Patterns of Resistance to Exonuclease Digestion of Oligonucleotides

to Exonuclease Digestion of Oligonucleotides Containing Polycyclic Aromatic ... of the exocyclic 6-amino group of dA by polycyclic aromatic hydroc...
0 downloads 0 Views 125KB Size
Download PDF
1330

Chem. Res. Toxicol. 2001, 14, 1330-1338

Patterns of Resistance to Exonuclease Digestion of Oligonucleotides Containing Polycyclic Aromatic Hydrocarbon Diol Epoxide Adducts at N6 of Deoxyadenosine Palanichamy Ilankumaran, Lewis K. Pannell, Petros Gebreselassie, Anthony S. Pilcher, Haruhiko Yagi, Jane M. Sayer, and Donald M. Jerina* Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases, The National Institutes of Health, Bethesda, Maryland 20892-0820 Received May 14, 2001

The effect of adduct stereochemistry on the susceptibility to hydrolysis by snake venom (VPD) and bovine spleen (SPD) phosphodiesterases was investigated with short deoxyoligonucleotides containing defined adducts derived from alkylation of the exocyclic 6-amino group of dA by polycyclic aromatic hydrocarbon diol epoxides (DEs). In accordance with several earlier reports, we have found that adducts with R configuration at the site of attachment of dA to the DE moiety derived from either benzo[a]pyrene (BaP) or benzo[c]phenanthrene (BcPh) are generally more resistant to hydrolysis by VPD than are their (S)-diastereomers. The reaction with VPD initially yields a fragment containing the adducted dA residue at its 3′-end, which slowly hydrolyzes to a dimer (pXpA*) with an intact 5′-phosphodiester bond to the adducted dA. With several of the adducts studied, this dimer underwent cleavage to release eventually the monomeric adduct p(dA*). Adducts derived from cis opening of the epoxide ring of both BaP and BcPh DEs were considerably more resistant to VPD than the corresponding trans-opened adducts. Although several previous investigations had suggested that oligonucleotides containing adducts which have S configuration at the site of attachment of the hydrocarbon to adenine are more resistant to cleavage by SPD than are their (R)-diastereomers, the present results with a more extensive set of oligonucleotides indicate that SPD, in contrast to VPD, exhibits little discrimination between adducts with R and S configuration at the site of attachment to the base. Notably, for both enzymes, the most resistant internucleotide linkage (the bond 3′sugar to phosphate for VPD and 5′-sugar to phosphate for SPD) is between the modified base and the base immediately 5′ to it, regardless of the configuration of the adduct. Polycyclic aromatic hydrocarbons are environmental contaminants that are metabolically activated to give mutagenic and carcinogenic bay- or fjord-region diol epoxides (1, 2). The initial step in tumorigenesis and mutagenesis induced by these diol epoxides most likely involves covalent modification of DNA. For a number of diol epoxides, the principal targets for alkylation in DNA have been shown to be the exocyclic amino groups, N6 of deoxyadenosine (dA) and N2 of deoxyguanosine (dG), which react at the benzylic position of the epoxide ring to give cis- and trans-opened adducts (3, 4). For a given hydrocarbon, mammalian metabolism produces four diol epoxide isomers [two enantiomers of diol epoxide-1 (DE1),1 in which the epoxide group and the benzylic hydroxyl are cis, and two enantiomers of diol epoxide-2 (DE-2), in which these groups are trans; see Figure 1]. Cis and trans addition of a given base to these DEs results in a total of eight possible isomeric adducts. Previous observations in our laboratory (5, 6) and elsewhere (7-9) have docu-

mented that the stereochemistry of these adducts has a marked influence on the ease of their release from DNA by exonucleases such as snake venom phosphodiesterase (VPD) and spleen phosphodiesterase (SPD). However, the effects of factors in addition to the configuration at the site of attachment of the hydrocarbon to the purine, namely, variation of the parent hydrocarbon as well as cis vs trans opening of the epoxide, on the enzymatic hydrolysis of deoxyoligonucleotides containing specific adducts have not been systematically addressed. Since VPD and SPD cleave single-stranded oligonucleotides in opposite directions [from the 3′- to 5′- and the 5′- to 3′ends, respectively (10)], it was anticipated (5, 9) that the relative sensitivities of adducted oligonucleotides to these two enzymes might help to elucidate the orientation of bulky hydrocarbon adducts based on the direction of cleavage.

* To whom correspondence should be addressed. Phone: (301) 4964560. Fax: (301) 402-0008. E-mail: [email protected]. 1 Abbreviations: BaP, benzo[a]pyrene; BcPh, benzo[c]phenanthrene; DE, diol epoxide; DE-1, diol epoxide in which the epoxide oxygen and the benzylic hydroxyl group are cis; DE-2, diol epoxide in which the epoxide oxygen and the benzylic hydroxyl group are trans; SPD, spleen phosphodiesterase; VPD, snake venom phosphodiesterase; TBDMS, tert-butyldimethylsilyl.

Chemical reagents obtained commercially were used without further purification. Snake venom phosphodiesterase from Crotalus adamanteus was purchased from USB, bovine spleen phosphodiesterase from Boehringer Mannheim (Indianapolis, IN) and Escherichia coli alkaline phosphatase from Sigma. We used the enzyme activities provided by the suppliers. The indicated enzyme activities were checked as follows. For VPD,

10.1021/tx010092l

Materials and Methods

This article not subject to U.S. Copyright. Published 2001 by the American Chemical Society Published on Web 08/30/2001

Exonuclease Digestion of Diol Epoxide Adducts

Chem. Res. Toxicol., Vol. 14, No. 9, 2001 1331 Table 1. Sequences of the R/S Pairs of Adducted Oligonucleotides Used in the Present Study 1 2 3 4 5 6 7 8 a

Figure 1. Structures of the hydrocarbons BaP and BcPh, their bay- and fjord-region DEs, and the adducts formed on cis and trans opening of the DEs by the exocyclic N6-amino group of deoxyadenosine. Metabolism to the DE occurs on the numbered ring (shown with heavy bonds) in each hydrocarbon structure. Note that only one enantiomer of each DE diastereomer is shown. Adducts are labeled based on the diol epoxide diastereomer from which they are derived (in DE-2, the epoxide oxygen and benzylic hydroxyl group are trans whereas in DE-1, these groups are cis). one unit is defined by USB as the amount of enzyme required to hydrolyze 1 µmol of thymidine 5′-p-nitrophenyl phosphate/ min at 25 °C. The activity of VPD measured at 37 °C (this work) with this substrate at pH 8.9 as described (10) but with the addition of MgCl2 (final concentration 10 mM) was approximately twice as great (note higher temperature) as that indicated by USB for the batch of enzyme used and was linear in enzyme concentration over the range of 0-0.4 unit/mL. For SPD, one unit is defined by Boehringer Mannheim as the amount of enzyme required to hydrolyze 1 µmol of thymidine 3′-p-nitrophenyl phosphate/min at 25 °C. The activity of SPD was measured (this work) by a continuous UV assay (330 nm;  5520 M-1cm-1) of the hydrolysis of thymidine 3′-p-nitrophenyl phosphate (0.4 mM) essentially as described (11), but without EDTA (10) in the assay mixtures. In our experience, the observed initial velocities did not depend linearly on the enzyme concentration, but tended to level off as the enzyme concentration was increased. At the lowest enzyme concentration used (2 × 10-3 units/mL), the activity per milliliter of enzyme solution determined by us was, however, in good agreement with the supplier’s stated value. Oligonucleotides. General synthetic procedures for the preparation of BcPh dA phosphoramidites used in the synthesis of the oligonucleotides below are provided in the Supporting Information. Phosphoramidites of adducts (see Figure 1) corresponding to cis and trans opening of DE-1 and trans opening of DE-2 were prepared by the method of Lakshman et al. (12) in which a silyl protected fluoropurine is coupled to an amino triol of the hydrocarbon. For cis opened DE-2, we utilized the aminohydroxylation procedure described by Pilcher et al. (13)

sequence

adduct

ref

CTCTCA*CTTCC TTTA*GAGTCTGCTCCC TTTA*GAGTCTGCTCCC TTTA*GAGTCTGCTCCC TTTA*GAGTCTGCTCCC TTTA*GAGTCTGCTCCC TTCA*GAGTCTGCTCCC TTGA*GAGTCTGCTCCC

BcPh DE-2 trans BcPh DE-2 trans BcPh DE-1 trans BaP DE-2 trans BcPh DE-2 cis BaP DE-2 cis BcPh DE-2 cis BcPh DE-2 cis

8 16 16 17 6 18 a a

This work.

in which dA, again with silyl protecting groups on the sugar, is oxidatively coupled to the BcPh trans-3,4-dihydrodiol. None of the product corresponding to cis opening of DE-1 was detected as was also the case for the aminohydroxylation of BaP trans7,8-dihydrodiol (13). This represents the first application of the dA aminohydroxylation procedure to a dihydrodiol with a fjordregion double bond. Chemical shifts and coupling constants for the hydrogens on the saturated benzo-ring of the disilyl triacetate adducts generated in the present study were consistent with values for the tetraol tetraacetates as well as the pentaacetates of the dA and dG adducts previously described (see Supporting Information). Neither acetates nor TBDMS protecting groups (on the hydrocarbon or sugar hydroxyl groups, respectively) substantially affect the conformation of the these adducts as shown by their CD spectra (Supporting Information). Thus, the CD spectra of the four diastereomeric 1R/1S pairs of TBDMS protected adducts could be used for assignment of their absolute configuration by comparison with the known (14) unprotected adducts. Diastereomeric pairs of adducted oligonucleotides used in the present study are listed in Table 1. Adducted 11-mer 1 (CTCTCA*CTTCC) containing trans-opened (1R)- and (1S)BcPh adducts at the central dA* was made as described (8) by direct reaction of the oligonucleotide with (()-BcPh DE-2 (15). Since in our experience, the yields on direct reaction of DEs with oligonucleotides were poor, larger quantities were prepared from the trans-BcPh DE-2/dA adducted phosphoramidite by a semiautomated method (cf. ref 12) using standard phosphoramidite methodology. All other oligonucleotides (Table 1) were prepared only by the phosphoramidite method. Oligonucleotide 1 was purified by HPLC at 60 °C on a Zorbax Eclipse XDB C18 column (4.6 × 250 mm) eluted at 1.5 mL/min with a linear gradient of CH3CN in 0.1 M (NH4)2CO3 that increased the proportion of CH3CN from 0 to 11% over 20 min, followed by a ramp to 50% CH3CN over the next 20 min: Tr (early isomer), 16.4 min; Tr (late isomer), 17.7 min. The early and late eluting isomers were assigned as R and S, respectively, on the basis of their CD spectra (8) as well as the CD spectra of the monomeric nucleoside adducts (14) isolated from the enzymatic hydrolysis mixtures (see below). Thus, the elution order under the present chromatographic conditions is the same as that reported by Laryea et al. (8) in a different HPLC system. Syntheses and purifications of the diastereomeric pairs of adducted 16-mers 2-6 (Table 1) have been described previously (6, 16-18). The diastereomeric pairs of adducted 16-mers 7 and 8 were synthesized by the same method (6, 12) and purified by HPLC on a Beckman Ultrasphere C18 column (10 × 250 mm) eluted at 3 mL/min with a linear gradient of acetonitrile in 0.1 M (NH4)2CO3, pH 7.5, that increased the acetonitrile composition from 5 to 11% over 40 min. In both cases, the early eluting isomer had (1R)-adduct configuration: for 7, Tr 36.5 min (1R) and 38.6 min (1S); for 8, 27.3 min (1R), and 29.2 min (1S). Absolute configurations of the diastereomers of 7 and 8 were assigned by enzymatic hydrolysis to the known (14) nucleoside adducts with 1R and 1S configuration at the site of attachment of the base to the hydrocarbon. Their CD spectra are shown in Figure 2. Enzymatic Digestion. SPD. Each single-stranded oligonucleotide (0.5A260 unit) was dissolved in 0.5 mL of 150 mM sodium acetate buffer, pH 6.0, in a 1.5-mL Eppendorf tube. To

1332

Chem. Res. Toxicol., Vol. 14, No. 9, 2001

Ilankumaran et al.

Figure 2. Circular dichroism spectra normalized to 1.0 absorbance at 260 nm (in 20 mM sodium phosphate, pH 7, ionic strength 100 mM [NaCl]) of the diastereomers of oligonucleotides 7 and 8. gradient of CH3CN in 0.01 M ammonium carbonate buffer, pH 7.4, that increased the proportion of CH3CN from zero to 15% over 25 min, followed by a ramp to 40% CH3CN over the next 20 min) with an interface to a Hewlett-Packard 1100 electrospray mass spectrometer. Masses were scanned from m/z 300 to 2500 every 4 s in negative ion mode. Molecular weights were calculated by deconvolution using the instrument’s software. All molecular weights were within 1 mass unit or less of the calculated (average mass) values. Mass data for selected partial hydrolysis products are given in the Supporting Information. VPD. Each single-stranded oligonucleotide (0.5A260 unit) was dissolved as above in 0.5 mL of 50 mM Tris-HCl buffer containing 10 mM MgCl2 at pH 8.5. To the solution was added 1 unit of VPD, and the resulting mixture was incubated at 37 °C. Quenching, alkaline phosphatase digestion, and HPLC and MS analyses were as described above for the products of the SPD incubations.

Results and Discussion

Figure 3. Chromatograms showing the time course of VPD digestion of the diastereomeric oligonucleotides 2 (TTTA*GAGTCTGCTCCC) containing trans-opened (1R) (left)- and (1S) (right)-adducts of BcPh at A*. Shaded peaks correspond to an impurity that did not contain the hydrocarbon or DNA chromophore. The peak designated a had a UV spectrum similar to those of the tetramer and dimer and may represent a minor trimer intermediate resulting from loss of the 5′-terminal thymidine residue. HPLC conditions: Zorbax Eclipse XDB C18 column (4.6 × 250 mm) eluted at 1.2 mL/min with a gradient of solvent B in solvent A that increased the percentage of B from 0 to 100% in 30 min, where solvent A is 50 mM sodium phosphate buffer (pH 7.0) containing 5% MeOH, and solvent B is 10% H2O in MeOH. Chromatograms were monitored at 260 nm. the solution was added 0.06-0.6 unit of SPD, and the resulting mixture was incubated at 37 °C. At appropriate time intervals, aliquots (150 µL) were quenched by heating in boiling water for 2 min. Alkaline phosphatase (2 units) was added to the quenched solution, and the mixture was incubated for another 5 h at 37 °C, at the end of which time 75-µL samples were analyzed by HPLC (see Figure 3). For identification of partial hydrolysis products, the appropriate HPLC peaks were collected, dried and redissolved in a minimal volume of H2O. Samples were analyzed by HPLC/MS (Zorbax 5 µm 300SB-C8 column, 2.1 × 150 mm, typically eluted at 0.3 mL/min with a linear

It is known that adducts with bulky hydrocarbon moieties retard digestion of the adducted DNA by exonucleases such as snake venom (VPD) and spleen (SPD) phosphodiesterases due to resistance to hydrolysis at or near the adduct site. When DNA or an oligonucleotide is treated with a DE to produce adducts at dA and dG and subsequently subjected to (partial) enzymatic hydrolysis by VPD or SPD, incomplete release of certain dA adducts is observed experimentally under conditions which release all the dG adducts (5, 19). Although most work has focused on the effect of the stereochemistry of dA adducts on the selectivity of their exonuclease digestion, diastereomer selectivity of these exonucleases has been documented for dG adducts as well (9). VPD hydrolyzes singlestranded DNA proceeding from the 3′-end toward the 5′end, with cleavage of the 3′-sugar to phosphate bond and release of nucleoside 5′-phosphates. In contrast, SPD cleaves oligonucleotides in the opposite direction, starting at the 5′- and proceeding toward the 3′-end with cleavage of the 5′-sugar to phosphate bond and releasing nucleoside 3′-phosphates. The majority of observations on single-stranded oligonucleotides containing dA adducts have been consistent with the interpretation that cleavage by VPD is retarded to a greater extent by adducts with R as compared to S configuration at the site of attachment to the N6-amino group of dA, whereas the opposite is true for SPD (5-7). These observations led to the suggestion (5) that dA adducts in single-stranded DNA orient so that the hydrocarbon portion of (R)adducts lies toward the 3′-end and preferentially blocks

Exonuclease Digestion of Diol Epoxide Adducts

access by VPD, whereas (S)-adducts lie toward the 5′end and preferentially inhibit SPD cleavage. Interestingly, these proposed orientations were opposite to those observed by NMR (20-25) for intercalated BcPh and BaP dA adducts in double-stranded DNA. The aim of the present study was to examine more systematically the effect of adduct configuration, as well as effects of the parent hydrocarbon and of cis vs trans ring opening, on susceptibility to VPD and SPD hydrolysis. Thus, we investigated several cis- and trans-dA adducts (A*) derived from BaP and BcPh DE within a single sequence context, TTTA*GAGTCTGCTCCC. Venom Phosphodiesterase Hydrolysis. Figure 3 shows partial HPLC traces obtained after alkaline phosphatase treatment of reaction mixtures quenched at different times during the VPD hydrolysis of the (R) (left side)- and (S) (right side)-adducted diastereomers of oligonucleotide 2 (trans-BcPh DE-2 dA adduct). The nucleosides dA, dT, dC, and dG eluted before 25 min (data not shown). After 15 min of digestion, both isomeric 16-mers (tR ) 22-23 min) had entirely disappeared, and the first HPLC trace obtained for each isomer showed a single hydrocarbon-containing species that eluted at ∼26 min. After 24 h, the peak at ∼26 min had disappeared, and two new products were observed at ∼28 and ∼29.5 min from both (1R)- and (1S)-isomers. After 48 h, the intermediate at ∼28 min still constituted a substantial fraction of the reaction mixture from the (1R)-isomer. The products that eluted at ∼29.5 min corresponded to the fully hydrolyzed nucleoside adducts, as shown by their CD spectra, which were identical to those of the known compounds (14). Mass spectral analysis indicated that the intermediates derived from both diastereomers and eluting at ∼26 min are tetramers, TTTA* (calcd, 1442; found, 1442) whereas the intermediates eluting at ∼28 min are dimers, TpA* (calcd, 833; found, 833). The first observable partial hydrolysis products from both (1R)and (1S)-adducted 2 thus corresponded to cleavage of the phosphodiester linkage immediately 3′ to the adduct. Adducted tetramer fragments with the same sequence were also identified by MS (see Supporting Information) upon enzymatic cleavage of the oligonucleotides containing the diastereomeric (1R) and (1S)-BcPh DE-2 cis dA adducts 5. The same result was also obtained with the corresponding BaP DE-dA adducts. After 15 min of hydrolysis by VPD, both the (10R)- and (10S)-trans (4) and cis (6) BaP DE-2 adducted oligonucleotides exhibited only one significant adduct-containing peak on HPLC (tR ≈ 27 min). In all four cases, this product had a mass corresponding to the tetramer TTTA*. Notably, no adducted pentamer, which would have arisen from stalling of the enzyme one nucleotide short of the adduct site, was detected. The initial blockage is thus at the 3′ O-P bond immediately 5′ to the adduct. A BaP adducted dimer (see below) was isolated after 6 h. In the face of blockage adjacent to the adduct, the enzyme could either react slowly at the highly resistant bond to release a monomeric 5′-phosphorylated adduct or move to the next phosphodiester linkage to give the dinucleotide adduct with a 5′-phosphate, which subsequently cleaves to give the expected monomeric adduct (Scheme 1). Accumulation and decay of a dinucleotide over the time course of the reaction (Figure 3) is qualitatively consistent with slow, endonucleolytic hydrolysis to the dinucleotide which then undergoes even slower cleavage to monomers. Possible cleavage products

Chem. Res. Toxicol., Vol. 14, No. 9, 2001 1333 Scheme 1

longer than dimers were at best very minor intermediates and did not accumulate to any significant extent (cf. Figures 3 and 6). The observation of a dinucleotide intermediate was initially surprising, since VPD generally digests linear oligonucleotides by stepwise cleavage of single oligonucleotide residues (10). Endonuclease activity of VPD is not unprecedented when the normal exonuclease cleavage is inhibited. “Skipping” of the 3′ O-P bond immediately 5′ to a DNA lesion site in favor of attack on the next phosphodiester bond (26), analogous to dimer formation in the present study, has been reported. When DNA fragments containing thymine photodimers (T*T*) were subjected to VPD hydrolysis, products resistant to further digestion were identified as trimers of the type pXpT*T*, where X represents any nucleic acid base. In a more recent study, VPD was observed to stall at the lesion site in oligonucleotides containing a formamido nucleoside lesion, an 8-oxo dG or both these lesions in tandem. Longer incubation times resulted in the appearance of multiple products, which arose from slow endonucleolytic cleavage of this initial product at various positions in the sequence 5′ to the lesion (27). Figure 4 illustrates the effect of adduct type and stereochemistry on the susceptibility of the 16-mer oligonucleotides to VPD hydrolysis. Hydrolysis to the dimer stage was complete after 48 h in all but three cases [all of which were (1R)-BcPh DE adducts]. The monomeric nucleotide adduct is released by cleavage of the internucleotide linkage immediately 5′ to the adduct in an even slower reaction. Thus, the extent of monomer formation at a given time should reflect the ease of cleavage at this most resistant position, regardless of whether its direct precursor is the dimer or the tetramer. In contrast to a previous study (7) in which relative sensitivities of (R)- and (S)-dA adducts to VPD were determined on the basis of rates of hydrolysis up to (but not beyond) the stall site, we would argue that the rate of hydrolysis of this most resistant internucleotide bond provides a better measure of the relative sensitivities of different adducted oligonucleotides to enzymatic digestion. All the (S)-adducts but one (BaP DE-2 cis) gave at least some monomer after 48 h. In contrast, only two of the seven 16-mers with R configuration, both containing DE-2 trans adducts, showed any monomer release after 48 h. For the BaP DE-2 cis adducts, both the (R)- and the (S)-isomers were completely resistant to hydrolysis to the monomeric adduct level under these conditions. Surprisingly, the BaP DE-2 trans adduct showed a reversal of selectivity in that more monomeric adduct was

1334

Chem. Res. Toxicol., Vol. 14, No. 9, 2001

Ilankumaran et al.

Figure 4. Distribution of hydrolysis products from 16-mer oligonucleotides (0.5 A260) containing different dA adducts of diol epoxides in a ∼TA*G∼ sequence context after 48 h of digestion with 1 unit of VPD followed by alkaline phosphatase. Bars represent percent of the total area of adduct-containing peaks measured at 260 nm, with cross-hatched areas corresponding to the monomer (nucleoside adduct) formed on complete hydrolysis and open areas corresponding to the dimer (TpA*) formed by hydrolysis when the phosphodiester bond immediately 5′ to the adduct is bypassed. Where the total height of the bar is less than 100%, the difference from 100% corresponds to residual tetramer in which the enzyme had stalled immediately 3′ to the adduct position. The insert shows the product distribution from three different sequence contexts containing the same dA adducts (BcPh DE-2 cis).

Figure 5. Time course for the digestion of BcPh DE-2 trans (1S)-2 and cis (1S)-5 adducted oligonucleotides by VPD (1 unit/0.5 A260). The ordinate axis represents integration areas of the respective peaks at 260 nm. Solid lines correspond to tetramer, dotted lines, to dimer and dashed lines, to monomer. The first time point measured was at 5 h. At this time no tetramer remained in the digestion of the trans isomer.

released from the (R)- than from the (S)-isomer in a given time (see below). For (S)-adducts with the partial sequence TTTA*G∼, relative extents of cleavage to monomer were in the order BcPh DE-2 trans > BcPh DE-1 trans > BcPh DE-2 cis > BaP DE-2 trans > BaP DE-2 cis. For these (S)-adducts, cleavage to the monomer was more difficult for BaP adducts than for the corresponding BcPh adducts and for cis relative to trans adducts. The effect of cis opening is particularly striking. Figure 5 illustrates the time course for hydrolysis of the BcPh DE-2 trans and cis (S)-adducts (oligonucleotides 2 and 5) with VPD. Notably, the halftime for monomer formation is less than 5 h for the trans isomer, but on the order of 35 h for the cis isomer. Even after 70 h, a significant amount (∼25%) of cis-adducted dimer is still present. The BcPh DE-2 cis (S)- and trans (R)-adducts are comparable in their susceptibility to VPD, because the unusual resistance of the cis adduct to VPD cleavage compensates for the normal preference of VPD for the S configuration. Similarly, for BaP DE-2 adducts, the retarding effect of the cis isomer is so large that the BaP DE-2 cis (S)-dA adduct is actually more resistant than the corresponding trans (R)-adduct. Under the standard set of conditions, neither the (10R)- nor the

(10S)-BaP DE-2 cis adducts gave any monomeric adduct even after 48 h, although at a higher VPD concentration (4 units/0.5A260 of oligonucleotide, 4 days) both diastereomeric oligonucleotides were completely hydrolyzed to the monomer level. The effect of the base 5′ to the adduct when A* was the (1S)-BcPh DE-2 cis adduct was small, and followed the order -CA*- ) -GA*- > -TA*- (insert, Figure 4). For the (1R)-BcPh DE-2 cis adduct, the sequence ∼CA*G∼ appeared to accelerate the cleavage 5′ to the C to give dimer, but no monomer was formed after 48 h from any of the sequences containing this cis adduct. In all but one of the diastereomeric pairs of 16-mers, the oligonucleotides containing (R)-adducts at dA were considerably more resistant to hydrolysis by VPD to the nucleotide level than the corresponding (S)-diastereomers. The exception was the pair of BaP DE-2 trans adducts where the (R)-adduct gave much more monomer compared to dimer than did the (S)-adduct (Figure 4). To preclude the possibility of error, the nucleoside adducts from these digestion mixtures were isolated. Their CD spectra (cf. ref 28) showed that the nucleoside adducts derived from the fast and slow reacting oligonucleotides in this case did in fact have the R and S configuration,

Exonuclease Digestion of Diol Epoxide Adducts

Figure 6. Chromatograms showing the time course of VPD digestion of the diastereomeric oligonucleotides 1 (CTCTCA*CTTCC) containing trans-opened (1R) (A)- and (1S) (B)adducts of BcPh at A*. The shaded peak corresponds to an impurity that did not contain the hydrocarbon or DNA chromophore. The peaks designated a and b had UV spectra intermediate between those of the hexamer and dimer and may correspond to minor cleavages of the hexamer to give sequences longer than the major dimer intermediate. (C) Normalized UV spectra of the partially hydrolyzed intermediates and the nucleoside (monomer) product derived from the (1S)-dA adduct. Solid line, hexamer (∼24 min); dotted line, dimer (∼27 min); dashed line, monomer (∼29 min). For HPLC conditions, see Figure 3.

respectively, confirming this unexpected result. We have no explanation for this reversal, and it should be noted that both isomers gave at least some monomer after 48 h; thus, the difference in reactivity between the two diastereomers in this case is not very large. Another apparent exception to the general trend of resistance to VPD by (R)-dA adducts was reported by Laryea et al. (8) for the oligonucleotide CTCTCA*CTTCC containing the diastereomeric BcPh DE-2 trans adducts at dA*. The early eluting adducted oligonucleotide with 1R configuration at the point of attachment to the N6 amino group of adenine was reported to be more easily digested than its diastereomeric (1S)-adduct by VPD and less easily digested by SPD. Because we observed a reversal in sensitivity to VPD only for the pair of diastereomeric BaP DE-2 trans adducts and not for the corresponding BcPh DE-2 trans adducts, we reinvestigated the apparent anomaly in the previous work. After 30 min digestion with VPD of oligonucleotide 1 (earlyeluting isomer) containing the BcPh DE-2 trans (1R)-dA adduct, a single peak at 24.5 min that did not correspond

Chem. Res. Toxicol., Vol. 14, No. 9, 2001 1335

to the intact 11-mer (22.3 min) was observed. This product was identified by MS as the hexamer CTCTCA* (see Supporting Information). Longer incubation times resulted in the disappearance of the hexamer with concomitant formation of a later eluting product (Figure 6A, ∼27-28 min) that was identified by MS as the modified dinucleotide dCpA*. This product in turn is eventually (66 h) largely converted (∼80%) to the monomeric adduct that elutes ∼30 min. For comparison, the time course for the VPD digestion of the diastereomeric (1S)-adducted oligonucleotide 1 (late-eluting isomer) is shown in Figure 6B. In this case, essentially complete conversion to the nucleotide adduct was observed after 6 h. UV spectra (Figure 6C) of the hexamer and dimer intermediates were consistent with the presence of the hydrocarbon chromophore (note 300 nm shoulder) and as expected the contribution of the hydrocarbon to the UV absorbance increases as the size of the fragment becomes smaller. The UV spectra of the dimer and the nucleoside adduct (cf. ref 29) are almost identical. Our present observation that the BcPh DE-2 trans (1R)-dA adduct is more resistant to VPD than its (1S)diastereomer in oligonucleotide 1 is opposite to that reported by Laryea et al. (8). To confirm that this discrepancy did not result from an incorrect assignment of the (1R)- and (1S)-adducts in the adducted oligonucleotides, we isolated these adducts from the two diastereomeric oligonucleotides 1 after alkaline phosphatase treatment of their digestion mixtures and verified their absolute configurations by measurement of their CD spectra. The spectra for the adducts obtained from our early-eluting and late-eluting oligonucleotides, respectively, were identical to those reported for authentic BcPh DE-2 trans (1R)- and (1S)-dA adducts, respectively (14). Thus, the previous stereochemical assignment (8) of the early- and late-eluting adducted oligonucleotides as R and S was correct and identical to ours, but in the present study, the susceptibility of the (R)- and (S)-adducted isomeric oligonucleotides 1 to VPD digestion is opposite to the previous report. The cause of this discrepancy is unclear, but we believe our present results to be correct. Consistent with these results, these same (1R)- and (1S)BcPh DE-2 trans dA adduct diastereomers also show the same relative susceptibility to VPD when incorporated into a different oligonucleotide sequence (see above). Notably, the previous report did not identify a dimeric intermediate. Spleen Phosphodiesterase Hydrolysis. We subsequently investigated the hydrolysis of the present series of BaP and BcPh DE-2 adducted nucleotides with SPD. Digestion of the late eluting isomer (1S)-2 resulted in the initial formation of a single product that eluted at about the same position as the starting 16-mer but was shown by MS to be the 14-mer TA*GAGTCTGCTCCC. Thus, the action of SPD, which digests from the 5′- to the 3′-end of the oligonucleotide sequence, stopped one base before reaching the adduct site, in contrast to VPD which digests from the opposite direction and stalled at the adduct itself. As shown in Scheme 1, the most resistant cleavage site is the same for both enzymes, and corresponds to the internucleotide bond between to the adducted dA* and its 5′-adjacent residue. In general, differences in sensitivity to SPD between the (R)- and (S)-isomers of a given adduct were quite small and did not correlate well with the configuration

1336

Chem. Res. Toxicol., Vol. 14, No. 9, 2001

Ilankumaran et al.

Table 2. Relative Susceptibility of dA-adducted Oligonucleotides to SPD Hydrolysisa percent nucleoside adductb

partial sequence

time (h)

adduct

2

∼TA*G∼

1

BcPh DE-2 trans

4

∼TA*G ∼

1

BaP DE-2 trans

5

∼TA*G ∼

1

BcPh DE-2 cis

6

∼TA*G ∼

24

BaP DE-2 cis

7

∼CA*G ∼

26

BcPh DE-2 cis

8

∼GA*G ∼

5.5

BcPh DE-2 cis

S R S R S R S R S R S R

100 51 24 31 37 56 0 0 84 47 18 26

a The oligonucleotide (0.5 A 260 unit) was treated with 0.6 unit of SPD as described in the text. b Percentages based on relative peak areas (integrated at 260 nm) for the monomer and 14-mer.

Table 3. Hydrolysis of the Diastereomeric dA-Adducted 11-mer Oligonucleotides 1 (CTCTCA*CTTCC) with SPDa enzyme (units)c

time (h)

fully hydrolyzed productb (1R isomer)

fully hydrolyzed productb (1S isomer)

1.2 0.60 0.30 0.06

0.5 1 1 3

100 65 21 23

100 80 25 35

a On treatment of 0.5A 260 units of oligonucleotide with the indicated amount of enzyme as described in the text. b Percentages based on relative peak areas (integrated at 260 nm) for the nucleoside adduct and the adducted 7-mer intermediate. c One unit hydrolyzes 1 µmol of thymidine 3′-p-nitrophenyl phosphate per min at 25 °C. The enzyme activity reported by the manufacturer using this substrate was verified (this work) at low enzyme concentrations (see Materials and Methods).

at the point of attachment of the hydrocarbon to the base (Table 2). Cis vs trans opening had a dramatic effect in the case of both diastereomeric BaP DE-2 adducts, a smaller effect with the (1S)-BcPh DE-2 adducts, and no effect with the (1R)-BcPh DE-2 adducts. As with VPD, the BaP cis adducts exhibited very low reactivity with SPD. For example, no monomeric adduct was released from either diastereomer of the BaP DE-2 cis-adducted oligomer after 24 h, even though significant hydrolysis of both diastereomeric trans adducts had occurred after 1 h. Interestingly, there was a substantial sequence dependence for the hydrolysis by SPD of the oligonucleotides containing BcPh DE-2 cis adducts, with the rate of adducted monomer release from the partial sequences following the order TA*G > GA*G g CA*G for both (1R)and (1S)-isomers (Table 2; note different times). This is consistent with the resistant phosphodiester bond being between the adducted dA* and the residue immediately 5′ to it. We also repeated the digestion of (R)- and (S)adducted 1 with SPD (Table 3). Under the present conditions, which were apparently similar to those in the previous study (8), complete digestion of both oligonucleotides with 3 units of the spleen enzyme required less than 30 min, in contrast with >12 h for the (1R)-adduct as reported previously. Since the unit definition for the SPD used in the previous work was not clear, we repeated the determination using less enzyme to obtain partial hydrolysis. The selectivity of SPD was in the same direction as that previously reported; namely, the (1R)isomer was marginally more resistant than the (1S)isomer to SPD digestion, but in contrast with the

previous work (8), the difference between the isomers in the present study was very small (Table 3). Unlike VPD, which gave a tetramer and a dimer, SPD digestion of either the (1R)- or the (1S)-isomer produced only one resistantfragmentwithamasscorrespondingtoCA*CTTCC (see Supporting Information). Thus, the resistant bond was on the 5′-side of the adduct, analogous to our observation with the 16-mer oligonucleotide 2. In summary, the present evidence strongly suggests that the internucleotide bond most resistant to cleavage by both VPD and SPD is immediately 5′ to the adduct (Scheme 1). VPD easily cleaves from the 3′-end up to the adduct and then skips this resistant linkage. The next phosphodiester bond is cleaved instead to give a dinucleotide intermediate. SPD cleaves from the 5′-end up to the base preceding the adduct on the 5′-side and then stalls. Skipping of the resistant bond to give a dimer was not observed with SPD. With the present set of substrates, the rate of SPD cleavage of the most resistant bond shows no consistent relationship with the configuration of the adduct. In contrast, VPD generally cleaves (S)-adducted oligonucleotides considerably faster than their (R)-adducted counterparts, as previously observed on the basis of a smaller number of adducts. Mao et al. (9) studied the enzymatic cleavage of oligonucleotides containing trans dG adducts derived from BaP DE-2 and observed that stall sites for both VPD and SPD were diastereomer dependent. VPD (proceeding from the 3′- to the 5′- end of the oligonucleotide) was slow or failed to hydrolyze the phosphodiester linkage on the 3′-side of trans (10R)-dG adducts and on the 5′-side of trans (10S)-dG adducts. Thus, the initially formed oligonucleotide fragment that is resistant to further VPD hydrolysis was one base longer for the (10R)- than for the (10S)-dG adducts. Conversely, SPD (proceeding from the 5′- to the 3′- end of the oligonucleotide) stalled before removing the unadducted residue 5′ to the trans (10S)dG adducts, and cleaved all the way up to the trans (10R)-dG adducts while being blocked by the next phosphodiester linkage. Thus, for both enzymes, the resistant internucleotide linkage was immediately 5′ to the (10S)dG adducts and immediately 3′ to the (10R)-dG adducts (Scheme 2). In the present study, the BcPh DE-dA adducts 1 with either 1R or 1S configuration, as well as (1S)-adducted 2, resulted in stalling of SPD to give fragments with one unmodified base 5′ to the adduct (Schemes 1 and 2). Since we saw little systematic variation in sensitivity to SPD hydrolysis between (R)- and (S)-adducts, the effect of adduct configuration on stall sites for this enzyme was not determined with the other oligonucleotides (Table 1). With VPD, more thorough investigation by MS of the partial hydrolysis products obtained clearly indicated that the stall site for this enzyme was not diastereomer dependent for dA (in contrast to dG) adducts: the most resistant linkage, immediately 5′ to the modified dA* (Schemes 1 and 2), was the same for the diastereomeric (R)- and (S)-dA adducts derived from both BaP and BcPh DE-2, as well as for adducts resulting from both cis and trans epoxide ring opening. Thus, in contrast to the dG adducts, both the (R)- and (S)-dA adducts appear to be “seen” by VPD and SPD as qualitatively blocking access to the 5′-internucleotide linkage, regardless of configuration. In contrast to dG, our present observations provide no clear evidence for different preferred orientations of the

Exonuclease Digestion of Diol Epoxide Adducts

Chem. Res. Toxicol., Vol. 14, No. 9, 2001 1337 Scheme 2

diastereomeric (R)- and (S)-adducts at dA relative to the direction of the oligonucleotide chain, and indeed suggest that when bound at the enzymes’ active sites, both (R)and (S)-dA adducts may orient in approximately the same direction (i.e., toward the 5′-end of the single-stranded oligonucleotide substrate). CD spectra (30) and computational studies (31, 32) suggest that nearly mirror-image orientations of the hydrocarbon relative to the base are preferred in the monomeric (R)- vs (S)-dA (31) as well as (R)- vs (S)-dG (32) adducts. The orientation of the base relative to the sugar, however, can be quite flexible, so that the hydrocarbon moiety of either dA adduct diastereomer could be oriented in either direction relative to the sugar with relatively little difference in energy, simply by rotating the base between the syn and anti conformations around the glycosidic bond, while maintaining the same preferred orientation of the hydrocarbon relative to the base. Single-stranded oligonucleotides lack the steric and hydrogen-bonding constraints imposed by the duplex structure of DNA, and thus the glycosidic torsion angles in the present single-stranded dA-adducted oligonucleotides may also be quite flexible, such that the position of the hydrocarbon could be primarily determined by intermolecular contacts with the enzymes in addition to, or instead of, intramolecular interactions.

Supporting Information Available: Synthetic procedures for BcPh dA phosphoramidites and CD spectra of the 1R/ 1S pairs of diastereomeric adducts in which the sugar hydroxyl groups are protected as their TBDMS derivatives, as well as CD spectra (for the cis BcPh DE-1 adduct derivatives) that show that hydroxyl group protection on either the hydrocarbon or the sugar has little effect on conformation are presented. Masses determined for selected partial hydrolysis products of 1, 2, 4, 5, and 6 are given. This material is available free of charge via the Internet at http://pubs.acs.org.

References (1) Jerina, D. M., Yagi, H., Thakker, D. R., Sayer, J. M., van Bladeren, P. J., Lehr, R. E., Whalen, D. L., Levin, W., Chang, R. L., Wood, A. W., and Conney, A. H. (1984) Identification of the ultimate carcinogenic metabolites of the polycyclic aromatic hydrocarbons: Bay region (R,S)-diol-(S,R)-epoxides. In Foreign Compound Metabolism (Caldwell, J., and Paulson, G. D., Eds.) pp 257-266, Taylor and Francis Ltd., London. (2) 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. (3) Jerina, D. M., Chadha, A., Cheh, A. M., Schurdak, M. E., Wood, A. W., and Sayer, J. M. (1991) Covalent bonding of bay-region diol epoxides to nucleic acids. In Biological Reactive Intermediates IV. Molecular and Cellular Effects and Their Impact on Human Health (Adv. Exp. Med. Biol. 283) (Witmer, C. M., Snyder, R., Jollow, D. J., Kalf, G. F., Kocsis, J. J., and Sipes, I. G., Eds.) pp 533-553, Plenum Press, New York.

1338

Chem. Res. Toxicol., Vol. 14, No. 9, 2001

(4) Szeliga, J., and Dipple, A. (1998) DNA adduct formation by polycyclic aromatic hydrocarbon dihydrodiol epoxides. Chem. Res. Toxicol. 11, 1-11. (5) Cheh, A. M., Yagi, H., and Jerina, D. M. (1990) Stereoselective release of polycyclic aromatic hydrocarbon-deoxyadenosine adducts from DNA by the 32P postlabeling and deoxyribonuclease I/snake venom phosphodiesterase digestion methods. Chem. Res. Toxicol. 3, 545-550. (6) Ponte´n, I., Sayer, J. M., Pilcher, A. S., Yagi, H., Kumar, S., Jerina, D. M., and Dipple, A. (1999) Sequence context effects on mutational properties of cis-opened benzo[c]phenanthrene diol epoxidedeoxyadenosine adducts in site-specific mutation studies. Biochemistry 38, 1144-1152. (7) Chary, P., and Lloyd, R. S. (1996) Impact of the stereochemistry of benzo[a]pyrene 7,8 dihydrodiol 9,10-epoxide-deoxyadenosine adducts on resistance to digestion by phosphodiesterases I and II and translesion synthesis with HIV-1 reverse transcriptase. Chem. Res. Toxicol. 9, 409-417. (8) Laryea, A., Cosman, M., Lin, J.-M., Liu, T., Agarwal, R., Smirnov, S., Amin, S., Harvey, R. G., Dipple, A., and Geacintov, N. E. (1995) Direct synthesis and characterization of site-specific adenosyl adducts derived from binding of a 3,4-dihydroxy-1,2-epoxybenzo[c]phenanthrene stereoisomer to an 11-mer oligodeoxyribonucleotide. Chem. Res. Toxicol. 8, 444-454. (9) Mao, B., Li, B., Amin, S., Cosman, M., and Geacintov, N. E. (1993) Opposite stereoselective resistance to digestion by phosphodiesterases I and II of benzo[a]pyrene diol epoxide-modified oligonucleotide adducts. Biochemistry 32, 11785-11793. (10) Razzell, W. E. (1963) Phosphodiesterases. In Methods in Enzymology, (Colowick, S. P., and Kaplan, N. O., Eds.) Vol. VI, pp 236258, Academic Press, New York. (11) Dolapchiev, L. B., and Bakalova, A. T. (1988) New spectrophotometric method for continuous recording of the spleen exonuclease activity. J. Biochem. Biophys. Methods 17, 207-214. (12) Lakshman, M. K., Sayer, J. M., Yagi, H., and Jerina, D. M. (1992) Synthesis and duplex-forming properties of a nonanucleotide containing an N6-deoxyadenosine adduct of a bay-region diol epoxide. J. Org. Chem. 57, 4585-4590. (13) Pilcher, A. S., Yagi, H., and Jerina, D. M. (1998) A novel synthetic method for cis-opened benzo[a]pyrene 7,8-diol 9,10-epoxide adducts at the exocyclic N6-amino group of deoxyadenosine. J. Am. Chem. Soc. 120, 3520-3521. (14) Agarwal, S. K., Sayer, J. M., Yeh, H. J. C., Pannell, L. K., Hilton, B. D., Pigott, M. A., Dipple, A., Yagi, H., and Jerina, D. M. (1987) Chemical characterization of DNA adducts derived from the configurationally isomeric benzo[c]phenanthrene-3,4-diol 1,2epoxides. J. Am. Chem. Soc. 109, 2497-2504. (15) Sayer, J. M., Yagi, H., Croisy-Delcey, M., and Jerina, D. M. (1981) Novel bay-region diol epoxides from benzo[c]phenanthrene. J. Am. Chem. Soc. l03, 4970-4972. (16) Ponte´n, I., Sayer, J. M., Pilcher, A. S., Yagi, H., Kumar, S., Jerina, D. M., and Dipple, A. (2000) Factors determining mutagenic potential for individual cis and trans opened benzo[c]phenanthrene diol epoxide-deoxyadenosine adducts. Biochemistry 39, 4136-4144. (17) Christner, D. F., Lakshman, M. K., Sayer, J. M., Jerina, D. M., and Dipple, A. (1994) Primer extension by various polymerases using oligonucleotide templates containing stereoisomeric benzo[a]pyrene-deoxyadenosine adducts. Biochemistry 33, 1429714305. (18) Page, J. E., Pilcher, A. S., Yagi, H., Sayer, J. M., Jerina, D. M., and Dipple, A. (1999) Mutational consequences of replication of M13mp7L2 constructs containing cis-opened benzo[a]pyrene 7,8diol 9,10-epoxide- deoxyadenosine adducts. Chem. Res. Toxicol. 12, 258-263. (19) Dipple, A., and Pigott, M. A. (1987) Resistance of 7,12-dimethylbenz[a]anthracene-deoxyadenosine adducts in DNA to hydrolysis by snake venom phosphodiesterase. Carcinogenesis 8, 491-493.

Ilankumaran et al. (20) Cosman, M., Fiala, R., Hingerty, B. E., Laryea, A., Lee, H., Harvey, R. G., Amin, S., Geacintov, N. E., Broyde, S., and Patel, D. J. (1993) Solution conformation of the (+)-trans-anti-[BPh]dA adduct opposite dT in a DNA duplex: intercalation of the covalently attached benzo[c]phenanthrene to the 5′-side of the adduct site without disruption of the modified base pair. Biochemistry 32, 12488-12497. (21) Cosman, M., Laryea, A., Fiala, R., Hingerty, B. E., Amin, S., Geacintov, N. E., Broyde, S., and Patel, D. J. (1995) Solution conformation of the (-)-trans-anti- benzo[c]phenanthrene-dA ([BPh]dA) adduct opposite dT in a DNA duplex: Intercalation of the covalently attached benzo[c]phenanthrenyl ring to the 3′-side of the adduct site and comparison with the (+)-trans-anti-[BPh]dA opposite dT stereoisomer. Biochemistry 34, 1295-1307. (22) Zegar, I. S., Kim, S. J., Johansen, T. N., Horton, P. J., Harris, C. M., Harris, T. M., and Stone, M. P. (1996) Adduction of the human N-ras codon 61 sequence with (-)-(7S,8R,9R,10S)-7,8-dihydroxy9,10-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene: Structural refinement of the intercalated SRSR(61, 2) (-)-(7S,8R,9S,10R)-N6-[10(7,8,9,10-tetrahydrobenzo[a]pyrenyl)]-2′-deoxyadenosyl adduct from 1H NMR. Biochemistry 35, 6212-6224. (23) Mao, B., Gu, Z., Gorin, A., Chen, J., Hingerty, B. E., Amin, S., Broyde, S., Geacintov, N. E., and Patel, D. J. (1999) Solution structure of the (+)-cis-anti-benzo[a]pyrene-dA ([BP]dA) adduct opposite dT in a DNA duplex. Biochemistry 38, 10831-10842. (24) Schurter, E. J., Sayer, J. M., Oh-hara, T., Yeh, H. J. C., Yagi, H., Luxon, B. A., Jerina, D. M., and Gorenstein, D. G. (1995) Nuclear magnetic resonance solution structure of an undecanucleotide duplex with a complementary thymidine base opposite a 10R adduct derived from trans addition of a deoxyadenosine N6-amino group to (-)-(7R,8S,9R,10S)-7,8-dihydroxy-9,10-epoxy-7,8,9,10tetrahydrobenzo[a]pyrene. Biochemistry 34, 9009-9020. (25) Pradhan, P., Tirumala, S., Liu, X., Sayer, J. M., Jerina, D. M., and Yeh, H. J. C. (2001) Solution structure of a trans-opened (10S)-dA adduct of (+)-(7S,8R,9S,10R)-7,8-dihydroxy-9,10-epoxy7,8,9,10-tetrahydrobenzo[a]pyrene in a fully complementary DNA duplex: Evidence for a major syn conformation. Biochemistry 40, 5870-5881. (26) Setlow, R. B., Carrier, W. L., and Bollum, F. J. (1964) Nuclease resistant sequences in ultraviolet irradiated deoxyribonucleic acid. Biochim. Biophys. Acta 91, 446-461. (27) Bourdat, A.-G., Gasparutto, D., and Cadet, J. (1999) Synthesis and enzymatic processing of oligodeoxynucleotides containing tandem base damage. Nucleic Acids Res. 27, 1015-1024. (28) Cheng, S. C., Hilton, B. D., Roman, J. M., and Dipple, A. (1989) DNA adducts from carcinogenic and noncarcinogenic enantiomers of benzo[a]pyrene dihydrodiol epoxide. Chem. Res. Toxicol. 2, 334-340. (29) Dipple, A., Pigott, M. A., Agarwal, S. K., Yagi, H., Sayer, J. M., and Jerina, D. M. (1987) Optically active benzo[c]phenanthrene diol epoxides bind extensively to adenine in DNA. Nature 327, 535-536. (30) Sayer, J. M., Chadha, A., Agarwal, S. K., Yeh, H. J. C., Yagi, H., and Jerina, D. M. (1991) Covalent nucleoside adducts of benzo[a]pyrene 7,8-diol 9,10-epoxides: Structural reinvestigation and characterization of a novel adenosine adduct on the ribose moiety. J. Org. Chem. 56, 20-29. (31) Tan, J., Geacintov, N. E., and Broyde, S. (2000) Principles governing conformations in stereoisomeric adducts of bay region benzo[a]pyrene diol epoxides to adenine in DNA: Steric and hydrophobic effects are dominant. J. Am. Chem. Soc. 122, 30213032. (32) Xie, X.-M., Geacintov, N. E., and Broyde, S. (1999) Stereocemical origin of opposite orientations in DNA adducts derived from enantiomeric anti-benzo[a]pyrene diol epoxides with different tumorigenic potentials. Biochemistry 38, 2956-2968.

TX010092L