Chem. Res. Toxicol. 2004, 17, 827-838
827
Identification and Characterization of Novel Stable Deoxyguanosine and Deoxyadenosine Adducts of Benzo[a]pyrene-7,8-quinone from Reactions at Physiological pH Narayanan Balu,† William T. Padgett, Guy R. Lambert, Adam E. Swank, Ann M. Richard, and Stephen Nesnow* Environmental Carcinogenesis Division, National Health and Environmental Effects Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, B143-06, Research Triangle Park, North Carolina 27711 Received October 6, 2003
Benzo[a]pyrene (B[a]P) is an archetypal member of the family of polycyclic aromatic hydrocarbons (PAHs) and is a widely distributed environmental pollutant. B[a]P is known to induce cancer in animals, and B[a]P-containing complex mixtures are human carcinogens. B[a]P exerts its genotoxic and carcinogenic effects through metabolic activation forming reactive intermediates that damage DNA. DNA adduction by B[a]P is a complex phenomenon that involves the formation of both stable and unstable (depurinating) adducts. One pathway by which B[a]P can mediate genotoxicity is through the enzymatic formation of B[a]P-7,8-quinone (BPQ) from B[a]P-7,8-diol by members of the aldo-keto-reductase (AKR) family. Once formed, BPQ can act as a reactive Michael acceptor that can alkylate cellular nucleophiles including DNA and peptides. Earlier studies have reported on the formation of stable and depurinating adducts from the reaction of BPQ with DNA and nucleosides, respectively. However, the syntheses and characterization of the stable adducts from these interactions have not been addressed. In this study, the reactivity of BPQ toward 2′-deoxyguanosine (dG) and 2′-deoxyadenosine (dA) nucleosides under physiological pH conditions is examined. The identification and characterization of six novel BPQ-nucleoside adducts obtained from the reaction of BPQ and dG or dA in a mixture of phosphate buffer and dimethylformamide are reported. The structures of these adducts were determined by ultraviolet spectroscopy, electrospray mass spectrometry, and NMR experiments including 1H, 13C, two-dimensional COSY, one-dimensional NOE, ROESY, HMQC, HSQC, and HMBC. The reaction of BPQ with dG afforded four unique Michael addition products: two diastereomers of 8-N1,9-N2-deoxyguanosyl-8,10-dihydroxy-9,10-dihydrobenzo[a]pyren-7(8H)-one (BPQ-dG1,2) and two diastereomers of 10-(N2-deoxyguanosyl)-9,10-dihydro9-hydroxybenzo[a]pyrene-7,8-dione (BPQ-dG3,4). The BPQ-dG1,2 adducts suggest a 1,6-Michael addition reaction of dG, an oxidation of the hydroquinone to the quinone, a 1,4-Michael addition of water, and an internal cyclization. The BPQ-dG3,4 adducts suggest a 1,4-Michael addition reaction of dG, an oxidation of the hydroquinone to the quinone, and a 1,6-Michael addition of water. Under similar but extended reaction conditions, the reaction of BPQ with dA produced only one diastereomeric pair of adducts identified as 8-N6,10-N1-deoxyadenosyl-8,9-dihydroxy9,10-dihydrobenzo[a]pyren-7(8H)-one (BPQ-dA1,2). The BPQ-dA1,2 adducts suggest a 1,4-Michael addition reaction of dA, an oxidation of the hydroquinone to the quinone, a 1,6-Michael addition of water, and an internal cyclization. As considerable efforts have been placed in documenting the genotoxic effects of BPQ, this first report of the identification and characterization of these stable adducts of BPQ formed under physiological pH conditions is expected to contribute significantly to the area of BPQ-mediated genotoxicity and carcinogenesis.
Introduction B[a]P1 (benzo[def]chrysene) is the archetypal member of the family of PAHs that are a group of widely distributed environmental pollutants (1). B[a]P is an * To whom correspondence should be addressed. Fax: 919-541-0694. E-mail:
[email protected]. † N.B. is a National Research Council Associate. 1 Abbreviations: AKR, aldo-keto-reductases; B[a]P, benzo[a]pyrene; BPDE, B[a]P-7,8-diol-9,10-epoxide; B[a]P-7,8-diol, (()-trans-7,8-dihydroxy-7,8-dihydrobenzo[a]pyrene; BPQ, B[a]P-7,8-quinone; dA, 2′deoxyadenosine; dG, 2′-deoxyguanosine; P450, cytochrome P450; DMF, dimethylformamide; ESI, electrospray ionization; PAH, polycyclic aromatic hydrocarbon.
anthropogenic PAH formed from the incomplete combustion processes of fossil fuels and emitted from a variety of sources such as automobile exhaust fumes, diesel exhaust fumes, and tobacco smoke (1-3). B[a]P is a potent carcinogen in many species of rodents, and B[a]Pcontaining complex environmental mixtures are known human respiratory carcinogens (4-7). The major metabolic pathways by which B[a]P forms reactive metabolites that induce a series of cellular DNA modifications have been identified and well-reviewed (3). In mammalian cells and tissues, B[a]P is mainly metabolized to epoxides, dihydrodiols, dihydrodiol epoxides, tetraols, quinones,
10.1021/tx034207s CCC: $27.50 © 2004 American Chemical Society Published on Web 05/15/2004
828
Chem. Res. Toxicol., Vol. 17, No. 6, 2004
and conjugates of glucuronic acid, sulfuric acid, and glutathione. The specific cytochromes involved in these processes have also been well-characterized (8, 9). The major biological effects in mammalian cells associated with B[a]P exposure are mutation and cancer, and these generally are attributed, at least in part, to the formation of BPDE. One of the primary B[a]P metabolites is B[a]P-7,8-diol, derived from successive P450 and epoxide hydrolase enzymatic processes (10-16). The enzymatic action of P450 on B[a]P-7,8-diol produces BPDE. BPDE is considered to be one of the ultimate carcinogenic metabolites of B[a]P as it binds covalently to DNA and induces mutations, DNA breakage, and cancer (3). Several other B[a]P metabolic pathways have been identified that lead to DNA damage and genotoxic effects including radical cation processes, Kregion bond activation, and the formation of BPQ. B[a]P can form a radical cation through one electron oxidation producing depurinating adducts (e.g., N7-guanine, C8-guanine, N7-adenine, and N3-adenine) and apurinic sites (17). The K-region dihydrodiol metabolite of B[a]P, B[a]P-4,5-diol, has been shown to be a potent genotoxic agent in mammalian cells inducing both morphological cell transformation and DNA damage (18). Another important pathway by which B[a]P can mediate genotoxicity is through the formation of BPQ (19, 20). BPQ is formed metabolically from B[a]P-7,8-diol by members of the AKR family. The AKR1C isoforms have been postulated to contribute to the carcinogenesis process by oxidizing PAH trans-dihydrodiols to reactive o-quinones through the intermediary hydroquinone (21). These enzymes are ubiquitous in mammals (22), and recently, four homogeneous human recombinant AKRs (AKR1C1-AKR1C4) have been identified (23). Once formed, the PAH oquinones can act as reactive Michael acceptors and have the ability to alkylate a variety of cellular nucleophiles, including DNA, N-acetylcysteine, and glutathione (24-26). They may also undergo nonenzymatic two electron redox processes in the presence of cellular [NAD(P)H] or enzymatic one electron reduction with microsomal or mitochondrial enzymes, leading to the generation of reactive oxygen species (ROS) and eventually inducing oxidative damage and strand scission of DNA (20). This BPQ-induced DNA damage has been illustrated by the formation of stable DNA adducts, unstable N7- and N3depurinating adducts, oxidized bases (e.g., 8-oxo-dG), and strand scission (27-29). The ability of BPQ to act as a Michael acceptor is an important feature as it can initiate a new set of reaction profiles leading to cellular DNA modifications. A preliminary investigation into such reactions provided mainly the dG adducts both in stable and in depurinated forms depending on the reaction conditions. In one study, [1,3-3H2]BPQ was reported to yield stable uncharacterized dG adducts with calf thymus and plasmid DNA (26). However, a complete characterization of these important stable BPQ-nucleoside adducts has not been reported. In primary rat hepatocytes treated with [1,3-3H2]BPQ, the isolated DNA exhibited BPQbound radioactivity. However, this radioactivity did not survive the enzymatic degradation to nucleosides suggesting that these adducts were unstable or not subject to digestion (30). More recently, another study reported the detection and characterization by LC/MS and collision-induced dissociation of depurinating N7-guanine adducts of BPQ with dG produced under acid-assisted Michael type reaction conditions (28).
Balu et al.
In a continuation of our research on the carcinogenic effects of B[a]P metabolites (18, 31), we have further investigated the reaction of BPQ with deoxynucleosides with a major emphasis on physiological pH conditions to clarify the possible role of BPQ in B[a]P genotoxicity and carcinogenesis. This initial report represents the results of our efforts to synthesize and characterize novel stable adducts of BPQ with dG and dA in phosphate buffer/DMF mixtures (Scheme 1).
Materials and Methods Caution: B[a]P and all of its derivatives are classified as potentially hazardous materials and should be handled in accordance with the National Cancer Institute’s guidelines for the use of chemical carcinogens. Chemicals. B[a]P (99%), dG, dA, and DMF were purchased from Sigma-Aldrich Co. (St. Louis, MO). HPLC/spectrophotometric grade acetonitrile and methanol were obtained from Fisher Scientific (Pittsburgh, PA). All other chemicals were reagent grade and obtained from commercial sources. Instrumentation. UV spectra were recorded on a Beckman model DU-70 spectrometer. 1H NMR spectra were recorded on a Bruker AVANCE DRX 300 instrument at 300.13 MHz. 13C NMR spectra were recorded at 75.49 MHz and used for further structural analyses by HMQC and HSQC NMR experiments. Chemical shifts were reported as parts per million referenced to tetramethylsilane. Two-dimensional COSY, HMBC, and ROESY spectra and one-dimensional NOE difference spectra were recorded according to established manufacturer’s protocols. HPLC/ESI-MS were obtained on a Hewlett-Packard model 1100 HPLC system with a mass selective detector and an electrospray interface. Analytical TLC was carried out on fluorescent silica gel plates or C18 reverse phase plates. Bands were visualized with short- and long-wavelength UV lamps. HPLC was conducted on a Shimadzu model 10A liquid chromatograph at 254 nm or on a Hewlett-Packard model 1050 system connected to a Hewlett-Packard model 1050 diode array detector. Effluents were monitored over the range of 220-550 nm. Synthesis of BPQ. BPQ was prepared essentially by the method of Harvey and Fu (32) from trans-7,8-dihydroxy-7,8,9,10-tetrahydroB[a]P that was obtained from the hydrolysis of the Prevost reaction product of 9,10-dihydroB[a]P. Briefly, in a typical preparation, trans-7,8-dihydroxy-7,8,9,10-tetrahydroB[a]P (1.0 g, 3.47 mmol) was suspended in 100 mL of methylene chloride and 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (4.0 g, 17 mmol) was added. The reaction was stirred at room temperature for 24 h. It was then cooled on ice and filtered to remove the hydroquinone formed. The solvent was exchanged to methanol and cooled to -20 °C on a salt-ice bath. The precipitate was filtered and washed with cold methanol. The product was recrystallized with methylene chloride and methanol to give 421 mg (43% yield) of BPQ as dark purple to black needles. The product was confirmed by MS and NMR spectroscopy by comparison to known literature values (33). General Procedure for HPLC Chromatography and Adduct Purification. 1. Analytical HPLC Conditions. Crude reaction mixtures were analyzed on a Luna C-18 reverse phase column (4.6 mm × 250 mm) (Phenomenex, Torrance, CA) using a gradient mobile phase of 50% 3:1 methanol-acetonitrile/ 50% deionized water (solvent A) and 3:1 methanol-acetonitrile (solvent B). The gradient was initiated with 100% solvent A for 5 min followed by a linear increase to 25% solvent A, 75% solvent B in 65 min. This was followed by 100% solvent B and immediately returned back to 100% solvent A in 5 min. The flow rate was maintained at 2 mL/min. 2. Preparative HPLC Conditions. Adducts BPQ-dG1-4 were separated on a Luna C-18 reverse phase semipreparative column (10 mm × 250 mm) by using an isocratic system consisting of 50% water and 50% 3:1 methanol-acetonitrile with a flow rate of 5 mL/min. For BPQ-dA1,2 adducts, a C18 YMC
Novel, Stable Deoxyguanosine and Deoxyadenosine Adducts
Chem. Res. Toxicol., Vol. 17, No. 6, 2004 829
Scheme 1. Reaction Products of BPQ with dG and dA in Phosphate Buffer (pH 7.5)/DMF
column (10 mm × 250 mm) (Waters, Milford, MA) was used with a mobile phase of 100% solvent A (50% 3:1 methanol-acetonitrile/50% deionized water) for 5 min, increasing linearly to 13% solvent A and 87% solvent B (3:1 methanol-acetonitrile) in 40 min. The samples were collected from multiple HPLC analyses. Each individual peak was pooled, and the organic solvent was removed by rotary evaporation in vacuo. The water was removed by lyophylization. The residue was dissolved in methanol and diluted with 5× water. The residue solution was applied onto a C18 Analtech Spice solid phase extraction cartridge (Rainen, Woburn, MA). After 20 mL of water was passed through the cartridge, nitrogen gas was applied to the column to dry it. The adducts were eluted with 5 mL of methanol under a stream of nitrogen and further dried by lyophylization to obtain the adducts for analyses. Reaction of BPQ with dG. A solution of BPQ (20 mg, 0.07 mmol) in anhydrous DMF (1 mL) was stirred with a suspension of dG (350 mg, 1.31 mmol) in 8 mL of a solution of 1:1 DMF-sodium phosphate buffer (pH 7.5, 0.25 M) at 50-60 °C for 6 h. The reaction turned dark reddish-brown during this period. The solvents were removed by lyophylization, and the residue was dissolved in DMSO:H2O (1:1), was filtered, and was subjected to analytical and preparative HPLC chromatography (see above). Four peaks were isolated and subjected to spectroscopic analyses: Rt 13.5 min (adduct BPQ-dG1); Rt 15 min (adduct BPQ-dG2); Rt 22.8 min (adduct BPQ-dG3); and Rt 24.2 min (adduct BPQ-dG4). The approximate yields of the adducts were as follows: BPQ-dG1,2 (3.5% each) and BPQ-dG3,4 (2.5% each). Residual BPQ was still present at the completion of the reaction; however, longer reaction times did not give increased yields. BPQ-dG1. 1H NMR (DMSO-d6): ppm 2.00-2.07 (m, 1H, 2′-dG), 2.32-2.40 (m, 1H, 2′′-dG), 3.26-3.33 (m, partially obscured, 2H, 5′ and 5′′-dG), 3.56-3.60 (m, 1H, 4′-dG), 4.15-4.17 (m, 1H, 3′-dG), 4.68 (pseudo t, 1H, 5′OH-dG, D2O exchangeable), 5.09 (d, J ) 2.99 Hz, 1H, H9), 5.16 (d, J ) 4.12 Hz, 1H, 3′OHdG, D2O exchangeable), 5.80 (pseudo t, 1H, 1′-dG), 6.64 (d, J ) 3.13 Hz, 1H, OH10, D2O exchangeable), 6.78 (pseudo t, 1H, H10), 7.51 (bs, 1H, OH8, D2O exchangeable), 7.83 (s, 1H, H8dG), 8.15 (t, J ) 7.68 Hz, 1H, H2), 8.27-8.38 (m, 6H, H1, H3, H4, H5, H11, H12), 8.65 (s, 1H, H6), 9.51 (bs, 1H, N2H, D2O exchangeable). 13C NMR (DMSO-d6): ppm 39.2 (2′C-dG), 59.6 (C10), 61.5 (5′C-dG), 68.1 (C9), 70.7 (3′C-dG), 82.9 (1′C-dG), 84.9 (C8), 87.9 (4′C-dG), 122.1-130.8 (C1-5,11,12), 124.9 (C6), 136.2
(C8-dG), 154.2 (C2-dG), 155.5 (C6-dG), 192.0 (C7). Retention time, 13.5 min. BPQ-dG2. 1H NMR (DMSO-d6): ppm 1.88-1.96 (m, 1H, 2′-dG), 2.16-2.24 (m, 1H, 2′′-dG), 3.33-3.45 (m, partially obscured, 2H, 5′ and 5′′-dG), 3.68-3.69 (m, 1H, 4′-dG), 4.104.20 (m, 1H, 3′-dG), 4.81 (pseudo t, 1H, 5′OH-dG, D2O exchangeable), 5.11 (d, J ) 2.98 Hz, 1H, H9), 5.12 (d, J ) 4.17 Hz, 1H, 3′OH-dG, D2O exchangeable), 5.81 (pseudo t, 1H, 1′-dG), 6.64 (d, J ) 2.88 Hz, 1H, OH10, D2O exchangeable), 6.77 (pseudo t, 1H, H10), 7.52 (bs, 1H, OH8, D2O exchangeable), 7.84 (s, 1H, H8-dG), 8.15 (t, J ) 7.65 Hz, 1H, H2), 8.27-8.39 (m, 6H, H1, H3, H4, H5, H11, H12), 8.65 (s, 1H, H6), 9.53 (bs, 1H, N2H, D2O exchangeable). 13C NMR (DMSO-d6): ppm 40.1 (2′C-dG), 59.4 (C10), 61.8 (5′C-dG), 68.3 (C9), 71.0 (3′C-dG), 82.5 (1′C-dG), 84.9 (C8), 87.9 (4′C-dG), 122.1-130.8 (C1-5,11,12), 123.8 (C6), 136.0 (C8dG), 154.0 (C2-dG), 155.5 (C6-dG), 192.0 (C7). Retention time, 15 min. BPQ-dG3. 1H NMR (DMSO-d6): ppm 2.05-2.16 (m,1H, 2′-dG), 2.41-2.50 (m, 1H, 2′′-dG), 3.40-3.43 (m, partially obscured, 2H, 5′and 5′′-dG), 3.66-3.70 (m, 1H, 4′-dG), 4.26 (m, 1H, 3′-dG), 4.50 (m, 1H, H9), 4.83 (pseudo t, 1H, 5′OH-dG, D2O exchangeable), 5.21 (d, J ) 3.96 Hz, 1H, 3′OH-dG, D2O exchangeable), 5.96 (pseudo t, 1H, 1′-dG), 6.60 (d, J ) 3.55 Hz, 1H, OH9, D2O exchangeable), 6.98 (s, 1H, NH2, D2O exchangeable), 7.23 (d, J ) 4.19 Hz, 1H, H10), 7.93 (s, 1H, H8-dG), 8.18 (t, J ) 7.64 Hz, 1H, H2), 8.28-8.41 (m, 5H, H1, H3, H4, H5, H12), 8.97 (s, 1H, H6), 9.00 (s, 1H, N1H, D2O exchangeable), 9.29 (d, J ) 9.66 Hz, 1H, H11). 13C NMR (DMSO-d6): ppm 39.6 (2′C-dG), 48.1 (C10), 61.8 (5′C-dG), 68.6 (C9), 72.1 (3′C-dG), 82.9 (1′C-dG), 87.8 (4′C-dG), 125.5-130.4 (C1-5,12), 123.2 (C6), 125.1 (C11), 136.8 (C8-dG), 150.1 (C2-dG), 155.9 (C6-dG), 176.1 (C8), 192.7 (C7). Retention time, 22.8 min. BPQ-dG4. 1H NMR (DMSO-d6): ppm 1.96-2.03 (m,1H, 2′dG), 2.27-2.38 (m, 1H, 2′′-dG), 3.38-3.50 (m, partially obscured, 2H, 5′ and 5′′-dG), 3.72-3.74 (m, 1H, 4′-dG), 4.22-4.23 (m, 1H, 3′-dG), 4.51 (m, 1H, H9), 4.84 (pseudo t, 1H, 5′OH-dG, D2O exchangeable), 5.18 (d, J ) 3.82 Hz, 1H, 3′OH-dG, D2O exchangeable), 5.96 (pseudo t, 1H, 1′-dG), 6.61 (d, J ) 3.18 Hz, 1H, OH9, D2O exchangeable), 6.99 (s, 1H, NH2, D2O exchangeable), 7.23 (d, J ) 4.27 Hz, 1H, H10), 7.94 (s, 1H, H8-dG), 8.18 (t, J ) 7.65 Hz, 1H, H2), 8.28-8.41 (m, 5H, H1, H3, H4, H5, H12), 8.97 (s, 1H, H6), 9.00 (s, 1H, N1H, D2O exchangeable), 9.29 (d, J ) 9.69 Hz, 1H, H11). 13C NMR (DMSO-d6): ppm 40.1 (2′C-dG), 48.1 (C10), 61.8 (5′C-dG), 68.7 (C9), 70.8 (3′C-dG), 82.1
830
Chem. Res. Toxicol., Vol. 17, No. 6, 2004
(1′C-dG), 87.9 (4′C-dG), 125.4-129.9 (C1-5,12), 123.1 (C6), 125.5 (C11), 136.7 (C8-dG), 150.0 (C2-dG), 155.9 (C6-dG), 176.1 (C8), 192.7 (C7). Retention time, 24.2 min. Reaction of BPQ with dA. A solution of BPQ (18 mg, 0.064 mmol) in anhydrous DMF (1 mL) was stirred with a suspension of dA (86 mg, 0.342 mmol) in 8 mL of a solution of 1:1 DMFsodium phosphate buffer (pH 7.5, 0.25 M) at 50-60 °C for 18 h. The reaction turned dark reddish-brown during this period. The solvents were removed by rotary evaporation in vacuo followed by lyophylization. The residue was dissolved in a minimal volume of methanol-acetonitrile-water (1:1:1), was filtered, and was subjected to analytical and preparative HPLC chromatography (see above). Two peaks were isolated and subjected to spectroscopic analyses: Rt 19.4 min (adduct BPQ-dA1) and Rt 21.4 min (adduct BPQ-dA2). The yields of the adducts were as follows: BPQ-dA1, 5%, and BPQ-dA2, 4%. Residual BPQ was still present at the completion of the reaction; however, longer reactions times did not give increased yields. BPQ-dA1. 1H NMR (DMSO-d6): ppm 2.10-2.18 (m, 1H, 2′dA), 2.40-2.60 (m, partially obscured, 1H, 2′′-dA), 3.30-3.50 (m, partially obscured, 2H, 5′ and 5′′-dA), 3.70-3.74 (m, 1H, 4′-dA), 4.27-4.30 (m, 1H, 3′-dA), 4.48 (m, 1H, H9), 4.82 (pseudo t, 1H, 5′OH-dA, D2O exchangeable), 5.24 (d, J ) 4.08, 1H, 3′OHdA, D2O exchangeable), 6.07 (pseudo t, 1H, 1′-dA), 6.53 (bs, 1H, OH8, D2O exchangeable), 6.81 (d, J ) 3.78 Hz, 1H, H10), 6.38 (bs, 1H, OH9, D2O exchangeable), 8.07 (s, 1H, H8-dA), 8.21 (t, J ) 7.37 Hz, 1H, H2), 8.28 (d, J ) 9.09 Hz, 1H, H4), 8.35 (d, J ) 9.09 Hz, 1H, H5), 8.40 (d, J ) 7.37 Hz, 1H, H3), 8.46 (d, J ) 7.37 Hz, 1H, H1), 8.51 (d, J ) 9.45 Hz, 1H, H12), 8.75 (s, 1H, H2-dA), 8.95 (s, 1H, H6), 9.13 (d, J ) 9.45 Hz, 1H, H11). 13C NMR (DMSO-d6): ppm 39.6 (2′C-dA), 55.5 (C10), 61.6 (5′C-dA), 67.5 (C9), 70.9 (3′C-dA), 83.9 (1′C-dA), 85.5 (C8), 88.1 (4′C-dA), 123.3 (C11), 125.6-129.5 (C1-5,12), 138.6 (C8-dA), 140.1 (C6-dA), 141.9 (C4-dA), 146.1 (C2-dA), 193.3 (C7). Retention time, 19.4 min. BPQ-dA2. 1H NMR (DMSO-d6): ppm 2.02-2.10 (m,1H, 2′dA), 2.28-2.39 (m, 1H, 2′′-dA), 3.30-3.50 (m, 2H, 5′ and 5′′dA), 3.74-3.78 (m, 1H, 4′-dA), 4.22-4.23 (m, 1H, 3′-dA), 4.47 (pseudo t, 1H, H9), 4.88 (pseudo t, 1H, 5′OH-dA, D2O exchangeable), 5.21 (d, J ) 4.07, 1H, 3′OH-dA, D2O exchangeable), 6.07 (pseudo t, 1H, 1′-dA), 6.50 (s, 1H, OH8, D2O exchangeable), 6.80 (d, J ) 4.53 Hz, 1H, H10), 6.37 (d, 1H, OH9, D2O exchangeable), 8.07 (s, 1H, H8-dA), 8.21 (t, J ) 7.42 Hz, 1H, H2), 8.28 (d, J ) 9.09 Hz, 1H, H4), 8.35 (d, J ) 9.09 Hz, 1H, H5), 8.40 (d, J ) 7.42 Hz, 1H, H3), 8.46 (d, J ) 7.42 Hz, 1H, H1), 8.51 (d, J ) 9.51 Hz, 1H, H12), 8.73 (s, 1H, H2-dA), 8.95 (s, 1H, H6), 9.14 (d, J ) 9.51 Hz, 1H, H11). 13C NMR (DMSO-d6): ppm 40.2 (2′C-dA), 55.2 (C10), 61.5 (5′C-dA), 67.7 (C9), 70.9 (3′C-dA), 83.5 (1′C-dA), 85.5 (C8), 88.2 (4′C-dA), 123.4 (C11), 125.5-130.0 (C1-5,12), 138.1 (C8-dA), 140.1 (C6-dA), 141.9 (C4-dA), 146.1 (C2-dA), 193.3 (C7). Retention time, 21.4 min.
Balu et al.
Figure 1. Reverse phase HPLC profile of the reaction products of BPQ and dG (A) and BPQ and dA (B).
Results BPQ-dG Adducts. Reactions of BPQ and dG in buffer/ DMF solutions gave a number of products that were separable by reverse phase HPLC (Figure 1A). The peaks eluting prior to 12 min were determined to be dG and products arising from side reactions of BPQ. Residual BPQ was present after the incubations. Four peaks were collected for subsequent spectroscopic analyses (BPQ-dG1, BPQ-dG2, BPQ-dG3, and BPQ-dG4). The UV spectra (Figure 2) of BPQ-dG1,2 were identical exhibiting major absorbances at λmax 255 (sh), 278, 326, and 342 nm suggesting both dG and BPQ moieties present. The ESI mass spectra of BPQ-dG1,2 indicated two adducts with similar molecular ions and fragments (Table 1 and Chart 1). The molecular ion of m/z ) 564 (M - 1) in the negative mode and m/z ) 566 (M + 1) in the positive mode suggested that each product had a molecular weight of 565 and was comprised of 1 mol of each: dG, BPQ, and
Figure 2. UV spectral comparison of six BPQ-nucleoside adducts. Adducts BPQ-dG3,4 and BPQ-dA1,2 showed a pronounced bathochromic shift as compared to adducts BPQ-dG1,2 (indicated by broken line).
an atom of oxygen from a water molecule. Confirming the presence of these components were the observations of the following fragments. For both BPQ-dG1,2, the ESI negative mode mass spectra gave a loss of water (m/z ) 546). The ESI positive mode mass spectra gave two ions in addition to the molecular ion. The ion at m/z ) 450 represents the formation of a molecule with the elements of BPQ, guanine, and an oxygen atom (a loss from the molecular ion of 2-(hydroxymethyl)-2,3-dihydrofuran-3ol from deoxyribose). The ion at m/z ) 432 represents a loss of 2-(hydroxymethyl)-2,3-dihydrofuran-3-ol and water from the molecular ion. The proton NMR spectra of BPQ-dG1,2 were virtually identical, with some minor
Novel, Stable Deoxyguanosine and Deoxyadenosine Adducts Table 1. HPLC/ESI-MS of BPQ-dG and BPQ-dA Adducts adduct
ESI negative mode m/z (relative abundance)
ESI positive mode m/z (relative abundance)
BPQ-dG1
564 [M - 1] (100), 546 (72)
BPQ-dG2
564 [M - 1] (100), 546 (47)
BPQ-dG3 BPQ-dG4 BPQ-dA1
564 [M - 1] (100), 297 (66) 564 [M - 1] (100), 297 (27) 548 [M - 1] (100), 297 (65), 269 (69), 250 (19) 548 [M - 1] (100), 297 (50), 269 (79), 250 (22)
566 [M + 1] (12), 450 (100), 432 (40) 566 [M + 1] (13), 450 (100), 432 (43) 566 [M + 1] (10), 450 (100) 566 [M + 1] (11), 450 (100) 550 [M + 1] (100), 434 (18)
BPQ-dA2
550 [M + 1] (100), 434 (20)
differences, and provided critical information used in the structure identification of these adducts. In the proton NMR spectra, all of the resonances expected from the deoxyribose moiety (H1, H2, H2′, H3, H4, H5, H5, H3′-OH, and H5′-OH) were present as well as the H8-dG resonance (Figure 3). The BPQ aromatic resonances exhibited as a low field singlet at 8.65 ppm (H6), a multiplet at 8.278.38 ppm (H1,3-5,11,12), a triplet at 8.15 ppm (H2), and resonances at 5.09 (d), 6.64 (d), and 6.78 ppm (pseudo triplet). Two-dimensional COSY connectivity studies identified that the resonance at 6.78 ppm was coupled
Chem. Res. Toxicol., Vol. 17, No. 6, 2004 831
to both the 5.09 and the 6.64 ppm resonances and that these three protons were not coupled to other protons. This suggested that the resonances at 5.09, 6.64, and 6.78 ppm were H9, H10-OH, and H10, respectively (Table 2A). The D2O exchange experiment indicated five exchangeable protons, two from the deoxyribose, one from H10-OH, and two subsequently identified as H8-OH and N2H-dG. The key feature in the 13C NMR spectrum (Table 2B) was the presence of one carbonyl resonance at 192.0 ppm. On the basis of the 13C NMR spectrum of BPQ (C7, 180.6 ppm; C8, 179.5 ppm; data not shown), the 192 ppm resonance in BPQ-dG1,2 was assigned as C7. A new resonance was also observed at 84.9 ppm. This carbon resonance was found not to bear a hydrogen by the HSQC experiment, and its chemical shift was consistent with a carbon bearing at least two heteroatoms. These data suggested that one of the dG nitrogens had cyclized at the C8 carbonyl of BPQ generating H8-OH. HMBC experiments indicated cross-correlations between C2-dG and H9, as well a cross-correlation between C8 and H10-OH. ROESY studies confirmed these assignments as crosscorrelations were observed as follows: H8-OH with H9; H10
Chart 1. Proposed ESI-MS Fragments of BPQ-dG and BPQ-dA Adducts
a
Value obtained from negative mode ESI. bValue obtained from positive mode ESI.
832
Chem. Res. Toxicol., Vol. 17, No. 6, 2004
Figure 3.
1H
Balu et al.
NMR spectrum (300 MHz) of BPQ-dG1 recorded in DMSO-d6.
Table 2. Key Proton NMR Resonances Used in Adduct Structure Identification and Critical BPQ-Nucleoside Adductsa
13C
Resonances from
Section A adductb
H2-dA
N1H-dG
BPQ-dG1
H8-OH
9.51, bs
BPQ-dG3 BPQ-dA1
N2H-dG
9.00, s
H9
7.51, bs
d, J9,10 ) 2.99 4.50,c m
6.98, s
8.75, s
H9-OH
H10
5.09,c
4.48,c m
6.53, bs
6.78,c
t
H10-OH 6.64, d, J10-OH,10 ) 3.13
7.23,c d, J10,9 ) 4.19 6.81,c d, J10,9 ) 3.78
6.60, d, J9-OH,9 ) 3.55 6.38, bs
Section B adductd/carbon no.
C7
C8
C9
C10
BPQ-dG1
192.0 192.7 193.3
84.9 176.1 85.5
68.1 68.6 67.5
59.6 48.1 55.5
BPQ-dG3 BPQ-dA1
a Data: chemical shifts in ppm; multiplicity and coupling constants in Hertz. b The corresponding diastereomeric adducts showed similar resonance patterns. c COSY experiments reveal that H9 and H10 are coupled. d The corresponding diastereomeric adducts each exhibited similar 13C resonances.
with H11, H9; and H10-OH with H9, H11. ROESY experiments also resolved the position of attachment of the two dG nitrogens as weak cross-correlations were observed between N2H-dG and H9 indicating a structure with N2dG bonded to C9 and N1-dG bonded to C8 (Figure 4). This experiment was repeated with different samples with the same results. One-dimensional NOE difference studies confirmed the proximity of these protons as irradiation at the 5.09 ppm resonance (H9) gave responses at the 7.51 ppm (H8-OH) and a weak response at the resonance at 9.51 (N2H-dG). On the basis of these data, BPQ-dG1,2 were assigned structures as diastereomers of 8-N1,9-N2deoxyguanosyl-8,10-dihydroxy-9,10-dihydrobenzo[a]pyren7(8H)-one (Scheme 1).
Similar to the previous adducts, the structural characterizations of BPQ-dG3,4 were based on UV, mass spectra, and NMR spectroscopy data. The UV spectra of BPQ-dG3,4 revealed identical spectra with aborbances at λmax 278, 330, and 346 nm, demonstrating slight bathochromic shifts as compared to the UV spectra of BPQdG1,2 and suggesting that both BPQ and dG chromophores were present (Figure 2). The mass spectra of BPQ-dG3,4 were also identical showing molecular ions at m/z ) 564 (M - 1) in the negative mode and m/z ) 566 (M + 1) in the positive mode, which represented a molecule with the elements of BPQ, dG, and an atom of oxygen from a molecule water (molecular weight ) 565) (Table 1 and Chart 1). In negative mode ESI-MS, only
Novel, Stable Deoxyguanosine and Deoxyadenosine Adducts
Figure 4. Partial ROESY spectrum of BPQ-dG1 showing interactions of H9 with H10-OH, H10, and N2H-dG.
one major fragmentation was recorded, a loss of dG (molecular weight ) 267) giving an ion that can be represented as 9-hydroxyBPQ. In positive mode ESI-MS, a loss of 116 [2-(hydroxymethyl)-2,3-dihydrofuran-3-ol] from deoxyribose was observed. The observed ion (m/z ) 450) was consistent with 9,10-dihydro-10-(N2-guanyl)-9hydroxybenzo[a]pyrene-7,8-dione. NMR spectroscopy was used to further identify BPQ-dG3,4. The proton NMR spectra of both BPQ-dG3,4 were quite comparable. All of the resonances expected from the deoxyribose moiety were observed (Figure 5). Unlike the proton NMR spectra of BPQ-dG1,2, where the lowest field BPQ proton was H6, the lowest field proton in BPQ-dG3,4 was H11, absorbing at 9.29 ppm. Other downfield BPQ protons were H6 (8.97 ppm), H2 (8.18 ppm), and a complex multiplet at 8.28-8.41 ppm (H1,3-5,12). The purine proton (H8-dG) presented as a singlet at 7.93 ppm. The proton absorbances at 9.0 and 6.98 ppm were assigned to the N1H-dG and N2H-dG, respectively. The remaining proton absorbances at 7.23 (d), 6.60 (d), and 4.50 (m) ppm were assigned to H10, H9-OH, and H9 of the BPQ moiety, respectively, based on the following observations. The chemical shift of H10 was consistent with those recorded for the related structures, the peracetylated benzo ring diol epoxide PAH adducts of dG. These H10 proton chemical shifts ranged from 6.12 to 6.64 ppm for trans adducts and 6.80-7.03 ppm for cis adducts (34). However, the BPQ-dG3,4 proton chemical shifts were somewhat different than those observed with BPQ-dG1,2 (Table 2A). Two-dimensional COSY experiments indicated that the 4.50 ppm (H9) resonance was coupled to protons at 6.60 (H9-OH) and 7.23 ppm (H10) and these three protons were not coupled to other protons. The D2O exchange experiment indicated five exchangeable protons, two from the deoxyribose, one from H9-OH, and two identified as N1H-dG and N2H-dG. In addition to the expected resonances, the 13C NMR spectrum revealed differences when compared to the BPQ-dG1,2 adducts (Table 2B). Two carbonyl resonances were detected at 192.7 and 176.1 ppm, representing C7 and C8, respectively. The ROESY experiment gave the following crosscorrelations: H11 with H10, H12; H10 with H9-OH, H9; and H9 with N2H-dG.). On the basis of these data, BPQ-dG3,4 were assigned structures as diastereomers of 10-(N2deoxyguanosyl)-9,10-dihydro-9-hydroxybenzo[a]pyrene7,8-dione (Scheme 1). BPQ-dA Adducts. The crude mixture from the relatively longer reaction of BPQ with dA produced two
Chem. Res. Toxicol., Vol. 17, No. 6, 2004 833
major adducts, BPQ-dA1 and BPQ-dA2, eluting at 19.4 and 21.4 min, respectively, along with some minor side products (Figure 1B). Both BPQ-dA1,2 were collected, and an examination of the UV spectra revealed the existence of two identical adducts with major absorbances at λmax 265 (sh), 285, 333, and 348 nm. Furthermore, the observed bathochromic shift in the UV spectra of BPQ-dA1,2 as compared to the data of the BPQ-dG adducts helped in the preliminary identification of an adenosine moiety (Figure 2). The ESI-MS analysis of BPQ-dA1,2 indicated that both fragmented in a similar manner in both positive and negative ESI modes (Table 1 and Chart 1). Both adducts gave the molecular ions of m/z ) 548 (M - 1) in the negative mode and m/z ) 550 (M + 1) in the positive mode. This suggested that each adduct had a molecular weight of 549 and contained 1 mol each of BPQ, dA, and an atom of oxygen from a water molecule as observed in the case of BPQ-dG adducts. However, the fragmentation patterns observed for BPQ-dA1,2 were different than those of the dG adducts. The negative mode ESI-MS of both of these adducts showed three fragments at m/z ) 297, 250, and 269. Fragments m/z ) 297 (M - 1 - dA) and 250 (dA) were indicative of the loss of dA from the molecular ion (m/z ) 548). This retroaddition of nucleoside in the fragmentation is similar to that observed with the BPQdG3,4. An additional ion at m/z ) 269 suggested the elimination of CO from the M - 1 - dA ion (m/z ) 297). Furthermore, the positive mode ESI spectra of both adducts BPQ-dA1,2 gave an ion at m/z ) 434 indicating the loss of a deoxyribose moiety [2-(hydroxymethyl)-2,3dihydrofuran-3-ol] from the molecular ion peak at m/z ) 550. The proton NMR spectra of BPQ-dA1,2 were almost identical with very minor variations. All of the expected resonances for the deoxyribose, adenine, and BPQ moieties were identified (Figure 6). Some key features of the 1 H and 13C NMR resonances, which assisted in assigning structures to these adducts, are highlighted in Table 2A,B. The D2O exchange experiment indicated a total of four exchangeable protons, two from the deoxyribose, one from H9-OH, and one from the subsequently identified H8-OH. A two-dimensional COSY study identified the resonances at 4.48, 6.38, and 6.81 ppm as H9, H9-OH, and H10, respectively. In summary, the resonance at 6.81 ppm (H10) was correlated with the resonance at 4.48 ppm (H9), and the resonance at 6.38 ppm (H9-OH) was correlated with the resonance at 4.48 ppm (H9), and these three resonances were not coupled to other protons. Interestingly, the key C13 resonances for C7 (193.3 ppm) and C8 (85.5 ppm) were found analogous to the observed data of BPQ-dG1,2 indicating saturation at the C8 carbon. The HSQC experiment confirmed that the C8 carbon (85.5 ppm) had no correlated protons. Thus, the data clearly indicated a bonding between C8 and dA through one of the latter’s nitrogens with the generation of C8-OH. HMBC experiments indicated the following cross-correlations: C2-dA with H10, C6-dA with H10, and C10 with H2-dA suggesting covalent bonding between the dA and the BPQ molecules. The most definitive data were obtained from ROESY NMR experiments. The strong ROESY cross-peaks observed between the H10, the H11, and the H2-dA clearly indicated the close proximity of these three nuclei in space (Figure 7). These data were verified by three-dimensional molecular modeling studies, which confirmed that all of the above nuclei are in close proximity (
