Sequence Distribution of Acetaldehyde-Derived - ACS Publications

Chem. Res. Toxicol. 2007, 20, 1379–1387

1379

Sequence Distribution of Acetaldehyde-Derived N2-Ethyl-dG Adducts along Duplex DNA Brock Matter,† Rebecca Guza,† Jianwei Zhao,‡ Zhong-ze Li,§ Roger Jones,‡ and Natalia Tretyakova*,† UniVersity of Minnesota Cancer Center and Department of Medicinal Chemistry, MMC 860, 420 Delaware Street Southeast, Minneapolis, Minnesota 55455, Department of Chemistry, Rutgers UniVersity, Piscataway, New Jersey 08854, and Biostatistics Core, UniVersity of Minnesota Cancer Center, Minneapolis, Minnesota 55455 ReceiVed April 12, 2007

Acetaldehyde (AA) is the major metabolite of ethanol and may be responsible for an increased gastrointestinal cancer risk associated with alcohol beverage consumption. Furthermore, AA is one of the most abundant carcinogens in tobacco smoke and induces tumors of the respiratory tract in laboratory animals. AA binding to DNA induces Schiff base adducts at the exocyclic amino group of dG, N2ethylidene-dG, which are reversible on the nucleoside level but can be stabilized by reduction to N2ethyl-dG. Mutagenesis studies in the HPRT reporter gene and in the p53 tumor suppressor gene have revealed the ability of AA to induce G f A transitions and A f T transversions, as well as frameshift and splice mutations. AA-induced point mutations are most prominent at 5′-AGG-3′ trinucleotides, possibly a result of sequence specific adduct formation, mispairing, and/or repair. However, DNA sequence preferences for the formation of acetaldehyde adducts have not been previously examined. In the present work, we employed a stable isotope labeling-HPLC-ESI+-MS/MS approach developed in our laboratory to analyze the distribution of acetaldehyde-derived N2-ethyl-dG adducts along double-stranded oligodeoxynucleotides representing two prominent lung cancer mutational “hotspots” and their surrounding DNA sequences. 1,7,NH2-15N-2-13C-dG was placed at defined positions within DNA duplexes derived from the K-ras protooncogene and the p53 tumor suppressor gene, followed by AA treatment and NaBH3CN reduction to convert N2-ethylidene-dG to N2-ethyl-dG. Capillary HPLC-ESI+-MS/MS was used to quantify N2-ethyl-dG adducts originating from the isotopically labeled and unlabeled guanine nucleobases and to map adduct formation along DNA duplexes. We found that the formation of N2-ethyl-dG adducts was only weakly affected by the local sequence context and was slightly increased in the presence of 5-methylcytosine within CG dinucleotides. These results are in contrast with sequence-selective formation of other tobacco carcinogen–DNA adducts along K-ras- and p53-derived duplexes and the preferential modification of endogenously methylated CG dinucleotides by benzo[a]pyrene diol epoxide and acrolein. Introduction Lung cancer is responsible for 30% of all cancer deaths worldwide and is expected to kill over 160000 Americans this year (1). An estimated 87% of total lung cancer cases are the result of cigarette smoking (2). Binding of tobacco carcinogens to genomic DNA is considered critical for lung cancer initiation in smokers (3, 4). The resulting chemically modified nucleobases (DNA adducts), if not repaired prior to DNA replication, can cause polymerase errors and induce mutations in critical genes (5, 6). In addition, chemical modification of gene promoter regions by carcinogen metabolites can lead to heritable changes of gene expression (7–10). Smoking-related lung tumors contain characteristic mutations in the p53 tumor suppressor gene and the K-ras protooncogene, suggesting that these genetic changes may result from exposure to tobacco carcinogens (11, 12). Smoking-associated K-ras mutations are mostly G fT transversions and G fA transitions * To whom correspondence should be addressed. Tel: 612-626-3432. Fax: 612-626-5135. E-mail: [email protected]. † University of Minnesota Cancer Center and Department of Medicinal Chemistry. ‡ Rutgers University. § Biostatistics Core, University of Minnesota Cancer Center.

found specifically at codon 12 (GGT f GTT, TGT, GAT), which have been shown to activate the K-ras protooncogene, leading to uncontrolled cell growth (13). The p53 mutations found in lung tumors of smokers are primarily G fT transversions concentrated at the endogenously methylated CG dinucleotides within exons 5–8 (14). It has been proposed that these genetic changes result from the preferential binding of diol epoxide metabolites of bay region polycyclic aromatic hydrocarbons (PAH)1 present in tobacco smoke, for example, benzo[a]pyrene diol epoxide (BPDE), to methylated CG sites and p53 codons 157, 158, 248, and 273 (15, 16). However, in addition to BPDE and other PAH, tobacco smoke contains over 60 other carcinogens, all of which may target MeCG dinucleotides, contributing to mutagenesis at these sites. This is especially true for carcinogen metabolites that form N2-guanine adducts, because the same factors that increase the reactivity of the N2 position of dG towards BPDE may enhance its reactivity towards other electrophiles. For example, acrolein has been recently found to preferentially bind to p53 mutational “hotspots”, 1 Abbreviations: AA, acetaldehyde; BPDE, benzo[a]pyrene diol epoxide; N2-Et-dG , N2-ethyl-dG; 15N3,13C1-dG, 1,7-NH2-15N-2-13C-dG; NaBH3CN, sodium cyanoborohydride; NNK, 4-(methylnitrosamino)-1-(3-pyridyl)-1butanone; PAH, polycyclic aromatic hydrocarbons.

10.1021/tx7001146 CCC: $37.00  2007 American Chemical Society Published on Web 09/15/2007

1380 Chem. Res. Toxicol., Vol. 20, No. 10, 2007

Communications

Scheme 1. Formation of N2-Ethylidene-dG Adducts and Their Reduction to Stable N2-Et-dG Adducts

Table 1. DNA Sequences Selected for This Study molecular weight sequence 15

13

5′-G1G2A[ N3, C1-G3]CTG4G5TG6G7CG8TAG9G10C-3′ 5′-G1G2AG3CT[15N3,13C1-G4]G5TG6G7CG8TAG9G10C-3′ 5′-G1G2AG3CTG4[15N3,13C1-G5]TG6G7CG8TAG9G10C-3′ 5′-G1G2AG3CTG4G5T[15N3,13C1-G6]G7CG8TAG9G10C-3′ 5′-G1G2AG3CTG4G5TG6[15N3,13C1-G7]CG8TAG9G10C-3′ 5′-G1G2AG3CTG4G5TG6G7C[15N3,13C1-G8]TAG9G10C-3′ 5′-GCCTACGCCACCAGCTCC-3′ 5′-CCMeC[15N3,13C1-G1]G2CACCMeCG3MeCG4TCMeCG5MeCG6-3′ 5′-CCMeCG1[15N3,13C1-G2]CACCMeCG3MeCG4TCMeCG5MeCG6-3′ 5′-CCMeCG1G2CACCMeC[15N3,13C1-G3]MeCG4TCMeCG5MeCG6-3′ 5′-CCMeCG1G2CACCMeCG3MeC[15N3,13C1-G4]TCMeCG5MeCG6-3′ 5′-CCMeCG1G2CACCMeCG3MeCG4TCMeC[15N3,13C1-G5]MeCG6-3′ 5′-MeCGMeCGGAMeCGMeCGGGTGCMeCGGG-3′ 5′-ATGGGC[15N3,13C1-G]GCATGAAC-3′ 5′-ATGGGMeC[15N3,13C1-G]GCATGAAC-3′ 5′-GTTCATGCCGCCCAT-3′ 5′-GTTCATGCMeCGCCCAT-3′

including codons 152, 154, 156, 157, and 158 in exon 5, codons 248 and 249 in exon 7, and codons 273 and 282 in exon 8 (17). Acetaldehyde (AA) is among the most abundant carcinogens in tobacco smoke, with concentrations in cigarette smoke 1000fold greater than those of PAHs and tobacco specific nitrosamines (3). AA induces mutations, sister chromatid exchanges, micronuclei, and aneuploidy in vitro and in vivo (18). It is carcinogenic in laboratory animals, causing tumors of the respiratory tract (19), and is classified by the U.S. Department of Health as “reasonably anticipated to be a human carcinogen” (20). Oxidative metabolism of ethanol to AA is associated with alcohol-related cancers of the upper GI tract in alcohol users and in aldehyde dehydrogenase-deficient individuals (21). Like BPDE, AA modifies the N2 position of guanine in DNA (Scheme 1) (22). The resulting imine adducts (N2-ethylidenedG) can undergo reduction to irreversible N2-ethyl-dG (N2-EtdG) lesions (22). In addition, AA is capable of inducing several other DNA lesions, including 1,N2-propano-dG and interstrand DNA cross-links (23). N2-Ethylidene-dG adducts appear to be the predominant AA–DNA lesions in humans but are too unstable to be analyzed directly and must be reduced to N2-EtdG in the presence of NaBH3CN prior to HPLC-MS/MS analysis (23–25). N2-Et-dG itself has been detected in the DNA of peripheral blood cells of alcohol abusers, in human buccal cell DNA following exposure to AA, and in human urine (22, 26, 27). These adducts are hypothesized to be derived by in vivo reduction of N2-ethylidene-dG mediated by vitamin C, glutathione or other yet unidentified reducing agents (22, 27). While the biological effects of N2-ethylidene-dG have not been analyzed because of their limited stability, N2-Et-dG has been shown to induce G to C transversions during DNA synthesis by Escherichia coli Pol I (28). In contrast, studies with human DNA polymerases have shown that N2-Et-dG strongly blocks synthesis catalyzed by human replicative polymerase R (29), while human translesional polymerases η and ι bypass the adduct in an error-free manner (30, 31). Site-specific mutagenesis

ID 15

13

(+) K-ras -[ N3, C1-G3] (+) K-ras -[15N3, 13C1-G4] (+) K-ras -[15N3, 13C1-G5] (+) K-ras -[15N3, 13C1-G6] (+) K-ras -[15N3, 13C1-G7] (+) K-ras -[15N3, 13C1-G8] (-) K-ras p53 exon 5-[15N3, 13C1-G1] p53 exon 5-[15N3, 13C1-G2] p53 exon 5-[15N3, 13C1-G3] p53 exon 5-[15N3, 13C1-G4] p53 exon 5-[15N3, 13C1-G5] (-) p53 exon 5 (+) codon 245 (+) codon 245 MeC (-) codon 245 (-) codon 245 MeC

calculated

observed

5640.7 5640.7 5640.7 5640.7 5640.7 5640.7 5365.5 5785.7 5785.7 5785.7 5785.7 5785.7 5981.9 4646.1 4660.1 4504.0 4518.0

5639.7 5640.0 5639.9 5640.0 5640.0 5640.0 5365.0 5786.0 5785.9 5785.8 5784.9 5785.9 5981.4 4645.9 4659.7 4503.1 4517.4

experiments in the human cell line indicate that N2-Et-dG blocks translesion DNA synthesis and induces deletion mutations (32). Matsuda et al. have shown that N2-Et-dGTP was efficiently incorporated opposite dC in DNA polymerization catalyzed by mammalian Pol R and δ, revealing another possible source of N2-Et-dG in genomic DNA (26). Little is known about DNA sequence effects on the formation of AA–DNA adducts. Because N2-ethylidene-dG and N2-EtdG lesions are not recognized by the UvrABC nuclease and are not cleavable by hot piperidine, their sequence locations cannot be readily determined by methods that rely on gel electrophoresis separation of the DNA fragments generated by enzymatic or chemical strand cleavage at the adducted nucleotides. In the present work, a new approach based on stable isotope labeling of DNA was developed to map the formation of N2-Et-dG adducts along short DNA duplexes representing frequently mutated regions of the K-ras and p53 genes following AA treatment and quantitative reduction of N2-ethylidene-dG to N2-Et-dG in the presence of NaBH3CN.

Experimental Procedures Chemicals. Ammonium acetate, acetonitrile, methanol, zinc chloride, phosphate, potassium hydroxide, tris-HCl, and sodium hydroxide were from Fisher Scientific (Hanover Park, IL). AA, nuclease P1 and alkaline phosphatase were from Sigma-Aldrich (Milwaukee, WI). N2-Et-dG and 15N5-N2-Et-dG standards were generously provided by Professor Stephen Hecht (University of Minnesota). DNA Oligodeoxynucleotides. DNA duplexes representing major K-ras and p53 lung cancer mutational “hotspots” and surrounding sequences selected for this study (Table 1) have been prepared previously as described (33, 34). For each sequence, a series of DNA strands was generated containing the 15N3, 13C1 stable isotope tag at a different guanine base (Table 1). All synthetic oligodeoxynucleotides were synthesized by standard phosphoramidite chemistry using an Applied

Communications

Biosystems ABI 394 instrument (Foster City, CA). 15N3, 13C1dG was introduced by using 1,7,NH2-15N-2-13C-dG (15N3, 13C1dG) phosphoramidite (35, 36). Both labeled and unlabeled DNA oligomers were purified by HPLC and structurally characterized as described previously (33). The identity and purity of each strand was confirmed by HPLC-ESI--MS (Table 1). Each isotopically tagged DNA oligomer was annealed to an equimolar amount of the corresponding unlabeled complementary strand. AA Treatment of DNA and Reduction of N2-Ethylidene-dG to N2-Et-dG. Isotopically tagged or unlabeled DNA duplexes (2 nmol, in triplicate) were dissolved in 20 µL of 50 mM sodium phosphate buffer, pH 7 to yield a final concentration of 100 µM and treated with 4 mM AA for 96 h at 4 °C. To convert N2-ethylidene-dG to N2-Et-dG, AAtreated DNA was subjected to reduction in the presence of 80 mM NaBH3CN (25 °C for 5 h). These conditions have been reported to completely convert N2-ethylidene-dG to N2-Et-dG (23, 24). Following reduction, DNA was precipitated with cold ethanol. The same results were obtained if the reduction and enzymatic digestion steps were performed at the same time (data not shown). Enzymatic Hydrolysis of AA-Treated DNA and Isolation of N2-Et-dG. DNA samples (2 nmol) were dissolved in 20 µL of 25 mM ammonium acetate/2.5 mM ZnCl2 buffer (pH 5.3). Unlabeled duplexes used for the concentration dependence studies were spiked with 15N5-N2-Et-dG (800 fmol, internal standard for mass spectrometry). DNA hydrolysis was performed in the presence of Nuclease P1 (8 U) and alkaline phosphatase (22 U) (1 h at 37 °C). The completeness of enzymatic hydrolysis was confirmed by HPLC-UV (37). N2Et-dG and its isotope analogs were isolated by solid phase extraction on Oasis mixed mode solid phase extraction cartridges using published methods (24, 25). Capillary HPLC-ESI-MS/MS. Quantitative analysis of N2Et-dG and 15N3, 13C1-N2-Et-dG in DNA hydrolysates was performed by capillary HPLC-ESI+-MS/MS. A TSQ Quantum Ultra mass spectrometer interfaced with an Agilent Technologies 1100 capillary HPLC was employed in all analyses. An Agilent Extend C18 column (0.5 × 100 mm, 3.5 µm) was eluted at a flow rate of 13 µL/min with a gradient of acetonitrile (solvent B) in 15 mM ammonium acetate (solvent A). HPLC solvent composition was gradually changed as follows: 0 min, 7% B; 13 min, 17.9% B; 14 min, 7% B; 25 min, 7% B. The mass spectrometer was operated in the ESI+ MS/MS mode. Typically, the source temperature was 250 °C, and the spray voltage was set to 2.9 kV. Quantitative analyses were performed in the selected reaction monitoring mode. The first quadrupole was set to isolate the protonated molecules ([M + H]+) of N2-EtdG (m/z 296.1), 15N3, 13C1-N2-Et-dG (m/z 300.1), and 15N5N2-Et-dG (m/z 301.1), and their fragmentation was induced in the second quadrupole serving as a collision cell. Typical collision energy was 14 V, and a collision gas pressure (Ar) was 1 mTorr. The third quadrupole was set to detect the product ions corresponding to a neutral loss of deoxyribose ([M + 2H - dR]+): m/z 180.2 for N2-Et-dG, m/z 184.2 for 15N3, 13C1-N2Et-dG, and m/z 185.2 for 15N5-N2-Et-dG (Figure 1). The lower limit of detection for N2-Et-dG standard was 300 amol (S/N ) 10), comparable to the previously reported values (24, 25). HPLC-ESI+-MS/MS calibration curves for N2-Et-dG using 15 N5-N2-Et-dG as an internal standard were linear from 5 fmol to 10 pmol N2-Et-dG on column (See Supplement S-1). The extent of N2-Et-dG formation at the 15N3,13C1-labeled guanine (X) was calculated from the corresponding amounts of N2-Et-dG using the following equation:

Chem. Res. Toxicol., Vol. 20, No. 10, 2007 1381

% reaction at X ) AEt-dX/(AEt-dX + AEt-dG) × 100% (1) where AEt-dG and AEt-dX are the areas under the HPLC-ESIMS/MS peaks corresponding to the unlabeled and [15N3, 13C1]labeled N2-Et-dG, respectively. Statistical Analyses of the Data. All statistical analyses were performed at the Biostatistics Core of the University of Minnesota Cancer Center. Analysis of variance (ANOVA) (38) was used to investigate the overall effect of DNA sequence on AA adduct formation at a given site. Subsequent pair-wise differences were tested, and the Bonferroni method was applied to adjust p values of pair-wise comparisons. To compare the reactivity at a specified guanine nucleobase with the theoretical “random” reactivity value, the following statistic was used, (µi – c)/[MSE(1/ni)]1/2, where µi is the mean reactivity at position i, c is the theoretical reactivity value, and MSE is the mean squared error from ANOVA. The Bonferroni method was used to maintain the overall level of significance for multiple comparisons. All statistical analyses were conducted in SAS (Statistical Analysis Software) version 9.1. The significance level was set up to 5%.

Results Stable Isotope Labeling Approach. Our experimental strategy is based on the stable isotope labeling HPLC-MS/MS approach developed in our laboratory (Scheme 2) (39, 40). DNA duplexes were constructed representing small regions of K-ras exon 1 (containing frequently mutated codon 12) and p53 exon 5 (containing lung cancer mutational “hotspots” at codons 157 and 158) (Table 1). The formation of N2-Et-dG at a given guanine was investigated by inserting a 15N3,13C1-labeled guanine at that site (Table 1). For each target sequence, a series of DNA oligomers were generated in which one of the guanine bases was replaced with 15N3,13C1-Gua (Table 1). Isotopically labeled oligomers were annealed to the equimolar amounts of the complementary strands. Following treatment with AA, N2ethylidene-dG, adducts were stabilized as N2-Et-dG under conditions that have been reported to result in a quantitative reduction (23, 24). Alkylated DNA was enzymatically digested to deoxyribonucleosides, and HPLC-ESI+-MS/MS was used to establish the extent of adduct formation at the 15N3,13C1-labeled guanine (e.g., Figure 2). By preparing several oligomers of the same sequence but with different label positions (Table 1), the pattern of adduct formation within each sequence of interest was determined. Our previous experiments employing stable isotope labeling methodology demonstrate that the oligodeoxynucleotides selected for our studies do not exhibit any significant “end effects” as a result of local duplex melting at the ends of the sequence (33, 40). To ensure that DNA remained in a double-stranded form, AA treatment was performed at a reduced temperature (4 °C). A representative HPLC-ESI+-MS/MS trace corresponding to 15 N3,13C1-N2-Et-dG and N2-Et-dG originating from an 15N3,13C1dG-containing DNA duplex following AA treatment and reduction with NaBH3CN is shown in Figure 2. HPLC-ESI+-MS/ MS quantification of N2-Et-dG is based on the loss of deoxyribose (M ) 116) from the protonated molecules of the adducted nucleoside (m/z 296.1, [M + H]+) under collisioninduced dissociation conditions, leading to the predominant fragment at m/z 180.2 (MS/MS transition: m/z 296.1 [M + H]+f 180.2 [M + 2H – dR]+) (Figures 1A and 2A). Adducts originating from 15N3,13C1-dG contain the 15N3,13C1 isotope tag and thus undergo a +4 mass shift (m/z 300.1 [M + H]+f 184.2

1382 Chem. Res. Toxicol., Vol. 20, No. 10, 2007

Figure 1. ESI+-MS/MS spectra of N2-Et-dG (A) and

Communications

15

N3,13C1-N2-Et-dG (B).

Scheme 2. Strategy for Quantitation of AA-Induced N2-Ethyl-dG Lesions Originating from Specific Sites within DNA Sequence

[M + 2H – dR]+) (Figures 1B and 2B). The extent of adduct formation at 15N3,13C1-dG was calculated directly from the ratios of the areas under the HPLC-ESI MS/MS peaks corresponding to 15N3,13C1-labeled and unlabeled N2-BPDE-dG adducts, respectively (Figure 2). A similar procedure was followed for oligomers containing 15N3,13C1-dG at a different site, making it possible to map the patterns of adduct formation along each DNA duplex. Optimization of AA Reaction Conditions and Sample Processing. Our initial studies employed unlabeled DNA duplexes to establish concentration dependence for N2-Et-dG

adduct formation and to identify optimal AA exposure conditions. Our goal was to induce sufficient numbers of N2-Et-dG adducts to enable their detection at each guanine base but without a loss of sequence selectivity. Double-stranded DNA 18-mers were treated with AA at various concentrations (0.5–16 mM), temperatures (4–37 °C), and reaction times (4–96 h), and the amounts of N2-Et-dG adducts present were determined by isotope dilution HPLC-ESI-MS/MS with the 15N5-N2-Et-dG internal standard. We found that AA-dG adduct yields increased in a dose-dependent manner as a function of AA concentration and treatment times (Supporting Information, Figures S-2 and

Communications

Figure 2. HPLC-ESI+-MS/MS analysis of N2-Et-dG (A) and 15 N3,13C1-dG-containing DNA duplex.

Chem. Res. Toxicol., Vol. 20, No. 10, 2007 1383

15

N3,13C1-N2-Et-dG (B) following AA treatment and NaBH3CN reduction of the

S-3). Higher adduct yields were observed at lower temperatures, probably because of the low boiling temperature of AA (21 °C) and the limited stability of N2-ethylidene-dG (23). Therefore, reduced temperature conditions (4 °C) were chosen to maximize AA-dG adduct formation and to help maintain DNA duplex integrity. Interestingly, N2-Et-dG adducts yields increased when the reaction mixtures were frozen prior to reduction, suggesting that AA reaction with DNA can take place at -20 °C (results not shown). Higher numbers (∼10×) of N2-Et-dG adducts were detected if the reducing agent (NaBH3CN) was added before DNA precipitation with ethanol, consistent with previous studies that have shown that AA adducts are reversible and can be readily lost during DNA handling (t1/2 ) 5 min on the nucleoside level and 24 h in DNA at 37 °C) (23, 24). Our optimized reaction conditions include AA treatment of DNA at 4 °C for 96 h, followed by reduction with NaBH3CN to quantitatively convert N2-ethylidene-dG to N2-Et-dG (23, 24), DNA precipitation, and enzymatic hydrolysis. The same results were obtained if NaBH3CN treatment and enzymatic hydrolysis were performed at the same time (data not shown), indicating that under our conditions, reduction efficiency is not influenced by DNA sequence context. The overall numbers of N2-Et-dG adducts in DNA samples treated with 0.5–16 mM AA (96 h at 4 °C) ranged between 0.5 and 16 pmol/nmol DNA (Supporting Information, Figure S-2), indicating that no more than one adduct per DNA molecule was formed (“single hit” conditions). Treatment with 4 mM AA was selected for our adduct distribution studies because it produced ∼4 pmol N2-Et-dG/nmol DNA (0.2–0.4 pmol N2-Et-dG per G for our DNA duplexes that contain between 11 and 16 guanine residues) (Table 1). Given our detection limits for N2-Et-dG in the mid-attamolar range, we estimated that when starting with 2 nmol of DNA per sample, this treatment produced sufficient amounts of adducts at each guanine to be accurately quantified by HPLC-ESI-MS/MS. Our method (Scheme 2) requires that AA-treated DNA is quantitatively digested to deoxynucleosides. To ensure complete digestion, small aliquots of hydrolysates were taken and checked by HPLC with UV detection. The digests were considered complete if the only HPLC peaks observed were the signals of dA, dC, dG, and dT, and the molar ratios of released nucleosides were consistent with the oligonucleotide sequence. We found that the use of nuclease P1 (8 U) and alkaline phosphatase (22

Figure 3. Distribution of AA-induced N2-Et-dG along the K-ras-derived DNA duplex (N ) 4).

U) resulted in a complete digestion following incubation for 1 h at 37 °C. Distribution of N2-Et-dG Adducts along K-ras-Derived DNA Duplex. We first examined the formation of N2-Et-dG adducts of AA at specific guanine nucleobases within a region of the K-ras gene containing codon 12, a major mutational “hotspot” in smoking-induced cancer (Figure 3) (13). A series of DNA 18-mers, 5′-G1G2AG3CTG4G5TG6G7CG8TAG9G10C3′ (codon 12 ) G4G5T), were prepared, each containing a single [15N3,13C1]-dG at G3, G4, G5, G6, G7, or G8 (Table 1). Each oligomer was annealed to the complementary strand, and the resulting duplexes were treated with AA. Following reduction to convert N2-ethylidene-dG to N2-Et-dG, the extent of reaction at each of the labeled guanines was established by isotope ratio HPLC-ESI-MS/MS as described above (Scheme 2). In the absence of any sequence effects on AA-dG adduct formation, the calculated adduct amounts originating from each guanine are 7.7% (100%/13). The actual reactivities of G3–G8 determined from HPLC-ESI-MS/MS peak ratios varied between 5.8 and 7.5% (Figure 3). Despite slight variations of N2-Et-dG adduct yields at different guanines, statistical analyses indicated that none of them were significantly different from uniform reactivity value (7.7%). Furthermore, there was no significant difference between the reactivities of different guanines as indicated by statistical comparisons using pair-wise tests (p > 0.1). Taken together, these results indicate that the formation of N2-Et-dG within the K-ras-derived sequence is consistent with a random distribution, ruling out any sequence effects on

1384 Chem. Res. Toxicol., Vol. 20, No. 10, 2007

Figure 4. Distribution of AA-induced N2-Et-dG along the p53 exon 5-derived DNA duplex (N ) 3).

AA-dG adduct yields within this DNA duplex. This is in contrast with our earlier data for other tobacco carcinogens that revealed selective modification of G5 by diazohydroxide metabolites of the tobacco-specific nitrosamine, 4-(methylnitrosamino)-1-(3pyridyl)-1-butanone (NNK) (12–17% of total O6-dG adducts) (40) and G4 by BPDE (17% of total N2-BPDE-dG adducts) (39). Distribution of N2-Et-dG Adducts along p53 Exon 5-Derived DNA Duplex. Codons 157 and 158 of the p53 tumor suppressor gene are frequently mutated in lung cancer (GTC f TGC and CGC f CTC, respectively) (41) but not in other tumor types, suggesting that they are induced by exposure to tobacco carcinogens (42). We investigated the distribution of AA-induced dG adducts along p53 exon 5-derived sequence, 5′-CCMeCG1G2CACCMeCG3MeCG4TCMeCG5MeCG6-3′, containing lung “hotspots” at codons 157 (G4TC) and 158 (MeCG5Me C) (Figure 4). A series of DNA duplexes were prepared in which one of the highlighted guanines (G1, G2, G3, G4, or G5) was replaced with [15N3,13C1]G (Table 1). 5-Methylcytosine (MeC) was introduced at all physiologically methylated sites (43). Following AA treatment, the extent of N2ethylidene-dG formation at each site was established from isotope ratios as described above. Because our p53 exon 5-derived duplex contains a total of 17 guanines, uniform reaction would lead to ∼5.9% of reaction at each of them (100%/17 ) 5.88%). Unlike our results for the K-ras-derived duplex (Figure 3), AA-dG adduct formation pattern within the p53 exon 5 sequence was not completely random (Figure 4). The highest numbers of N2-Et-dG adducts originated from the two guanine bases that are flanked by methylated cytosines (G3 and G5, MeCGMeC) (5.2–5.9 % of total adducts), suggesting that cytosine methylation may have a stimulating effect on the formation of N2-ethylidene-dG. In contrast, G1, G2, and G4 exhibited lower than expected reactivity (3.1–4.5 %). Pair-wise comparisons indicated that, with the exception of G1 and G2, AA adduct numbers formed at different guanines were significantly different from each other. However, sequence selectivity for AA-dG adduct formation in this duplex is modest in comparison with other tobacco carcinogen metabolites investigated by our group. For example, NNK-derived diazohydroxides specifically target G2 (G1G2MeC) (12% of total O6-Me-dG adducts) (34), while BPDE-dG adducts are overproduced at G3 (MeC G3MeC, 18% of total N2-BPDE-dG adducts) (33). Effect of Cytosine Methylation on N2-Et-dG Adduct Formation. Enzymatic methylation of cytosine to produce 5-methylcytosine (MeC) takes place at 60–80% of CpG sites in the human genome, including all CpG dinucleotides along exons 5–8 of the p53 tumor suppressor gene (43, 44). Once established early in the development, DNA methylation patterns are faithfully maintained by the action of methylase enzymes (45).

Communications

However, many tumor types, including smoking-induced lung cancer, are characterized by aberrant DNA methylation patterns, resulting in silencing of tumor suppressor genes, activation of oncogenes, and decreasing chromosome stability (46). The presence of a 5-methyl group on cytosine within MeCG sites induces a small change in DNA structure and dynamics, leading to altered DNA–protein interactions and chromatin remodeling (47–49). It is also capable of increasing the reactivity of guanine bases in MeCG dinucleotides towards carcinogens (16, 50–58). We have previously shown that a neighboring MeC can have opposite effects on the reactivity of guanine bases towards two prominent tobacco carcinogens, a tobacco-specific nitrosamine (NNK) and a PAH, benzo[a]pyrene (B[a]P) (33, 59). While B[a]P diol epoxides preferentially modify guanine bases within endogenously methylated MeCG dinucleotides along p53derived DNA sequences (33), the same sites are protected against alkylation by NNK metabolites (59). Our results for AAmediated modification of the p53 exon 5-derived duplex (Figure 4) suggest that cytosine methylation may facilitate the formation of N2-Et-dG adducts at neighboring guanines. Therefore, a more systematic study was initiated to examine the effects of MeC on AA-dG adduct formation. Synthetic oligomers containing isotopically labeled guanine within p53 codon 245 sequence (5′-ATG1G2G3X[15N3,13C1G4]G5CATG6AAC-3′, X ) C or MeC) were used to establish the effects of cytosine methylation on N2-ethylidene-dG adduct yields at the neighboring guanine (Table 1). Codon 245 (G4G5C) contains a major lung cancer mutational “hotspot” preceded by an endogenously methylated C (MeCG4) (60). Four DNA duplexes were prepared, where 15N3,13C-G (G) was placed in unmethylated, hemimethylated, or fully methylated CG sequence:

Following AA treatment, the extent of adduct formation at N3,13C1-G in duplexes with different cytosine methylation status was determined by the stable isotope-labeling HPLCESI-MS/MS approach described above (Scheme 2). We found that N2-Et-dG adduct formation was significantly increased in the presence of neighboring MeC (p < 0.0001), with a maximum effect (∼60% increase) achieved in fully methylated CG dinucleotides (Figure 5). Similarly, a 24–26% increase in adduct yields was observed for MeC introduced at the CG dinucleotide within p53 codon 157 (Supporting Information, Figure S-4). These results indicate that endogenous cytosine methylation has a slight stimulating effect on the formation of AA-dG adducts. However, a MeC-associated increase in AA adduct yields is much more modest than that previously observed for N2-BPDE-dG adducts (33). 15

Discussion The initiation of smoking-induced lung cancer is thought to be triggered by the covalent binding of tobacco carcinogens and their metabolites to DNA nucleobases (4). If not repaired, the resulting DNA adducts can be misread by DNA polymerases, giving rise to heritable mutations in critical genes. Because tobacco smoke contains over 60 known carcinogens that are capable of forming a variety of DNA adducts, it is difficult to draw a link between the mutations observed in lung tumors of smokers and specific tobacco carcinogen–DNA adducts. A thorough examination of the patterns of tobacco carcinogen–DNA adduct formation along critical genes, for example,

Communications

Chem. Res. Toxicol., Vol. 20, No. 10, 2007 1385

Figure 5. Effect of cytosine methylation within p53 codon 245 on N2-Et-dG adduct formation at G4 (N ) 4).

the K-ras protooncogene and the p53 tumor suppressor gene, may provide insight into the origins of mutational “hotspots” identified in lung tumors of smokers. Our previous studies employed a stable isotope-labeling HPLC-ESI+-MS/MS methodology to analyze the distribution of N2-BPDE-dG lesions of a potent tobacco carcinogen, B[a]P, within physiologically methylated double-stranded DNA sequences derived from the frequently mutated regions of p53 exons 5, 7, and 8 and K-ras exon 1 (33, 39). We found that that the presence of 5-methylcytosine (MeC) increased the reactivity of BPDE towards neighboring guanine bases (33, 39). These results were consistent with the earlier findings of Denissenko et al., who employed nuclease excision assay in combination with ligation-mediated PCR (15). However, more recent studies by the Tang group revealed that acrolein targets the same sites within the p53 gene (17), raising the possibility that other components of tobacco smoke may preferentially bind to p53 and K-ras mutational “hotspots”. The goal of the present study was to identify the sites of the preferential formation of AA-dG adducts along the regions of the K-ras and p53 genes that are targets for smoking-induced mutagenesis. AA is present in tobacco smoke at a very high concentrations (4) and is capable of inducing mutations and cancer (18, 19). However, DNA sequence preferences for the formation of AA-DNA adducts have not been previously investigated, probably because N2-Et-dG lesions cannot be readily converted to DNA strand breaks for analysis by gel electrophoresis-based methods. Therefore, a novel methodology based on stable isotope-labeling HPLC-ESI-MS/MS (39) was employed (Scheme 2). DNA sequences selected for this study represented major K-ras and p53 lung cancer mutational “hotspots” (K-ras codon 12 and p53 codons 157 and 158) and their surrounding sequences (Table 1). Following AA treatment and quantitative reduction of the resulting N2-ethylidene-dG lesions to N2-EtdG, DNA was digested to free deoxynucleosides, and the relative extent of reaction at the isotopically labeled position was established by mass spectrometry (Figure 2). Because the molecular weight of AA lesions originating from the 15N3,13C1labeled guanine is increased by 4 mass units, they can be readily distinguished from the corresponding adducts formed at unlabeled guanines elsewhere in the sequence (Scheme 2). The use of synthetic DNA oligomers as a model for genomic DNA in these studies is justified because the patterns of DNA modification by tobacco carcinogens are controlled by the local DNA sequence and the presence of endogenous modifications, rather than chromosomal structure (16, 61, 62). We found that the distribution of N2-Et-dG adducts along K-ras- and p53-derived

duplexes was only weakly influenced by DNA sequence context (Figures 3 and 4). This is in contrast with sequence-dependent formation of guanine adducts of BPDE, NNK, and reactive oxygen species along the same DNA duplexes (33, 34, 40). Furthermore, AA-dG adduct formation was only weakly stimulated by the presence of a neighboring MeC (Figure 5 and Supporting Information, Figure S-4). The observed lack of sequence specificity for AA-dG adduct formation is in contrast with the nonrandom adduct formation by PAH diol epoxides (15, 33, 63), tobacco nitrosamine-derived alkyl diazohydroxides (40, 64, 65), and reactive oxygen species (66). Because our experiments were performed at carefully controlled conditions, these results cannot be explained by overproduction of AA adducts, resulting in random reactivity patterns. A more likely explanation has to do with the dynamic nature of the AA-dG Shiff base linkages. It is possible that the initially formed N2-ethylidene-dG adducts can redistribute between various DNA sites before they are reductively stabilized as N2-Et-dG. Therefore, our data may not reveal the initial alkylation sites but rather reflect the distribution of adducts after the system has reached an equilibrium (thermodynamic rather than kinetic control of product distribution). Furthermore, while our study employed reduction with excess NaBH3CN that leads to the quantitative conversion of N2-ethylidene-dG to N2-EtdG (23, 24), a greater complexity can be expected for a cellular system, where vitamin C, glutathione, or other antioxidants act as reducing agents. Therefore, although our results presented herein provide initial evidence against the preferential formation of AA-induced N2-Et-dG adducts at lung cancer mutational “hotspots” within the p53 tumor suppressor gene and at CGrich gene promoter sequences, further studies are needed to determine whether the distribution of N2-Et-dG adducts is also random in human genomic DNA. Additionally, while our study focused on the dominant AA-DNA adduct, N2-ethylidene-dG and its reduced form, N2-Et-dG, DNA sequence preferences for the formation of other AA-induced lesions, for example, 1,N2propano-dG and DNA–DNA cross-links (23), remain to be established. Acknowledgment. This study was supported by a grant from the National Cancer Institute (CA-095039). R.G. is a trainee of the NIH Chemistry–Biology Interface Training Grant (T32GM08700). We thank Professor Stephen Hecht for adduct standard and for critical review of the manuscript. Supporting Information Available: Figures of HPLC-MS/ MS calibration curve for N2-Et-dG, dose-dependent formation of N2-Et-dG in a DNA duplex, effect of incubation time on

1386 Chem. Res. Toxicol., Vol. 20, No. 10, 2007

N2-Et-dG adduct yields, and effect of cytosine methylation on N2-Et-dG formation in p53 codon 157. This material is available free of charge via the Internet at http://pubs.acs.org.

References (1) American Cancer Society (2007) Cancer Facts and Figures 2007, American Cancer Society, Atlanta. (2) International Agency for Research on Cancer (2004) Tobacco Smoke and InVoluntary Smoking, IARC, Lyon, France. (3) Hoffmann, D., and Hoffmann, I. (1997) The changing cigarette, 1950– 1995. J. Toxicol. EnViron. Health 50, 307–364. (4) Hecht, S. S. (1999) Tobacco smoke carcinogens and lung cancer. J. Natl. Cancer Inst. 91, 1194–1210. (5) Hussain, S. P., and Harris, C. C. (1999) p53 mutation spectrum and load: the generation of hypotheses linking the exposure of endogenous or exogenous carcinogens to human cancer. Mutat. Res. 428, 23–32. (6) Malejka-Giganti, M., and Tretyakova, N. (2006) Molecular mechanisms of carcinogenesis. In Carcinogenic and Anticarcinogenic Food Components (Baer-Dubowska, W., Bartoszek, A., and Malejka-Giganti, D., Eds.) pp 13–36, CRC Press, Boca Raton, FL. (7) Valinluck, V., Liu, P., Kang, J. I., Jr., Burdzy, A., and Sowers, L. C. (2005) 5-halogenated pyrimidine lesions within a CpG sequence context mimic 5-methylcytosine by enhancing the binding of the methyl-CpG-binding domain of methyl-CpG-binding protein 2 (MeCP2). Nucleic Acids Res. 33, 3057–3064. (8) Valinluck, V., Wu, W., Liu, P., Neidigh, J. W., and Sowers, L. C. (2006) Impact of cytosine 5-halogens on the interaction of DNA with restriction endonucleases and methyltransferase. Chem. Res. Toxicol. 19, 556–562. (9) Valinluck, V., Tsai, H. H., Rogstad, D. K., Burdzy, A., Bird, A., and Sowers, L. C. (2004) Oxidative damage to methyl-CpG sequences inhibits the binding of the methyl-CpG binding domain (MBD) of methyl-CpG binding protein 2 (MeCP2). Nucleic Acids Res. 32, 4100– 4108. (10) Baskunov, V. B., Subach, F. V., Kolbanovskiy, A., Kolbanovskiy, M., Eremin, S. A., Johnson, F., Bonala, R., Geacintov, N. E., and Gromova, E. S. (2005) Effects of benzo[a]pyrene-deoxyguanosine lesions on DNA methylation catalyzed by EcoRII DNA methyltransferase and on DNA cleavage effected by EcoRII restriction endonuclease. Biochemistry 44, 1054–1066. (11) Hollstein, M., Rice, K., Greenblatt, M. S., Soussi, T., Fuchs, R., Sorlie, T., Hovig, E., Smith-Sorensen, B., Montesano, R., and Harris, C. C. (1994) Database of p53 gene somatic mutations in human tumors and cell lines. Nucleic Acids Res. 22, 3551–3555. (12) Westra, W. H., Baas, I. O., Hruban, R. H., Askin, F. B., Wilson, K., Offerhaus, G. J., and Slebos, R. J. (1996) K-ras oncogene activation in atypical alveolar hyperplasias of the human lung. Cancer Res. 56, 2224–2228. (13) Westra, W. H., Slebos, R. J., Offerhaus, G. J., Goodman, S. N., Evers, S. G., Kensler, T. W., Askin, F. B., Rodenhuis, S., and Hruban, R. H. (1993) K-ras oncogene activation in lung adenocarcinomas from former smokers. Evidence that K-ras mutations are an early and irreversible event in the development of adenocarcinoma of the lung. Cancer 72, 432–438. (14) Pfeifer, G. P. (2000) p53 mutational spectra and the role of methylated CpG sequences. Mutat. Res. 450, 155–166. (15) Denissenko, M. F., Pao, A., Tang, M., and Pfeifer, G. P. (1996) Preferential formation of benzo[a]pyrene adducts at lung cancer mutational hotspots in p53. Science 274, 430–432. (16) Denissenko, M. F., Chen, J. X., Tang, M. S., and Pfeifer, G. P. (1997) Cytosine methylation determines hot spots of DNA damage in the human p53 gene. Proc. Natl. Acad. Sci. U.S.A. 94, 3893–3898. (17) Feng, Z., Hu, W., Hu, Y., and Tang, M. S. (2006) Acrolein is a major cigarette-related lung cancer agent: Preferential binding at p53 mutational hotspots and inhibition of DNA repair. Proc. Natl. Acad. Sci. U.S.A. 103, 15404–15409. (18) Obe, G., Natarajan, A. T., Meyers, M., and Hertog, A. D. (1979) Induction of chromosomal aberrations in peripheral lymphocytes of human blood in Vitro, and of SCEs in bone-marrow cells of mice in vivo by ethanol and its metabolite acetaldehyde. Mutat. Res. 68, 291– 294. (19) Woutersen, R. A., Appelman, L. M., Garderen-Hoetmer, A., and Feron, V. J. (1986) Inhalation toxicity of acetaldehyde in rats. III. Carcinogenicity study. Toxicology 41, 213–231. (20) Department of Health and Human Services (2000) The Ninth Report on Carcinogens, Department of Health and Human Services, U.S. Public Health Service, National Toxicology Program, Research Triangle Park, NC. (21) Brooks, P. J., and Theruvathu, J. A. (2005) DNA adducts from acetaldehyde: Implications for alcohol-related carcinogenesis. Alcohol 35, 187–193.

Communications (22) Vaca, C. E., Fang, J. L., and Schweda, E. K. (1995) Studies of the reaction of acetaldehyde with deoxynucleosides. Chem.-Biol. Interact. 98, 51–67. (23) Wang, M., McIntee, E. J., Cheng, G., Shi, Y., Villalta, P. W., and Hecht, S. S. (2000) Identification of DNA adducts of acetaldehyde. Chem. Res. Toxicol. 13, 1149–1157. (24) Wang, M., Yu, N., Chen, L., Villalta, P. W., Hochalter, J. B., and Hecht, S. S. (2006) Identification of an acetaldehyde adduct in human liver DNA and quantitation as N2-ethyldeoxyguanosine. Chem. Res. Toxicol. 19, 319–324. (25) Chen, L., Wang, M., Villalta, P. W., Luo, X., Feuer, R., Jensen, J., Hatsukami, D. K., and Hecht, S. S. (2007) Quantitation of an acetaldehyde adduct in human leukocyte DNA and the effect of smoking cessation. Chem. Res. Toxicol. 20, 108–113. (26) Matsuda, T., Terashima, I., Matsumoto, Y., Yabushita, H., Matsui, S., and Shibutani, S. (1999) Effective utilization of N2-ethyl-2′deoxyguanosine triphosphate during DNA synthesis catalyzed by mammalian replicative DNA polymerases. Biochemistry 38, 929–935. (27) Fang, J. L., and Vaca, C. E. (1997) Detection of DNA adducts of acetaldehyde in peripheral white blood cells of alcohol abusers. Carcinogenesis 18, 627–632. (28) Terashima, I., Matsuda, T., Fang, T. W., Suzuki, N., Kobayashi, J., Kohda, K., and Shibutani, S. (2001) Miscoding potential of the N2ethyl-2′-deoxyguanosine DNA adduct by the exonuclease-free Klenow fragment of Escherichia coli DNA polymerase I. Biochemistry 40, 4106–4114. (29) Perrino, F. W., Blans, P., Harvey, S., Gelhaus, S. L., McGrath, C., Akman, S. A., Jenkins, G. S., LaCourse, W. R., and Fishbein, J. C. (2003) The N2-ethylguanine and the O6-ethyl- and O6-methylguanine lesions in DNA: Contrasting responses from the “bypass” DNA polymerase η and the replicative DNA polymerase R. Chem. Res. Toxicol. 16, 1616–1623. (30) Choi, J. Y., and Guengerich, F. P. (2005) Adduct size limits efficient and error-free bypass across bulky N2-guanine DNA lesions by human DNA polymerase η. J. Mol. Biol. 352, 72–90. (31) Choi, J. Y., and Guengerich, F. P. (2006) Kinetic evidence for inefficient and error-prone bypass across bulky N2-guanine DNA adducts by human DNA polymerase ι. J. Biol. Chem. 281, 12315– 12324. (32) Upton, D. C., Wang, X., Blans, P., Perrino, F. W., Fishbein, J. C., and Akman, S. A. (2006) Replication of N2-ethyldeoxyguanosine DNA adducts in the human embryonic kidney cell line 293. Chem. Res. Toxicol. 19, 960–967. (33) Matter, B., Wang, G., Jones, R., and Tretyakova, N. (2004) Formation of diastereomeric benzo[a]pyrene diol epoxide-guanine adducts in p53 gene-derived DNA sequences. Chem. Res. Toxicol. 17, 731–741. (34) Rajesh, M., Wang, G., Jones, R., and Tretyakova, N. (2005) Stable isotope labeling-mass spectrometry analysis of methyl- and pyridyloxobutyl-guanine adducts of 4-(methylnitrosamino)-1-(3-pyridyl)-1butanone in p53-derived DNA sequences. Biochemistry 44, 2197– 2207. (35) Zhao, H., Pagano, A. R., Wang, W., Shallop, A., Gaffney, B., and Jones, R. A. (1997) Use of a 13C atom to differentiate two 15N-labeled nucleosides. Syntheses of [15NH2]-adenosine, [1,NH2-15N2]- and [2-13C,1,NH2-15N2]-guanosine, and [1,7,NH2-15N3]- and [2-13C,1,7,NH215 N3]-2′-deoxyguanosine. J. Org. Chem. 62, 7832–7835. (36) Shallop, A. J., Gaffney, B. L., and Jones, R. A. (2003) Use of 13C as an indirect tag in 15N specifically labeled nucleosides. Syntheses of [8-13C-1,7,NH2-15N3]adenosine, -guanosine, and their deoxy analogues. J. Org. Chem. 68, 8657–8661. (37) Matter, B., Malejka-Giganti, D., Csallany, A. S., and Tretyakova, N. (2006) Quantitative analysis of the oxidative DNA lesion, 2,2-diamino4-(2-deoxy-β-D-erythro-pentofuranosyl)amino]-5(2H)-oxazolone (oxazolone), in Vitro and in ViVo by isotope dilution-capillary HPLCESI-MS/MS. Nucleic Acids Res. 34, 5449–5460. (38) Sheskin, D. J. (2000) The between-subjects factorial analysis of variance. In Handbook of Parametric and Nonparametric Statistical Procedures, pp 705–755, CRC Press, New York. (39) Tretyakova, N., Matter, B., Jones, R., and Shallop, A. (2002) Formation of benzo[a]pyrene diol epoxide-DNA adducts at specific guanines within K-ras and p53 gene sequences: Stable isotope-labeling mass spectrometry approach. Biochemistry 41, 9535–9544. (40) Ziegel, R., Shallop, A., Jones, R., and Tretyakova, N. (2003) K-ras gene sequence effects on the formation of 4-(methylnitrosamino)-1(3-pyridyl)-1-butanone (NNK)-DNA adducts. Chem. Res. Toxicol. 16, 541–550. (41) Hainaut, P., and Pfeifer, G. P. (2001) Patterns of p53 GfT transversions in lung cancers reflect the primary mutagenic signature of DNAdamage by tobacco smoke. Carcinogenesis 22, 367–374. (42) Greenblatt, M. S., Bennett, W. P., Hollstein, M., and Harris, C. C. (1994) Mutations in the p53 tumor suppressor gene: Clues to cancer etiology and molecular pathogenesis. Cancer Res. 54, 4855–4878.

Communications (43) Tornaletti, S., and Pfeifer, G. P. (1995) Complete and tissueindependent methylation of CpG sites in the p53 gene: Implications for mutations in human cancers. Oncogene 10, 1493–1499. (44) Ehrlich, M., Gama-Sosa, M. A., Huang, L. H., Midgett, R. M., Kuo, K. C., McCune, R. A., and Gehrke, C. (1982) Amount and distribution of 5-methylcytosine in human DNA from different types of tissues of cells. Nucleic Acids Res. 10, 2709–2721. (45) Newell-Price, J., Clark, A. J., and King, P. (2000) DNA methylation and silencing of gene expression. Trends Endocrinol. Metab. 11, 142– 148. (46) Momparler, R. L., and Bovenzi, V. (2000) DNA methylation and cancer. J. Cell Physiol. 183, 145–154. (47) Hausheer, F. H., Rao, S. N., Gamcsik, M. P., Kollman, P. A., Colvin, O. M., Saxe, J. D., Nelkin, B. D., McLennan, I. J., Barnett, G., and Baylin, S. B. (1989) Computational analysis of structural and energetic consequences of DNA methylation. Carcinogenesis 10, 1131–1137. (48) Norberg, J., and Vihinen, M. (2001) Molecular dynamics simulation of the effects of cytosine methylation on structure of oligonucleotides. J. Mol. Structure (THEOCHEM) 546, 51–62. (49) Zacharias, W., O’Connor, T. R., and Larson, J. E. (1988) Methylation of cytosine in the 5-position alters the structural and energetic properties of the supercoil-induced Z-helix and of B-Z junctions. Biochemistry 27, 2970–2978. (50) Mathison, B. H., Said, B., and Shank, R. C. (1993) Effect of 5-methylcytosine as a neighboring base on methylation of DNA guanine by N-methyl-N-nitrosourea. Carcinogenesis 14, 323–327. (51) Weisenberger, D. J., and Romano, L. J. (1999) Cytosine methylation in a CpG sequence leads to enhanced reactivity with benzo[a]pyrene diol epoxide that correlates with a conformational change. J. Biol. Chem. 274, 23948–23955. (52) Sendowski, K., and Rajewsky, M. F. (1991) DNA sequence dependence of guanine-O6 alkylation by the N-nitroso carcinogens N-methyland N-ethyl-N-nitrosourea. Mutat. Res. 250, 153–160. (53) Pradhan, P., Graslund, A., Seidel, A., and Jernstrom, B. (1999) Implications of cytosine methylation on (+)-anti-benzo[a]pyrene 7,8 dihydrodiol 9,10-epoxide N2-dG adduct formation in 5′-d(CGT), 5′d(CGA), and 5′-d(CGC) sequence contexts of single- and doublestranded oligonucleotides. Chem. Res. Toxicol. 12, 816–821. (54) Mathur, P., Xu, J., and Dedon, P. C. (1997) Cytosine methylation enhances DNA damage produced by groove binding and intercalating enediynes: Studies with esperamicins A1 and C. Biochemistry 36, 14868–14873. (55) Das, A., Tang, K. S., Gopalakrishnan, S., Waring, M. J., and Tomasz, M. (1999) Reactivity of guanine at m5CpG steps in DNA: Evidence for electronic effects transmitted through the base pairs. Chem. Biol. 6, 461–471.

Chem. Res. Toxicol., Vol. 20, No. 10, 2007 1387 (56) Ross, M. K., Mathison, B. H., Said, B., and Shank, R. C. (1999) 5-Methylcytosine in CpG sites and the reactivity of nearest neighboring guanines toward the carcinogen aflatoxin B1-8,9-epoxide. Biochem. Biophys. Res. Commun. 254, 114–119. (57) Chen, J. X., Zheng, Y., West, M., and Tang, M. S. (1998) Carcinogens preferentially bind at methylated CpG in the p53 mutational hot spots. Cancer Res. 58, 2070–2075. (58) Li, V. S., Reed, M., Zheng, Y., Kohn, H., and Tang, M. (2000) C5 cytosine methylation at CpG sites enhances sequence selectivity of mitomycin C-DNA bonding. Biochemistry 39, 2612–2618. (59) Ziegel, R., Shallop, A., Upadhyaya, P., Jones, R., and Tretyakova, N. (2004) Endogenous 5-methylcytosine protects neighboring guanines from N7 and O6-methylation and O6-pyridyloxobutylation by the tobacco carcinogen 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone. Biochemistry 43, 540–549. (60) Harris, C. C. (1996) Structure and function of the p53 tumor suppressor gene: Clues for rational cancer therapeutic strategies. J. Natl. Cancer Inst. 88, 1442–1455. (61) Cloutier, J. F., Drouin, R., and Castonguay, A. (1999) Treatment of human cells with N-nitroso(acetoxymethyl)methylamine: Distribution patterns of piperidine-sensitive DNA damage at the nucleotide level of resolution are related to the sequence context. Chem. Res. Toxicol. 12, 840–849. (62) Cloutier, J. F., Drouin, R., Weinfeld, M., O’Connor, T. R., and Castonguay, A. (2001) Characterization and mapping of DNA damage induced by reactive metabolites of 4-(methylnitrosamino)-1-(3pyridyl)-1-butanone (NNK) at nucleotide resolution in human genomic DNA. J. Mol. Biol. 313, 539–557. (63) Smith, L. E., Denissenko, M. F., Bennett, W. P., Li, H., Amin, S., Tang, M., and Pfeifer, G. P. (2000) Targeting of lung cancer mutational hotspots by polycyclic aromatic hydrocarbons. J. Natl. Cancer Inst. 92, 803–811. (64) Wurdeman, R. L., and Gold, B. (1988) The effect of DNA sequence, ionic strength, and cationic DNA affinity binders on the methylation of DNA by N-methyl-N-nitrosourea. Chem. Res. Toxicol. 1, 146–147. (65) Gold, B., Marky, L. M., Stone, M. P., and Williams, L. D. (2006) A review of the role of the sequence-dependent electrostatic landscape in DNA alkylation patterns. Chem. Res. Toxicol. 19, 1402–1414. (66) Margolin, Y., Cloutier, J. F., Shafirovich, V., Geacintov, N. E., and Dedon, P. C. (2006) Paradoxical hotspots for guanine oxidation by a chemical mediator of inflammation. Nat. Chem. Biol. 2, 365–366.

TX7001146