Effect of Substitution Site upon the Oxidation Potentials of Alkylanilines

Yukari Totsuka, Takeji Takamura-Enya, Nobuo Kawahara, Rena Nishigaki, Takashi ... Takeji Enya, Masanobu Kawanishi, Hitomi Suzuki, Saburo Matsui, and ...
0 downloads 0 Views 205KB Size
Download PDF
1266

Chem. Res. Toxicol. 1997, 10, 1266-1274

Effect of Substitution Site upon the Oxidation Potentials of Alkylanilines, the Mutagenicities of N-Hydroxyalkylanilines, and the Conformations of Alkylaniline-DNA Adducts M. Matilde Marques,*,† Luı´sa L. G. Mourato,† M. Teresa Amorim,† M. Ame´lia Santos,† William B. Melchior, Jr.,‡ and Frederick A. Beland*,‡ Centro de Quı´mica Estrutural, Complexo I, Instituto Superior Te´ cnico, Av. Rovisco Pais, 1096 Lisboa Codex, Portugal, and Division of Biochemical Toxicology, National Center for Toxicological Research, Jefferson, Arkansas 72079 Received June 17, 1997X

Carcinogenic arylamines typically undergo metabolic activation via N-hydroxylation followed in most instances by O-esterification. In this study, the ability of methyl-, dimethyl-, and ethylaniline constituents of tobacco smoke to undergo oxidation at the nitrogen atom was analyzed. In addition, the mutagenicity of the corresponding N-hydroxyalkylanilines and the conformational properties of the DNA adducts generated from their N-acyloxy derivatives were investigated. All the arylamines underwent irreversible electrochemical N-oxidation at potentials higher than those observed for the oxidation of carcinogenic polynuclear aromatic amines. There were minor differences in the oxidation potentials, which were consistent with the position and electron-donating abilities of the alkyl substituents; however, the differences appeared to be too small to account for the range of genotoxic effects among the alkylanilines. N-Hydroxyarylamines containing p-alkyl substituents had increased mutagenicities in Salmonella typhimurium TA100, which was attributed to their higher efficiencies of adduct formation. Increased mutagenicities were also observed upon o-alkyl substitution; however, this property was not related to a greater ability of the ortho-substituted species to form DNA adducts, which suggested that adducts from ortho-substituted alkylanilines may be intrinsically more mutagenic than their meta- and para-substituted analogues. In all instances, N-(acyloxy)arylamines generated from the N-hydroxyarylamines reacted with dG, dG nucleotides, and DNA to yield C8-substituted dG derivatives as the major adducts. The alkylaniline-dG adducts displayed distinct conformational trends that were determined by the location of the alkyl substituents. Spectroscopic data indicated higher percentages of low-energy syn conformers in the adducts that contained alkyl groups ortho to the arylamine nitrogen as opposed to adducts not bearing ortho substituents. The data strongly suggest that the conformational properties of the DNA adducts, in particular their ability to adopt syn conformations, may be determinant factors for the genotoxic responses elicited by certain alkylanilines (e.g., 2-methylaniline and 2,6-dimethylaniline).

Introduction Aromatic amines and amides are chemical carcinogens to which human exposure from diverse origins is welldocumented (1-4). Cigarette smoke is a significant source of primary aromatic amines, including the human urinary bladder carcinogens 2-naphthylamine and 4-aminobiphenyl. Since no other class of compounds identified in cigarette smoke has been implicated in human bladder carcinogenesis (5), aromatic amines are thought to be responsible for the causal relationship observed between cigarette consumption and the onset of bladder cancer (5-7). Aromatic amines have also been suggested to play a role in the induction of breast cancer in postmenopausal women who smoke (8). The major contributors to the arylamine fraction of cigarette smoke are single-ring arylamines such as aniline and its methyl, dimethyl, and ethyl derivatives (2, 3). These compounds have recently been detected in † ‡ X

Instituto Superior Te´cnico. National Center for Toxicological Research. Abstract published in Advance ACS Abstracts, November 1, 1997.

S0893-228x(97)00104-5 CCC: $14.00

human milk from women smokers (9). In contrast to their binuclear analogues, mononuclear aromatic amines are generally considered to be weak carcinogens (1012); however, the risks of chronic exposure to some aniline derivatives may be underestimated. 2-Methylaniline (o-toluidine), for example, has long been recognized as a rodent carcinogen (13) and may account for the increased risk of bladder cancer observed in rubber industry workers (14). 2,6-Dimethylaniline (2,6-xylidine), which exists in cigarette smoke and also as a major metabolite of the potent anesthetic and antiarrhythmic drug lidocaine (15), is carcinogenic in rats (16). The potential carcinogenicity of this arylamine to humans remains uncertain (17). Aromatic amine carcinogens are known to exert their biological effects upon metabolic activation to reactive electrophilic intermediates that bind to DNA, yielding N-(deoxyguanosin-8-yl)arylamines (dG-C8-Ar)1 as major persistent adducts (18-20). In most instances, the initial step in the activation involves cytochrome P450-mediated N-hydroxylation to N-hydroxyarylamines, which may be further activated via enzymatic O-esterification. Previous studies (21-25) have shown that synthetic analogues © 1997 American Chemical Society

Alkylamine Oxidation, Mutagenicity, and DNA Adducts

of plausible electrophilic metabolites [i.e., N-(acyloxy)arylamines] derived from a series of single-ring aromatic amines react in vitro with 2′-deoxyguanosine (dG), dG nucleotides, or DNA to give dG-C8-Ar adducts as the predominant products. In addition, a large number of mononuclear arylamines, including the alkylanilines present in tobacco smoke, have been shown to form hemoglobin adducts in rodents and humans (26-28), which implies that N-hydroxylation of these arylamines must occur in vivo (29). Interestingly, while the consistent detection of hemoglobin-alkylaniline adducts in human blood samples is an indicator of widespread exposure to the parent anilines, some of these adducts (e.g., from 2- and 4-methylaniline, 2,4-dimethylaniline, and 2-ethylaniline) have been found at significantly higher levels in smokers than in nonsmokers (26, 27). Similarly, substantial increases in 2,6-dimethylanilinehemoglobin adducts have been reported in patients treated with therapeutic doses of lidocaine (30). Taken together, these observations suggest that mononuclear arylamines, like their carcinogenic binuclear analogues, undergo bioactivation to species that have the potential to bind to DNA in vivo. Considerable efforts have been directed at elucidating the structural factors responsible for the genotoxicity of arylamines. In a comprehensive study, Sabbioni (31) investigated possible correlations between the experimental half-wave oxidation potentials (E1/2) of several single-ring arylamines and their mutagenic and carcinogenic properties. Similar analyses were conducted using electronic descriptors of the arylamines and their corresponding nitrenium ions, derived by semiempirical methods. With both approaches, mutagenicity and carcinogenicity increased with the oxidizability of the mononuclear arylamines, which was consistent with N-oxidation being a requisite for the bioactivation of this class of compounds. Nonetheless, oxidizability per se was insufficient for predicting biological effects (31-33), thus implying that other structural factors, subsequent events, or both must be determinants for the genotoxicity of each particular arylamine. We recently reported (21) the synthesis and conformational analysis of dG-C8-Ar adducts from a number of methylated single-ring aromatic amines (2-, 3-, and 4-methylaniline, 2,3- and 2,4-dimethylaniline) present in tobacco smoke. The results indicated that adducts containing a methyl substituent ortho to the arylamine nitrogen had a substantial proportion of low-energy conformers with the dG oriented syn about the glycosyl bond. Since ortho-methylated binuclear arylamines tend to be more mutagenic (34, 35) and carcinogenic (13, 36) than their analogues not bearing an o-methyl substituent, we suggested that the syn conformer may contribute to the biological activity of arylamines containing omethyl substituents. We have now extended our study to the entire series of methyl-, dimethyl-, and ethylanilines. Specifically, we have examined the effects of the alkyl substitution site upon the oxidation potentials 1 Abbreviations: Ar, arylamine; Bis-Tris, bis(2-hydroxyethyl)iminotris(hydroxymethyl)methane; diMeA, dimethylaniline; dG, 2′-deoxyguanosine; dG3′p, 2′-deoxyguanosine 3′-phosphate; dG5′p, 2′-deoxyguanosine 5′-phosphate; dG3′,5′p, 2′-deoxyguanosine 3′,5′-bisphosphate; dG-C8-Ar, N-(deoxyguanosin-8-yl)arylamine; DEPT, distortionless enhancement by polarization transfer; DMF, N,N-dimethylformamide; dR, 2′-deoxyribosyl; dR5′p, 2′-deoxyribosyl 5′-phosphate; EI, electron impact; EtA, ethylaniline; FAB, fast atom bombardment; MeA, methylaniline; NOE, nuclear Overhauser effect; SCE, sodium-saturated calomel electrode.

Chem. Res. Toxicol., Vol. 10, No. 11, 1997 1267

of the alkylanilines, the mutagenicities of the N-hydroxyarylamines, and the conformations of the DNA adducts formed from their reactive N-(acyloxy)arylamine derivatives.

Materials and Methods Instrumentation. Melting temperatures were measured with a Leica Galen III hot-stage apparatus and are uncorrected. Reversed-phase HPLC analyses and separations were conducted using a µBondapak C18 column (0.39 × 30 cm; Waters Associates, Milford, MA) on a Waters Associates system consisting of two model 510 pumps, a U6K injector, and a model 660 automated gradient controller. Peaks were monitored at 280 nm with a Hewlett-Packard 1050 diode array spectrophotometric detector. Infrared spectra were recorded on a Perkin-Elmer 683 spectrophotometer. Solid samples were dispersed in KBr pellets, and oils were inserted between two KCl disks. UV spectra were recorded with either a Beckman DU-40 or a Shimadzu 1202 UV/ vis spectrophotometer. Mass spectra were obtained on a VG Trio-2000 or a Finnigan TSQ 70 instrument. Electron impact (EI) spectra were generated at 70 eV. For mass spectra recorded in the fast atom bombardment (FAB) mode, the samples were dispersed in a matrix of either thioglycerol or 3-nitrobenzyl alcohol; source temperatures were 90 or 120 °C. 1H NMR spectra were recorded on a Varian Unity 300 or a Bruker AM500 spectrometer operating at 300 and 500 MHz, respectively. 13C NMR spectra were recorded on a Varian Unity 300 spectrometer operating at 75.43 MHz. N-Hydroxyarylamines were dissolved in either Me2CO-d6 or CDCl3, using tetramethylsilane as an internal standard. Adduct samples were dissolved in Me2SO-d6 (dG adducts) or Me2SO-d6/D2O [2′deoxyguanosine 5′-phosphate (dG5′p) adducts]. The 1H and 13C chemical shifts for each adduct were referenced to the Me2SO proton (2.49 ppm) and carbon (39.50 ppm) signals, respectively. 1H NMR assignments for the adducts were based on comparison with literature data for dG and related dG adducts (21, 24, 37, 38), combined with homonuclear decoupling experiments, chemical exchange of the labile protons with D2O, and observation of the nuclear Overhauser effect (NOE) enhancement patterns. 13C NMR assignments for the adducts were based on published data for polysubstituted aniline derivatives (39), dG (40), and related dG adducts (21, 41), as well as on the use of a distortionless enhancement by polarization transfer (DEPT) pulse sequence and analysis of the 13C-1H coupling patterns. Cyclic voltammetry was conducted in a PAR 173 instrument, with a three-electrode cell containing a platinum wire working electrode, a platinum counter electrode, and a sodium-saturated calomel reference electrode (SCE). Chemicals. dG was obtained from US Biochemicals (Cleveland, OH). 2′-Deoxyguanosine 3′-phosphate (dG3′p), dG5′p, 2′deoxyguanosine 3′,5′-bisphosphate (dG3′,5′p), salmon testes DNA, bis(2-hydroxyethyl)iminotris(hydroxymethyl)methane (BisTris), and the enzymes used in DNA hydrolyses and dephosphorylation reactions of arylamine-nucleotide adducts were purchased from Sigma Chemical Co. (St. Louis, MO). All other commercially available reagents were obtained from E. Merck (Darmstadt, Germany), Fluka Chemie AG (Buchs, Switzerland), Aldrich Chemical Co. (Milwaukee, WI), or Sigma-Aldrich Quı´mica, S.A. (Madrid, Spain) and were used as received. Whenever necessary, solvents were purified by standard methods (42). Caution: Nitroarenes, N-arylamines, N-hydroxyarylamines, and N-(acyloxy)arylamines are potentially carcinogenic. They should be handled with protective clothing, in a well-ventilated fumehood. Syntheses. Starting Materials. 1. Nitroarenes. 2,5Dimethylnitrobenzene [bp 238-240 °C, lit. (43) bp 239.5-241 °C] was prepared by nitration of p-xylene as described in Bowen et al. (44). 3-Ethylnitrobenzene [bp 244-245 °C, lit. (45) bp 242-243 °C/752 mmHg] was synthesized in 56% yield by thermal decomposition of 4-ethyl-2-nitrobenzenediazonium sul-

1268 Chem. Res. Toxicol., Vol. 10, No. 11, 1997 fate, which was prepared in situ from 4-ethyl-2-nitroaniline following the procedure described for the synthesis of 3-nitrotoluene (46). 4-Ethyl-2-nitroaniline [mp 46-48 °C, lit. (47) mp 47-48.5 °C) was obtained by a sequence of steps that involved quantitative acetylation of 4-ethylaniline with acetic anhydride to yield 4-ethylacetanilide [mp 95-97 °C, lit. (47) mp 96-97.5 °C] followed by nitration to 4-ethyl-2-nitroacetanilide [mp 4546 °C, lit. (48) mp 45-47 °C] as described by Case et al. (48). Hydrolysis of the latter in 70% sulfuric acid afforded the aniline in quantitative yield. All the remaining nitroarenes are commercially available. 2. N-Hydroxyarylamines. The N-hydroxyarylamines were synthesized in 50-60% yields by reduction of the corresponding nitroarenes (10 mmol) at 50 °C with zinc dust (20-30 mmol) in aqueous ammonium chloride containing 30% ethanol (49, 50). The following products were obtained: N-hydroxy-2,5-dimethylaniline, mp 90-92 °C [lit. (50) mp 91.5 °C]; N-hydroxy-2,6dimethylaniline, mp 97-99 °C [lit. (50) mp 98.5 °C]; N-hydroxy3,4-dimethylaniline, mp 101-103 °C [lit. (50) mp 101 °C). N-Hydroxy-3,5-dimethylaniline: mp 94-96 °C; IR (KBr) νmax 3300-3060 (b, NH + OH) cm-1; 1H NMR (CDCl3) δ 2.20 (6H, s, 3-CH3 + 5-CH3), 6.29 (2H, bs, NH + OH), 6.55 (3H, bs, H2 + H4 + H6); 13C NMR (CDCl3) δ 21.38 (3-CH3 + 5-CH3), 112.71 (C2 + C6), 124.35 (C4), 138.74 (C3 + C5), 149.50 (C1). Anal.: C, 69.72; H, 8.11; N, 10.07 (C8H11NO requires C, 70.04; H, 8.08; N, 10.21). N-Hydroxy-2-ethylaniline: mp 185 °C dec; UV (EtOH) λmax 278 nm (log  4.36); 1H NMR (Me2SO-d6) δ 2.02 (1H, m, H2′′), 2.10 (3H, s, 2-CH3), 2.22 (3H, s, 5-CH3), 2.68 (1H, m, H2′), 3.65 (2H, m, H5′,5′′), 3.87 (1H, m, H4′), 4.36 (1H, m, H3′), 5.30 (1H, bs, 3′OH), 5.48 (1H, bs, 5′-OH), 6.24 (1H, m, H1′), 6.29 (2H, s, NH2),

Marques et al. 6.79 (1H, d, J ) 7.2 Hz, ArH4), 7.04 (1H, d, J ) 7.2 Hz, ArH3), 7.09 (1H, s, ArH6), 7.85 (1H, s, ArNH), 10.46 (1H, s, N1H); 13C NMR (Me2SO-d6) δ 17.46 (2-CH3), 20.69 (5-CH3), 37.98 (C2′), 61.23 (C5′), 71.04 (C3′), 82.86 (C1′), 87.18 (C4′), 112.41 (C5), 123.77 (ArC4/6), 124.02 (ArC4/6), 127.36 (ArC2), 130.18 (ArC3), 134.98 (ArC5), 138.95 (C8), 145.23 (ArC1), 149.83 (C4), 152.49 (C2), 155.60 (C6); MS m/z (FAB) 409 [(M + Na)+, 10], 387 (MH+, 40), 293 [(MH + Na - dR)+, 6], 271 [(MH2 - dR)+, 100]. N-(Deoxyguanosin-8-yl)-2,6-dimethylaniline (dG-C8-2,6diMeA): UV (H2O/MeOH, 1/1) λmax 271 nm; 1H NMR (Me2SOd6) δ 2.02 (1H, m, H2′′), 2.13 (6H, s, 2-CH3 + 6-CH3), 2.71 (1H, m, H2′), 3.67 (2H, m, H5′,5′′), 3.91 (1H, m, H4′), 4.40 (1H, m, H3′), 5.38 (2H, bs, 3′-OH + 5′-OH), 6.32 (1H, m, H1′), 6.33 (2H, s, NH2), 7.04 (3H, m, ArH3 + ArH4 + ArH5), 7.84 (1H, s, ArNH), 10.56 (1H, s, N1H). N-(Deoxyguanosin-8-yl)-3,4-dimethylaniline (dG-C8-3,4diMeA): η 3%; mp >205 °C dec; UV (EtOH) λmax 282 nm (log  4.21); 1H NMR (Me2SO-d6) δ 1.98 (1H, m, H2′′), 2.14 (3H, s, 4-CH3), 2.17 (3H, s, 3-CH3), ∼2.50 (1H, m, H2′, partially obscured by the solvent resonance), 3.74 (2H, m, H5′,5′′), 3.90 (1H, m, H4′), 4.40 (1H, m, H3′), 5.34 (1H, bs, 3′-OH), 5.90 (1H, bs, 5′-OH), 6.30 (1H, m, H1′), 6.35 (2H, s, NH2), 6.99 (1H, d, J ) 8.4 Hz, ArH5), 7.40 (1H, s, ArH2), 7.52 (1H, d, J ) 8.4 Hz, ArH6), 8.45 (1H, s, ArNH), 10.54 (1H, s, N1H); 13C NMR (Me2SO-d6) δ 18.64 (4-CH3), 19.70 (3-CH3), 38.28 (C2′), 61.24 (C5′), 71.23 (C3′), 82.71 (C1′), 87.09 (C4′), 112.13 (C5), 114.92 (ArC6), 118.64 (ArC2), 128.11 (ArC4), 129.39 (ArC5), 135.97 (ArC3), 138.59 (C8), 143.59 (ArC1), 149.47 (C4), 152.71 (C2), 155.64 (C6); MS m/z (FAB) 387 (MH+, 43), 271 [(MH2 - dR)+, 100]. N-(Deoxyguanosin-8-yl)-3,5-dimethylaniline (dG-C8-3,5diMeA): η 3%; mp >295 °C dec; UV (EtOH) λmax 284 nm (log  4.35); 1H NMR (Me2SO-d6) δ 1.98 (1H, m, H2′′), 2.21 (6H, s, 3-CH3 + 5-CH3), ∼2.50 (1H, m, H2′, partially obscured by the solvent resonance), 3.73 (2H, m, H5′,5′′), 3.89 (1H, m, H4′), 4.40 (1H, m, H3′), 5.33 (1H, bs, 3′-OH), 5.90 (1H, bs, 5′-OH), 6.30 (1H, m, H1′), 6.35 (2H, s, NH2), 6.52 (1H, s, ArH4), 7.31 (2H, s, ArH2 + ArH6), 8.44 (1H, s, ArNH), 10.58 (1H, s, N1H); 13C NMR (Me2SO-d6) δ 21.29 (3-CH3 + 5-CH3), 38.27 (C2′), 61.23 (C5′), 71.21 (C3′), 82.68 (C1′), 87.09 (C4′), 112.19 (C5), 115.06 (ArC2+6), 122.34 (ArC4), 137.41 (ArC3+5), 140.71 (C8), 143.33 (ArC1), 149.50 (C4), 152.80 (C2), 155.71 (C6); MS m/z (FAB) 387 (MH+, 94), 293 [(MH + Na - dR)+, 9], 271 [(MH2 - dR)+, 100]. N-(Deoxyguanosin-8-yl)-2-ethylaniline (dG-C8-2-EtA): η 2%; mp >171 °C dec; UV (EtOH) λmax 279 nm (log  4.16); 1H NMR (Me2SO-d6) δ 1.12 (3H, t, J ) 7.5 Hz, CH3), 2.04 (1H, m, H2′′), 2.56 (2H, m, CH2), 2.69 (1H, m, H2′), 3.65 (2H, m, H5′,5′′), 3.87 (1H, m, H4′), 4.38 (1H, m, H3′), 5.26 (1H, bs, 3′-OH), 5.46 (1H, bs, 5′-OH), 6.25 (1H, m, H1′), 6.27 (2H, s, NH2), 7.04 (1H, dt, Jo ) 7.6 Hz, Jm ) 1.2 Hz, ArH4), 7.12 (1H, dt, Jo ) 7.6 Hz, Jm ) 1.4 Hz, ArH5), 7.20 (1H, dd, Jo ) 7.6 Hz, Jm ) 1.4 Hz, ArH3), 7.27 (1H, dd, Jo ) 7.8 Hz, Jm ) 1.2 Hz, ArH6), 7.87 (1H, s, ArNH), 10.52 (1H, s, N1H); 13C NMR (Me2SO-d6) δ 13.82 (CH3), 23.64 (CH2), 38.07 (C2′), 61.26 (C5′), 71.10 (C3′), 82.92 (C1′), 87.20 (C4′), 112.39 (C5), 123.88 (ArC3/4/5/6), 124.26 (ArC3/4/5/6), 125.98 (ArC3/4/5/6), 128.27 (ArC3/4/5/6), 136.72 (ArC2/C8), 138.51 (C8/ArC2), 145.73 (ArC1), 149.81 (C4), 152.49 (C2), 155.57 (C6); MS m/z (FAB) 409 [(M + Na)+, 13], 387 (MH+, 38), 293 [(MH + Na - dR)+, 13], 271 [(MH2 - dR)+, 100]. N-(Deoxyguanosin-8-yl)-3-ethylaniline (dG-C8-3-EtA): η 1%; mp >190 °C dec; UV (EtOH) λmax 283 nm (log  4.16); 1H NMR (Me2SO-d6) δ 1.17 (3H, t, J ) 7.8 Hz, CH3), 1.98 (1H, m, H2′′), ∼2.50 (1H, m, H2′, partially obscured by the solvent resonance), 2.52 (2H, q, J ) 7.8 Hz, CH2), 3.74 (1H, m, H5′,5′′), 3.90 (1H, m, H4′), 4.38 (1H, m, H3′), 5.33 (1H, bs, 3′-OH), 5.89 (1H, bs, 5′-OH), 6.28 (1H, m, H1′), 6.33 (2H, s, NH2), 6.74 (1H, d, J ) 7.5 Hz, ArH4), 7.14 (1H, t, J ) 7.5 Hz, ArH5), 7.47 (1H, s, ArH2), 7.59 (1H, d, J ) 7.5 Hz, ArH6), 8.53 (1H, s, ArNH), 10.54 (1H, s, N1H); 13C NMR (Me2SO-d6) δ 15.72 (CH3), 28.42 (CH2), 38.35 (C2′), 61.24 (C5′), 71.19 (C3′), 82.73 (C1′), 87.13 (C4′), 112.13 (C5), 114.83 (ArC6), 116.73 (ArC2), 120.08 (ArC4), 128.38 (ArC5), 140.75 (C8/ArC3), 143.35 (ArC3/C8), 144.06 (ArC1), 149.43 (C4), 152.78 (C2), 155.62 (C6); MS m/z (FAB) 387 (MH+, 19), 293 [(MH + Na - dR)+, 26], 271 [(MH2 - dR)+, 100].

Alkylamine Oxidation, Mutagenicity, and DNA Adducts N-(Deoxyguanosin-8-yl)-4-ethylaniline (dG-C8-4-EtA): η 2%; mp >190 °C dec; UV (EtOH) λmax 281 nm (log  4.12); 1H NMR (Me2SO-d6) δ 1.14 (3H, t, J ) 7.5 Hz, CH3), 1.98 (1H, m, H2′′), ∼2.50 (1H, m, H2′, partially obscured by the solvent resonance), 2.52 (2H, q, J ) 7.5 Hz, CH2), 3.73 (1H, m, H5′,5′′), 3.90 (1H, m, H4′), 4.40 (1H, m, H3′), 5.33 (1H, bs, 3′-OH), 5.89 (1H, bs, 5′-OH), 6.30 (1H, m, H1′), 6.36 (2H, s, NH2), 7.08 (2H, d, J ) 8.3 Hz, ArH3 + ArH5), 7.61 (2H, d, J ) 8.3 Hz, ArH2 + ArH6), 8.54 (1H, s, ArNH), 10.55 (1H, s, N1H); 13C NMR (Me2SO-d6) δ 15.94 (CH3), 27.48 (CH2), 38.33 (C2′), 61.24 (C5′), 71.23 (C3′), 82.75 (C1′), 87.11 (C4′), 112.10 (C5), 117.52 (ArC2/6), 127.67 (ArC3/5), 135.91 (ArC4), 138.49 (C8), 143.55 (ArC1), 149.41 (C4), 152.73 (C2), 155.60 (C6); MS m/z (FAB) 409 [(M + Na)+, 4], 387 (MH+, 70), 271 [(MH2 - dR)+, 100]. 4. N-(Deoxyguanosin-8-yl)arylamine n-Phosphates (n ) 3′p, 5′p, or 3′,5′p). Each nucleotide (dG3′p, dG5′p, or dG3′,5′p) was reacted with each N-(acyloxy)arylamine, and the resulting adduct was isolated as detailed previously for other phosphate and bisphosphate arylamine-dG adducts (21). Briefly, dG5′p adducts were purified using Waters Sep-Pak C18 reversedphase cartridges with a 0-50% step gradient of acetonitrile in 100 mM ammonium acetate (pH 5.7). dG3′p and dG3′,5′p adducts were isolated by reversed-phase HPLC using a 30-min linear gradient of 5-60% acetonitrile in either 100 mM ammonium acetate, pH 5.7 (dG3′p adducts), or 100 mM ammonium acetate, 10 mM ammonium phosphate, pH 5.7 (dG3′,5′p adducts). dG5′p adducts were characterized by 1H NMR and mass spectral analyses as outlined below. dG3′p and dG3′,5′p adducts were identified through comparison of their UV spectra with the spectra of the corresponding dG and dG5′p adducts. In addition, all nucleotide adducts were treated with alkaline phosphatase for 3 h at 37 °C, and the HPLC retention times and UV spectra of the resultant nucleoside adducts were compared to those of the corresponding dG adducts. N-(Deoxyguanosin-8-yl)-2,5-dimethylaniline 5′-phosphate (dG5′p-C8-2,5-diMeA): 1H NMR (Me2SO-d6/D2O) δ 2.05 (1H, m, H2′′), 2.10 (3H, s, 2-CH3), 2.19 (3H, s, 5-CH3), 2.86 (1H, m, H2′), 3.71 (1H, m, H5′′), 3.81 (1H, m, H4′), 4.21 (1H, m, H5′), 4.64 (1H, m, H3′), 6.06 (1H, m, H1′), 6.69 (1H, dd, Jo ) 7.7 Hz, Jm ) 1.3 Hz, ArH4), 6.99 (1H, d, J ) 7.7 Hz, ArH3), 7.02 (1H, s, ArH6); MS m/z (FAB) 489 (M + Na)+, 467 (MH+), 271 (MH2 - dR5′p)+. N-(Deoxyguanosin-8-yl)-3,4-dimethylaniline 5′-phosphate (dG5′p-C8-3,4-diMeA). 1H NMR (Me2SO-d6/D2O) δ 2.02 (1H, m, H2′′), 2.12 (3H, s, 4-CH3), 2.16 (3H, s, 3-CH3), 2.79 (1H, m, H2′), 3.79 (1H, m, H5′′), 3.84 (1H, m, H4′), 4.18 (1H, m, H5′), 4.66 (1H, m, H3′), 6.18 (1H, m, H1′), 6.96 (1H, d, J ) 8.3 Hz, ArH5), 7.35 (1H, d, J ) 2.3 Hz, ArH2), 7.39 (1H, dd, Jo ) 8.1 Hz, Jm ) 2.3 Hz, ArH6); MS m/z (FAB) 489 (M + Na)+, 467 (MH+), 271 (MH2 - dR5′p)+. N-(Deoxyguanosin-8-yl)-3,5-dimethylaniline 5′-phosphate (dG5′p-C8-3,5-diMeA): 1H NMR (Me2SO-d6/D2O) δ 2.03 (1H, m, H2′′), 2.20 (6H, s, 3-CH3 + 5-CH3), 2.77 (1H, m, H2′), 3.78 (1H, m, H5′′), 3.82 (1H, m, H4′), 4.20 (1H, m, H5′), 4.67 (1H, m, H3′), 6.16 (1H, m, H1′), 6.48 (1H, s, ArH4), 7.18 (2H, s, ArH2 + ArH6); MS m/z (FAB) 489 (M + Na)+, 467 (MH+), 271 (MH2 - dR5′p)+. N-(Deoxyguanosin-8-yl)-2-ethylaniline 5′-phosphate (dG5′p-C8-2-EtA): 1H NMR (Me2SO-d6/D2O) δ 1.14 (3H, t, J ) 7.6 Hz, CH3), 2.04 (1H, m, H2′′), 2.55 (2H, m, CH2), 2.93 (1H, m, H2′), 3.75 (1H, m, H5′′), 3.86 (1H, m, H4′), 4.17 (1H, m, H5′), 4.55 (1H, m, H3′), 6.10 (1H, m, H1′), 6.92 (1H, dt, Jo ) 7.5 Hz, Jm ) 1.1 Hz, ArH4), 7.08 (1H, dt, Jo ) 7.5 Hz, Jm ) 1.3 Hz, ArH5), 7.14 (1H, dd, Jo ) 7.5 Hz, Jm ) 1.3 Hz, ArH3), 7.25 (1H, d, J ) 7.5 Hz, ArH6); MS m/z (FAB) 467 (MH+), 271 (MH2 dR5′p)+. N-(Deoxyguanosin-8-yl)-3-ethylaniline 5′-phosphate (dG5′p-C8-3-EtA): 1H NMR (Me2SO-d6/D2O) δ1.15 (3H, t, J ) 7.7 Hz, CH3), 2.04 (1H, m, H2′′), 2.54 (2H, q, J ) 7.7 Hz, CH2), 2.82 (1H, m, H2′), 3.78 (1H, m, H5′′), 3.84 (1H, m, H4′), 4.20 (1H, m, H5′), 4.67 (1H, m, H3′), 6.18 (1H, m, H1′), 6.70 (1H, d, J ) 7.6 Hz, ArH4), 7.11 (1H, t, J ) 7.8 Hz, ArH5), 7.39 (1H, s,

Chem. Res. Toxicol., Vol. 10, No. 11, 1997 1269 ArH2), 7.43 (1H, dd, Jo ) 7.8 Hz, Jm ) 1.9 Hz, ArH6); MS m/z (FAB) 467 (MH+), 271 (MH2 - dR5′p)+. N-(Deoxyguanosin-8-yl)-4-ethylaniline 5′-phosphate (dG5′p-C8-4-EtA): 1H NMR (Me2SO-d6/D2O) δ 1.14 (3H, t, J ) 7.6 Hz, CH3), 2.03 (1H, m, H2′′), 2.51 (2H, q, J ) 7.6 Hz, CH2), 2.79 (1H, m, H2′), 3.81 (1H, m, H5′′), 3.85 (1H, m, H4′), 4.17 (1H, m, H5′), 4.65 (1H, m, H3′), 6.19 (1H, m, H1′), 7.05 (2H, d, J ) 8.5 Hz, ArH3 + ArH5), 7.56 (1H, d, J ) 8.5 Hz, ArH2 + ArH6); MS m/z (FAB) 489 (M + Na)+, 467 (MH+), 271 (MH2 dR5′p)+. DNA Modification Reactions. Chemical modification of DNA with each arylamine was conducted by reacting the corresponding N-acetoxyarylamine with salmon testes DNA as described previously (21). Following standard enzymatic hydrolysis of the DNA to nucleosides (51), the HPLC retention times and UV spectra of arylamine-DNA adducts were compared to those of the corresponding dG adducts. Electrochemical Studies. All the arylamines used in the electrochemical studies were obtained commercially. To prevent oxidation upon storage, each amine was converted to the corresponding hydrochloride salt by reaction with gaseous HCl. The free base was subsequently released in the electrochemical cell by addition of an excess of anhydrous pyridine. A 100 mM solution of tetraethylammonium perchlorate in anhydrous DMF was used as supporting electrolyte. The tetraethylammonium perchlorate was obtained by reacting tetraethylammonium bromide with sodium perchlorate and purified by recrystallization from 95% ethanol. All solutions were degassed with nitrogen. The measurements were conducted at 25.0 ( 0.1 °C using a voltage scan rate of 50 mV/s. Salmonella Mutagenicity Assays. Mutagenicity assays were conducted in the absence of exogenous metabolic activation (i.e., S9) using Salmonella typhimurium TA100 as described by Maron and Ames (52). Triplicate plates were used at each dose, with doses of 0, 10, 30, 60, and 100 µg/plate. Theoretical Simulation Studies. Molecular orbital calculations for the alkylated anilines and their oxidation products were performed using the MOPAC-PM3 program (53) included in the Insight II (version 2.2.0)/Discover (version 2.9) software package from Biosym Technologies Inc. (San Diego, CA). Mulliken atomic charges, which were calculated by the program through simple subtraction of the atomic electron densities from the nuclear charge, were used. Energy minimization calculations for the dG-C8-Ar adducts were conducted with the same software package as described previously (21), using the AMBER force field together with MOPAC charges and distancedependent dielectric constants to simulate the solvent.

Results and Discussion Electrochemical Studies on the Alkylanilines. Any attempt to rationalize the relative contributions of arylamine oxidation and adduct formation to the potential genotoxicity of specific alkylanilines must take into consideration the structural factors ruling the oxidizability of arylamines. Half-wave oxidation potentials (E1/2) for several monosubstituted and disubstituted anilines have been reported in aqueous systems (31, 54). In these studies, ring substitution by methyl and ethyl groups was found to produce cathodic shifts of E1/2 compared to aniline. Except for some o-substituted anilines, these shifts were generally consistent with a simple interpretation based on the electron-releasing capabilities of substituents in different ring positions, as measured by the Hammett σ parameters (54). Not unexpectedly, the halfwave oxidation potentials for methyl-, dimethyl-, and ethylanilines were also shown to correlate inversely with the stability of the corresponding nitrenium ions, calculated by semiempirical methods (31). To gain further insight into the mechanisms involved in the oxidation process, representative methyl-, dimeth-

1270 Chem. Res. Toxicol., Vol. 10, No. 11, 1997

Marques et al. Table 2. Mutagenicity of N-Hydroxyarylamines (N-OH-Ar) in S. typhimurium TA100 Ar

revertants/nmola

aniline 2-methylaniline 3-methylaniline 4-methylaniline 2,3-dimethylaniline 2,4-dimethylaniline 2,5-dimethylaniline 2,6-dimethylaniline 3,4-dimethylaniline 3,5-dimethylaniline 2-ethylaniline 3-ethylaniline 4-ethylaniline

0.25 ( 0.02b 0.49 ( 0.08 0.37 ( 0.02 1.22 ( 0.11 2.28 ( 0.43 5.92 ( 0.99 1.37 ( 0.16 11.99 ( 0.97 2.55 ( 0.15 0.75 ( 0.08 1.02 ( 0.17 0.51 ( 0.02 1.11 ( 0.10

a Mutagenicity was assessed in the absence of S9 using triplicate plates at each dose, with doses of 10, 30, 60, and 100 µg/plate. The data are from an increasing portion of the dose-response curve (i.e., 30-µg dose). b Data are presented as the mean ( standard deviation. Background values were at 79 ( 9 revertants/ plate.

Figure 1. Cyclic voltammograms for aniline and representative alkylanilines, recorded under basic conditions (excess pyridine) in anhydrous DMF, using 100 mM tetraethylammonium perchlorate as supporting electrolyte: (a) 3,5-dimethylaniline; (b) 3,4-dimethylaniline; (c) 4-methylaniline; (d) aniline. Table 1. Oxidation Potentials for Aniline and Representative Alkylanilines basic mediuma alkylaniline aniline 2-methylaniline 3-methylaniline 4-methylaniline 2,3-dimethylaniline 2,4-dimethylaniline 2,5-dimethylaniline 2,6-dimethylaniline 3,4-dimethylaniline 3,5-dimethylaniline 2-ethylaniline 3-ethylaniline 4-ethylaniline

EI

EII

∆EId

acidic mediumb aqueousc EI

EII ∆EId

0.97 1.23 0.00 1.06 1.30 0.00 0.86 1.20 0.11 f 1.30 0.91 1.19 0.06 1.03 1.29 0.03 0.91 0.90 0.83 0.92 0.95 0.94 0.86

1.20 1.18 1.22 1.20 1.18 1.17 1.20

0.06 0.07 0.14 0.05 0.02 0.03 0.11

1.01 0.96 f 1.05 1.04 1.05 0.92

1.30 1.30 1.31 1.30 1.30 1.29 1.26

0.05 0.10 0.01 0.02 0.01 0.14

∆E1/2e 0.00 0.01 0.00 0.09 0.12 0.06 0.08 0.12 0.03 0.04 0.02 0.07

a Obtained in 100 mM tetraethylammonium chloride in dry DMF containing excess pyridine. Values are referenced versus SCE. b Obtained in 100 mM tetraethylammonium chloride in dry DMF containing excess perchloric acid. Values are referenced versus SCE. c Values are from Sabbioni (31) and were obtained by HPLC in water/methanol. d ∆EI ) EI(aniline) - EI(alkylaniline). e ∆E1/2 ) E1/2 (aniline) - E1/2 (alkylaniline). f Poorly defined peak.

yl-, and ethylanilines were oxidized using cyclic voltammetry in anhydrous DMF. Typical voltammograms, recorded under basic conditions (excess pyridine), are shown in Figure 1, and the wave potentials are summarized in Table 1. Double anodic waves (I and II) were observed in all the voltammograms, at potentials of 0.81.0 and 1.2-1.3 V, respectively (Figure 1). There were no cathodic counterparts. Higher oxidation potentials were consistently observed in an acidic medium (excess perchloric acid, Table 1), compared with those obtained in the presence of base, as would be expected from oxidation at the nitrogen atom. These results are in agreement with the mechanism proposed by Sharma et al. (55), which involves an irreversible one-electron

transfer at wave I to form a cation radical. This is followed by another one-electron oxidation to a dication that subsequently undergoes rapid chemical and electrochemical transformations. Further support for a preferential oxidation on the arylamine nitrogen, as opposed to ring oxidation, was provided by semiempirical calculations on the anilines and the corresponding radical cations using the MOPACPM3 program. In all instances, the calculations confirmed that 40-50% of the positive charge density developed on the nitrogen atoms upon oxidation, which was considerably higher than the charge density on any of the carbon atoms, including those in resonance with the nitrogen (i.e., at the ortho and para positions). The effects of the alkyl substituents on the oxidation potentials were more apparent on the first oxidation peak (wave I). The cathodic shifts (Table 1) tended to be more pronounced for the para-substituted anilines (e.g., 4methyl- and 4-ethylaniline) as compared to meta-substituted anilines (e.g., 3-ethylaniline). Nonetheless, these shifts were modest (|∆E| E 0.14 V), such that the observed differences appear unlikely to account for the range of alkylaniline-induced genotoxic responses. Similar trends have been detected in aqueous media (Table 1; also ref 54). For comparative purposes, it should be noted that much higher cathodic shifts relative to aniline (|∆E| ∼ 0.4-0.5 V) have been reported for carcinogenic polynuclear aromatic amines, such as 2-naphthylamine, 4-aminobiphenyl, and 2-aminofluorene (56). Mutagenicity of N-Hydroxylarylamines in S. typhimurium TA100. The relationship between the mutagenicity and the location of the alkyl substituent in the N-hydroxyarylamines was analyzed using S. typhimurium TA100. Differences in the mutagenic response for the entire series of N-hydroxyarylamines investigated could stem from variations in the efficiency of O-acylation of the N-hydroxyarylamines, in the reactivities of the N-(acyloxy)arylamines, and/or in the mutagenicities of the adducts. N-Hydroxyarylamines with only m-alkyl substituents (i.e., N-hydroxy-3-methylaniline, N-hydroxy-3-ethylaniline, and N-hydroxy-3,5-dimethylaniline) tended to give the weakest mutagenic response (Table 2). The introduction of a p-alkyl substituent increased the mutagenicity (cf. 3-methyl versus 4-methyl, 3-ethyl versus 4-ethyl, and 3,5-dimethyl versus 3,4-dimethyl, Table 2).

Alkylamine Oxidation, Mutagenicity, and DNA Adducts

Using semiempirical methods, Sabbioni (28) found lower enthalpies of formation for nitrenium ions derived from para-substituted alkylanilines, as compared to their meta-substituted isomers. Thus, it is conceivable that a higher electron-donating ability of p-alkyl substituents will contribute to stabilize the putative nitrenium/carbenium ion intermediates, thereby increasing the efficiency of adduct formation and the subsequent mutagenic response. This interpretation is consistent with the observation that N-(acyloxy)-4-alkylanilines reacted with dG to a greater extent (2-7%) than their 3-substituted analogues [1-3% (21, 57); see also Materials and Methods]. The introduction of an o-alkyl substituent increased the mutagenicity of the N-hydroxyarylamines compared to their meta-substituted isomers (cf. 2-methyl versus 3-methyl, 2-ethyl versus 3-ethyl, Table 2). This effect does not appear to be associated with an increased reactivity of the N-(acyloxy)-2-alkylanilines [2-3% yield; (21, 25); see also Materials and Methods], and may be due to an intrinsically higher mutagenic efficiency of adducts containing o-alkyl substituents. When the orthosubstituted N-hydroxyarylamines contained two methyl groups, the general trend of increased mutagenicity upon para-substitution was also observed. Thus, N-hydroxy2,4-dimethylaniline was more mutagenic than its 2,3- and 2,5-dimethyl analogues (Table 2). It should be noted, however, that part of the increase in activity observed for N-hydroxy-2,4-dimethylaniline can be attributed to a higher adduct-forming efficiency of its N-acyloxy derivative (9%), compared to that of the N-(acyloxy)-2,3- and 2,5-dimethylaniline isomers [4% and 2%, respectively (21, 25); see also Materials and Methods]. In agreement with a previous study by Nohmi et al. (58), N-hydroxy-2,6dimethylaniline was the most mutagenic of the compounds examined (Table 2). This increased response does not seem to result from a higher efficiency of adduct formation [2% (21, 25); see also Materials and Methods], which further supports the suggestion that orthosubstituted adducts may have greater intrinsic mutagenicity. The mutagenicity of N-hydroxy-3-methylaniline did not differ appreciably from that of N-hydroxy-3-ethylaniline (Table 2). The same trend was observed with the 4-substituted isomers. Since N-(acyloxy)-3- and 4-methylanilines gave higher yields of adduct formation than their ethylated counterparts [3-7% versus 1-2%, respectively (21); see also Materials and Methods], this observation suggests that the DNA adducts resulting from ethylated anilines may be inherently more mutagenic than their methylated analogues, perhaps as a result of the greater bulkiness of the ethyl group (vide infra). This is also consistent with the greater mutagenicity of N-hydroxy-2-ethylaniline compared to N-hydroxy-2-methylaniline (Table 2), because the corresponding N-acyloxy derivatives yielded comparable adduct levels (2-3%; ref 21 and Materials and Methods). Synthesis and Characterization of AlkylanilineDNA Adducts. N-Acetoxy- and N-(pivaloyloxy)arylamines (21-25) were used as electrophilic synthons for the preparation of alkylaniline-DNA adducts. Due to their instability, these reactive intermediates were prepared in situ from the corresponding N-hydroxyarylamines by reaction at low temperature (-30 to -40 °C) with acetyl cyanide or pivaloyl cyanide in dry THF. The N-(acyloxy)arylamines were then reacted immediately

Chem. Res. Toxicol., Vol. 10, No. 11, 1997 1271 Scheme 1. Synthesis of Arylamine-DNA Adductsa

a R ) methyl, dimethyl, or ethyl; R ) methyl or tert-butyl; R 1 2 ) H or PO32-; Nu ) dG, dGnp (n ) 3′, 5′, or 3′,5′p), or DNA. The glycosidic torsion angle is defined as χ ) O4′-C1′-N9-C4.

with dG, dGnp (n ) 3′, 5′, and 3′,5′), and DNA (Scheme 1). The adducts formed upon reaction with dG or the dG nucleotides were isolated by chromatography, after solvent extractions to remove hydrolysis products of the N-(acyloxy)arylamines. Adducts formed upon reaction of the N-(acyloxy)arylamines with DNA were identified after enzymatic hydrolysis of the modified DNA to nucleosides. A single major adduct, which was characterized as a dG-C8-Ar species on the basis of spectroscopic evidence, was detected in all instances. With the exception of the adduct derived from 2,6-dimethylaniline (25), none of the adducts has been reported previously. FAB mass spectral analyses of the products obtained from reaction with dG were fully consistent with formation of a covalent bond between dG and each of the arylamines. Thus, strong MH+ ions were typically detected, accompanied in some instances by sodiated ions [(M + Na)+]. In addition, the cleavage of the glycosidic bond consistently gave rise to the most prominent fragment ion, (MH2 - dR)+. The UV spectra of the adducts had bathochromic shifts of 17-30 nm compared to dG, which was in the range detected previously with other monoarylamine-dG adducts (21, 25, 57) and reflected the formation of extended chromophores. These bathochromic shifts were less pronounced in the adducts derived from ortho-substituted anilines, presumably due to decreased coplanarity between the arylamine and the dG chromophores, as a result of steric hindrance introduced by the alkyl substituents. The 1H NMR spectra confirmed the formation of a covalent bond between the arylamine nitrogen and the C8 atom of the guanine ring. This conclusion was based upon the absence of the nonexchangeable H8 resonance from dG at ∼8.0 ppm, while all the remaining resonances from dG and the arylamine were observed. 13C NMR spectra further confirmed the presence of all the expected carbon resonances. All the dG5′p adducts were characterized as dG5′pC8-Ar species on the basis of FAB mass spectral analyses and 1H NMR spectroscopy. These adducts had UV spectra identical to those of the corresponding dG adducts, into which they were converted upon treatment with alkaline phosphatase. This same treatment converted all the adducts isolated from reactions with dG3′p and dG3′,5′p into species that had HPLC retention times and UV spectra identical to those of the corresponding dG adducts. Identification of the arylamine-DNA adducts was achieved after enzymatic hydrolysis of the DNA to nucleosides and was based upon the chromatographic characteristics and UV spectra of the resulting adducts, which were identical to those of the products obtained from reaction with dG.

1272 Chem. Res. Toxicol., Vol. 10, No. 11, 1997

Marques et al.

Table 3. Selected NMR Data for the N-(Deoxyguanosin-8-yl)arylamine (dG-C8-Ar) and N-(Deoxyguanosin-8-yl)arylamine 5′-Phosphate (dG5′p-C8-Ar) Adductsa δ (ppm) dG5′p′

δG adduct, Ar noneb 2-methylanilineb 3-methylanilineb 4-methylanilineb 2,3-dimethylanilineb 2,4-dimethylanilineb 2,5-dimethylaniline 2,6-dimethylaniline 3,4-dimethylaniline 3,5-dimethylaniline 2-ethylaniline 3-ethylaniline 4-ethylaniline

H2′

C2′

H2′

∼2.50c

40.33 38.02 38.34 38.40 38.04 38.00 37.98 NDd 38.28 38.27 38.07 38.35 38.33

∼2.56c 2.90 2.78 2.80 2.93 2.90 2.86 NDd 2.79 2.77 2.93 2.82 2.79

2.69 ∼2.50c ∼2.50c 2.66 2.69 2.68 2.71 ∼2.50c ∼2.50c 2.69 ∼2.50c ∼2.50c

a The selected resonances correspond to the deoxyribose protons and carbons that are sensitive to the glycosidic torsion angle. The spectra were recorded at 22 °C in Me2SO-d6 (dG adducts) or Me2SO-d6/D2O (dG5′p adducts). Proton chemical shifts were referenced to the residual Me2SO resonance at 2.49 ppm and carbon chemical shifts to the solvent resonance at 39.5 ppm. b Data from Marques et al. (21). c Partially obscured by the solvent resonance. d ND, not determined.

Conformational Properties of the Adducts. A comparative analysis of the NMR spectral data for the dG and dG5′p adducts extended the evidence of conformational trends detected in our previous work (21). Thus, the H2′ resonances of the dG adducts containing an o-alkyl substituent (e.g., dG-C8-2,5-diMeA) were shifted downfield by 0.16-0.21 ppm (Table 3), as compared to the same resonances in dG and in the dG adducts not containing an o-alkyl substituent (e.g., dGC8-3,4-diMeA). Likewise, compared to dG5′p, the H2′ resonances of the dG5′p adducts containing o-alkyl substituents were shifted downfield to a greater extent than the corresponding resonances of the dG5′p adducts having no ortho substituents in the arylamine moiety (Table 3). Such downfield shifts of the H2′ resonance are generally considered good indicators of changes in the dynamic anti h syn equilibria toward increased populations of syn conformers (37, 41, 59-61). The C2′ resonances of dG adducts containing o-methyl or o-ethyl substituents were shielded ∼0.3 ppm compared to the corresponding resonances in adducts not containing an ortho substituent (Table 3). Although the magnitude of this shielding was modest, it should be noted that upfield shifts of the C2′ resonance have also been associated with greater proportions of syn conformers (37, 60, 61). The resolution of the proton-coupled 13C NMR spectra was insufficient to allow accurate measurements of the C8H1′ and C4-H1′ coupling constants, which prevented a quantitative evaluation of the relative populations of anti and syn conformers for specific adducts based on Karplustype relationships (41, 61). Nonetheless, the 1H and 13C NMR data for the entire series of alkylaniline-dG adducts strongly suggest that increased syn populations are associated with ortho substitution on the arylamine fragment. The conformational trends were further probed using molecular dynamics and molecular mechanics in order to search for the minimum energy geometries in the entire series of dG-C8-Ar adducts, as well as the relative contributions of anti (χ ) 200-235°) and syn (χ ) 30-

55°) conformers for each adduct. All the adducts proved to be very flexible, with numerous conformers having energies within 3 kcal/mol of the global minimum. In addition, with the exception of the 4-methylaniline adduct, all the nucleoside adducts were found to have a global minimum in the syn domain. Typically, the total energy of the global minimum was higher in the adducts containing o-alkyl substituents (e.g., dG-C8-2-EtA) than in their meta- and para-substituted analogues (e.g., dGC8-3-EtA and dG-C8-4-EtA). When the percent occupancy of anti and syn states was calculated on the basis of their relative energies, we found a consistent and unexpectedly high incidence of syn contributors (e.g., 71% for the dG-C8-4-EtA and 96% for the dG-C8-2,3-diMeA at 300 K), which appeared to be associated mainly with the occurrence of short hydrogen bonds (E1.96 Å) between the guanine N3 and the sugar 5′-OH. Since the AMBER force field tends to overestimate electrostatic interactions, and therefore hydrogen bonds (62), other interactions prevailing in solution are likely to have been underestimated in the calculations. This precluded a detailed analysis of the conformational properties for the different adducts. Nonetheless, it should be noted that adducts with ethyl substituents (e.g., dG-C8-4-EtA) tended to have increased populations of syn conformers compared to their methylated analogues (e.g., dG-C8-4MeA), presumably as a consequence of greater steric hindrance introduced by the ethyl group.

Conclusions Carcinogenic arylamines typically undergo metabolic activation via N-hydroxylation followed in most instances by O-esterification (18-20). In this study, we analyzed the ability of methyl-, dimethyl-, and ethylaniline constituents of tobacco smoke to undergo oxidation at the nitrogen atom. In addition, we investigated both the mutagenicity of the corresponding N-hydroxyalkylanilines and the conformational properties of the DNA adducts generated from their N-acyloxy derivatives. All the arylamines underwent irreversible N-oxidation at potentials 0.4-0.5 V higher than those observed for the oxidation of carcinogenic polynuclear aromatic amines (56). There were minor differences (E0.14 V) in the oxidation potentials, which were consistent with the position and electron-donating abilities of the alkyl substituents. However, the differences appeared to be too small to account for the range of genotoxic effects among the alkylanilines, which suggested that subsequent events may be significant factors in the process. Using S. typhimurium TA100 as the tester strain, we found increased mutagenicities in N-hydroxyarylamines containing p-alkyl substituents. This was attributed in part to higher efficiencies of adduct formation with the para-substituted compounds, due to greater stabilities, and thus greater half-lifes, of the corresponding nitrenium ion intermediates. Increased mutagenicities were also observed upon o-alkyl substitution, to the extent that N-hydroxy-2,6-dimethylaniline was the most mutagenic of all the N-hydroxyalkylanilines. However, this property was not related to a greater ability of the orthosubstituted species to form DNA adducts, which suggested that adducts from ortho-substituted alkylanilines may be intrinsically more mutagenic than their metaand para-substituted analogues. In addition, the mutagenicity data indicated that ethyl substituents may induce higher mutagenic responses than their methyl analogues.

Alkylamine Oxidation, Mutagenicity, and DNA Adducts

In all instances, N-(acyloxy)arylamines generated from the N-hydroxyarylamines reacted with dG, dG nucleotides, and DNA to yield C8-substituted dG derivatives as the major adducts. This is consistent with what has been observed with carcinogenic aromatic amines, such as 2-aminofluorene and 4-aminobiphenyl (18-20). Despite their structural similarity, the alkylaniline-dG adducts displayed distinct conformational trends that were determined by the location of the alkyl substituents. Spectroscopic data indicated higher percentages of lowenergy syn conformers in the adducts that contained alkyl groups ortho to the arylamine nitrogen than in the adducts not bearing ortho substituents. Increased populations of syn conformers were also apparent for adducts containing ethyl substituents compared to their methylated analogues. Taken together, the data strongly suggest that the conformational properties of the DNA adducts, in particular their ability to adopt syn conformations, may be determinant factors for the genotoxic responses elicited by certain alkylanilines (e.g., 2-methylaniline and 2,6-dimethylaniline).

Acknowledgment. This work was supported in part by NATO (Collaborative Research Grant 910561), by Junta Nacional de Investigac¸ a˜o Cientı´fica e Tecnolo´gica, Portugal (Contract PBIC/C/CEN/1154/92), and by a Postgraduate Research Program administered by Oak Ridge Institute for Science and Education. We thank Joanna Deck for some of the NMR spectra, J. Pat Freeman and Indale´cio Marques for the mass spectra, Vı´tor Pereira for the elemental analyses, and Cindy Hartwick for helping prepare this manuscript. Part of this study was presented at the Sixth International Conference on Carcinogenic/Mutagenic N-Substituted Aryl Compounds (63).

References (1) Searle, C. E., and Teale, O. J. (1990) Occupational carcinogens. In Handbook of Experimental Pharmacology. Chemical Carcinogenesis and Mutagenesis (Cooper, C. S., and Grover, P. L., Eds.) Vol. 94/I, pp 103-151, Springer-Verlag, Heidelberg. (2) Patrianakos, C., and Hoffmann, D. (1979) Chemical studies on tobacco smoke LXIV. On the analysis of aromatic amines in cigarette smoke. J. Anal. Toxicol. 3, 150-154. (3) Luceri, F., Pieraccini, G., Moneti, G., and Dolara, P. (1993) Primary aromatic amines from side-stream cigarette smoke are common contaminants of indoor air. Toxicol. Ind. Health 9, 405413. (4) Felton, J. S., Knize, M. G., Roper, M., Fultz, E., Shen, N. H., and Turteltaub, K. W. (1992) Chemical analysis, prevention, and lowlevel dosimetry of heterocyclic amines from cooked foods. Cancer Res. 52 (Suppl.), 2103s-2107s. (5) Hoffmann, D., and Hecht, S. S. (1990) Advances in tobacco carcinogenesis. In Handbook of Experimental Pharmacology. Chemical Carcinogenesis and Mutagenesis (Cooper, C. S., and Grover, P. L., Eds.) Vol. 94/I, pp 63-102, Springer-Verlag, Heidelberg. (6) Mommsen, S., and Aagaard, J. (1983) Tobacco as a risk factor in bladder cancer. Carcinogenesis 4, 335-338. (7) Vineis, P. (1992) Epidemiological models of carcinogenesis: the example of bladder cancer. Cancer Epidemiol. Biomarkers Prev. 1, 149-153. (8) Ambrosone, C. B., Freudenheim, J. L., Graham, S., Marshall, J. R., Vena, J. E., Brasure, J. R., Michalek, A. M., Laughlin, R., Nemoto, T., Gillenwater, K. A., Harrington, A. M., and Shields, P. G. (1996) Cigarette smoking, N-acetyltransferase 2 genetic polymorphisms, and breast cancer risk. J. Am. Med. Assoc. 276, 1494-1501. (9) DeBruin, L. S., Josephy, P. D., Kingsmill, C. A., Pawliszyn, J., and Atkinson, S. A. (1997) Detection of aromatic amines in human milk. Proc. Am. Assoc. Cancer Res. 38, 352. (10) Garner, R. C., Martin, C. N., and Clayson, D. B. (1984) Carcinogenic aromatic amines and related compounds. In Chemical Carcinogens (Searle, C. E., Ed.) ACS Monograph 182, pp 175276, American Chemical Society, Washington, DC.

Chem. Res. Toxicol., Vol. 10, No. 11, 1997 1273 (11) Environmental Criteria and Assessment Office. (1988) Health and environmental effects profile for toluidines. NTIS/PB88-130075, 77 pp, National Technical Information Service, Springfield, VA. (12) Environmental Criteria and Assessment Office. (1989) Health and environmental effects profile for 2,4-dimethylaniline and 2,4dimethylaniline hydrochloride. NTIS/PB89-123004, 70 pp, National Technical Information Service, Springfield, VA. (13) Weisburger, E. K., Russfield, A. B., Homburger, F., Weisburger, J. H., Boger, E., Van Dongen, C. G., and Chu, K. C. (1978) Testing of twenty-one environmental aromatic amines or derivatives for long-term toxicity or carcinogenicity. J. Environ. Pathol. Toxicol. 2, 325-356. (14) Ward, E., Carpenter, A., Markowitz, S., Roberts, D., and Halperin, W. (1991) Excess number of bladder cancers in workers exposed to ortho-toluidine and aniline. J. Natl. Cancer Inst. 83, 501-506. (15) Keenaghan, J. B., and Boyes, R. N. (1972) The tissue distribution, metabolism and excretion of lidocaine in rats, guinea pigs, dogs and man. J. Pharmacol. Exp. Ther. 180, 454-463. (16) U.S. National Toxicological Program. (1990) Toxicology and carcinogenesis studies of 2,6-xylidine (2,6-dimethylaniline) (CAS No. 87-62-7) in Charles River CD rats (feed studies). Technical Report Series No. 278, NTP, Research Triangle Park, NC. (17) IARC. (1993) 2,6-Dimethylaniline (2,6-xylidine). IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Occupational Exposures of Hairdressers and Barbers and Personal Use of Hair Colourants; Some Hair Dyes, Cosmetic Colourants, Industrial Dyestuffs and Aromatic Amines, Vol. 57, pp 323-335, International Agency for Research on Cancer, Lyon. (18) Beland, F. A., and Kadlubar, F. F. (1990) Metabolic activation and DNA adducts of aromatic amines and nitroaromatic hydrocarbons. In Handbook of Experimental Pharmacology. Chemical Carcinogenesis and Mutagenesis (Cooper, C. S., and Grover, P. L., Eds.) Vol. 94/I, pp 267-325, Springer-Verlag, Heidelberg. (19) Kadlubar, F. F. (1994) DNA adducts of carcinogenic aromatic amines. In DNA Adducts: Identification and Biological Significance (Hemminki, K., Dipple, A., Shuker, D. E. G., Kadlubar, F. F., Segerba¨ck, D., and Bartsch, H., Eds.) IARC Scientific Publications No. 125, pp 199-216, International Agency for Research on Cancer, Lyon. (20) Delclos, K. B., and Kadlubar, F. F. (1997) Carcinogenic aromatic amines and amides. In Comprehensive Toxicology, Volume 12: Chemical Carcinogens and Anticarcinogens (Bowden, G. T., and Fischer, S. M., Eds.) pp 141-170, Elsevier Science, Ltd., Oxford. (21) Marques, M. M., Mourato, L. L. G., Santos, M. A., and Beland, F. A. (1996) Synthesis, characterization, and conformational analysis of DNA adducts from methylated anilines present in tobacco smoke. Chem. Res. Toxicol. 9, 99-108. (22) Famulok, M., and Boche, G. (1989) Formation of N-(deoxyguanosin-8-yl)aniline in the in vitro reaction of N-acetoxyaniline with deoxyguanosine and DNA. Angew. Chem., Int. Ed. Engl. 28, 468469. (23) Meier, C., and Boche, G. (1990) N-Aryl-O-(R-aminoacyl)hydroxylamine: Modellreaktionen mit Desoxyguanosin, Guanosin und 5′-Guanosinmonophosphat zur Aktivierung monocyclischer aromatischer Amine (z.B. Phenacetin) zu ultimaten Carcinogenen. [N-Aryl-O-(R-aminoacyl)hydroxylamines: Model reactions with deoxyguanosine, guanosine and 5′-guanosine monophosphate for the activation of monocyclic aromatic amines (e.g. phenacetin) into ultimate carcinogens.] Chem. Ber. 123, 1699-1705. (24) Meier, C., and Boche, G. (1991) The modification of guanine nucleosides and nucleotides by the borderline arylamine carcinogens 4-methyl- and 4-methoxyaniline: chemistry and structural characterization. Carcinogenesis 12, 1633-1640. (25) Beyerbach, A., Farmer, P. B., and Sabbioni, G. (1996) Synthesis and analysis of DNA adducts of arylamines. Biomarkers 1, 9-20. (26) Stillwell, W. G., Bryant, M. S., and Wishnok, J. S. (1987) GC/MS analysis of biologically important aromatic amines. Application to human dosimetry. Biomed. Environ. Mass Spectrom. 14, 221227. (27) Bryant, M. S., Vineis, P., Skipper, P. L., and Tannenbaum, S. R. (1988) Hemoglobin adducts of aromatic amines: Associations with smoking status and type of tobacco. Proc. Natl. Acad. Sci. U.S.A. 85, 9788-9791. (28) Sabbioni, G. (1992) Hemoglobin binding of monocyclic aromatic amines: Molecular dosimetry and quantitative structure activity relationships for the N-oxidation. Chem.-Biol. Interact. 81, 91117. (29) Ringe, D., Turesky, R. J., Skipper, P. L., and Tannenbaum, S. R. (1988) Structure of the single stable hemoglobin adduct formed by 4-aminobiphenyl in vivo. Chem. Res. Toxicol. 1, 22-24. (30) Bryant, M. S., Simmons, H. F., Harrell, R. E., and Hinson, J. A. (1994) 2,6-Dimethylaniline-hemoglobin adducts from lidocaine in humans. Carcinogenesis 15, 2287-2290.

1274 Chem. Res. Toxicol., Vol. 10, No. 11, 1997 (31) Sabbioni, G. (1994) Hemoglobin binding of arylamines and nitroarenes: Molecular dosimetry and quantitative structureactivity relationships. Environ. Health Perspect. 102 (Suppl. 6), 61-67. (32) Sabbioni, G., and Sepai, O. (1995) Comparison of hemoglobin binding, mutagenicity, and carcinogenicity of arylamines and nitroarenes. Chimia 49, 374-380. (33) Debnath, A. K., Debnath, G. D., Shusterman, A. J., and Hansch, C. (1992) A QSAR investigation of the role of hydrophobicity in regulating mutagenicity in the Ames test: 1. Mutagenicity of aromatic and heteroaromatic amines in Salmonella typhimurium TA98 and TA100. Environ. Mol. Mutagen. 19, 37-52. (34) Hecht, S. S., El-Bayoumy, K., Tulley, L., and LaVoie, E. (1979) Structure-mutagenicity relationships of N-oxidized derivatives of aniline, o-toluidine, 2′-methyl-4-aminobiphenyl, and 3,2′-dimethyl4-aminobiphenyl. J. Med. Chem. 22, 981-987. (35) El-Bayoumy, K., LaVoie, E. J., Tulley-Freiler, L., and Hecht, S. S. (1981) Effects of ortho-methyl substituents on the mutagenicity of aminobiphenyls and aminonaphthalenes. Mutat. Res. 90, 345354. (36) Nussbaum, M., Fiala, E. S., Kulkarni, B., El-Bayoumy, K., and Weisburger, J. H. (1983) In vivo metabolism of 3,2′-dimethyl-4aminobiphenyl (DMAB) bearing on its organotropism in the Syrian golden hamster and the F344 rat. Environ. Health Perspect. 49, 223-231. (37) Evans, F. E., Miller, D. W., and Beland, F. A. (1980) Sensitivity of the conformation of deoxyguanosine to binding at the C-8 position by N-acetylated and unacetylated 2-aminofluorene. Carcinogenesis 1, 955-959. (38) Kadlubar, F. F., Beland, F. A., Beranek, D. T., Dooley, K. L., Heflich, R. H., and Evans, F. E. (1982) Arylamine-DNA adduct formation in relation to urinary bladder carcinogenesis and Salmonella typhimurium mutagenesis. In Environmental Mutagens and Carcinogens (Sugimura, T., Kondo, S., and Takebe, H., Eds.) pp 385-396, Liss, New York. (39) Breitmeier, E., and Voelter, W. (1987) Carbon-13 NMR Spectroscopy, 3rd ed., 515 pp, VCH, Weinheim. (40) Jones, A. J., Grant, D. M., Winkley, M. W., and Robins, R. K. (1970) Carbon-13 magnetic resonance. XVII. Pyrimidine and purine nucleosides. J. Am. Chem. Soc. 92, 4079-4087. (41) Evans, F. E., Miller, D. W., and Levine, R. A. (1984) Conformation and dynamics of the 8-substituted deoxyguanosine 5′-monophosphate adduct of the carcinogen 2-(acetylamino)fluorene. J. Am. Chem. Soc. 106, 396-401. (42) Perrin, D. D., and Armarego, W. L. F. (1988) Purification of Laboratory Chemicals, 3rd ed., 391 pp, Pergamon, Oxford. (43) Brand, K., and Mahr, J. (1935) Reduction of nitro and polynitro compounds. XIV. The reduction of aromatic mono- and polynitro compounds. J. Prakt. Chem. 142, 153-176; Chem. Abstr. 29, 29309, 1935. (44) Bowen, D. M., Belfit, R. W., Jr., and Walser, R. A. (1953) The synthesis and nitration of 2,6- and 2,7-dimethylquinoline and of 2,5,8-trimethylquinoline. J. Am. Chem. Soc. 75, 4307-4311. (45) Kochergin, P. M. (1957) Synthesis of p-, o- and m-nitroethylbenzenes. Zh. Obshch. Khim. 27, 3204-3207; Chem. Abstr. 52, 8987h, 1958. (46) Clarke, H. T., and Taylor, E. R. (1923) m-Nitrotoluene. Org. Synth. 3, 91-93.

Marques et al. (47) Sargent, L. J. (1954) The synthesis of 1-ethyl-, 2-ethyl-, 3-ethyland 4-ethylacridine. J. Org. Chem. 19, 599-607. (48) Case, F. H., Jacobs, Z. B., Cook, R. S., and Dickstein, J. (1957) Substituted 1,10-phenanthrolines. X. Ethyl derivatives. J. Org. Chem. 22, 390-393. (49) Bamberger, E. (1895) Ueber die Reduction aromatischer Nitroverbindungen. (On the reduction of aromatic nitrocompounds.) Ber. 28, 245-251. (50) Bamberger, E., and Rising, A. (1901) Einfluss von Methyl auf die Reactionsgeschwindigkeit der Arylhydroxylamine. (Influence of methyl on the reactivity of arylhydroxylamines.) Liebigs Ann. Chem. 316, 257-311. (51) Heflich, R. H., Morris, S. M., Beranek, D. T., McGarrity, L. J., Chen, J. J., and Beland, F. A. (1986) Relationships between the DNA adducts and the mutations and sister-chromatid exchanges produced in Chinese hamster ovary cells by N-hydroxy-2-aminofluorene, N-hydroxy-N′-acetylbenzidine and 1-nitrosopyrene. Mutagenesis 1, 201-206. (52) Maron, D. M., and Ames, B. N. (1983) Revised methods for the Salmonella mutagenicity test. Mutat. Res. 113, 173-215. (53) Stewart, J. J. P. (1989) Optimization of parameters for semiempirical methods. I. Method. J. Comput. Chem. 10, 209-220. (54) Suatoni, J. C., Snyder, R. E., and Clark, R. O. (1961) Voltammetric studies of phenol and aniline ring substitution. Anal. Chem. 33, 1894-1897. (55) Sharma, L. R., Manchanda, A. K., Singh, G., and Verma, R. S. (1982) Cyclic voltammetry of aromatic amines in aqueous and non-aqueous media. Electrochim. Acta 27, 223-233. (56) Kadlubar, F. F., Fu, P. P., Jung, H., Shaikh, A. U., and Beland, F. A. (1990) The metabolic N-oxidation of carcinogenic arylamines in relation to nitrogen charge density and oxidation potential. Environ. Health Perspect. 87, 233-236. (57) Schu¨tze, D., Sagelsdorff, P., Sepai, O., and Sabbioni, G. (1996) Synthesis and quantification of DNA adducts of 4,4′-methylenedianiline. Chem. Res. Toxicol. 9, 1103-1112. (58) Nohmi, T., Yoshikawa, K., Nakadate, M., Miyata, R., and Ishidate, M., Jr. (1984) Mutations in Salmonella typhimurium and inactivation of Bacillus subtilis transforming DNA induced by phenylhydroxylamine derivatives. Mutat. Res. 136, 159-168. (59) Jordan, F., and Niv, H. (1977) C8-Amino purine nucleosides. A well-defined steric determinant of glycosyl conformational preferences. Biochim. Biophys. Acta 476, 265-271. (60) Stolarski, R., Dudycz, L., and Shugar, D. (1980) NMR studies of the syn-anti dynamic equilibrium in purine nucleosides and nucleotides. Eur. J. Biochem. 108, 111-121. (61) Stolarski, R., Hagberg, C.-E., and Shugar, D. (1984) Studies on the dynamic syn-anti equilibrium in purine nucleosides and nucleotides with the aid of 1H and 13C NMR spectroscopy. Eur. J. Biochem. 138, 187-192. (62) Weiner, S. J., Kollman, P. A., Case, D. A., Singh, U. C., Ghio, C., Alagona, G., Profeta, S., Jr., and Weiner, P. (1984) A new force field for molecular mechanical simulation of nucleic acids and proteins. J. Am. Chem. Soc. 106, 765-784. (63) Beland, F. A., Melchior, W. B., Jr., Mourato, L. L. G., Santos, M. A., and Marques, M. M. (1997). Arylamine-DNA adduct conformation in relation to mutagenesis. Mutat. Res. 376, 13-19.

TX970104W