Synthesis of Depurinating DNA Adducts Formed by One-Electron

Chem. Res. Toxicol. 1997, 10, 225-233

225

Synthesis of Depurinating DNA Adducts Formed by One-Electron Oxidation of 7H-Dibenzo[c,g]carbazole and Identification of These Adducts after Activation with Rat Liver Microsomes Liang Chen,† Prabu D. Devanesan,† Jaeman Byun,‡ Jonathan K. Gooden,‡ Michael L. Gross,‡ Eleanor G. Rogan,† and Ercole L. Cavalieri*,† Eppley Institute for Research in Cancer, University of Nebraska Medical Center, 600 South 42nd Street, Omaha, Nebraska 68198-6805, and Department of Chemistry, Washington University, One Brookings Drive, St. Louis, Missouri 63130 Received August 19, 1996X

It is hypothesized that 7H-dibenzo[c,g]carbazole (DBC) is metabolically activated by oneelectron oxidation in accordance with its propensity to be easily oxidized to its radical cation. Iodine oxidation of DBC produces a radical cation that subsequently binds to nucleophilic groups of dG or Ade. Oxidation of DBC in the presence of dG produces three adducts: DBC-5-N7Gua, DBC-6-N7Gua, and DBC-6-C8Gua, whereas in the presence of Ade, four adducts are obtained: DBC-5-N7Ade, DBC-5-N3Ade, DBC-5-N1Ade, and DBC-6-N3Ade. Formation of these adducts demonstrates that the DBC radical cation reacts at C-5 or C-6 with the reactive nucleophiles N-7 and C-8 of dG and N-7, N-3, and N-1 of Ade. Formation of DNA adducts by DBC was studied by using horseradish peroxidase or 3-methylcholanthrene-induced rat liver microsomes for activation. Identification of the biologically-formed depurinating adducts was achieved by comparison of their retention times on HPLC in two different solvent systems and by matrix-assisted laser desorption ionization (MALDI) mass spectrometry. Quantitation of the adducts formed by rat liver microsomes shows that 96% are depurinating adducts, DBC5-N7Gua (11%), DBC-6-N7Gua (32%), and DBC-5-N7Ade (53%), and 4% are unidentified stable adducts. Activation of DBC by horseradish peroxidase affords 32% stable unidentified adducts and 68% depurinating adducts: 19% DBC-5-N7Gua, 13% DBC-6-N7Gua, 27% DBC-5-N7Ade, and 9% DBC-5-N3Ade. Thus, activation of DBC by cytochrome P450 predominantly forms depurinating adducts by one-electron oxidation.

Introduction Elucidation of the mechanisms of metabolic activation of polycyclic aromatic hydrocarbons (PAH)1 is essential for understanding their mechanisms of tumor initiation. Two major pathways of activation are involved in the formation of the ultimate carcinogenic metabolites of PAH: one-electron oxidation to form intermediate radical cations and monooxygenation to produce bay-region diol epoxides (1-4). The predominant mechanism of activation for some potent carcinogenic PAH appears to be oneelectron oxidation (5-11). Activation of PAH by formation of radical cations can be predicted on the basis of two major factors: a low oxidation potential that allows easy removal of one electron by cytochrome P450 or peroxidases (1, 2, 12) and sufficient charge localization that allows efficient and specific reaction of the radical cations with the nucleo* To whom correspondence should be addressed. † Eppley Institute for Research in Cancer. ‡ Washington University. X Abstract published in Advance ACS Abstracts, January 15, 1997. 1 Abbreviations: BCC, 4-(benzyloxy)-R-cyanocinnamic acid; BP, benzo[a]pyrene; CA, collisional activation; CAD, collisional activation decomposition(s); COSY, two-dimensional chemical shift correlation spectroscopy; DBA, dibenz[a,j]acridine; DB[a,l]P, dibenzo[a,l]pyrene; DBC, 7H-dibenzo[c,g]carbazole; DMF, dimethylformamide; FAB MS/ MS, fast atom bombardment tandem mass spectrometry; FWHM, full width at half-maximum; Gly/TFA, glycerol/1% trifluoroacetic acid; HRP, horseradish peroxidase; MALDI-TOF, matrix-assisted laser desorption ionization time-of-flight; 3-MC, 3-methylcholanthrene; PAH, polycyclic aromatic hydrocarbon(s); PDA, photodiode array; PSD, post source decay(s).

S0893-228x(96)00149-X CCC: $14.00

philic groups of DNA (1, 2). These rules have been applied for predicting the occurrence of this mechanism for alternate PAH (condensed aromatic rings containing an even number of carbon atoms), but the same rules are not applicable to nonalternate PAH (condensed aromatic rings containing one or more rings with an odd number of carbon atoms, e.g., cyclopenta[c,d]pyrene) or heterocyclic PAH. 7H-Dibenzo[c,g]carbazole (DBC), an N-heterocyclic PAH, is a potent environmental carcinogen that has been shown to induce tumors in dogs, rats, hamsters, and mice (13, 14). Both local and systemic induction of tumors has been observed in mouse skin, liver, and lung (14-16). DBC has a relatively low oxidation potential (anodic peak potential ) 1.07 eV), that is similar to that of 3-methylcholanthrene (3-MC, 1.08 eV) (12) and lower than that of benzo[a]pyrene (BP, 1.12 eV) (12). Thus, DBC can be easily oxidized by both cytochrome P450 and peroxidases. The low value of the oxidation potential is necessary, although not sufficient, for metabolic activation by one-electron oxidation. To assess the importance of this mechanism for DBC, this compound was oxidized by iodine in the presence of dG or Ade. Investigation of the adducts formed by one-electron oxidation of DBC serves several purposes: (1) to determine the specific reactivity of one or more positions in the DBC radical cation, (2) to determine the nucleophilic groups in the nucleic acid bases that participate in adduct formation, (3) to provide evidence for the mechanism of adduct © 1997 American Chemical Society

226 Chem. Res. Toxicol., Vol. 10, No. 2, 1997

Chen et al.

formation, and (4) to provide standards for identifying and quantifying these adducts formed in biological systems. This article reports the synthesis of adducts obtained by reaction of the DBC radical cation with dG and Ade, and the identification and quantitation of adducts formed in reactions catalyzed by 3-MC-induced rat liver microsomes or by horseradish peroxidase (HRP) in the presence of DNA.

Table 1. HPLC Separation of Standards Used to Identify Biologically-Formed DBC-DNA Adducts

Experimental Section

a DBC-5-N1Ade was found to behave differently from other adducts, eluting as a broad peak with no definite retention time in both CH3CN/H2O and CH3OH/H2O gradients. It was isolated at a consistent retention time by eluting with 65% C2H5OH/CH3CN (3:1) in H2O for 15 min, followed by a linear gradient to 100% C2H5OH/CH3CN (3:1) in 40 min at a flow rate of 1 mL/min. b Gradient A: 40% CH CN in H O for 5 min, followed by a linear 3 2 (CV6) gradient to 100% CH3CN in 65 min at a flow rate of 1.0 mL/min. c Gradient B: 40% CH3OH in H2O for 5 min, followed by a convex (CV5) gradient to 100% CH3OH in 65 min at a flow rate of 1.0 mL/min.

Caution: DBC is hazardous and should be handled carefully in accordance with NIH guidelines (17). (A) General Procedures. (1) UV. UV absorbance spectra were recorded with a Waters 996 photodiode array (PDA) detector during HPLC using a CH3OH/H2O or CH3CN/H2O gradient. (2) NMR. Proton and homonuclear two-dimensional chemical shift correlation spectroscopy (COSY) NMR spectra were recorded on a Varian Unity 500 at 499.835 MHz in deuterated dimethylsulfoxide (Me2SO-d6) at 25 °C. Chemical shifts (δ) are reported relative to Me2SO (2.49 ppm), and the coupling constants (J) are given in Hz. (3) (a) Fast Atom Bombardment Tandem Mass Spectrometry (FAB MS/MS). Collisionally activated decomposition (CAD) spectra were obtained by using a VG ZAB-T (Manchester, U.K.), a four-sector tandem mass spectrometer of BEBE design (18). MS1 is a standard high-resolving power, double-focusing mass spectrometer (ZAB) of reverse geometry. MS2 has a prototype Mattauch-Herzog-type design, incorporating a standard magnet and a planar electrostatic analyzer having an inhomogeneous electric field, a single-point and an array detector. Samples were dissolved in 10 µL of Me2SO, and a 1-µL aliquot was loaded on the probe along with 1 µL of the matrix, a mixture of glycerol, and 1% trifluoroacetic acid (Gly/TFA) for fast atom bombardment (FAB) analysis. A Cs+ ion gun operated at 30 keV was used to desorb the ions, which were accelerated to 8 kV. CAD spectra were obtained after precursor ion activation in the third field-free region (between MS1 and MS2). Helium was added to the collision cell (floated at 4 kV) to attenuate the ion beam by 50%. MS1 was operated at a resolving power of 1000; resolving power of MS2 was set to 1000, full width at halfmaximum (FWHM) definition. Ten to fifteen 15-s scans were signal-averaged for each spectrum. Data acquisition and workup were carried out with a DEC Alpha 3000 workstation equipped with OPUS V 3.1X software and interfaced with the mass spectrometer via a VG SIOS I unit. Exact mass measurements were conducted with a Kratos MS50 triple analyzer tandem mass spectrometer equipped with a standard Kratos FAB source (19). The atom beam was 6-7 keV argon atoms at a total current of 1 mA at the cathode of the gun. A mixture of CsI and glycerol was used to produce reference-mass ions for the peak match mode. (b) Matrix-Assisted Laser Desorption Ionization Timeof-Flight (MALDI-TOF) MS. The MALDI-TOF experiments were carried out on a PerSeptive Biosystems, Inc., Voyager RP mass spectrometer (Cambridge, MA). A nitrogen laser (337nm, 20-kW peak laser power, 2-ns pulse width) was used to desorb the sample ions. The instrument was operated in the reflectron mode with an accelerating potential of 25 kV. Post source decay (PSD) (20) experiments were carried out under the same conditions. The synthesis of the MALDI matrix, 4-(benzyloxy)-R-cyanocinnamic acid (BCC), was described elsewhere (21). A solution of 0.25 M d-fucose in 1% TFA and 50% CH3CN/H2O was saturated with BCC, mixed with the analyte in a 1:1 ratio on the sample plate, and allowed to air-dry. MALDITOF spectra were summed from 40-60 laser pulses. PSD spectra were summed from 50-200 laser pulses. Raw data were acquired with a Tektronix 520A digitizing oscilloscope before being transferred to a PC equipped with GRAMS/386 software (Galactic Industries).

retention time, min adducta

Ab

Bc

DBC-5-N7Gua DBC-5-N7Ade DBC-6-N7Gua DBC-6-C8Gua DBC-5-N3Ade DBC-6-N3Ade

24.3 29.4 30.5 34.8 36.8 46.1

42.6 43.8 40.1 44.7 46.1 47.1

(4) HPLC. Analytical HPLC was conducted on a Waters 600E (Milford, MA) solvent delivery system equipped with a Waters 700 WISP autoinjector. Effluents were monitored for UV absorbance (254 nm) with a Waters 996 PDA detector, and the data were collected on a NEC Powermate 486/33i computer. Runs were conducted on a YMC (YMC, Wilmington, NC) ODSAQ 5 µm column (6.0 × 250 mm). For HPLC separation of the adducts obtained by iodine oxidation of DBC, the column was eluted with 30% CH3CN in H2O for 5 min, followed by a linear gradient to 100% CH3CN in 75 min. Preparative HPLC was conducted on a Waters 600E solvent delivery system coupled with a Waters 996 PDA detector. Runs were conducted on a YMC ODS-AQ 5 µm column (20 × 250 mm) at a flow rate of 6 mL/min. Separation of adducts obtained from biological reactions was carried out with two different solvent systems: 40% CH3CN in H2O for 5 min, followed by a linear (CV6) gradient to 100% CH3CN in 65 min, or 40% CH3OH in H2O for 5 min, followed by a convex (CV5) gradient to 100% CH3OH in 65 min at 1 mL/min. When enzymatically-formed adducts were analyzed by HPLC, the effluent was also passed through a Jasco FP-920 fluorescence detector. The details of column elution conditions for each adduct are listed in Table 1. (5) Materials. DBC was synthesized according to a published procedure (22) and purified prior to use by chromatography on a silica gel column (benzene/cyclohexane, 1:1) and by recrystallization from benzene/hexane (mp 154 °C). It was >99% pure by HPLC analysis. Ade and dG (Aldrich, Milwaukee, WI) were desiccated over P2O5 under vacuum at 110 °C for 48 h prior to use. Anhydrous dimethylformamide (DMF, Aldrich) and HPLC grade organic solvents (EM Science, Gibbstown, NJ) were used. Iodine was purchased from Aldrich and was used as received. (B) One-Electron Oxidation of DBC in the Presence of dG or Ade. To a solution of DBC (1 mmol) and dG (10 mmol) or Ade (5 mmol) in dry DMF (15 mL), iodine (3 mmol) in 1 mL of dry DMF was added dropwise, and the mixture was heated to 50 °C for 4 h under nitrogen. The mixture was then cooled to room temperature, and aqueous sodium thiosulfate was added until the iodine color disappeared. The solvent was evaporated and the residue was extracted three times with a solvent mixture of ethanol/chloroform/acetone (2:1:1). The resulting extract was filtered, and the combined solvent mixture was evaporated under vacuum. The residue was dissolved in Me2SO (4 mL), passed through a 0.45-µm filter, and analyzed by HPLC by using the CH3CN/H2O gradient (see above). Purification of the adducts was conducted by preparative HPLC. The products isolated from the reaction of DBC with dG were DBC-5-N7Gua (0.9%), DBC-6-N7Gua (2.8%), and DBC-6-C8Gua (0.2%) (Scheme 1). The reaction between DBC and Ade yielded four products: DBC-5-N7Ade (1.5%), DBC-5-N3Ade (3.6%), DBC-5-N1Ade (18%), and DBC-6-N3Ade (1.4%) (Scheme 1).

DBC Adducts Formed by One-Electron Oxidation

Chem. Res. Toxicol., Vol. 10, No. 2, 1997 227

Scheme 1. Iodine Oxidation of DBC in the Presence of dG or Ade

7H-Dibenzo[c,g]carbazole (DBC): UV, λmax (nm) 245, 268, 280, 301, 334, 348, 366; 1H NMR, δ 7.49 (dd, 2H, 3-H,11-H), 7.69 (dd, 2H, 2-H,12-H), 7.81 (d, 2H, 6-H,8-H), 7.91 (d, 2H,5H,9-H), 8.08 (d, 2H, 4-H,10-H), 9.04 (d, 2H, 1-H,13-H), 12.32 (s, 1H, 7-H), J1,2 ) 8.5 Hz, J2,3 ) 7.5 Hz, J3,4 ) 7.5 Hz, J5,6 ) 8.5 Hz. DBC-5-N7Gua: UV, λmax (nm) 243, 279, 303, 337, 353, 371; 1H NMR, δ 6.19 [s, 2H, 2-NH (Gua)], 7.38 (d, 1H, 4-H), 7.49 2 (dd, 1H, 3-H), 7.54 (dd, 1H, 11-H), 7.73 (dd, 1H, 12-H), 7.76 (dd, 1H, 2-H), 7.87 (d, 1H, 8-H), 7.96 (s, 1H, 6-H), 7.98 (d, 1H, 9-H), 8.12 (d, 1H, 10-H), 8.25 [s, 1H, 8-H (Gua)], 9.07 (d, 1H, 13-H), 9.16 (d, 1H, 1-H), 10.80 [s, 1H, 1-NH (Gua)], 12.54 (s, 1H, 7-H), J1,2 ) 8.0 Hz, J2,3 ) 7.5 Hz, J3,4 ) 7.5 Hz, J8,9 ) 8.5 Hz, J10,11 ) 8.5 Hz, J11,12 ) 7.5 Hz, J12,13 ) 7.0 Hz. MS: [M + 1]+ C25H17ON6, calcd 417.1464, found 417.1465. DBC-6-N7Gua: UV, λmax (nm) 243, 279, 303, 333, 348, 366; 1H NMR, δ 6.22 [s, 2H, 2-NH (Gua)], 7.53 (dd, 1H, 11-H), 7.58 2 (dd, 1H, 3-H), 7.71 (dd, 1H, 12-H), 7.75 (d, 1H, 8-H), 7.78 (dd, 1H, 2-H), 7.93 (d, 1H, 9-H), 8.04 (s, 1H, 5-H), 8.10 (d, 1H, 10H), 8.15 (d, 1H, 4-H), 8.33 [s, 1H, 8-H (Gua)], 9.07 (d, 1H, 13H), 9.12 (d, 1H, 1-H), 10.80 [s, 1H, 1-NH (Gua)], 12.13 (s, 1H, 7-H), J1,2 ) 8.0 Hz, J2,3 ) 7.5 Hz, J3,4 ) 7.5 Hz, J8,9 ) 8.5 Hz, J10,11 ) 8.0 Hz, J11,12 ) 7.5 Hz, J12,13 ) 7.0 Hz. MS: [M +1]+ C25H17ON6, calcd 417.1464, found 417.1468. DBC-6-C8Gua: UV, λmax (nm) 243, 278, 300, 333, 348, 366; 1H NMR, δ 6.43 [s, 2H, 2-NH (Gua)], 7.53 (dd, 1H, 11-H), 7.58 2 (dd, 1H, 3-H), 7.72 (dd, 1H, 12-H), 7.76 (d, 1H, 8-H), 7.78 (dd, 1H, 2-H), 7.94 (d, 1H, 9-H), 8.07 (s, 1H, 5-H), 8.10 (d, 1H, 10H), 8.17 (d, 1H, 4-H), 9.06 (d, 1H, 13-H), 9.12 (d, 1H, 1-H), 9.58 [s, 1H, 7- or 9-NH (Gua)], 10.68 [s, 1H, 1-NH (Gua)], 12.19 (s, 1H, 7-H), J1,2 ) 8.0 Hz, J2,3 ) 7.5 Hz, J3,4 ) 7.0 Hz, J8,9 ) 9.0 Hz, J10,11 ) 7.5 Hz, J11,12 ) 7.5 Hz, J12,13 ) 7.0 Hz. MS: [M + 1]+ C25H17ON6, calcd 417.1464, found 417.1466. DBC-5-N7Ade: UV, λmax (nm) 243, 278, 303, 337, 355, 373; 1H NMR, δ 5.82 [s, 2H, 6-NH (Ade)], 7.24 (d, 1H, 4-H), 7.52 2 (dd, 1H, 3-H), 7.55 (dd, 1H, 11-H), 7.74 (dd, 1H, 12-H), 7.81 (dd, 1H, 2-H), 7.88 (d, 1H, 8-H), 8.01 (d, 1H, 9-H), 8.13 (d, 1H, 10H), 8.16 (s, 1H, 6-H), 8.33 [s, 1H, 2-H (Ade)], 8.61 [s, 1H, 8-H (Ade)], 9.10 (d, 1H, 13-H), 9.23 (d, 1H, 1-H), 12.65 (s, 1H, 7-H), J1,2 ) 9.0 Hz, J2,3 ) 7.0 Hz, J3,4 ) 7.0 Hz, J8,9 ) 9.0 Hz, J10,11 ) 7.0 Hz, J11,12 ) 7.5 Hz, J12,13 ) 7.0 Hz. MS: [M + 1]+ C25H17N6, calcd 401.1515, found 401.1517. DBC-5-N3Ade: UV, λmax (nm) 241, 256, 275, 303, 336, 353, 371; 1H NMR, δ 7.32 (d, 1H, 4-H), 7.42 [s, 2H, 6-NH2 (Ade)],

7.47 (dd, 1H, 3-H), 7.55 (dd, 1H, 11-H), 7.74 (dd, 1H, 12-H), 7.78 (dd, 1H, 2-H), 7.88 (d, 1H, 8-H), 7.99 (d, 1H, 9-H), 8.04 (s, 1H, 6-H), 8.07 [s, 1H, 8-H (Ade)], 8.13 (d, 1H, 10-H), 8.49 [s, 1H, 2-H (Ade)], 9.08 (d, 1H, 13-H), 9.20 (d, 1H, 1-H), 12.59 (s, 1H, 7-H), J1,2 ) 8.5 Hz, J2,3 ) 7.0 Hz, J3,4 ) 7.5 Hz, J8,9 ) 9.0 Hz, J10,11 ) 8.0 Hz, J11,12 ) 7.0 Hz, J12,13 ) 7.0 Hz. MS: [M + 1]+ C25H17N6, calcd 401.1515, found 401.1510. DBC-5-N1Ade: UV, λmax (nm) 241, 275, 302, 336, 354, 371; NMR, δ 7.29 (d, 1H, 4-H), 7.46 (dd, 1H, 3-H), 7.55 (dd, 1H, 11-H), 7.64 [s, 1H, 8-H (Ade)], 7.74 (dd, 1H, 12-H), 7.78 (dd, 1H, 2-H), 7.89 (d, 1H, 8-H), 8.00 (d, 1H, 9-H), 8.14 (d, 1H, 10-H), 8.17 (s, 1H, 6-H), 8.19 [s, 1H, 6-NH (Ade)], 8.23 [s, 1H, 6-NH (Ade)], 8.61 [s,1H, 2-H (Ade)], 9.08 (d, 1H, 13-H), 9.20 (d, 1H, 1-H), 12.67 (s, 1H, 7-H), J1,2 ) 8.5 Hz, J2,3 ) 8.5 Hz, J3,4 ) 7.0 Hz, J8,9 ) 8.5 Hz, J10,11 ) 8.0 Hz, J11,12 ) 7.5 Hz, J12,13 ) 7.5 Hz. MS: [M + 1]+ C25H17N6, calcd 401.1515, found 401.1518. 1H

DBC-6-N3Ade: UV, λmax (nm) 241, 276, 300, 333, 348, 366; NMR, δ 7.45 [s, 2H, 6-NH2 (Ade)], 7.54 (dd, 1H, 11-H), 7.60 (dd, 1H, 3-H), 7.73 (d, 1H, 8-H), 7.74 (dd, 1H, 12-H), 7.80 (dd, 1H, 2-H), 7.94 (d, 1H, 9-H), 8.10 (s, 1H, 5-H), 8.11 (d, 1H, 10H), 8.12 [s, 1H, 8-H (Ade)], 8.18 (d, 1H, 4-H), 8.57 [s, 1H, 2-H (Ade)], 9.09 (d, 1H,13-H), 9.15 (d, 1H, 1-H), 12.15 (s, 1H, 7-H), J1,2 ) 8.0 Hz, J2,3 ) 7.5 Hz, J3,4 ) 7.5 Hz, J8,9 ) 9.0 Hz, J10,11 ) 8.5 Hz, J11,12 ) 7.5 Hz, J12,13 ) 7.0 Hz. MS: [M + 1]+ C25H17N6, calcd 401.1515, found 401.1515. 1H

Binding of DBC to DNA. DBC (80 µM) was bound to DNA in reactions catalyzed either by 3-MC-induced rat liver microsomes or by HRP in 15-mL incubation mixtures as described previously (23). All reaction mixtures were incubated for 30 min at 37 °C. After incubation, 1 mL of the mixture was taken out and used to determine the stable DNA adducts by the 32Ppostlabeling method after purification of the DNA, as previously described (23). The amount of each adduct was determined by liquid scintillation counting. The DNA was precipitated from the rest of the incubation mixture with 2 volumes of absolute ethanol, and the depurinating DNA adducts were isolated and quantified from the supernatant by HPLC in two different solvent systems, by using CH3CN/H2O (see Table 1) and CH3OH/H2O gradients with a Waters 996 PDA detector and a Jasco FP-920 fluorescence detector in series. Proof of the depurinating adducts formed in the biological systems was obtained by mass spectrometry.


Chen et al.

Results Synthesis and Structure Determination of Adducts Formed upon Iodine Oxidation of DBC and Reaction with dG or Ade. Electrochemical oxidation of DBC in the presence of deoxyribonucleosides under the conditions used with other PAH (24-26) inexplicably did not yield any products. The ability of iodine to effect a one-electron oxidation of PAH to radical cations is wellestablished (27-31). Recently, PAH-nucleoside adducts have been synthesized in good yields in our laboratory by iodine oxidation in DMF or Me2SO.2 The results parallel those achieved by electrochemical oxidation of PAH (24-26). Iodine oxidation of DBC in the presence of dG afforded the three depurinating adducts DBC-5-N7Gua (0.9%), DBC-6-N7Gua (2.8%), and DBC-6-C8Gua (0.2%) (Scheme 1); the remainder of the mixture was starting material. The products (Scheme 1) were separated and purified by HPLC. Formation of DBC-6-C8Gua as a depurinating adduct is consistent with the similar C8Gua adducts formed by BP (24) and dibenzo[a,l]pyrene (DB[a,l]P) (26). In these cases, the loss of deoxyribose is rationalized by the coplanarity of the PAH ring with the Gua ring. The driving force to destabilize the deoxyribose is protonation at the N-7 of the Gua moiety, with formation of a carbenium ion at C-8 that is stabilized by the PAH moiety. This coplanarity cannot be achieved with the ribose in guanosine because of the presence of the hydroxy group at C-2 in the sugar. In this case, the C8 adduct retains the ribose, forming a relatively stable nucleoside adduct (24). Oxidation of DBC by iodine in the presence of deoxyadenosine did not afford any products. When the reaction between DBC and Ade was carried out, four adducts were obtained: DBC-5-N7Ade (1.5%), DBC-5-N3Ade (3.6%), DBC-5-N1Ade (18%), and DBC-6-N3Ade (1.4%) (Scheme 1). This indicates that iodine oxidation of DBC produces a radical cation with charge localization mainly at C-5, but also at C-6. The most reactive nucleophilic groups are N-7 and C-8 for Gua and N-1, N-3, and N-7 for Ade. NMR. Proton and COSY NMR techniques have been used to elucidate the structure of all adducts. The C2 axis of symmetry of the DBC molecule is reflected in its NMR spectrum, in which the paired protons have the same chemical shifts (Figure 1A). In contrast, the adducts reveal an asymmetric pattern of proton splitting. One of the characteristic features of the NMR spectra of these adducts is the singlet with a chemical shift of 8.08.2 ppm, which corresponds to the 6-H or 5-H in the 5-substituted or 6-substituted adducts. A second one is the shielding of the resonance corresponding to the doublet signal of proton 4-H from 8.2 ppm in DBC and the 6-substituted adducts to 7.3-7.4 ppm in the 5-substituted adducts (peri effect). The perpendicular orientation of the base with respect to the DBC is responsible for this effect. DBC-5-N7Gua, DBC-6-N7Gua, and DBC-6-C8Gua. These adducts do not contain the deoxyribose moiety, as evidenced by the mass spectrometric molecular mass and the absence of its proton resonances in the aliphatic region of the spectra (not shown). The spectra of DBC5-N7Gua (Figure 1B) and DBC-6-N7Gua (Figure 2A) contain the characteristic sharp singlet resonating at 8.25 2

Unpublished results (A. Hanson et al.).

Figure 1. NMR spectra of (A) DBC and (B) DBC-5-N7Gua.

and 8.33 ppm, respectively, which corresponds to the C-8 proton of Gua and indicates that no detectable substitution occurs at this position. Substitution at the amino group of Gua is also ruled out by the presence of the singlet signal at 6.2 ppm in both spectra, corresponding to the two amino protons, which were exchanged with D2O. In the spectrum of DBC-5-N7Gua (Figure 1B), the doublet signal of 4-H is shielded at 7.38 ppm, and this resonance couples with that of 3-H at 7.49 ppm, as evidenced by its COSY spectrum (not shown). The doublet resonance of 4-H at 8.17 ppm in Figure 2A (no shielding effect) and its coupling with that of 3-H at 7.58 ppm suggest that substitution in this adduct takes place at C-6. These combined data support assignment of the structures of these adducts. The spectrum of DBC-6C8Gua (Figure 2B) resembles that of DBC-6-N7Gua (Figure 2A), except for the absence of the C-8 proton signal of Gua. DBC-5-N7Ade, DBC-5-N3Ade, DBC-5-N1Ade, and DBC-6-N3Ade. The upfield shifts of the doublet signal 4-H in the spectra of DBC-5-N7Ade (Figure 3A), DBC5-N3Ade (Figure 3B), and DBC-5-N1Ade (Figure 4A) indicate that substitution occurs in these adducts at C-5 of DBC. The broad singlet resonance at 5.82 ppm (Figure 3A) (exchangeable with D2O) corresponds to the NH2 protons of Ade. This shielded NH2 signal is typical of unsubstituted PAH that form adducts at the N-7 of Ade (24, 26). The deshielded singlet at 8.61 ppm is assigned to the C-8 proton of Ade. It is known that ortho substitution of alkyl groups at N-9, as in deoxyadenosine (8.34 ppm, not shown), or at N-7, as in benzo[a]pyrene6-N7Ade (8.68 ppm (24), deshields the 8-H proton resonance of Ade (8.07 ppm, not shown).



Figure 3. NMR spectra of (A) DBC-5-N7Ade and (B) DBC-5N3Ade. Figure 2. NMR spectra of (A) DBC-6-N7Gua and (B) DBC-6C8Gua.

Similar deshielding effects are observed for the 2-H resonance of Ade in the N-3 (8.49 ppm, Figure 3B) and N-1 (8.61 ppm, Figure 4A) substituted adducts compared to the corresponding signal of Ade itself (8.07 ppm). The spectrum of DBC-5-N1Ade is also characterized by a downfield shift of the 6-NH2 resonances (8.19 and 8.23 ppm) of the Ade moiety compared to the corresponding resonances of DBC-5-N3Ade (7.42 ppm) and Ade (7.10 ppm) and by the splitting of the two proton signals in the DBC-5-N1Ade adduct. The bulky group of the DBC moiety substituted at the position ortho to the amino group in the N1Ade adduct results in a rotational hindrance of the latter so that the two protons are in different magnetic environments. The two signals merge into one singlet when the temperature is raised to 50 °C. A similar effect was also observed for the N1Ade adducts of BP (7) and dibenzo[a,l]pyrene (unpublished results).3 The spectrum in Figure 4B does not display a shielding effect of the doublet signal 4-H, suggesting that substitution of the DBC moiety takes place at C-6. The similar chemical shift of the 6-NH2 in DBC-6-N3Ade (7.45 ppm, Figure 4B) and DBC-5-N3Ade (7.42 ppm, Figure 3B) indicates that the same nitrogen of Ade is substituted in both adducts. Mass Spectrometry. (a) FAB MS and MS/MS. Low resolution FAB experiments in various matrices were performed to support the structure determination of synthetic DBC-DNA base adducts. Although the 3

Unpublished results (K.-M. Li et al.).

abundances of the molecular ions from all adducts were low, the best signal-to-noise ratio was achieved when GLY/TFA was used as the matrix. FAB of DBC-Ade adducts produces an [M + H]+ ion of m/z 401 and of the DBC-Gua adducts gives an [M + H]+ ion of m/z 417. The FAB mass spectra of all DBC-DNA base adducts show no detectable fragment ions. The exact masses of protonated DBC-Ade and DBCGua adducts, as determined by high resolution peak matching experiments, are within 1.2 part-per-million (ppm) average error of the theoretical values for the formulae C25H17ON6 and C25H17N6 of DBC-Gua and DBC-Ade adducts, respectively. CAD spectra of the [M + H]+ ions were investigated to obtain molecular structure information. Upon collisional activation (CA), the [M + H]+ ion of DBC-6C8Gua, for example, decomposes to form major fragment ions of m/z 400, 307, 292, 266, and 252 (Figure 5A). The most abundant fragment ion of m/z 266, which is formed by the C-N bond cleavage between the Gua and the DBC moiety, is strong evidence for the presence of DBC. The formation of the fragment ion of m/z 400 occurs by the elimination of NH3, presumably upon protonation on the NH2 group. The ion of m/z 292 is [DBC - CN]•+. The ion of m/z 307 is formed by charge-remote cleavages of the C5-N7 and C4-N9 bonds. The CAD spectra of the [M + H]+ ions of the other DBC-Gua adducts contain the same fragment ions, but the relative abundances are slightly different. The C-8 and N-7 adduct isomers of this heterocyclic aromatic are significantly more difficult to distinguish by tandem mass spectrometry than are those of BP and DB[a,l]P (24, 26, 32, 33). Normally,


Chen et al.

Figure 5. Comparison of (A) the CAD spectrum taken on the four-sector tandem and (B) the PSD spectrum taken on the MALDI-TOF of the DBC-6-C8Gua. The precursor [M + H]+ is not shown.

Figure 4. NMR spectra of (A) DBC-5-N1Ade and (B) DBC-6N3Ade.

when a DNA base is substituted at C-8, a facile cycloreversion occurs upon CA to form ArCN•+ (32, 33). For this C-8 DBC-Gua adduct, the cycloreversion would give a product at m/z 293 that is more abundant than the ions in the m/z 264-267 range. This is not seen possibly because the lower ionization energy of the heterocycle compared to those of PAH decreases the energy for cleavage of the bond joining the heterocycle to the DNA base. As a result, the cycloreversion reaction no longer competes for DBC adducts. For DBC-Ade adducts, the [M + H]+ ions collisionally decompose in tandem mass spectrometry experiments to produce major fragment ions of m/z 384 [M + H - NH3]+, m/z 292 [C20H12NNC]+, m/z 265 [C20H11N]+ (where C20H13N is DBC), and m/z 239 [C18H9N]+. For three DBC-Ade adducts, the most abundant fragment ion is of m/z 265 [C20H11N]+, which may be produced by loss of two hydrogens from the DBC moiety, accompanied by formation of the C1-C13 bond, and followed by cleavage of the C-N bond between Ade and DBC. The relative abundances of m/z 384 [M + H - NH3]+ and m/z 267 [C20H13N]•+ ions are of particular interest for distinguishing the different nitrogen-substituted structural isomers. The relative abundance of m/z 384 is the largest for the isomer in which the N-7 position of Ade is bonded to DBC, whereas the relative abundance is lowest for the isomer in which N-3 is bonded. It is suggested that facile formation of m/z 384 ions is promoted by anchimeric assistance by the DBC ring system to expel NH3 when the NH2 group is located on the PAH side. (b) MALDI-TOF MS. The matrix (BCC) was developed to increase the limit of detection of PAH-DNA base

adducts by MALDI-TOF (21). A further improvement in detection was achieved by using the comatrix d-fucose (34). The addition of the d-fucose also allows the laser power needed for threshold-ion production to be lowered and the extent of matrix adduct formation to be decreased. The MALDI-TOF spectra of the synthetic DBC adducts show a strong [M + H]+ ion peak with mass resolving power of approximately 400 FWHM. For approximately a 1-pmol loading, the signal-to-noise for [M + H]+ was typically greater than 25:1, and there were few detectable matrix ions within (50 u of the molecular ion. The PSD of the [M + H]+ ion of DBC-5-C8Gua (Figure 5B) occur to give a fragmentation pattern that is very similar to that seen in the CAD spectrum (Figure 5A). The PSD spectrum of the [M + H]+ ion of DBC-5-N3Ade shows a fragmentation pattern very similar to that seen in its CAD spectrum; the relative intensities of the fragment ion peaks are nearly identical. The PSD spectrum of the [M + H]+ ion of DBC-5-N7Ade, however, shows differences in the relative intensities of fragment ion peaks at m/z 357, 373, and 384. A more detailed discussion of the ion chemistry will be presented in a sequel article.4 Identification and Quantitation of DBC-DNA Adducts Formed in Vitro. The DBC-DNA adducts formed by cytochrome P450 in 3-MC-induced rat liver microsomes were compared to those formed in the HRP activating system. The stable adducts were quantitated by the 32P-postlabeling technique (23). The depurinating adducts were identified by comparison with authentic synthesized adduct standards. Preliminary identification of these adducts was made on the basis of coelution with 4

Unpublished results (J. K. Gooden, et al.).



Table 2. Quantitation of Biologically-Formed DBC-DNA Adductsa biological system MC-induced microsomes HRP

total adducts, µmol/mol of DNA-P

stable adducts, µmol/mol of DNA-P

DBC-5N7Gua

8.7

0.3 (4)b 9.1 (32)

1.0 (11) 5.5 (19)

28.5

depurinating adducts, µmol/mol of DNA-P DBC-6- DBC-5- DBC-5- total depurinating N7Gua N7Ade N3Ade adducts 2.8 (32) 3.8 (13)

4.6 (53) 7.6 (27)

Synthesis of Depurinating DNA Adducts Formed by One-Electron

Recommend Documents