Implications of Cytosine Methylation on - American Chemical Society

816

Chem. Res. Toxicol. 1999, 12, 816-821

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 Double-Stranded Oligonucleotides† Padmanava Pradhan,‡ Astrid Gra¨slund,‡ Albrecht Seidel,§ and Bengt Jernstro¨m*,| Department of Biophysics, Arrhenius Laboratory, Stockholm University, S-106 91 Stockholm, Sweden, Biochemical Institute of Environmental Carcinogens, Prof. Dr. Gernot Grimmer Foundation, Lurup 4, D-22927 Grosshansdorf, Germany, and Institute of Environmental Medicine, Division of Biochemical Toxicology, Karolinska Institutet, S-171 77 Stockholm, Sweden Received October 16, 1998

Covalent binding of (+)-anti-benzo[a]pyrene 7,8-dihydrodiol 9,10-epoxide (anti-BPDE) to the N2-amino group of deoxyguanine in the oligonucleotides 5′-d(CCTATCGXTATCC) and 5′d(CCTATm5CGXTATCC) (X being T, A, or C) has been studied. The extent of formation of the (+)-trans-anti-BPDE-N2-dG adduct in single-stranded 13-mer oligonucleotides with 5′-d(m5CGT) and 5′-d(m5CGA) sequence contexts was significantly higher (1.5- and 2.4-fold, respectively) relative to that of the nonmethylated sequences. With the 5′-d(CGC) sequence context, m5dC had no significant effect on adduct formation. When the reaction was allowed to proceed in the presence of oligonucleotide duplexes (composed of a 13-mer parent strand and a 9-mer complement), a significant increase in the extent of adduct formation was obseved with 5′d(m5CGT)/d(CGA) and 5′-d(m5CGA)/d(CGT), but not with 5′-d(CGC)/d(GCG), relative to those of the nonmethylated duplexes. Independent of sequence context, no clear effect of m5dC on diol epoxide binding to the opposite dG in the complementary strand was observed. The level of diol epoxide binding to the dG target in the 13-mer oligonucleotides is in general higher in single-stranded sequences than in the duplexes. With 5′-d(CGA) and 5′-d(m5CGA), for instance, adduct yields were 3- and 4-fold higher, respectively. The thermodynamic stability of the (+)trans-anti-BPDE-N2-dG adduct in the 5′-d(m5CGT)-containing duplex (composed of a 13-mer parent strand and a full complement) was substantially higher than in the 5′-d(CGT)/d(GCA) sequence context. The stimulating effect of cytosine methylation on the formation of DNA adducts of anti-BPDE has previously been demonstrated in other experimental systems. The increase in yield could possibly be rationalized in terms of prestacking of the pyrenyl ring system with the nucleobases prior to the nucleophilic addition reaction of the exocyclic amino group. The results from induced circular dichroism studies with the (+)-trans-anti-BPDEN2-dG adduct in the 5′-d(m5CGT)-containing duplex are consistent with substantial heterogeneity of adduct conformations, including both external minor groove-localized and intercalated structures.

Introduction Primary tumors in the respiratory tract in humans are closely related to cigarette smoking and to the extent of smoke exposure (1). The content of polycyclic aromatic hydrocarbons (PAH1), such as benzo[a]pyrene (BP), in tobacco smoke is generally considered to contribute to the tumorigenic process (1, 2). Historically, BP has served as an important model substrate for studying mutagenic † Part of this work was presented at the 16th International Symposium on Polycyclic Aromatic Compounds, November 4-8, 1997, Charlotte, NC. * To whom correspondence should be addressed: Institute of Environmental Medicine, Division of Biochemical Toxicology, Karolinska Institutet, Box 210, S-17177 Stockholm, Sweden. Phone: ++468-7287576. Fax: ++46-8-334467. E-mail: [email protected]. ‡ Stockholm University. § Prof. Dr. Gernot Grimmer Foundation. | Karolinska Institutet.

and carcinogenic properties of PAH. The biological activity of BP requires metabolic activation to reactive electrophiles and their subsequent covalent binding to DNA (3, 4). Metabolites meeting these requirements have been identified as the different stereoisomers of benzo[a]pyrene 7,8-dihydrodiol 9,10-epoxide (BPDE) (5, 6). Specifically, the biological activity of BP appears to be causally linked to the covalent binding of (+)-anti-BPDE to DNA which occurs almost exclusively at the exocyclic amino group of deoxyguanosine (dG) (7). Stereochemically favored is the trans opening of the oxirane ring by nucleophilic attack of the amino group at the benzylic C-10 1 Abbreviations: BP, benzo[a]pyrene; BPDE, trans-7,8-dihydroxy9,10-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene; (+)-anti-BPDE, benzo[a]pyrene (7R,8S)-diol (9S,10R)-oxide; (-)-anti-BPDE, benzo[a]pyrene (7S,8R)-diol (9R,10S)-oxide; (+)-syn-BPDE, benzo[a]pyrene (7S,8R)diol (9S,10R)-oxide; (-)-syn-BPDE, benzo[a]pyrene (7R,8S)-diol (9R,10S)-oxide.

10.1021/tx980230f CCC: $18.00 © 1999 American Chemical Society Published on Web 08/07/1999

Effect of m5dC on dG-BPDE Adduct Formation

position (8). Exposure of mammalian cells, including those of human origin, to (+)-anti-BPDE results in the induction of different types of mutations, including different types of point mutations and deletions. However, independent of cell origin, the G f T transversion point mutation has been found to greatly predominate (9). Most human lung cancers demonstrate mutations in the p53 tumor suppressor gene, in particular at codons 157, 248, and 273, and the majority of these mutations are G f T transversions (10-13). In fact, codon 157 seems to be a mutational hotspot for lung cancer (10, 13). Recently, Denissenko et al. (13) studied the distribution of adducts derived from racemic anti-BPDE in exons 5-8 of the p53 gene in human cells. Interestingly, a preference for adduct formation at guanines in d(CpG) sequences in codons 157, 248, and 273 was observed. The d(CpG) sites in exons 5-8 are methylated at C-5 of dC (m5dC) (14), thus indicating that this modification increases the reactivity of the adjacent dG. This observation is consistent with previous work by Geacintov et al. (15). An increased level of noncovalent interaction of racemic anti-BPDE with poly(dG-m5dC)(dG-m5dC) and subsequent adduct formation relative to the nonmethylated polynucleotide were observed. In a more recent study, Denissenko et al. (16) clearly demonstrated the role of m5dC in stimulating preferential binding of the diol epoxide to dG. To provide further information about the effect of methylation of dC on adduct formation, we have studied the binding of (+)-anti-BPDE to the oligonucleotides 5′d(CCTATCGXTATCC) and 5′-d(CCTATm5CGXTATCC) (X being T, A, or C) in the single-stranded or duplexed form. In addition, thermal stability studies and optical spectroscopic methods have been used to gain information about adopted adduct conformations.

Materials and Methods Caution: Benzo[a]pyrene diol epoxides are mutagenic and carcinogenic agents and should be handled with care, as outlined in the National Cancer Institute Guidelines. Chemicals. The (+)-anti-BPDE enantiomer was synthesized according to literature methods (17). The oligonucleotides 5′d(CCTATCGXTATCC) and 5′-d(CCTATm5CGXTATCC) (X being T, A, or C) and their fully or partially complementary sequences 5′-d(GGATAYCGATAGG) and 5′-d(ATAYCGATA) (Y being A, T, or G) were obtained from Cybergene AB (Stockholm, Sweden). Prior to use, the purity of the oligonucleotides was checked by HPLC (see below). Synthesis and Isolation of Oligonucleotide-BPDE Adducts. Oligonucleotides in single- or double-stranded form were dissolved in 50 mM Tris-HCl, 5 mM EDTA, and 100 mM NaCl (pH 7.5) (TE buffer) to a final concentration of 40 µM together with a 5-fold excess of the (+)-anti-BPDE, added as a single portion dissolved in (CH3)2SO (final concentration of 5%). The reaction was allowed to proceed on ice in the dark. To estimate completion of the reaction, aliquots were removed at different time points (15 min to 5 h), and the reaction was terminated by addition of alkaline mercaptoethanol (18) to trap remaining diol epoxide. The analysis of adducts was performed with HPLC, in principle as described previously (19, 20). Prior to circular dichroism measurements, the (+)-anti-BPDE-modified oligonucletides were bubbled with N2 to remove the acetonitrile and then passed through Sep-Pak C18 cartridges (Waters Associates, Milford, MA) previously washed with methanol (6 mL) and equilibrated with distilled water (6 mL). Following application of the sample, 3 mL of water was allowed to pass through the cartridge followed by 2 mL of methanol for eluting the (+)-anti-

Chem. Res. Toxicol., Vol. 12, No. 9, 1999 817 BPDE-modified oligonucleotide. The extent of overall modification was calculated from the maximal absorbance of the BPDE chromophore at 346-354 nm using an  of 28.5 mM-1 cm-1. Light Absorbance and Circular Dichroism. The UV spectra of (+)-anti-BPDE-modified oligonucleotides in singlestranded or duplexed form (dissolved in TE buffer) and the melting curves were recorded on a Varian CARY 45 spectrophotometer equipped with a Eurotherm RM6 LAUDA temperature control device. The CD spectra of the same samples were recorded on a JASCO model 720 spectropolarimeter at 5 °C. The spectra of the duplexes were monitored after storage for at least 1 h following addition of the complementary strand. The net contribution of the (+)-anti-BPDE adduct in duplexes to the ICD was obtained by subtracting the intrinsic CD of the corresponding BPDE tetraol (21). Thermal Stability of Oligonucleotide Duplexes. Unmodified or (+)-trans-anti-BPDE-N2-dG-modified 5′-d(CCTATCGTTATCC) or 5′-d(CCTATm5CGTTATCC) was annealed to the 9-mer sequence 5′-d(ATAACGATA) or the full complement 5′d(GGATAACGATAGG) by mixing at 60 °C for 10 min followed by slow cooling to 4 °C. The stability of the duplexes (2 µM in TE buffer) was studied by continuously monitoring the change in absorption at 260 nm as a function of temperature (5 to 75 °C over the course of 140 min).

Results and Discussion Adduct Formation of (+)-anti-BPDE. The (+)-antiBPDE was reacted with single-stranded 5′-d(CCTATCGXTATCC) and 5′-d(CCTATm5CGXTATCC) (X being T, A, or C) or with the oligonucleotides in duplex with the 9-mer complement 5′-d(ATAXCGATA) (X being A, T, or G). The reason for using a 9-mer rather than the entire complement with four additional dGs was to minimize complications arising from adduct formation at multiple dG sites. The overall yields of dG adducts were estimated from HPLC analysis. Examples of the results are shown in panels A and B of Figure 1. The reaction of (+)-antiBPDE with single-stranded 5′-d(CCTATCGTTATCC) or the methylated analogue gives rise to several products with both UV-absorbing and fluorescent properties (not shown). The UV-absorbing but nonfluorescent peaks eluting at about 13 min correspond to the unmodified oligonucleotides, and the UV-absorbing and fluorescent peaks eluting beween 17 and 26 min correspond to BPDE-modified oligonucleotides. As revealed by circular dichroism measurements (21), the major peaks eluting at about 19 min are identical to (+)-trans-anti-BPDEN2-dG adducts, whereas one of the peaks eluting at 1718 min most likely is the corresponding cis adduct (22). The chromatographic and fluorescent properties of the later eluting peaks are compatible with dA adducts (23). If the reaction of (+)-anti-BPDE is allowed to take place with the 5′-d(CCTATCGTTATCC)/d(ATAGCAATA) duplex or the methylated one, the complex adduct distribution disappeared and the major trans adducts were almost exclusively formed (tR ≈ 19 min) (panels C and D of Figure 1). This observation is fully consistent with previous results of Meehan and Straub (24). It should be pointed out that a complex adduct distribution is observed with all oligonucleotides studied and that the complex pattern disappears upon duplex formation. The peaks eluting at about 22 min in panels C and D of Figure 1 correspond to the (+)-trans-anti-BPDE-N2-dG adducts in 5′-d(ATAACGATA). With all the oligonucleotide duplexes that were examined, the (+)-trans-anti-BPDE-N2-dG adducts in the 9and 13-mer oligonucleotides are easily separated by

818 Chem. Res. Toxicol., Vol. 12, No. 9, 1999

Pradhan et al. Table 1. Modification of Single- and Double-Stranded Oligonucleotides with (+)-anti-BPDE and Yields of trans-anti-BPDE-N2-dG Adducts sequence context

yield (%)a

sequence context

5′-d(CGT)

8.92 ( 1.84

5′-d(CGT)/d(G6CA)

5′-d(m5CGT)

13.5 ( 0.9

5′-d(m5CGT)/d(GCA)

5′-d(CGA)

13.6 ( 3.2

5′-d(CGA)d(GCT)

5′-d(m5CGA)

32.6 ( 4.6

5′-d(m5CGA)d(GCT)

5′-d(CGC)

14.9 ( 0.9

5′-d(CGC)/d(G6CG4)

5′-d(m5CGC)

16.1 ( 2.7

5′-d(m5CGC)/d(GCG)

yield (%)a 6.73 ( 1.21 8.57 ( 1.36b 10.7 ( 2.3 11.3 ( 0.9b 4.70 ( 0.53 11.6 ( 1.3b 7.55 ( 3.01 13.2 ( 2.11b 8.81 ( 0.57 10.0 ( 0.64b 5.22 ( 0.23c 11.1 ( 2.75 9.14 ( 0.74b 8.02 ( 0.32c

a The yields of adducts were calculated from the combined area of the peaks corresponding to the unmodified and the (+)-antiBPDE-modified oligonucleotides. The yields of adducts in duplexes have been corrected for the different contribution of absorbance of the 13- and 9-mer oligonucleotides, respectively. The results are expressed as the mean ( SD (n g 3). b Yields of adduct formation on G6 in the complementary strand. c Yields of adduct formation on G4 in the complementary strand.

Figure 1. HPLC elution profiles of reaction products of (+)anti-BPDE and single-stranded 5′-d(CCTATCGTTATCC) (A), single-stranded 5′-d(CCTATm5CGTTATCC) (B), 5′-d(CCTATCGTTATCC) in duplex with 5′-d(ATAACGATA) (C), and 5′d(CCTATm5CGTTATCC) in duplex with 5′-d(ATAACGATA) (D). The effluent was monitored via UV absorbance at 260 nm.

HPLC, thus allowing estimation of both the extent of overall adduct formation and the effect of sequence context on adduct formation. The extent of modification was calculated from the combined area of the peaks corresponding to the unmodified and the modified oligonucleotides. We have previously shown that incubation of (+)-antiBPDE with 5′-d(AGA)- and 5′-d(TGC)-containing oligonucleotides under conditions similar to those used here results in maximal adduct yields after incubation for less than 5 h (25). Here we have repeated the experiments with 5′-d(CCTATCGTTATCC) and the corresponding methylated oligonucleotide and conclude that the reaction is completed after incubation for approximately 2 h (results not shown). Calculation of the optimal extent of trans-anti-BPDEN2-dG adduct formation (incubation time of >5 h) in single-stranded 5′-d(CCTATCGXTATCC) (X being T, A, or C) or the methylated analogues (Table 1) showed that m5C significantly stimulated diol epoxide binding to the adjacent dG in 5′-d(CGT) and 5′-d(CGA) sequence contexts but not in 5′-d(CGC). The highest extent of binding (about 33% adducted oligonucleotides) and the most pronounced effect of methylation (>2-fold) were observed with 5′-d(CGA). A dA located on the 3′-side of a dG has previously been shown to stimulate adduct formation at this site (25). The (+)-trans-anti-BPDE-N2-dG adduct yields obtained with duplexed oligonucleotides (composed of a 13mer parent strand and a 9-mer complement) are also shown in Table 1. As is evident, the extent of adduct formation in the 13-mers was lower, in particular with the 5′-d(CGA)/d(GCT) duplexes, than those observed with the single-stranded sequences. With the 5′-d(CGT)/

Table 2. Melting Points of Unmodified and (+)-trans-anti-BPDE-N2-dG-Modified Oligonucleotide Duplexes sequence context

complement

Tm (°C)a

5’-d(CGT) 5′-d(m5CGT) 5′-d(CG*T)b 5′-d(m5CG*T) 5′-d(CGT) 5′-d(m5CGT) 5′-d(CG*T) 5′-d(m5CG*T)

5′-d(ACG), 9-mer 5′-d(ACG), 9-mer 5′-d(ACG), 9-mer 5′-d(ACG), 9-mer 5′-d(ACG), 13-mer 5′-d(ACG), 13-mer 5′-d(ACG), 13-mer 5′-d(ACG), 13-mer

28.5 29.5 not well-defined not well-defined 47.5 49.5 32.0 39.0

a The melting points were obtained from plots of change in absorption at 260 nm as a function of temperature. Measurements were carried out in TE buffer at oligonucleotide duplex concentrations of 2 µΜ. b G* denotes the position of the (+)-anti-BPDEN2-dG adduct.

d(GCA) and 5′-d(CGA)/d(GCT) duplexes but not with 5′d(CGC)/d(GCG), methylation significantly increased the yield (about 1.6-fold) of the adducts. Since the chromatographic method allowed calculation of the amount of adducts formed in the complementary 9-mers, these results are also included in Table 1. With 5′-d(CGT)/ d(GCA) but not with the 5′-d(CGA)/d(GCT) and 5′d(CGC)/d(GCG) duplexes, the presence of m5dC slightly increased the extent of adduct formation on the opposite dG. As suggested by the results with the 5′-d(CGC)/ d(GCG) duplex, cytosine methylation may also stimulate diol epoxide binding to a dG remote from m5dC and located on the complementary strand. Thermal Stability. The 5′-d(CCTATCGTTATCC)/ d(ATAGCAATA) and the methylated analogue were studied with respect to thermal stability. The transition from the duplex to single-stranded form as a function of temperature was monitored. The melting temperature (Tm) was estimated from the first-order derivative of the corresponding absorbance-temperature plot, and the results that were obtained are compiled in Table 2. The unmodified oligonucleotides in a duplex with the 9-mer complement exhibited Tm values of 350 nm that exhibit either positive or negative contributions to the ICD have intercalated binding (type I) as the predominant conformation (29). For the samples studied here, the following spectroscopic observations were made.

Chem. Res. Toxicol., Vol. 12, No. 9, 1999 819

Figure 2. Absorption spectra of (+)-trans-anti-BPDE-N2-dGmodified 5′-d(CCTATm5CGTTATCC) in single-stranded form (‚ ‚‚) (A and B) or duplexed with 5′-d(ATAACGATA) (s) (A) or 5′d(CCATAACGATACC) (s) (B). ICD spectra of (+)-trans-antiBPDE-N2-dG-modified 5′-d(CCTATm5CGTTATCC) in singlestranded form (‚‚‚) or duplexed with 5′-d(ATAACGATA) (s) (C). ICD spectra of (+)-trans-anti-BPDE-N2-dG-modified 5′-d(CCTATm5CGTTATCC) or 5′-d(CCTATCGTTATCC) in singlestranded form (1 and 2, respectively) or duplexed with 5′-d(CCATAACGATACC) (3 and 4, respectively) (D). The adduct concentration was 2-7 µM. The measurements were carried out in TE buffer at 5 °C.

(1) Absorption Spectroscopy. The (+)-trans-antiBPDE-N2-dG adduct in the single-stranded 5′-d(CCTATm5CGTTATCC) or 5′-d(CCTATCGTTATCC) exhibited an absorption maximum at 350-351 nm. Upon addition of the 9-mer complement, a small shift of the maximum (0.5-1 nm) toward shorter wavelengths was observed for both duplexes (Figure 2A). However, upon addition of the 13-mer complement, the absorption maximum shifted 3-4 nm toward shorter wavelengths in both cases, thus consistent with type II binding characteristics (Figure 2B). (2) Induced Circular Dichroism. The (+)-trans-antiBPDE-N2-dG adduct in single-stranded 5′-d(CCTATm5CGTTATCC) exhibited a negative ICD in the 320-360 nm region (Figure 2C). Upon duplex formation with the 9-mer complement, a reduced intensity but no change in the sign of the ICD were observed. Similar experiments with the 5′-d(CCTATCGTTATCC) oligonucleotide yielded almost identical results except that the net duplex contribution to the ICD of the (+)-trans-anti-BPDE-N2dG adduct in the methylated 13-mer/9-mer duplex was slightly more negative than that of the adduct in the nonmethylated duplex (not shown). This suggests retained adduct-oligonucleotide base interactions probably due to a less ordered helical structure of the 13-mer/9mer duplexes. However, upon duplex formation with the 13-mer complement, the ICD spectrum of the nonmethylated duplex exhibited opposite polarity and a positive ICD in the 320-360 nm region (Figure 2D). With the methylated duplex, the ICD spectrum was similar, although it had a significantly lower amplitude (Figure 2D). With both oligonucleotides, duplex formation was associated with a 2 nm shift in the |λmax| toward shorter wavelengths. The ICD results obtained with the fully

820 Chem. Res. Toxicol., Vol. 12, No. 9, 1999

Pradhan et al.

Table 3. ICD (∆E in M-1 cm-1) of trans-anti-BPDE-N2-dG Oligonucleotide Adducts modified sequence 5′-d(CGT) 5′-d(m5CGT)

single-stranded |λmax| ICD 349 349

-17.7 -19.5

duplex |λmax| ICD 347 347

9.0 6.6

duplex contributiona ICD 12.1 9.7

a Obtained by subtracting the intrinsic CD of the corresponding BPDE-tetraol adduct (-3.1 M-1 cm-1) from the observed ICD of the modified duplex (21).

complementary duplexes are summarized in Table 3. As shown in Table 3, the duplex contribution to the overall ICD was significantly larger in the nonmethylated duplex. Taken together, the results suggest a heterogeneous distribution of adduct conformations in both duplexes. In the nonmethylated one, the major adduct has minor groove localization, whereas in the duplex with the m5dC adjacent to the (+)-trans-anti-BPDE-N2-dG adduct, the distribution is likely to be different and to have a greater contribution of type I adducts. Previous results using alternating or nonalternating polydeoxyribonucleotides have shown that the distribution of external and intercalative adduct localization is strongly dependent on sequence context. For instance, in 5′-d(GGG) and 5′d(CGC) sequence contexts external localization greatly dominates, whereas in 5′-d(AGA) and 5′-d(TGT) the contribution of intercalative binding becomes increasingly more important (32). The preferred adopted conformation of PAH diol epoxide adducts has major implications for their recognition by repair proteins (33-35). For example, investigations with mouse skin using fluorescence line narrowing spectroscopy have shown that external localized antiBPDE-DNA adducts are more readily repaired than intercalated adducts (34). These findings are supported by a more recent study also with mouse skin in which the repair of diol epoxide-DNA adducts formed by dibenzo[a,l]pyrene was assessed (35). Therefore, the increase in the level of type I adducts observed in this study in the duplex where dC has been replaced with m5dC at the 5′-side of the (+)-trans-anti-BPDE-N2-dG adduct strongly suggests that the repair of this adduct is less efficient if it is formed in the 5′-d(m5dCGT) sequence context. An increase in persistence in vivo may also readily explain a higher frequency with which mutations occur at this site. (3) Effect of Temperature. Figure 3 shows the effect of temperature on the ICD of the (+)-trans-anti-BPDEN2-dG adduct in fully duplexed 5′-d(m5CGT). The decrease in the magnitude of ICD with increasing temperature up to 30 °C (Tm ) 39 °C) reflects the change in conformational equilibrium (more type I adducts than type II adducts at higher temperatures). Recent studies using ICD (21) and NMR2 with the (+)-trans-anti-BPDEN2-dG adduct in a 5′-d(AGA)-containing duplex demonstrated a similar temperature-dependent equilibrium of adduct conformations. The recent studies by Denissenko et al. and Chen et al. (13, 16, 36) have focused on the importance of a 5′flanking m5dC in stimulating the formation of adducts between (()-anti-BPDE and the adjacent dG. The mechanistic details underlying this phenomenon have not been 2 P. Pradhan, B. Jernstro ¨ m, A. Seidel, and A. Gra¨slund, unpublished observations.

Figure 3. CD spectra of 5′-d(CCTATm5CGTTATCC)/5′-d(CCATAACGATACC) as a function of temperature: 5 (1), 15 (2), 25 (3), 30 (4), 35 (5), and 45 °C (6). The adduct concentration was 4 µΜ. The measurements were carried out in TE buffer.

elucidated. However, a previous study by Geacintov et al. (15) may provide some clues. Replacing dC in poly(dG-dC)(dG-dC) with m5dC increased the intercalative association constant of (()-anti-BPDE more than 5-fold and significantly stimulated covalent adduct formation. It was suggested that since the methyl group in dC is localized in the major groove and interacts with water, the destabilization of the B-form duplex might enhance the conditions for intercalation of the diol epoxide. However, this explanation is not fully applicable in the case of short oligonucleotides since the presence of m5dC adjacent to dG in 5′-d(CCTATCGTTATCC) seems to increase duplex stability, although moderately. At present, we lack information about the effect of m5dC on the stability of duplexes with 5′-d(CGA)/GCT and 5′-d(CGC)/ GCG sequence contexts. However, as indicated for the 5′-d(CGT)/GCA duplex, the presence of m5dC may indeed promote the interaction of BPDE and increase adduct distribution heterogeneity in these sequences as well. With regard to the stimulating effect of m5dC on adduct formation in 5′-d(CGT)GCA and 5′-d(CGA)GCT, it is possible that the minor groove-localized adducts are formed via a bimolecular transition state complex which differs from that leading to the intercalated adduct and that the m5dC (or dT/dA) promotes formation of the latter (37). In conclusion, replacing dC with m5dC stimulates (+)anti-BPDE-N2-dG adduct formation in both the singleand double-stranded oligonucleotides with 5′-d(CGT) and 5′-d(CGA) sequence contexts. Cytosine methylation at sites adjacent to the 5′-side of the dG residue seems to increase the conformational heterogeneity of adducts toward a larger proportion of intercalated structures. The increased level of adduct formation may in part be explained by an increased level of noncovalent intercalative binding of the diol epoxide and enhanced conditions for formation of transition state complexes that yield intercalated adducts.

References (1) International Agency for Research on Cancer (1986) Tobacco Smoking. IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans, Vol. 38, International Agency for Research on Cancer, Lyon, France. (2) Hoffmann, D., and Hecht, S. S. (1990) Advances in tobacco carcinogenesis. In Chemical Carcinogenesis and Mutagenesis I (Cooper, C. S., and Grover, P. L., Eds.) pp 63-102, SpringerVerlag, Berlin.

Effect of m5dC on dG-BPDE Adduct Formation (3) Gelboin, V. H. (1980) Benzo[a]pyrene metabolism, activation, and carcinogenesis: role and regulation of mixed function oxidases and related enzymes. Physiol. Rev. 60, 1107-1166. (4) Cooper, C. S., Grover, P. L., and Sims, P. (1983) The metabolism and activation of benzo[a]pyrene. Prog. Drug Metab. 7, 295-396. (5) Thakker, D. R., Yagi, H., Levin, W., Wood, A. W., Conney, A. H., and Jerina, D. M. (1985) Polycyclic aromatic hydrocarbons: metabolic activation to ultimate carcinogens. In Bioactivation of Foreign Compounds (Anders, M. W., Ed.) pp 177-242, Academic Press, New York. (6) Harvey, R. G. (1991) Polycyclic Aromatic Hydrocarbons: Chemistry and Carcinogenicity, Cambridge University Press, Cambridge, U.K. (7) Jeffrey, A. M. (1985) Polycyclic aromatic hydrocarbon-DNA adducts: formation, detection and characterization. In Polycyclic Hydrocarbons and Carcinogenesis (Harvey, R. G., Ed.) ACS Symposium Series 283, pp 187-208, American Chemical Society, Washington, DC. (8) Cheng, S. C., Hilton, B. D., Roman, J. M., and Dipple, A. (1989) DNA adducts from carcinogenic and non-carcinogenic enantiomers of benzo[a]pyrene diol epoxide. Chem. Res. Toxicol. 2, 334340. (9) Jernstro¨m, B., and Gra¨slund, A. (1994) Covalent binding of benzo[a]pyrene 7,8-dihydrodiol 9,10-epoxides to DNA: Molecular structures, induced mutations and biological consequences. Biophys. Chem. 49, 185-199. (10) Hollstein, M., Sidranski, D., Vogelstein, B., and Harris, C. C. (1991) p53 mutations in human cancers. Science 253, 49-53. (11) 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. (12) Hollstein, M., Shomer, B., Greenblatt, M., Soussi, T., Hovig, E., Montesano, R., and Harris, C. C. (1996) Somatic point mutations in the p53 gene of human tumors and cell lines: updated compilation. Nucleic Acids Res. 24, 141-146. (13) Denissenko, M. F., Pao, A., Tang, M.-s., and Pfeifer, G. P. (1996) Preferential formation of benzo[a]pyrene adducts at lung cancer mutational hotspots in p53. Science 274, 430-432. (14) 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. (15) Geacintov, N. E., Shahbaz, M., Ibanez, V., Moussaoui, K., and Harvey, R. G. (1988) Base-sequence dependence of noncovalent complex formation and reactivity of benzo[a]pyrene diol epoxide with polynucleotides. Biochemistry 27, 8380-8387. (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, 38933898. (17) Jernstro¨m, B., Funk, M., Frank, H., Mannervik, B., and Seidel, A. (1996) Glutathione S-transferase A1-1 catalysed conjugation of bay and fjord region diol epoxides of polycyclic aromatic hydrocarbons with glutathione. Carcinogenesis 17, 1491-1498. (18) Michaud, D. P., Gupta, S. C., Whalen, D. L., Sayer, J. M., and Jerina, D. M. (1983) Effects of pH and salt concentration on the hydrolysis of a benzo[a]pyrene 7,8-diol-9,10-epoxide catalyzed by DNA and polyadenylic acid. Chem.-Biol. Interact. 44, 41-52. (19) Ponte´n, I., Seidel, A., Gra¨slund, A., Norde´n, B., and Jernstro¨m, B. (1994) Spectroscopic studies of of the trans adducts derived from the (+)- and (-)-anti-benzo[a]pyrene 7,8-dihydrodiol 9,10epoxide and the oligonucleotide 5′-d(CCTATAGATATCC). Carcinogenesis 15, 2007-2213. (20) Ponte´n, I., Seidel, A., Gra¨slund, A., and Jernstro¨m, B. (1996) Synthesis and characterization of adducts derived from the syndiastereomer of benzo[a]pyrene 7,8-dihydrodiol 9,10-epoxide and the 5′-d(CCTATAGATATCC) oligonucleotide. Chem. Res. Toxicol. 9, 188-196. (21) Pradhan, P., Jernstro¨m, B., Seidel, A., Norde´n, B., and Gra¨slund, A. (1998) Induced circular dichroism of benzo[a]pyrene 7,8dihydrodiol 9,10-epoxide stereoisomers covalently bound to deoxyribooligonucleotides exploited to probe equilibrium distribution between groove binding and intercalative adduct conformations. Biochemistry 37, 4664-4673.

Chem. Res. Toxicol., Vol. 12, No. 9, 1999 821 (22) Geacintov, N. E., Cosman, M., Mao, B., Alfano, A., Hingerty, B. E., Ibanez, V., and Harvey, R. G. (1991) Spectroscopic characteristics and site I/site II classification of cis- and trans-benzo[a]pyrene diol epoxide enantiomer-guanosine adducts in oligonucleotides and polynucleotides. Carcinogenesis 12, 2099-2108. (23) Cosman, M., Ibanez, V., Geacintov, N. E., and Harvey, R. G. (1990) Preparation and isolation of adducts in high yield derived from the binding of two benzo[a]pyrene-7,8-dihydroxy-9,10-oxide stereoisomers of the oligonucleotide d(ATATGTATA). Carcinogenesis 11, 1667-1672. (24) Meehan, T., and Straub, K. (1979) Double stranded DNA stereoselectively binds benzo[a]pyrene diol epoxides. Nature 277, 410412. (25) Funk, M., Ponte´n, I., Seidel, A., and Jernstro¨m, B. (1997) Critical parameters for adduct formation of the carcinogen (+)-anti-benzo[a]pyrene-7,8-dihydrodiol 9,10-epoxide with oligonucleotides. Bioconjugate Chem. 8, 310-317. (26) Ya, N.-Q., Smirnov, S., Cosman, M., Bhanot, S., Ibanez, V., and Geacintov, N. E. (1994) Thermal stabilities of benzo[a]pyrene diol epoxide-modified oligonucleotide duplexes. Effects of mismatched complementary strands and bulges. In Structural Biology: The state of the Art. Proceedings of the 8th Conversation (Sarma, R. H., and Sarma, M. H., Eds.) pp 349-366, Adenine Press, Schenectady, NY. (27) Cosman, M., Geacintov, N. E., and Amin, S. (1993) Thermodynamics of helix-coil transitions in stereospecific benzo[a]pyrene diol epoxide-oligonucleotide adducts. In Polycyclic Aromatic Compounds (Garrigues, P., and Lamotte, M., Eds.) pp 1151-1158, Gordon and Breach Science, Lausanne, Switzerland. (28) Geacintov, N. E. (1985) Mechanisms of interaction of polycyclic aromatic diol epoxides with DNA and structures of adducts. In Polycyclic hydrocarbons and carcinogenesis (Harvey, R. G., Ed.) ACS Symposium Series 283, pp 107-124, American Chemical Society, Washington, DC. (29) Pradhan, P., Jernstro¨m, B., Seidel, A., Norde´n, B., and Gra¨slund, A. (1999) Studies on the adduct heterogeneity of benzo[a]pyrene 7,8-dihydrodiol 9,10-epoxide stereoisomers covalently bound to deoxyribooligonucleotides by induced circular dichroism and light absorption spectroscopy. Chem. Res. Toxicol. 12, 403-411. (30) Fountain, M. A., and Krugh, T. R. (1995) Structural characterization of a (+)-trans-anti-benzo[a]pyrene-DNA adduct using NMR, restrained energy minimization, and molecular dynamics. Biochemistry 34, 3152-3161. (31) Geacintov, N. E., Cosman, M., Hingerty, B. E., Amin, S., Broyde, S., and Patel, D. J. (1997) NMR solution structures of stereomeric polycyclic aromatic carcinogen-DNA adducts: Principles, patterns and diversity. Chem. Res. Toxicol. 10, 111-146. (32) Roche, C. J., Jeffrey, A. M., Mao, B., Alfano, A., Kim, S. K., Ibanez, V., and Geacintov, N. E. (1991) Dependence of conformations of benzo[a]pyrene diol epoxide-DNA adducts derived from stereoisomers of different tumorigenicities on base sequence. Chem. Res. Toxicol. 4, 311-317. (33) Hess, M. T., Gunz, D., Luneva, N., Geacintov, N. E., and Naegeli, H. (1997) Base pair conformation-dependent excision of benzo[a]pyrene diol epoxide-guanine adducts by human nucleotide excision repair enzymes. Mol. Cell. Biol. 17, 7069-7076. (34) Suh, M., Ariese, F., Small, G. J., Jankowiak, R. J., Hewer, A., and Phillips, D. H. (1995) Formation and persistence of benzo[a]pyrene DNA adducts in mouse epidermis in vivo: importance of adduct conformation. Carcinogenesis 16, 2561-2569. (35) Jankowiak, R., Ariese, F., Hewer, A., Luch, A., Zamzow, D., Hughes, N. C., Phillips, D., Seidel, A., Platt, K. L., Oesch, F., and Small, G. J. (1998) Structure, conformations, and repair of DNA adducts from dibenzo[a,l]pyrene: 32P-postlabeling and fluorescence studies. Chem. Res. Toxicol. 11, 674-685. (36) 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. (37) Chen, F.-M. (1986) Binding of enantiomers of trans-7,8-dihydroxyanti-benzo[a]pyrene to polynucleotides. J. Biomol. Struct. Dyn. 4, 401-418.

TX980230F