order kinetics. The calculated rate constants for the formation of G-1, G-2, G-3, G-4, G-5, G-6, G-7, and G-8 were 2.63 × 10-2, 2.61 × 10-2, 2.48 × 10-2, 3.26 × 10-3, 2.54 × 10-2, 3.23 × 10-3, 4.73 × 10-3, and 3.53 × 10-3 h-1, respectively. The stability of all of the guanosine adducts was investigated by determining the percent loss of the parent adduct after incubation in phosphate buffer adjusted to

130 Chem. Res. Toxicol., Vol. 9, No. 1, 1996

Selzer and Elfarra

Figure 3. Chemical structures of deuterated (A) N7-(1-hydroxy3-buten-2-yl)guanosine (G-2 and G-5), (B) N7-(2-hydroxy-3buten-1-yl)guanosine (G-1 and G-3), (C) N1-(1-hydroxy-3-buten2-yl)guanosine (G-6 and G-8), and (D) N2-(1-hydroxy-3-buten2-yl)guanosine (G-4 and G-7). Proton numbering is based on chemical shifts of the first spectrum solved (G-5).

Figure 2. 500 MHz proton NMR spectra for (A) N7-(2-hydroxy3-buten-1-yl)guanosine (G-1), (B) N7-(1-hydroxy-3-buten-2-yl)guanosine (G-5), (C) N2-(1-hydroxy-3-buten-2-yl)guanosine (G7), and (D) N1-(1-hydroxy-3-buten-2-yl)guanosine (G-6). Peak numbers correspond to assignments in Figure 3.

pH 7.4 at 37 °C. The N1- and N2-adducts appeared to be completely stable for the 192 h of incubation, while N7adducts had half-lives of approximately 50 h for G-1 and G-3 and 90 h for G-2 and G-5 (Figure 8). Hydrolysis of N7-adducts to the guanine adduct can be enhanced by high temperatures and acidic conditions. For example, boiling of the N7-guanosine adducts for 30 min at pH 7.4 resulted in partial loss (65%) of the sugar moiety. When the guanosine adducts were dissolved with 0.1 N HCl and samples were heated for 5 min at 80 °C, virtually no guanine adduct was produced. However, the guanosine adducts were fully converted to the guanine adducts by incubation with 1 N HCl at 80 °C for 90 min.

Discussion

Figure 4. FAB/MS of the diastereomers of N7-(2-hydroxy-3buten-1-yl)guanosine (G-1 and G-3) and N7-(1-hydroxy-3-buten2-yl)guanosine (G-2 and G-5). The peak at m/z 222 corresponds to the peak expected from the sugar-cleaved guanine adduct: (A) G-1; (B) G-3; (C) G-2; (D) G-5.

The results presented in this article demonstrate that racemic BM reacted with guanosine to yield diastereomeric pairs of N7-(2-hydroxy-3-buten-1-yl)guanosine (G-1 and G-3), N7-(1-hydroxy-3-buten-2-yl)guanosine (G-2 and

G-5), N2-(1-hydroxy-3-buten-2-yl)guanosine (G-4 and G-7), and N1-(1-hydroxy-3-buten-2-yl)guanosine (G-6 and G-8). The N7-adducts exhibited half-lives ranging from 48 to 96 h at 37 °C in phosphate buffer (pH 7.4), which is

Butadiene Monoxide-Guanosine Adducts

Figure 5. FAB/MS of the diastereomers of N2-(1-hydroxy-3buten-2-yl)guanosine (G-4 and G-7) and N1-(1-hydroxy-3-buten2-yl)guanosine (G-6 and G-8). The peak at m/z 354 corresponds to the expected M + 1 peak: (A) G-4; (B) G-7; (C) G-6; (D) G-8.

Chem. Res. Toxicol., Vol. 9, No. 1, 1996 131

Figure 7. Rates of BM-guanosine adduct formation at pH 7.4 at 37 °C. To guarantee pseudo-first-order kinetic conditions, the BM concentration (750 mM) was much higher than the guanosine concentrations (A, 3.1 mM; B, 9.6 mM). Values represent the results obtained from several experiments.

Figure 8. Stability of BM-guanosine adducts in 100 mM phosphate buffer containing 100 mM KCl (pH 7.4) at 37 °C.

Figure 6. Effect of BM concentration on the formation of eight BM-guanosine adducts. BM was reacted with 10 mM guanosine for 60 min at 37 °C in phosphate buffer (pH 7.4). Values represent the means ( SD of the results obtained from three experiments.

similar to the 50 h half-life reported by Citti et al. for the spontaneous depurination of these guanine adducts in DNA (17). The N1- and N2-adducts remained stable under these conditions, as is typical for alkylation products formed at these positions. When adduct formation was monitored over a range of BM concentrations, N7-adducts were detected at equimolar BM and guanosine concentrations. The N2adducts were also detected under these conditions, but the N1-guanosine adducts were not detected until BM

concentrations were 5-fold greater than that of guanosine. The roughly linear formation of adducts over the BM concentration range tested and the consistent ratios of formation of the individual adducts suggest that there is no threshold BM concentration for N1-, N2-, and N7adduct formation near the range tested. The inability to detect the N1-adducts at BM concentrations lower than 50 mM is a consequence of the detection limit of the assay. Adduct formation in the presence of excess BM exhibited pseudo-first-order kinetics, but different rates were observed for the formation of N7- vs N1- and N2-adducts. Similar to the results obtained at lower BM concentrations (Figure 6), the N7-adduct rates of formation were approximately 10-fold higher than those of the N1- or N2adducts (Figure 7). These results provide evidence for multiple reaction sites for BM with guanosine and

132 Chem. Res. Toxicol., Vol. 9, No. 1, 1996

indicate similar chemical reactivity of the two BM enantiomers. While the N7-adducts were the major adducts formed when guanosine was reacted with BM, the formation of N1- and N2-adducts may also be of toxicological significance. The N1 and N2 positions of guanosine/2′-deoxyguanosine are involved in the hydrogen bonding of nucleic acids. In addition, our results showed that these adducts are far more stable than the N7-adducts, in the absence of enzymatic repair. For many chemicals, studies have demonstrated the rapid repair of N7- and N2-adducts in human cells (25, 26). However, much less has been reported about the repair of N1-guanosine adducts. Thus, it is possible that the formation of N2- and N1- BMguanosine adducts could be at least partially responsible for the mutagenicity/carcinogenicity of 1,3-butadiene. Detection of N7-guanosine-BM adducts by 32P-postlabeling methods has been unsuccessful (27). Thus, the identification of N1- and N2-guanosine adducts may also be important for the use of these sensitive biomonitoring techniques. In conclusion, this article presents conclusive evidence for guanosine alkylation at multiple sites by BM. While the N7-adducts were formed preferably to the N1- and N2adducts, the N1- and N2-adducts were much more stable than the N7-adducts. Identification of these adducts may contribute to a better understanding of the molecular mechanisms of 1,3-butadiene-induced carcinogenicity.

Acknowledgment. The authors thank Dr. Laura Lerner for helpful comments. This research was supported by NIH Grant ES06841 from the National Institute of Environmental Health Sciences. R.R.S. was supported by a National Science Foundation Graduate Research Fellowship. NMR spectra were determined at the National Magnetic Resonance Facility at Madison, WI, which is supported in part by NIH Grant RR02301.

References (1) Melnick, R. L., Huff, J., Chou, B. J., and Miller, R. A. (1990) Carcinogenicity of 1,3-butadiene in C57BL/6 × CH3 F1 mice at low exposure concentrations. Cancer Res. 50, 6592-6599. (2) Owen, P. E., Glaister, J. R., Gaunt, I. F., and Pullinger, 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. (3) International Agency for Research on Cancer (1987) IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Supplement 7, World Health Organization, Lyon. (4) Santos-Burgoa, C., Matanoski, G. M., Zeger, S., and Schwartz, L. (1992) Lymphohematopoietic cancer in styrene-butadiene polymerization workers. Am. J. Epidemiol. 136, 843-854. (5) Ward, J. B., Ammenheuser, M. M., Bechtold, W. E., Whorton, E. B., and Legator, M. S. (1994) hprt mutant lymphocyte frequencies in workers at a 1,3-butadiene production plant. Environ. Health Perspect. 102 (Suppl. 9), 79-85. (6) Bolt, H. M., Schmiedel, G., Filser, J., Rolzhauser, H., Lieser, K., Wistuba, D., and Schurig, V. (1983) Biological activation of 1,3butadiene to vinyl oxirane by rat liver microsomes and expiration of the reactive metabolite by exposed rats. J. Cancer Res. Clin. Oncol. 106, 112-116. (7) Elfarra, A. A., Duescher, R. J., and Pasch, C. M. (1991) Mechanisms of 1,3-butadiene oxidations to butadiene monoxide and crotonaldehyde by mouse liver microsomes and chloroperoxidase. Arch. Biochem. Biophys. 286, 244-251. (8) Sharer, J. E., Duescher, R. J., and Elfarra, A. A. (1992) Species and tissue differences in the microsomal oxidation of 1,3-butadiene and the glutathione conjugation of butadiene monoxide in mice and rats. Possible role in 1,3-butadiene-induced toxicity. Drug Metab. Dispos. 20, 658-664. (9) Duescher, R. J., and Elfarra, A. A. (1992) 1,3-Butadiene oxidation by human myeloperoxidase. Role of chloride ion in catalysis of divergent pathways. J. Biol. Chem. 267, 19859-19865. (10) Duescher, R. J., and Elfarra, A. A. (1994) Human liver microsomes are efficient catalysts of 1,3-butadiene oxidation: evidence for

Selzer and Elfarra major roles by cytochromes P450 2A6 and 2E1. Arch. Biochem. Biophys. 311, 342-349. (11) Schurig, V., and Wistuba, D. (1984) Asymmetric microsomal epoxidation of simple prochiral olefins. Angew. Chem., Int. Ed. Engl. 23, 796-797. (12) Gervasi, P. G., Citti, L., Del Monte, M., Longo, V., and Benetti, D. (1985) Mutagenicity and chemical reactivity of epoxidic intermediates of isoprene metabolism and other structurally related compounds. Mutat. Res. 156, 77-82. (13) Van Duuren, P. L., Nelson, N., Orris, L., Palmes, E. D., and Schmitt, F. L. (1963) Carcinogenicity of epoxides, lactones, and peroxy compounds. J. Natl. Cancer Inst. 31, 41-55. (14) de Meester, C., Poncelet, F., Roberfroid, M., and Mercier, M. (1978) Mutagenicity of butadiene and butadiene monoxide. Biochem. Biophys. Res. Commun. 80, 298-305. (15) Cochrane, J. E., and Skopek, T. R. (1994) Mutagenicity of butadiene and its epoxide metabolites: II. Mutational spectra of butadiene, 1,2-epoxybutane and diepoxybutane at the hprt locus in splenic T cells from exposed B6C3F1 mice. Carcinogenesis 15, 719-723. (16) Goodrow, T., Reynolds, S., Maronpot, R., and Anderson, M. (1990) Activation of K-ras by codon 13 mutations in C57BL/6 × C3H F1 mouse tumors induced by exposure to 1,3-butadiene. Cancer Res. 50, 4818-4823. (17) Citti, L., Gervasi, P. G., Turchi, 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. (18) Singer, B., and Grunberger, D. (1983) Molecular Biology of Mutagens and Carcinogens, Plenum Press, New York. (19) Sadtler Research Laboratories Inc. (1978) Standard Proton NMR Collection, Spectra 27772M, Philadelphia, PA. (20) Tomasz, M. (1970) Extreme lability of the C-8 proton: a consequence of 7-methylation of guanine residues in model compounds and in DNA and its analytical application. Biochim. Biophys. Acta 199, 18-28. (21) Moore, P. D., Koreeda, M., Wislocki, P. G., Levin, W., Conney, A. H., Yagi, H., and Jerina, D. M. (1977) In vitro reactions of the diastereomeric 9,10-epoxides of (+)- and (-)-trans-7,8-dihydroxy7,8-dihydrobenzo[a]pyrene with polyguanylic acid and evidence for formation of an enantiomer of each diastereoisomeric 9,10epoxide from benzo[a]pyrene on mouse skin. In Drug Metabolism Concepts, ACS Symposium Series No. 44 (Jerina, D. M., ed.) pp 127-154, American Chemical Society, Washington, DC. (22) Kaur, S., Pongracz, K., Bodell, W. J., and Burlingame, A. L. (1993) Bis(hydroxyphenylethyl)deoxyguanosine adducts identified by [32P]-postlabeling and four-sector tandem mass spectrometry: unanticipated adducts formed upon treatment of DNA with styrene 7,8-oxide. Chem. Res. Toxicol. 6, 125-132. (23) Latif, F., Moschel, R. C., Hemminki, K., and Dipple, A. (1988) Styrene oxide as a stereochemical probe form the mechanism of aralkylation at different sites on guanosine. Chem. Res. Toxicol. 1, 364-369. (24) Pongracz, K., Kaur, S., Burlingame, A. L., and Bodell, W. J. (1992) Identification of N2-substituted 2′-deoxyguanosine-3′-phosphate adducts detected by 32P-postlabeling of styrene-oxide-treated DNA. Carcinogenesis 13, 315-319. (25) Stierum, R. H., van-Herwijnen, M. H., Maas, L. M., Hageman, G. J., and Kleinjans, J. C. (1994) Measurement by 32P-postlabeling of (+/-)anti-benzo[a]pyrene-diolepoxide-N2-deoxyguanosine adduct persistence in unstimulated human peripheral blood lymphocytes. Mutat. Res. 325, 31-37. (26) Mah, M. C., Boldt, J., Culp, S. J., Maher, V. M., and McCormick, J. J. (1991) Replication of acetylaminofluorene-adducted plasmids in human cells. Spectrum of base substitutions and evidence of excision repair. Proc. Natl. Acad. Sci. U.S.A. 88, 10193-10197. (27) Leuratti, C., Marafante, E., Jones, N. J., and Waters, R. (1993). Detection by HPLC-32P-postlabelling of DNA adducts formed after exposure to epoxybutene and diepoxybutane. IARC Sci. Pub. 124, 277-282.

TX950101O