454
Chem. Res. Toxicol. 1998, 11, 454-463
Reactivity of Haloketenes and Halothioketenes with Nucleobases: Chemical Characterization of Reaction Products Michael Mu¨ller,† Gerhard Birner,* and Wolfgang Dekant Institut fu¨ r Toxikologie, Universita¨ t Wu¨ rzburg, Versbacher Strasse 9, D-97078 Wu¨ rzburg, Germany Received August 15, 1997
Halothioketenes and haloketenes are postulated as intermediates in haloolefin bioactivation. Little is known about the interactions of these reactive intermediates with macromolecules such as DNA. DNA binding, however, may be relevant in the toxicity of the parent olefins since they or their proximate metabolites are genotoxic. This prompted us to elucidate the structures and properties of potential DNA adducts formed. Adenine, cytosine, guanine, and thymine were reacted with chloro- and dichlorothioketene, chloro- and dichloroketene, and chloro- and dichloroacyl chloride. While thymine did not react, adenine and cytosine formed stable DNA base adducts with all reaction partners as demonstrated by HPLC analysis. Guanine yielded only products with chloroketene and chloroacetyl chloride. The pH-dependent UV spectra, 1H and 13C NMR, FT-IR, and elemental analysis showed (i) nucleophilic attack of the exocyclic amino groups of the DNA bases yielded haloacyl (thio)amides with all reactants as clearly demonstrated by the FT-IR spectra; (ii) the sulfur in the initial thioamides seems to be rapidly exchanged with oxygen; (iii) the acyl chlorides form identical products but in lower yields as compared to the haloketenes. Reactions of the nucleosides with haloketenes showed the formation of similar nucleoside adducts upon HPLC and MS analysis. Beside the modification of the base moieties, additional peaks in the reaction mixtures analyzed suggested acylation of the deoxyribose hydroxyl groups. In aqueous solutions at pH 7 N6-(chloroacetyl)adenine, N4-(chloroacetyl)cytosine, and N2-(chloroacetyl)guanine are not stable and cleaved to the original base or form 1,N6-acetyladenine, 3,N4-acetylcytosine, 1,N2-acetylguanine, and N2,3acetylguanine. Under the same conditions, N6-(dichloroacetyl)adenine and N4-(dichloroacetyl)cytosine were completely hydrolyzed to adenine and cytosine, respectively. All haloacyl DNA base adducts proved to be stable at pH 5 but were rapidly degraded at neutral or alkaline pH. The compounds with an additional five-membered ring remained unchanged after 1 week at room temperature. All synthesized DNA base adducts except N2-(chloroacetyl)guanine and 1,N2-acetylguanine were fluorescent. The characterized compounds, especially the etheno () base adduct-related derivatives, may represent potential DNA adducts formed as a consequence of haloolefin bioactivation.
Introduction Halothioketenes, haloketenes, and haloacyl chlorides have been postulated to be metabolites formed in haloolefin bioactivation. There is substantial evidence suggesting halothioketenes as ultimate toxic products of glutathione conjugation and subsequent enzymatic processing of trichloroethene, tetrachloroethene, and dichloroacetylene. Haloketenes and haloacyl chlorides are likely biological reactive intermediates formed by cytochrome P450-mediated oxidation of tri- and perchloroethene (haloacyl chlorides) or haloethynes (haloketenes) (1). Until recently, little was known about the interactions of these highly reactive intermediates with macromolecules such as proteins or DNA. The toxic and tumorigenic dichloroacetylene is bioactivated by cytochrome * Address correspondence to Dr. G. Birner. Tel: +49 (0)931 201 3990. Fax: +49 (0)931 201 3446. E-mail:
[email protected]. † Present address: Georg-August-Universita ¨ t Go¨ttingen, Abteilung fu¨r Arbeits- und Sozialmedizin, Waldweg 37, 37073 Go¨ttingen, Germany.
P450 to give a dichloroketene as a presumed reactive intermediate (2, 3); dichloroacetylene and several nephrotoxic haloolefins such as perchloroethene are thought to be metabolized to acyl halides and thioketenes (Scheme 1) (4-6). Previous studies (7) have characterized some of these interactions with proteins and identified amino acid adducts in vitro as well as in the target organs rat liver and kidney in vivo. Amino acid modifications in perchloroethene-treated rats were identified as N(trichloroacetyl)-L-lysine and N-(dichloroacetyl)-L-lysine. N-(Trichloroacetyl)-L-lysine in proteins is thought to be formed by the reaction of lysine residues in proteins with trichloroacetyl chloride; N-(dichloroacetyl)-L-lysine in proteins is a likely product of dichlorothioketene formed by cysteine conjugate β-lyase-mediated metabolism of the perchloroethene metabolite S-(1,2,2-trichlorovinyl)-L-cysteine. With respect to DNA similar data are not available, although a variety of positive genotoxicity test results in vitro and radioactive binding experiments in rats in vivo suggest the presence of covalently modified DNA (8). This prompted us to elucidate the structures and physicochemical properties of such potential DNA
S0893-228x(97)00143-4 CCC: $15.00 © 1998 American Chemical Society Published on Web 04/29/1998
Synthesis of Ketene DNA Adducts Scheme 1. Formation of a Ketene and a Thioketene as Reactive Intermediates by Bioactivation of Dichloroacetylene
adducts formed in chemical reactions between nucleobases and ketenes or thioketenes. Ketenes were generated by HCl elimination from acyl chloride and thioketenes by cleavage of R-halovinyl 2-nitrobenzyl disulfides (9). The elucidated structures and the information on the stability of the base modifications should be the basis for development of methods to detect and quantify keteneand thioketene-derived DNA adducts in biological systems.
Experimental Procedures Chemicals. All chemicals and solvents were purchased from Aldrich Chemical Co. (Steinheim, FRG) or Sigma Chemical Co. (Deisenhofen, FRG) in the highest purity available. Spectroscopy. UV spectra were recorded on a Uvikon 860 instrument (Kontron, Neufahrn, FRG), and the reported λmax values were determined using the peak finder program. In the stability experiments, the absorption decrease at 300 nm of each DNA adduct was monitored with the kinetics program, collecting data every 10 min for a 3-h period. Fluorescence emission spectra were obtained with an SFM 25 instrument (Kontron, Neufahrn, FRG). Excitation wavelengths as suggested by Barrio et al. (10) were 245 and 300 nm, respectively; response times were set to 2 or 8 s. 1H and 13C NMR spectra (broadband proton-decoupled and “off-resonance“)1 of the samples dissolved in [2H6]dimethyl sulfoxide were run on a Bruker AM-400 instrument (Bruker, Karlsruhe, FRG). Tetramethylsilane served as an internal standard. Assignments of the 13C resonances were made on the basis of data reported for the unmodified nucleobases (11). Fourier transform infrared spectra were recorded with a Bruker IFS 88 instrument (Bruker, Karlsruhe, FRG) in a KBr matrix. Mass Spectrometry. Mass spectra were recorded with a Finnigan 4500 instrument (Finnigan MAT GmbH, Bremen, 1 The “off-resonance“ method is not very sensitive for quaternary carbon resonances; therefore, some of the 13C NMR spectra reported herein may be incomplete with respect to these resonances.
Chem. Res. Toxicol., Vol. 11, No. 5, 1998 455 FRG) with a thermospray interface or with a Trio 2000 mass spectrometer (Fisons Instruments, Mainz, FRG) with an electrospray interface in the positive and negative ion modes using NH4CH3CO2 buffer (pH 5, 10 mM). GC/MS data were acquired with a HP 5890 gas chromatograph coupled with a HP 5970 mass-selective detector (Hewlett-Packard, Avondale, PA) using electron-impact ionization (70 eV). Gas chromatographic separation was achieved on a DB-1 fused-silica gel capillary (10 m × 0.18 mm, film thickness 0.4 µm) (J & W Scientific, Folsom, CA). Physicochemical Analysis. Elemental analyses were performed at the Institut fu¨r Anorganische Chemie, University of Wu¨rzburg. Melting points (uncorrected) were determined with a Kofler heating block. High-Performance Liquid Chromatography (HPLC). HPLC was performed using two Waters M-6000 A pumps (Millipore GmbH, Eschborn, FRG) coupled with a gradient control unit. Steel columns filled with Partisil ODS III [5 µm, 8 × 250 mm (preparative) or 4 × 250 mm (analytical); Bischoff, Leonberg, FRG] were used for separation. Solvent systems were (i) solvent A, water, pH 5; solvent B, CH3OH; 0-100% B in 50 min, flow rate 1 mL/min (analytical) or 3 mL/min (preparative haloacetyl DNA adducts isolation); (ii) pH gradient elution method described for fluorescence detection by Bedell et al. (12); (iii) solvent A, 50 mM NH4HCO2, pH 5; solvent B, CH3OH; 0-50% B in 30 min, flow rate 3 mL/min (preparative cyclic DNA adducts isolation). The effluent was passed through a diode array detector (HP 1090 A) (Hewlett-Packard, Avondale, PA) and/or a programmable HP 1046 A fluorescence detector (Hewlett-Packard, Avondale, PA); λex 232 nm, λem 356 nm for detection of peaks. Analytical data were evaluated using a HP 9000 computer system (Hewlett-Packard, Avondale, PA). Syntheses. r-Haloalkenyl 2-Nitrophenyl Disulfides: 1,2,2-Trichlorovinyl 2-nitrophenyl disulfide (TCVDS)2 and 1,2dichlorovinyl 2-nitrophenyl disulfide (DCVDS) were prepared as described previously (4). Analytical data of TCVDS were identical to those reported. Characterization of 1,2-dichlorovinyl 2-nitrophenyl disulfide: GC/MS (EI, 70 eV) m/z (35Cl) 281 (M+), 154 (M+ - C2Cl2SH), 138 (C6H4SNO), 96 (M+ - C6H4NO2S2); 1H NMR (C2HCl3) δ 6.50 (s, 1H, H-1), 7.34 (td, 1H, H-6, 3J ) 7.7 Hz, 4J ) 1.2 Hz), 7.64 (td, 1H, H-7, 3J ) 8.3 Hz, 4J ) 1.3 Hz), 8.07 (dd, 1H, H-8, 2J ) 8.2 Hz, 3J ) 1.1 Hz), 8.21 (dd, 1H, H-5, 2J ) 8.2 Hz, 3J ) 1.3 Hz); 13C NMR δ 120.7 (d, C-1), 125.9 (d, C-5), 126.8 (d, C-8), 127.7 (d, C-6), 130.6 (s, C-2), 134.2 (s,C-3), 136.0 (d, C-7), 145.9 (s, C-4). DCVDS was kindly provided by Dr. C. Richling. Caution: The following chemicals are hazardous and should be handled carefully: dichloroacetyl chloride and chloroacetyl chloride. Corrosive! Lachrymator! Use a hood! Reaction of Ketenes and Thioketenes with Nucleobases. Haloketene DNA Adducts: DNA base or nucleoside (0.2 mmol) was dissolved at room temperature in 1 mL of dry dimethylformamide along with 0.5 mmol of 1,4-diazabicyclo[2.2.2]octane. Dichloro- or chloroketene was generated according to the method of Brady (13) by adding 50 µL of the appropriate acyl chlorides (0.5 mmol) in five portions of 10 µL (0.1 mmol) each in 10-min intervals with stirring. After the last addition stirring for another 20 min completed the reaction. The reaction mixture was centrifuged at (3 × 103)g for 15 min and the yellow-brown supernatant subjected to HPLC. Fractions were collected, methanol was removed under a stream of nitrogen, and the residues were lyophilized. Halothioketene DNA Base Adducts: DNA bases (10 µmol) and TCVDS or DCVDS (10 µmol) were dissolved in 200 µL of DMF and stirred for 1 h at room temperature. The reaction mixture was analyzed by HPLC as described above. 2 Abbreviations: 1,2,2-trichlorovinyl 2-nitrophenyl disulfide (TCVDS); 1,2-dichlorovinyl 2-nitrophenyl disulfide (DCVDS); 3H-7,8-dihydro-8oxoimidazo[1,2R]purine (1,N6-acetyladenine); 1,2,4,5-tetrahydro-2,5dioxoimidazo[1,2β]pyrimidine (3,N4-acetylcytosine); 3H-4,6,7,9-tetrahydro-6-oxoimidazo[1,2R]purin-9-one (1,N2-acetylguanine); 3H-5,6,8,9tetrahydro-6-oxoimidazo[1,2R]purin-9-one (N2,3-acetylguanine).
456 Chem. Res. Toxicol., Vol. 11, No. 5, 1998 Haloacyl Chloride DNA Base Adducts: Haloacyl chloride DNA base adducts were formed following the first procedure without the presence of 0.5 mmol of 1,4-diazabicyclo[2.2.2]octane, which would have caused the generation of haloketenes. Cyclic DNA Base Adducts: Haloacylated DNA base (100 µmol) was dissolved under stirring in 1 mL of water. The pH of the solution was adjusted to 9 with 1 N NaOH, and the pH was monitored and kept constant by further addition of 1 N NaOH for 1 h of hydrolysis. After neutralization with glacial acetic acid, the reaction mixture was separated by HPLC. Characterization of the DNA Base Adducts. All compounds were purified by the following procedures. (A) N6-(Dichloroacetyl)adenine. The fraction eluting at tR 23.0 min under HPLC conditions for preparative workup was collected and characterized: yield 47.9%; UV (H2O) λmax (pH 5) 215, 280 nm ( 16 370, 17 715 mol-1 cm-1), λmax (pH 7) 212, 270 nm ( 16 045, 10 550 mol-1 cm-1), λmax (pH 8) 216, 270 nm ( 16 270, 12 265 mol-1 cm-1); thermospray+ MS m/z 246 (20, [M]+), 136 (20, M - C2OCl2); mp >270 °C; 1H NMR [(C2H3)2SO)] δ 6.86 (s, 1H, H-10), 8.52 (s, 1H, H-2), 8.71 (s, 1H, H-5); 13C NMR [(C2H ) SO)] δ 66.5 (d, C-10), 145.8 (d, C-2), 151.0 (d, 3 2 C-5), 162.9 (s, C-9); FT-IR (KBr) ν 3340 cm-1 (NH), 1722 (CdO), 1631 (CdC), 1573 (NH), 801 (C-Cl). Anal. Calcd: C:H:N, 34.15:2.05:28.47. Found: C:H:N, 34.48:2.13:28.42. (B) N4-(Dichloroacetyl)cytosine. The fraction eluting at tR 20.8 min under HPLC conditions for preparative workup was collected and characterized: yield 44.1%; UV (H2O) λmax (pH 5) 213, 245, 273 nm ( 14 410, 9860, 6250 mol-1 cm-1), λmax (pH 7) 202, 267, 330 nm ( 14 290, 6835, 4300 mol-1 cm-1), λmax (pH 8) 201, 267, 331 nm ( 12 340, 5115, 2905 mol-1 cm-1); thermospray+ MS m/z 222 (100, [M]+); mp >235 °C; 1H NMR [(C2H3)2SO)] Z-isomer δ 6.61 (s, 1H, H-9), 6.97 (d, 1H, H-5, J ) 7.0 Hz), 7.92 (d, 1H, H-6, J ) 7.0 Hz), E-isomer δ 5.89 (d, 1H, H-5, J ) 7.3 Hz), 6.29 (s, 1H, H-9), 7.69 (d, 1H, H-6, J ) 7.3 Hz); ratio of isomers (determined by integration of signals) Z:E ) 3:2; 13C NMR [(C2H3)2SO)] Z-isomer δ 67.1 (d, C-9), 94.9 (d, C-5), 148.0 (d, C-6), 154.9 (s, C-4), 162.8 (s, C-8), 166.9 (s, C-2), E-isomer δ 68.7 (d, C-9), 92.8 (d, C-5), 145.7 (d, C-6), 150.2 (s, C-4), 162.3 (s, C-8), 164.8 (s, C-2); FT-IR (KBr) ν 3150 cm-1 (NH), 1749 (CdO), 1660 (CdC), 1616 (NH), 801 (C-Cl). Anal. Calcd: C:H: N, 32.44:2.27:18.93. Found: C:H:N, 32.43:2.60:19.42. (C) N6-(Chloroacetyl)adenine. The fraction eluting at tR 18.8 min under HPLC conditions for preparative workup was collected and characterized: yield 8.5%; UV (H2O) λmax (pH 5) 208, 278 nm ( 10 710, 7120 mol-1 cm-1), λmax (pH 7) 217, 269 nm ( 16 425, 11 925 mol-1 cm-1), λmax (pH 8) 212, 268 nm ( 14 890, 8800 mol-1 cm-1); electrospray- MS m/z 210 (70, [M H]-), 135 (100, M - C2OCl); mp >300 °C; 1H NMR [(C2H3)2SO)] δ 4.57 (s, 2H, H-10), 8.48 (s, 1H, H-2), 8.66 (s, 1H, H-5); 13C NMR [(C2H ) SO)] δ 43.3 (t, C-10), 144.2 (s, C-7), 145.5 (d, 3 2 C-2), 151.2 (d, C-5), 166.2 (s, C-9); FT-IR (KBr) ν 3348 cm-1 (NH), 1700 (CdO), 1630 (CdC), 1572 (NH), 800 (C-Cl). Anal. Calcd: C:H:N, 39.71:2.86:33.10. Found: C:H:N, 39.44:2.86: 31.17. (D) N4-(Chloroacetyl)cytosine. The fraction eluting at tR 15.1 min under HPLC conditions for preparative workup was collected and characterized: yield 29.1%; UV (H2O) λmax (pH 5) 212, 242, 295 nm ( 14 310, 10 280, 4280 mol-1 cm-1), λmax (pH 7) 212, 242, 295 nm ( 12 230, 8620, 3570 mol-1 cm-1), λmax (pH 8) 212, 242, 295 nm ( 12 640, 8910, 3650 mol-1 cm-1); electrospray- MS m/z 186 (40, [M - H]-), 111 (100, M - C2OCl); mp >300 °C; 1H NMR [(C2H3)2SO)] Z-isomer δ 4.37 (s, 2H, H-9), 7.03 (d, 1H, H-5, J ) 7.0 Hz), 7.87 (d, 1H, H-6, J ) 7.0 Hz), E-isomer δ 4.27 (s, 2H, H-9), 5.82 (d, 1H, H-5, J ) 7.2 Hz), 7.60 (d, 1H, H-6, J ) 7.2 Hz); ratio of isomers (determined by integration of signals) Z:E ) 4:1; 13C NMR [(C2H3)2SO)] Z-isomer δ 43.7 (t, C-9), 94.4 (d, C-5), 147.7 (d, C-6), 155.9 (s, C-4), 162.8 (s, C-8), 167.0 (s, C-2), E-isomer δ 41.5 (t, C-9), 92.7 (d, C-5), 145.5 (d, C-6), 155.9 (s, C-4), 162.3 (s, C-8), 167.0 (s, C-2); FT-IR (KBr) ν 3121 cm-1 (NH), 1736 (CdO), 1656 (CdC), 1610 (NH), 814 (C-Cl). Anal. Calcd: C:H:N, 38.84:3.22:22.4. Found: C:H:N, 38.90:3.40:21.8.
Mu¨ ller et al. (E) N2-(Chloroacetyl)guanine. The fraction eluting at tR 19.8 min under HPLC conditions for preparative workup was collected and characterized: yield 4.6%; UV (H2O) λmax (pH 5) 198, 260 nm ( 12 745, 8900 mol-1 cm-1), λmax (pH 7) 209, 263 nm ( 14 960, 13 995 mol-1 cm-1), λmax (pH 8) 207, 265 nm ( 12 055, 9120 mol-1 cm-1); electrospray+ MS m/z 228 (100, [M + H]+); mp >300 °C; 1H NMR [(C2H3)2SO)] δ 4.43 (s, 2H, H-10), 8.08 (s, 1H, H-2); 13C NMR [(C2H3)2SO)] δ 42.9 (t, C-10), 139.9 (d, C-2), 146.2 (s, C-5), 153.7 (s, C-7), 168.7 (s, C-9); FT-IR (KBr) ν 3234 cm-1 (NH), 1658 (CdO), 1585 (NH), 804 (C-Cl). Anal. Calcd: C:H:N, 36.92:2.66:30.78. Found: C:H:N, 36.99:2.87: 30.06. (F) 3H-7,8-Dihydro-8-oxoimidazo[1,2r]purine (1,N6Acetyl Adenine). The fraction eluting at tR 10.8 min under HPLC conditions for preparative workup was collected and characterized: yield 18.9%; UV (H2O) λmax (pH 5) 221, 270, 303 nm ( 14 400, 6070, 9890 mol-1 cm-1), λmax (pH 7) 221, 270, 303 nm ( 11 340, 4760, 7780 mol-1 cm-1), λmax (pH 8) 221, 270, 303 nm ( 14 080, 5900, 9750 mol-1 cm-1); electrospray+ MS m/z 176 (100, [M + H]+), 136 (45, M - C2O); mp >300 °C; 1H NMR [(C2H3)2SO)] δ 4.64 (s, 2H, H-7), 8.41 (s, 1H, H-2), 8.60 (s, 1H, H-5); 13C NMR [(C2H3)2SO)] δ 51.1 (t, C-7), 143.0 (d, C-2), 151.5 (d, C-5), 166.8 (s, C-8); FT-IR (KBr) ν 3109 cm-1 (NH), 2798 (acidic H), 1623 (CdO/CdC, broad signal), 1528, 1509, 1488 (aromatic ring system), 1204 (CdO). Anal. Calcd: C:H:N, 47.99:2.88:40.00. Found: C:H:N, 47.63:2.96:39.66. (G) 1,2,4,5-Tetrahydro-2,5-dioxoimidazo-[1,2β]pyrimidine (3,N4-Acetylcytosine). The fraction eluting at tR 7.2 min under HPLC conditions for preparative workup was collected and characterized: yield 11.9%; UV (H2O) λmax (pH 5) 213, 301 nm ( 4750, 10 280 mol-1 cm-1), λmax (pH 7) 214, 301 nm ( 5820, 11 693 mol-1 cm-1), λmax (pH 8) 221, 317 nm ( 7980, 15 080 mol-1 cm-1); electrospray+ MS m/z 152 (90, [M + H]+), 174 (50, M + Na), 112 (100, M + H - C2O); mp >300 °C; 1H NMR [(C2H3)2SO)] δ 4.15 (s, 2H, H-4), 6.14 (d, 1H, H-8, J ) 7.1 Hz), 7.75 (d, 1H, H-9, J ) 7.1 Hz); 13C NMR [(C2H3)2SO)] δ 48.3 (t, C-4), 94.2 (d, C-8), 146.2 (d, C-9), 146.8 (s, C-7), 172.7 (s, C-2), 184.9 (s, C-5); FT-IR (KBr) ν 3047 cm-1 (NH), 1739 (CdO), 1600, 1567, 1545 (aromatic ring system). Anal. Calcd: C:H:N, 47.67: 3.34:27.81. Found: C:H:N, 47.56:3.10:27.40. (H) 3H-4,6,7,9-Tetrahydro-6-oxoimidazo[1,2r]purin-9one (1,N2-Acetylguanine). The fraction eluting at tR 11.6 min under HPLC conditions for preparative workup was collected and characterized: yield 64.5%; UV (H2O) λmax (pH 7) 220, 260, 286 (shd) nm ( 3000, 4000, 1700 mol-1 cm-1); electrospray- MS m/z 190 (41, [M - H]-), 150 (100, M - C2O); mp >300 °C; 1H NMR [(C2H3)2SO)] δ 4.45 (s, 2H, H-7), 8.03 (s, 1H, H-2); 13C NMR [(C2H3)2SO)] δ 48.2 (t, C-7), 134.9 (d, C-2), 136.4 (s, C-4a), 169.3 (s, C-6), 171,2 (s, C-9); FT-IR (KBr) ν 3441 cm-1 (NH), 1755 (CdO), 1728 (CdO), 1627 (CdC), 1570, 1545, 1524 (aromatic ring system). Anal. Calcd: C:H:N, 43.97:2.64:36.65. Found: C:H:N, 43.89:2.44:36.34. (I) 3H-5,6,8,9-Tetrahydro-6-oxoimidazo[1,2r]purin-9one (N2,3-Acetylguanine). The fraction eluting at tR 10.4 min under HPLC conditions for preparative workup was collected and characterized: yield 3.3%; UV (H2O) λmax (pH 7) 239, 293 nm ( 12 500, 3000 mol-1 cm-1); electrospray- MS m/z 190 (39, [M - H]-), 150 (100, M - C2O); mp >300 °C; 1H NMR [(C2H3)2SO)] δ 4.45 (s, 2H, H-5), 8.03 (s, 1H, H-2).
Results DNA Base Adduct Formation and Structure Elucidation. Thioketenes were generated in situ from their corresponding disulfide precursors as described previously (4). Haloketenes are easily accessible in solution by the method of Brady et al. (13), while haloacyl chlorides were obtained commercially. To determine if side products arising either from polymerization of the halogenated specimen or from the disulfide breakdown during the thioketene release may interfere with the
Synthesis of Ketene DNA Adducts
Chem. Res. Toxicol., Vol. 11, No. 5, 1998 457
Figure 1. (A) Formation of N4-(chloroacetyl)cytosine from chlorothioketene. The electronic spectrum refers to the product peak eluting at tR 12.0 min (Experimental Procedures, HPLC system i, analytical column). Electrospray ionization-mass spectrometry data reveal mass fragments m/z 187 (M+) and 209 (M + Na+), indicative of a sulfur versus oxygen exchange. (B) Formation of N4-(dichloroacetyl)cytosine from dichlorothioketene. The electronic spectrum refers to the product peak eluting at tR 17.6 min (Experimental Procedures, HPLC system i, analytical column). Thermospray ionization-mass spectrometry data show mass fragments m/z 151 (M+ - Cl2) and 222 (M+), indicative of a sulfur versus oxygen exchange.
analytical methods, control reactions of the appropriate reagents without the presence of DNA bases were performed and used for background subtraction. Subsequently adenine, cytosine, guanine, and thymine were reacted with chlorothioketenes, generated from the corresponding 2-nitrobenzyl disulfides, with chloroketenes generated by base-catalyzed HCl elimination from acyl chlorides, and with chloroacyl and dichloroacyl chloride, respectively. HPLC analysis did not detect products of reaction between ketenes, thioketenes, or acyl chloride and thymine. In contrast, adenine, cytosine, and guanine each yielded one major product with thioketene, chloroketene, and chloroacetyl chloride. HPLC analysis of the reactions of adenine and cytosine with dichlorothioketene, dichloroketene, and dichloroacetyl chloride showed the presence of one additional peak, while no products were found with guanine. Typical HPLC chromatograms obtained and electronic spectra of the products are shown in Figure 1. Compared to the nucleobases, the products eluted at later tR and the UV maxima were shifted to higher wavelengths or gave additional maxima in the case of reactions with cytosine. These observations suggest a modification of the nucleobases. The formed products were isolated and subjected to MS analysis. Identical mass spectra were obtained from the
haloketene and haloacyl chloride reactions, respectively, demonstrating a chloroacetyl or dichloroacetyl residue attached to the individual DNA constituents. Halothioacetyl adducts were expected for the halothioketene reactions; instead molecular ions with 16 mass units less than calculated and identical fragmentation patterns as the corresponding products of the haloketene reactions were found (Figure 1). These observations suggest that initially formed sulfur-containing products rapidly exchange sulfur with oxygen in aqueous solution. Further evidence for the sulfur/oxygen exchange of the adducts was obtained by rechromatographing a mixture of aliquots of (chloroacetyl)adenine, derived from either chlorothioketene, chloroketene, or chloroacyl chloride reactions. Only a single peak judged pure by recording of several UV spectra during elution, indicative of one homogeneous product, was detected. Similar experiments were carried out with (dichloroacetyl)adenine, (chloro- and dichloroacetyl)cytosine, and (chloroacetyl)guanine. The identical results obtained confirm the initial mass spectrometrical findings (Scheme 2). A comparison of product yields by HPLC analysis established the efficiency of the reaction of the haloketenes with nucleobases. Reaction of the nucleobases with haloacyl chlorides, due to their lower reactivity (13, 14), yielded only ∼10% of the respective products, when
458 Chem. Res. Toxicol., Vol. 11, No. 5, 1998
Mu¨ ller et al.
Scheme 2. Cytosine Adducts Formed from Haloacyl Chlorides, Haloketenes, and Halothioketenesa
a Haloacyl chlorides (A) were either reacted directly with cytosine or used as precursors for the in situ generation of haloketenes (B) with DABCO (1,4-diazabicyclo[2.2.2]octane) in the reaction with cytosine. Halothioketenes were formed from disulfide precursors (C) in vitro or may be formed from (chlorovinyl)-S-cysteines (D) by a cysteine S-conjugate β-lyase-mediated reaction in vivo. N4-(Dichloroacetyl)cytosine and N4-(chloroacetyl)cytosine exist as E/Z-isomers; N4-(chloroacetyl)cytosine can further cyclize to 3,N4-acetylcytosine.
compared to the haloketenes under identical conditions. The reaction of bases with halothioketenes, hampered by the slow release of halothioketenes from the disulfide precursors, also generated only ∼5% of the desired materials, although halothioketenes are generally considered to be more potent electrophiles than haloketenes (15). As larger amounts of the haloacetyl DNA adducts were required for further structure elucidation and since identical compounds were formed from the thioketenes and the ketenes, all subsequent preparative work was focused on the high-yielding haloketene reactions. Elemental analyses as well as 1H and 13C NMR studies were clearly showing an attachment of chloro- or dichloroacetyl moieties to the nucleobases, but the actual site of modification in the purine and pyrimidine ring could not be unambiguously determined. From the fact that only DNA bases with the exocyclic amino group gave rise to adducts, while thymine did not react under these conditions, a nucleophilic attack of the exocyclic amino groups of the DNA bases on the (thio)carbonyl moieties of the halogenated electrophiles, yielding (thio)amides, was adopted as a working hypothesis for the structure determination. This assumption is in line with precedence demonstrating (thio)amide formation with other nucleophilic amino groups (13-15). Nevertheless, other nucleophilic sites such as the N1 position in adenine, N3 in cytosine, or N7 in guanine could be modified. FT-IR proved to be the spectroscopic method of choice to anwer this question. Cytosine showed two strong bands at 3380
and 3172 cm-1, assigned to a free exocyclic amino group (Figure 2). These bands completely disappeared in the dichloroacetyl adduct; instead an additional carbonyl band could be detected at 1749 cm-1, which confirmed the hypothesis of the amide formation. Identical data were gathered for the other haloacetyl adducts thus establishing the mechanism of nucleophilic attack of the exocyclic nitrogen. Adduct structures were assigned to be N6-(chloroacetyl)adenine (1), N4-(chloroacetyl)cytosine (2), and N2(chloroacetyl)guanine (3) (Scheme 3) as well as N6-(dichloroacetyl)adenine (1) and N4-(dichloroacetyl)cytosine (2) (Scheme 4). Each of the cytosine derivatives showed two complete sets of signals in the 1H and 13C NMR spectra, indicative of E/Z-isomers, a well-known feature of imino compounds. The Z-isomers were considered to be the preferred isomeric forms, since the carbonyl moiety of their haloacetyl residues could form a hydrogen bond with the N3 imino group. Alternatively, electrostatic interaction of the chlorine atom with the imino hydrogen may occur. The formation of the E-isomers is thus less favored, and they should be present in the equilibrium in lower concentrations. With the haloacetyl adenine and guanine adducts, no isomers were detected. Cyclization of Chloroacetyl DNA Base Derivatives. Upon standing in aqueous solutions at pH 7 or 8, N6-(dichloroacetyl)adenine and N4-(dichloroacetyl)cytosine were completely hydrolyzed to adenine and cytosine, respectively. The chloroacetyl compounds, how-
Synthesis of Ketene DNA Adducts
Chem. Res. Toxicol., Vol. 11, No. 5, 1998 459 Scheme 4. DNA Base Adducts Formed from the Dichlorinated Electrophiles
Figure 2. Evidence for the carbamide mechanism. FT-IR spectra of N4-(dichloroacetyl)cytosine (top panel) and cytosine (lower panel). Note the disappearance of the free amino group of cytosine (bands at 3380 and 3172 cm-1) and the new carbonyl band at 1749 cm-1 in the DNA base adduct.
ever, while hydrolysis was by far the dominant reaction, also yielded new products: one derived from N6-(chloroacetyl)adenine and N4-(chloroacetyl)cytosine, respectively, and two from N2-(chloroacetyl)guanine. All po-
tential adducts eluted after the parent DNA base but ahead of their chloroacetyl precursors. The electronic spectra of those compounds in general showed a shift to higher wavelengths and increased extinction coefficients for the individual maxima as compared to their starting materials (the haloacetyl compounds mentioned above). Mass spectrometrical analysis gave spectra consistent of structures having lost HCl from the starting material. This reaction could occur via a nucleophilic attack of the ring imino nitrogen on the chlorine-substituted carbon of the haloacetyl residue and subsequent ring closure to a new five-membered ring. Elemental analysis, 1H NMR, 13C NMR, and FT-IR substantiated the assignments of the following structures to those compounds: 1,N6-acetyladenine (4), 3,N4-acetylcytosine (5), 1,N2-acetylguanine (6), and N2,3-acetylguanine (7) (Scheme 3). In the case of the guanine adducts, distinction between the 1,N2 and N2,3 derivative was
Scheme 3. DNA Base Adducts Formed from the Chlorinated Electrophiles
460 Chem. Res. Toxicol., Vol. 11, No. 5, 1998
Mu¨ ller et al.
Figure 3. Thermospray ionization mass spectrum of the nucleoside adduct N4-(dichloroacetyl)deoxycytidine. The fragmentation pattern is consistent with the loss of the deoxyribose moiety (m/z 116) from mass fragment m/z 338 (M+), to yield the mass fragment m/z 222 (M+ of the DNA base adduct). This in turn loses two chlorine atoms to give the mass fragment m/z 151. Table 1. Electronic and Mass Spectral Data of the Haloacetyl Nucleoside Adductsa m/z (relative abundance) nucleoside N6-(chloroacetyl)deoxyadenosine N6-(dichloroacetyl)deoxyadenosine N4-(chloroacetyl)deoxycytidine N4-(dichloroacetyl)deoxycytidine N2-(chloroacetyl)deoxyguanosine a
UV, pH 5 (nm) λmax
λmin
nucleoside
haloacyl ester
haloacyl diester
291 (5) [M+ - HCl] 362 (30) [M+] 267 (5) [M+ - HCl] 338 (32) [M+] 307 (5) [M+ - HCl]
367 (5)
443 (10)
210, 275
234
472 (18)
582 (8)
215, 275
234
343 (5)
419 (5)
215, 247, 300
227, 271
448 (20)
558 (10)
215, 247, 303
231, 274
383 (5)
459 (9)
195, 255, 279 (shd)
223
UV spectra were recorded in water, pH 5.
based on product yields: the formation of the linear 1,N2acetylguanine should be thermodynamically favored over the angular N2,3-acetylguanine, resulting in higher yield of the first (16). More compelling evidence was gathered from the electronic spectra and fluorescence studies (see below). For the haloacetyl as well as for the cyclic DNA base adducts, keto/enol tautomers are possible. pH-Dependent electronic spectra of 3,N4-acetylcytosine acquired at pH 7 and 8 demonstrated a shift of the maximum from 301 to 317 nm and an increase of the molar extinction coefficient from 11 693 to 15 080 mol-1 cm-1, which could be explained by the formation of enolate ions at higher pH. 1,N6-Acetyladenine showed a similar effect at pH 10. Determination of the pKa and the keto/enol equilibrium of the compounds at different pH were not performed. Modification of Nucleosides by Haloketenes. To address the question of adduct formation on the nucleoside level, dA, dC, and dG were reacted with dichloroand chloroketene. Analogous to the DNA bases, dA and dC gave products with dichloroketene, while dA, dC, and dG reacted with chloroketene. HPLC analysis of each reaction mixture provided three distinct peaks with identical electronic spectra. The obtained spectra were
also identical to those of the corresponding DNA base adduct. As no change in the UV spectrum was observed, a second adduction at the base moiety of the molecule could be ruled out. Thus, it was assumed that a reaction at the deoxyribose had occurred, rendering the molecule more hydrophobic. The individual peaks were purified and subjected to mass spectrometrical analysis. The data acquired from the first eluting adduct peaks of each reaction showed the expected MH+ of a haloacylated nucleoside. Fragment patterns were consistent with the loss of the sugar moiety to yield the adducted base (Figure 3), which in the case of the monochloroacetyl adducts lost HCl to yield the ring-closed forms. The second and third peaks, respectively, had a second and third haloacetyl group in the molecule (Table 1). Upon the loss of mass fragments m/z 226 and 336, indicative of a mono- or dihaloacetylated deoxyribose moiety, the mass fragments of the haloacetylated DNA base adduct or its ring-closed derivative, formed through the harsh conditions of the thermospray, could be detected. This finding supports the haloacylation of the sugar portion by ester formation with the two free hydroxyl groups. N4-(Chloroacetyl)deoxycytidine, upon standing in aqueous solutions at pH 8, was mainly hydrolyzed to dC. Nevertheless it also formed a cyclic product as already
Synthesis of Ketene DNA Adducts
Chem. Res. Toxicol., Vol. 11, No. 5, 1998 461
Figure 4. Stability of N6-(chloroacetyl)adenine (A), N4-(chloroacetyl)cytosine (B), N6-(dichloroacetyl)adenine (C), and N4(dichloroacetyl)cytosine (D) in water, pH 5 (0), pH 7 (]), pH 8 (O), and in 0.1 M sodium phosphate buffer, pH 7.4 (∆), determined by measuring the UV absorption at 300 nm over 180 min. The following half-lives were calculated: N6-(chloroacetyl)adenine stable over 3 h (pH 5) and 20.8 (pH 7), 11.9 (pH 8), and 299 (pH 7.4) min; N4-(chloroacetyl)cytosine stable over 3 h (pH 5) and 44.3 (pH 7), 19.7 (pH 8), and 330 (pH 7.4) min; N6-(dichloroacetyl)adenine stable over 618 (pH 5), 45.9 (pH 7), 18.8 (pH 8), and 54.1 (pH 7.4) min; N4-(dichloroacetyl)cytosine stable over 135 (pH 5), 9.4 (pH 7), 8.2 (pH 8), and 48.1 (pH 7.4) min.
suggested by the thermal loss of HCl under thermospray conditions and now confirmed by HPLC analysis with comparison of the electronic spectrum to the analogous DNA base adduct. As the chemistry of the nucleoside adducts seemed to perfectly resemble that of their corresponding DNA bases, no further efforts were directed to a detailed characterization of these compounds. Chemical Stability of DNA Base Adducts. To be of biological significance for mutagenesis and carcinogenesis, DNA adducts are required to have a sufficient chemical stability. Therefore all DNA base adducts were investigated for their hydrolytic stability at 25 °C in water, pH 5, 7, 8, and 0.1 M sodium phosphate buffer, pH 7.4, to mimic physiological conditions. All haloacetyl adducts were stable at slightly acidic pH. Hydrolysis and half-lives of the adenine and cytosine derivatives were determined (Figure 4). The buffered solution showed a stabilizing effect over the pH-adjusted water. In general the dichloroacylated compounds were hydrolyzed faster than their monochlorinated counterparts. N6-(Chloroacetyl)adenine and N4-(chloroacetyl)cytosine appear to gain additional stability through hydrogen bonding to
water molecules by the hydrogen atom in their acetyl residue. Similar data for N2-(chloroacetyl)guanine could not be obtained, as it was immediately hydrolyzed or converted to the cyclic adducts at pHs g 7. The cyclic adducts proved to be highly stable under all investigated conditions. Upon standing in solution at room temperature for a week, each sample still contained more than 50% of the initial amount of the respective adduct as detected by HPLC analysis. Fluorescence Studies of the DNA Base Adducts. The enol tautomers of the cyclic DNA adducts display structural similarities to the well-known etheno () adducts of vinyl chloride, which are highly fluorescent (16). Therefore, 0.1 M solutions of the haloacetyl adducts and their cyclic products in water, pH 5, were tested for fluorescence, using λex 245 and 300 nm, as suggested by Barrio et al. (10), for the adducts. While no emission spectra could be recorded at λex 245 nm, spectra were acquired at λex 300 nm for N6-(chloroacetyl)adenine (λem 6 4 max 347), N -(dichloroacetyl)adenine (λem max 367), N 4 (chloroacetyl)cytosine (λem max 354), and N -(dichloroacetyl)cytosine (λem max 363). N2-(Chloroacetyl)guanine was
462 Chem. Res. Toxicol., Vol. 11, No. 5, 1998
Mu¨ ller et al.
Figure 5. Evidence for the fluorescence of the N2,3-acetylguanine. A mixture of 100 pmol of N2,3-acetylguanine and 100 pmol of 1,N2-acetylguanine was analyzed using a HPLC method developed by Bedell et al. (12) and diode array (B) and fluorescence detection (A). N2,3-Acetylguanine eluting at tR 10.4 min was highly fluorescent as compared to the 1,N2-acetylguanine (tR 11.6 min). UV spectrum C (pH 5) refers to N2,3-acetylguanine, and UV spectrum D (pH 5) refers to 1,N2-acetylguanine; both are consistent with the spectra of the -guanine adducts (16).
not fluorescent. As predicted 1,N6-acetyladenine and 3,N4-acetylcytosine showed an increase in fluorescence intensity at λex 300 nm compared to their haloacetyl precursors, most likely due to the addition of a fivemembered ring to the heteroaromatic ring system. Their λem max values were 342 and 340 nm, respectively. HPLC with fluorescence detection was evaluated as a potential detection system for the cyclic adducts in in vitro experiments. Using a pH gradient system developed by Bedell et al. (12) with λex 232 nm and λem 356, we were able to detect 1 pmol of 3,N4-acetylcytosine and 10 pmol of 1,N6-acetyladenine on the column. In addition the fluorescence of a mixture of 100 pmol of 1,N2acetylguanine and 100 pmol of N2,3-acetylguanine was investigated. While the angular N2,3-acetylguanine was highly fluorescent, the planar 1,N2-acetylguanine was not (Figure 5). This result and the recorded UV spectra were in excellent agreement with previous reports by Sattsangi et al. (16) on spectral properties of 1,N2--guanine and N2,3--guanine.
Discussion The DNA-alkylating potential of halothioketenes, haloketenes, and haloacyl chlorides, known metabolites of tri- and tetrachloroethene bioactivation, was investigated. As clearly demonstrated in the model reactions with nucleic acids and nucleosides, haloketenes, halothioketenes, and haloacyl chlorides react with the exocyclic amino group of adenine, cytosine, and guanine to give (thio)amides. (Thio)amide formation is a well-described reaction for these ketenes and thioketenes, often used to trap these highly reactive molecules (13-15). The observed exchange of sulfur versus oxygen in the thioamides has been previously demonstrated by Edward and
Wong (17) and appears to be an acid-catalyzed as well as a base-catalyzed reaction. N4-(Chloroacetyl)- and N4-(dichloroacetyl)cytosine have been synthesized from chloro- and dichloroacetic anhydride by Goody and Walker (18, 19). The analytical and stability data of the DNA adducts presented by these authors are consistent with our results. In addition we were able to extend their previous characterization. Similar to our findings they demonstrated the cyclization of N4-(chloroacetyl)cytosine to the stable 3,N4-acetylcytosine. When cytidine was reacted with chloroacetic anhydride, N4-(chloroacetyl)-2′,3′,5′-tris-O-(chloroacetyl)cytidine was recovered. This confirms our results of ester formation at the sugar moiety of the various nucleosides. Chheda and Hall (20-22) studied the basic chemistry of the labile N6-(R-aminoacyl)adenines, which had been found in enzymatic hydrolysates of yeast s-RNA. 1,N6Acetyladenine was identified as a stable degradation product of these compounds. Their identification was based on a comparison to 1,N6-acetyladenine, obtained from the alkaline hydrolysis of N6-(chloroacetyl)adenine. The analytical and stability data reported by the authors is in good agreement with our data on the identical compounds. 2-Chloroketene diethyl acetal was reacted with tri-Oacetyladenosine and tri-O-acetylcytidine, yielding 8-ethoxysubstituted -DNA adducts as reported by Leonard and Cruickshank (23). These compounds are the ethyl ethers of the enol tautomers of the DNA adducts described herein. The authors emphasize the extraordinary fluorescence of the modified nucleosides, which is confirmed by our findings. Moreover, 2-chloroketene diethyl acetal was shown to be a base-substitution mutagen after metabolic activation in the Ames test (23).
Synthesis of Ketene DNA Adducts
The literature cited clearly supports the DNA adduct structures of the reaction products formed from the investigated electrophiles and DNA bases. Their chemical properties, especially the stability of the cyclic compounds, and their close structural relationship to the known mutagenic -DNA adducts (24) suggest a genotoxic potential. The positive finding in the Ames test with 2-chloroketene diethyl acetal, a direct analogue of chloroketene, is the first indication for this potential in a biological system (23).
Acknowledgment. This work was supported by the Deutsche Forschungsgemeinschaft (SFB-172, A-3) and the Bundesministerium fu¨r Forschung und Technologie. The authors thank Dr. C. Richling for the synthesis of DCVDS, Dr. M. Sander (Anorganisch-chemisches Institut der Universita¨t Heidelberg) for the FT-IR analysis, and Mrs. J. Colberg for the mass spectrometrical analysis.
References (1) Henschler, D. (1994) Toxizita¨t chlororganischer Verbindungen: Einfluβ der Einfu¨hrung von Chlor in organische Moleku¨le. (Toxicity of chloro-organic compounds: the effect of a chloro-atom to organic molecules.) Angew. Chem. 106, 1997-2012. (2) Kanhai, W., Dekant, W., and Henschler, D. (1989) Metabolism of the nephrotoxin dichloroacetylene by glutathione conjugation. Chem. Res. Toxicol. 2, 51-56. (3) Kanhai, W., Koob, M., Dekant, W., and Henschler, D. (1991) Metabolism of 14C-dichloroethyne in rats. Xenobiotica 21, 905916. (4) Dekant, W., Urban, G., Go¨rsman, C., and Anders, M. W. (1991) Thioketene formation from R-haloalkenyl 2-nitrophenyl disulfides: models for biological reactive intermediates of cytotoxic S-conjugates. J. Am. Chem. Soc. 113, 5120-5122. (5) Bonse, G., and Henschler, D. (1976) Chemical reactivity, biotransformation, and toxicity of polychlorinated aliphatic compounds. Crit. Rev. Toxicol. 4, 395-409. (6) Miller, R. E., and Guengerich, F. P. (1982) Oxidation of trichloroethylene by liver microsomal cytochrome P-450: evidence for chlorine migration in a transition state not involving trichloroethylene oxide. Biochemistry 21, 1090-1097. (7) Birner, G., Richling, C., Henschler, D., Anders, M. W., and Dekant, W. (1994) Metabolism of tetrachloroethene in rats: identification of N-(dichloroacetyl)-L-lysine and N-(trichloroacetyl)-L-lysine as protein adducts. Chem. Res. Toxicol. 7, 724732. (8) Dekant, W., Anders, M. W., and Monks, J. (1993) Bioactivation of halogenated xenobiotics by S-conjugate formation. In Renal Disposition and Nephrotoxicity of Xenobiotics (Anders, M. W., Dekant, W., Henschler, D., Oberleithner, H., and Silbernagl, S., Eds.) pp 187-215, Academic Press, San Diego.
Chem. Res. Toxicol., Vol. 11, No. 5, 1998 463 (9) Dekant, W., and Henschler, D. (1983) New pathways of trichloroethylene metabolism. In Developments in the Science and Practice of Toxicology (Hayes, A. W., Schnell, R. C., and Miya, T. S., Eds.) pp 399-402, Elsevier Science Publishers, Amsterdam. (10) Barrio, J. R., Secrist, J. A. I., and Leonard, N. J. (1972) Fluorescent adenosine and cytidine derivatives. Biochem. Biophys. Res. Commun. 46, 597-604. (11) Kalinowski, H.-O., Berger, S., and Braun, S. (1984) Nucleoside, Nucleotide, zugeho¨rige Basen und verwandte Heterozyklen. 13C NMR-Spektroskopie, p 399, Georg Thieme Verlag, Stuttgart, New York. (12) Bedell, M. A., Dyroff, M. C., Doerjer, G., and Swenberg, J. A. (1986) Quantitation of etheno adducts by fluorescence detection. IARC Sci. Pub. 70, 425-434. (13) Brady, W. T. (1971) Halogenated ketenes: valuable intermediates in organic synthesis. Synthesis 415-422. (14) Bormann, D. (1968) Herstellung und Umwandlung von Ketenen. In Houben Weyl, Methoden der Organischen Chemie, 4th ed. (Mu¨ller, Eu., Ed.) VII/4, p 65, Georg Thieme Verlag, Stuttgart, New York. (15) Schaumann, E. (1988) The chemistry of thioketenes - Tetrahedron Report No. 231. Tetrahedron 44, 1827-1871. (16) Sattsangi, P. D., Leonard, N. J., and Frihart, C. R. (1977) 1,N2Ethenoguanine and N2,3-ethenoguanine. Synthesis and comparison of the electronic spectral properties of these linear and angular triheterocycles related to the Y bases. J. Org. Chem. 42, 3292-3296. (17) Edward, J. T., and Wong, S. C. (1979) Effect of acid concentration on the partitioning of the tetrahedral intermediate in the hydrolysis of thioacetanilide. J. Org. Chem. 101, 1807-1809. (18) Goody, R. S., and Walker, R. T. (1966) Same N6 acylated cytosines. Tetrahedron Lett. 4, 289-291. (19) Goody, R. S., and Walker, R. T. (1970) The preparation and properties of some cytosine derivatives. J. Org. Chem. 36, 727730. (20) Chheda, G. B., and Hall, R. H. (1968) Aminoacyl nucleosides. V. The mechanism of the rearrangement of N6-(R-aminoacyl)adenines into N-(6-purinyl)amino acids. J. Org. Chem. 34, 34983502. (21) Chheda, G. B., and Hall, R. H. (1966) Aminoacyl nucleosides. III. A novel rearrangement: conversion of N6-(R-aminoacyl)adenines into N-(6-purinyl)amino acids. Biochemistry 5, 2082-2091. (22) Chheda, G. B., and Hall, R. H. (1968) Aminoacyl nucleosides. IV. The rearrangement of N6-(R-aminoacyl)adenines to N-(6-purinyl)amino acids. Application to adenosine analogue. J. Org. Chem. 34, 2082-2091. (23) Leonard, N. J., and Cruickshank, K. A. (1986) Substituted ethenoadenosines and ethenocytidines. IARC Sci. Pub. 70, 3336. (24) Bartsch, H., Barbin, A., Marion, M.-J., Nair, J., and Guichard, Y. (1994) Formation, detection, and role in carcinogenesis of ethenobases in DNA. Drug Metab. Rev. 26, 349-372.
TX9701438
