TY - JOUR
AU - Broyde, Suse
AB - Abstract A key issue in the nucleotide excision repair (NER) of bulky carcinogen–DNA adducts is the ability of the NER machinery to recognize and repair certain adducts while failing to repair others. Unrepaired adducts can survive to cause mutations that initiate the carcinogenic process. Benzo[c]phenanthrene (B[c]Ph), a representative fjord region polycyclic aromatic hydrocarbon, can be metabolically activated to the enantiomeric benzo[c]phenanthrene diol epoxides (B[c]PhDEs), (+)‐(1S,2R,3R,4S)‐3,4‐ dihydroxy‐1,2‐epoxy‐1,2,3,4‐tetrahydrobenzo[c]phe nanthrene and the corresponding (–)‐(1R,2S,3S,4R) isomer. These react predominantly with adenine residues in DNA to produce the stereoisomeric 1R (+)‐ and 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA adducts. Duplexes containing the 1R (+) or 1S (–) B[c]Ph‐dA adduct in codon 61 of the human N‐ras mutational hotspot sequence CA*A, with B[c]Ph modification at A*, are not repaired by the human NER system. However, the analogous stereoisomeric DNA adducts of the bay region benzo[a]pyrene diol epoxide (B[a]PDE), 10S (+)‐ and 10R (–)‐trans‐anti‐B[a]P‐N6‐dA, are repaired in the same base sequence. In order to elucidate structural and thermodynamic origins of this phenomenon, we have carried out a 2 ns molecular dynamics simulation for the 1R (+)‐ and 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA adducts in an 11mer duplex containing the human N‐ras codon 61 sequence, and compared these results with our previous study of the B[a]P‐dA adducts in the same sequence. The molecular mechanics Poisson– Boltzmann surface area (MM‐PBSA) method was applied to calculate the free energies of the pair of stereoisomeric B[c]Ph‐dA adducts, and a detailed structural analysis was carried out. The different repair susceptibilities of the B[a]P‐dA adducts and the B[c]Ph‐dA adducts can be attributed to different degrees of distortion, stemming from combined effects of differences in the quality of Watson–Crick hydrogen bonding, unwinding, stretching and helix backbone perturbations. These differences are due to the different intrinsic topologies of the rigid, planar bay region adducts versus the twisted, sterically hindered fjord region adducts. Received March 29, 2002; Revised and Accepted May 22, 2002 INTRODUCTION Human nucleotide excision repair (NER) is the major mechanism for removing bulky DNA adducts (1–3) derived from reaction of DNA with metabolites of polycyclic aromatic hydrocarbons (PAHs) (4,5). Current understanding of the multi‐step process involved in mammalian NER is along the following lines (6–8): first, the damage recognition factor XPC‐hHR23B binds the lesion site and partially opens the DNA duplex around the adduct. This partial opening recruits the TFIIH complex, in concert with the entry of XPA and RPA, to form the fully opened pre‐incision complex. Upon verification of the damage, presumably by XPA, RPA and TFIIH, the XPG endonuclease incises the damaged strand on the 3′ side of the lesion, followed by ERCC1‐XPF endonuclease cleavage on the 5′ side. Finally, the 24–32 nt gap is patched by DNA polymerase δ or ϵ together with replication factor C and proliferating cell nuclear antigen. DNA ligase I then seals the nick. A key issue in the NER of a bulky carcinogen–DNA adduct is the ability of the NER machinery to recognize and repair certain adducts while failing to repair others. Unrepaired adducts have a greater chance to survive to the DNA replication cell cycle. Unfaithful translesion bypass could then give rise to mutations in cellular proto‐oncogenes or tumor suppressor genes and thereby initiate the carcinogenic process (9). For a lesion to become a substrate for NER, it is believed that both a distortion of the DNA helix and a modification to the DNA chemical structure are needed (1,6,10). DNA helix distortions, such as diminished Watson–Crick base pairing, DNA bending, and unwinding of the DNA strands, are thought to play an important role in the NER recognition process (7,10,11). These types of destabilizing distortions have also been described in terms of ‘thermodynamic probing’, in which the NER machinery tends to recognize lesions with helical instability (12). Fjord region PAHs, which have a twisted, sterically hindered aromatic region (13), are of great interest because of their significantly greater tumorigenic activities than those of the planar bay region derivatives (9,14,15). Benzo[c]phenanthrene (B[c]Ph), a representative fjord region PAH, can be metabolically activated to a pair of enantiomeric benzo[c]phenanthrene diol epoxides (B[c]PhDEs), (+)‐(1S,2R,3R,4S)‐3,4‐dihydroxy‐1,2‐epoxy‐1,2,3,4‐tetrahydrobenzo[c]phenanthrene, and its mirror image, the (–)‐(1R,2S,3S,4R) enantiomer (16) (Fig. 1). These intermediates react predominantly with adenine residues in DNA (17) to produce the stereoisomeric 1R (+)‐ and 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA adducts via trans epoxide opening (4). It has been found experimentally that both the 1R (+)‐ and 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA adducts in the human N‐ras mutational hotspot codon 61 sequence, one of the critical codons that has been found mutated in a variety of human cancers (18), are not repaired in human cell extracts (19). However, the 10S (+)‐ and 10R (–)‐trans‐anti‐B[a]P‐N6‐dA adducts, the analogous stereoisomeric DNA adducts of the bay region benzo[a]pyrene (B[a]P), are repaired in the same base sequence, with the 10S (+) B[a]P‐dA isomer more repair susceptible than the 10R (–) B[a]P‐dA isomer (19). This may account for the significantly higher tumorigenicity of the fjord region PAH diol epoxides as compared with their bay region counterparts. In order to elucidate the origins of the differences in repair susceptibilities of the B[c]Ph‐ and B[a]P‐dA adducts in terms of the thermodynamic and structural features of the adducts in double‐stranded DNA, we carried out an extensive thermodynamic analysis through investigating a large ensemble of structures generated from molecular dynamics (MD) simulations, and delineated in detail the structural features of the ensembles in relation to the elements the NER machinery is believed to recognize. We carried out a 2 ns MD simulation with explicit solvent and counterions using AMBER 5.0 for the 1R (+)‐ and 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA adducts in the human N‐ras codon 61 sequence context d(CGGACA*AGAAG)· d(CTTCTTGTCCG), with modification at A*. The molecular mechanics Poisson–Boltzmann surface area (MM‐PBSA) method (20–23) was applied to calculate relative free energies of the two stereoisomeric adducts. This methodology has recently been successfully employed to compute the structural and thermodynamic properties of biomacromolecules when benchmarked against an array of experimental data (20–25). We employed high resolution NMR solution structures in a different sequence context (26,27) as starting models for the B[c]Ph‐dA adducts. In these solution structures of the 1R (+)‐ and 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA adducts, the B[c]Ph moiety is intercalated into the DNA duplex without disruption of the modified base pairs. The 1R (+) B[c]Ph‐dA adduct is placed on the 5′ side of the modified adenine, while the 1S (–) B[c]Ph‐dA adduct is intercalated on the 3′ side. Our thermodynamic analyses of the dynamics trajectories of the 1R (+)‐ and 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA adducts reveal that they have similar stabilities in the human N‐ras codon 61 sequence context. Our structural analyses reveal minimal distortions to accommodate intercalation in both adducts, in terms of unwinding, stretching and distortions in the helix backbone, as well as disruption of Watson–Crick hydrogen bonding. In contrast, our previous structural and thermodynamic computations (28) showed that the corresponding 10R (–) and 10S (+) B[a]P‐dA adducts derived from the bay region benzo[a]pyrene diol epoxides (B[a]PDEs), which are also 5′‐ and 3′‐intercalated, respectively (29–33), are more distorted in terms of these structural features, with the S isomer more disturbed and also thermodynamically less stable than the R isomer. Thus, the structural distortions and thermodynamic stabilities parallel the susceptibility to NER. The differences between the B[c]Ph‐ and B[a]P‐dA adducts stem from the different intrinsic topologies of the bay and fjord region PAH diol epoxides. The bay region PAH aromatic region is planar, rigid and more extended. In contrast, the sterically hindered fjord PAH aromatic region is curved and twisted rather than planar (Fig. 1). In duplex DNA, this twist of the aromatic rings can adapt its direction to optimize stacking of the aromatic ring system at the intercalation site (26,27,34). These features of the B[c]Ph adducts allow for better fit in the intercalation pocket and hence less structural distortion. MATERIALS AND METHODS Starting structures and force field The high resolution NMR solution structures of the 1R (+)‐ and 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA adducts (A*) in an 11mer DNA duplex with sequence d(CTCTCA*CTTCC)· d(GGAAGTGAGAG) (26,27) were employed as the starting structures. We remodeled them to the human N‐ras codon 61 sequence context. The starting structure for the unmodified DNA duplex d(CTCTCACTTCC)·d(GGAAGTGAGAG) control in the human N‐ras codon 61 sequence context was an energy minimized B‐form DNA computed with the DUPLEX program (35) from a B‐DNA fiber diffraction model (36). Insight II 97.0 from Accelrys, Inc., a subsidiary of Pharmacopeia, Inc., was used for all modeling. Details regarding force field development for the adducts are given in Supplementary Material, Tables S1 and S2. Molecular dynamics and free energy calculations We carried out MD simulations using the AMBER 5.0 package (37) with Cornell et al. force field (38) and the parm98 parameter set (39) for the 1R (+) and 1S (–) B[c]Ph‐dA adducts and the unmodified control. The particle mesh Ewald method (40) was employed to calculate the long‐range electrostatic interactions, and a 12 Å cutoff was applied to the non‐bonded Lennard‐Jones interactions. We applied the SHAKE algorithm (41) to constrain all bonds involving hydrogen atoms with a tolerance of 0.0005 Å, and used a 1 fs time step in the MD simulations. We added 20 Na+ ions to the system for neutralization using the LEap module in AMBER 5.0 (37), and then solvated the whole system with a rectangular box of TIP3P water molecules (42) which extended ∼10 Å from the outmost DNA atoms in each direction. This yielded a periodic box size of about 52 × 50 × 65 Å for the 1R (+) and 1S (–) B[c]Ph‐dA adducts and the unmodified control. In all, 3943, 4111 and 3662 water molecules were added for the 1R (+) B[c]Ph‐dA adduct, the 1S (–) B[c]Ph‐dA adduct and the unmodified control, respectively. The minimization and equilibration protocols for all three systems were as follows: first, the water molecules and counterions were minimized for 1500 steps of steepest descent, followed by 50 ps MD with the DNA fully fixed to allow the solvent to relax. The whole system was then minimized for 1000 additional steps of steepest descent, followed by 3 ps MD with 25 kcal/mol restraints on the DNA, which further allowed the solvent to relax. Then the system was minimized for five rounds of 600 steps of steepest descent with restraints on the DNA reduced by 5 kcal/mol each round, from 20 to 0 kcal/mol. Finally, the whole system was heated from 10 to 300 K over 40 ps using the Berendsen coupling algorithm (43) with a coupling constant of 0.2 ps. Production simulation was then continued at atmospheric pressure with a 0.2 ps coupling parameter at 300 K for 2 ns. Free energy calculations using the MM‐PBSA methodology (20–23) were carried out with the exact protocol as described in detail in our earlier work (28). Briefly, the free energy (Gtot) was computed from the molecular mechanical energy (EMM), the solvation free energy (Gsolvation) and the solute entropic contributions to the free energy (ΔGtot = ΔEMM + ΔGsolvation – TΔS). The molecular mechanics energy (EMM) was calculated from the internal energy (Eint) stemming from deviations of the bonds (Ebonds), angles (Eangles) and dihedral angles (Edihedrals) from their equilibrium values, the van der Waals energy (EvdW) and the electrostatic energy (Eelectrostatic). The solvation free energy (Gsolvation) was approximated from the electrostatic solvation energy (GPB) calculated using the DelPhi program (44,45) and the nonpolar solvation energy (Gnonpolar); the latter was estimated as Gnonpolar = γSASA + b (γ = 0.00542 kcal/Å2, b = 0.92 kcal/mol), where SASA is the solvent‐accessible surface area (SASA) computed by Sanner’s algorithm implemented in the MSMS program (46). The solute entropic contributions to the free energies were approximated with normal mode calculations (21,28). Stability of the molecular dynamics simulation The root mean square deviations (RMSDs) of the 1R (+) B[c]Ph‐dA adduct, the 1S (–) B[c]Ph‐dA adduct and the unmodified control relative to the starting structures are shown in Figure S1. The mean RMSDs relative to the starting structures for the 1R (+) and 1S (–) B[c]Ph‐dA adducts are 2.9 ± 0.3 Å and 2.6 ± 0.3 Å over the first 500 ps, and 3.7 ± 0.5 Å and 3.3 ± 0.5 Å over the 500–2000 ps time frame, respectively. A small transition occurred at ∼500 ps to a state that then remained stably fluctuating for the remainder of the simulation. We then calculated an average structure for the 1R (+) B[c]Ph‐dA adduct, the 1S (–) B[c]Ph‐dA adduct and the unmodified control over the 500–2000 ps time frame, and computed the RMSDs relative to these average structures (Fig. S2). The mean value is 2.1 ± 0.2 Å for the 1R (+) B[c]Ph‐dA adduct, 1.5 ± 0.3 Å for the 1S (–) B[c]Ph‐dA adduct and 1.3 ± 0.2 Å for the unmodified control over the 500–2000 ps time frame, indicating reasonable stability. Distortion free energy calculation For each of the 150 frames retrieved at 10 ps intervals from the 1R (+) and 1S (–) B[c]Ph‐dA adduct MD trajectories, as well as from our previous trajectories for the 10S (+) and 10R (–) B[a]P‐dA adducts (28), we excised the modified B[c]Ph or B[a]P and replaced it with a hydrogen atom that is bonded with N6 of the adenine base. The bond length was adjusted to the value of the other N6–H bond. The chemical structure of this DNA duplex is thus the same as that of the unmodified DNA control duplex in our simulations. We then applied the MM‐PBSA method to compute the free energy for these distorted but now unmodified DNA duplexes, as well as for the normal unmodified control. We compared these data with the corresponding values for the unmodified duplex to estimate a distortion free energy (ΔGdistort) caused by the intercalation: where and and so on for the other terms. We then computed a relative distortion free energy: ΔΔGdistort = ΔGdistort – min(ΔGdistort) where min(ΔGdistort) is for the adduct with the lowest distortion free energy. Quality of Watson–Crick hydrogen bonding A hydrogen bond quality index (28,35), IH, was employed to quantitatively assess the deviation from ideal Watson–Crick hydrogen bonding distances and angles: where dDA is the instantaneous donor–acceptor distance, is an ideal donor–acceptor distance (47) [N6 (A) to O4 (T) is 2.95 Å, N1 (A) to N3 (T) is 2.82 Å] and γ is the instantaneous D–H···A bond angle with ideal value of 180°. The summation is over the two Watson–Crick hydrogen bonds in an A–T base pair. IH adopts a value of 0 when ideal Watson–Crick hydrogen bonding is maintained. The criteria for Watson– Crick hydrogen bond occupancy employed were <3.4 Å for donor to acceptor heavy atom distance and 140–180° for the D–H···A hydrogen bond angle. RESULTS Dynamics average structures Figure 2 shows stereo views of the central 5mer of the dynamics average structures for the 1R (+) B[c]Ph‐dA adduct, the 1S (–) B[c]Ph‐dA adduct and the unmodified control over the 500–2000 ps time frame. In the 1R (+) B[c]Ph‐dA adduct, the B[c]Ph is intercalated on the 5′ side of the modified adenine, between the C5–G18 and the A*6–T17 base pairs, with the fjord region twisting toward the 5′ direction. However, in the 1S (–) B[c]Ph‐dA adduct, the B[c]Ph residue is intercalated on the 3′ side of the modified adenine, between the A*6–T17 and A7–T16 base pairs, with the fjord region twisting to the 3′ direction. These dynamics simulation structures are consistent with the NMR solution structures of these adducts obtained in a different sequence context (26,27), which served as starting models for the simulations. The average value of the torsion angles α′, β′, γ′ and δ′ (Fig. 1) over the 500–2000 ps time frame for the 1R (+) and 1S (–) B[c]Ph‐dA adducts are in similar conformational domains as the corresponding data in the NMR solution structures (Table 1). We note that the δ′ angles of the two isomers, defining the direction of the fjord region twist, is opposite in sign; this reveals that the twist is opposite, which helps to optimize stacking on the 5′ side for the 1R (+) B[c]Ph‐dA adduct and on the 3′ side for the 1S (–) B[c]Ph‐dA adduct. The 1R (+) and 1S (–) B[c]Ph‐dA adducts exhibit negligible difference in free energy We calculated free energies for the 1R (+) and 1S (–) B[c]Ph‐dA isomeric adducts as detailed in Materials and Methods. As may be seen from Table 2, the enthalpic contributions to the free energy, comprised of the molecular mechanical energy (EMM) and the electrostatic solvation free energy (GPB) are very similar for the 1R (+) and 1S (–) B[c]Ph‐dA adducts, favoring the 1S (–) adduct by only 1.1 kcal/mol. However, the entropic contribution to the free energy favors the 1R (+) adduct by 1.3 kcal/mol. Therefore, the total free energy difference between the 1R (+) and 1S (–) B[c]Ph‐dA adducts is only 0.2 kcal/mol, which is negligible. Backbone and helicoidal parameter analysis shows similar modest distortions in the 1R (+) and 1S (–) B[c]Ph‐dA stereoisomers The backbone torsional parameters (defined in Fig. S3) for the 1R (+) and 1S (–) B[c]Ph‐dA adducts and the unmodified control duplex are shown in Figure S4. In both the 1R (+) and 1S (–) adducts, the backbone torsional parameters, α, β, γ and δ, are generally similar to those of the unmodified control duplex. However, distortion in the backbone parameters ϵ and ζ occurs at or near the intercalation pockets and beyond. The glycosidic torsion angles χ at C1 in the unmodified control and in the 1R (+) B[c]Ph‐dA adduct have a syn conformation instead of the normal anti conformation, due to end base‐pair fraying. The degree of distortion in the backbone torsional parameters, compared with the unmodified control, does not appear to differ much in the 1R (+) and 1S (–) B[c]Ph‐dA isomers. Helicoidal parameters (48) (defined in Fig. S5), calculated by the CURVES algorithm (49) in Dials and Windows (50), are shown in Table 1 and Figure S6. These give average values of the base pair and base pair step parameters at and near the lesion site for the 1R (+) and 1S (–) B[c]Ph‐dA adducts, and for the corresponding residues in the unmodified control. The differences in the helicoidal parameters relative to the unmodified control occur mainly at or near the intercalation pockets. The buckle and propeller of the 1R (+) and 1S (–) adducts at the lesion site A*6–T17 are in opposite directions, to permit the 5′ versus 3′ intercalation of the R and S stereoisomeric adducts, respectively. There is a large increase in rise, compared with the unmodified control for the C5–G18 to A*6–T17 base pair step in the 1R (+) adduct, and at the A*6–T17 to A7–T16 base pair step in the 1S (–) adduct. This accommodates the intercalated B[c]Ph, on the 5′ side in the 1R (+) adduct and on the 3′ side in the 1S (–) adduct. Significant deviations in tilt are also observed in or near the intercalation pockets in both adducts. The magnitude of propeller is significantly larger in the 1S (–) adduct than in the 1R (+) adduct. The origin of these differences is discussed below. The values of twist at and near the intercalation pocket in the 1R (+) and 1S (−) B[c]Ph‐dA adducts and in the unmodified control reveals that the total unwinding in the R and S isomers is about the same (∼18–19°) relative to the unmodified control (Table S3). Degree of B[c]Ph stacking in the intercalation pocket is similar in the 1R (+) and 1S (–) B[c]Ph‐dA adducts The overall stacking of the aromatic rings of B[c]Ph and the adjacent bases is quite similar in the 1R (+) and 1S (–) B[c]Ph‐dA adducts. Figure 2 shows the overlap of the aromatic moiety of B[c]Ph with the partner T17 of the modified residue A*6 and the adjacent base pairs C5–G18 for the 1R (+) B[c]Ph‐dA adduct and A7–T16 for the 1S (–) B[c]Ph‐dA adduct in the dynamics average structures. In the 1R (+) adduct, the aromatic rings partially overlap with all three adjacent bases, T17, C5 and G18; however, in the 1S (–) adduct they partially overlap with T17 and T16, but no obvious stacking with A7 is observed. This is due to the opposite directions of intercalation and fjord region twist, relative to the 5′ or 3′ direction, of the B[c]Ph in each stereoisomer. The van der Waals interaction energies also support the similar stacking in the 1R (+) and 1S (–) B[c]Ph‐dA adducts. We computed the van der Waals interaction energies between just the aromatic part of B[c]Ph and the adjacent bases T17, C5 and G18 for the 1R (+) B[c]Ph‐dA adduct, and T17, A7 and T16 for the 1S (–) B[c]Ph‐dA adduct (Table S4). In the 1R (+) adduct, B[c]Ph has similar van der Waals interactions with each of its adjacent bases: T17 (–8.8 kcal/mol), C5 (–9.1 kcal/mol) and G18 (–7.0 kcal/mol). In the 1S (–) adduct, the B[c]Ph has much weaker van der Waals interaction with A7 (–2.9 kcal/mol), consistent with our observation that there is no obvious stacking between B[c]Ph and A7 (Fig. 2). However, the van der Waals interactions of B[c]Ph with T17 (–10.6 kcal/mol) and B[c]Ph with T16 (–10.0 kcal/mol) in the 1S (–) adduct are slightly stronger than the van der Waals interactions of B[c]Ph with its adjacent bases in the 1R (+) adduct. Hence, the total van der Waals interaction energies in the intercalation pocket are quite similar for the two stereoisomers, with a value of –24.9 kcal/mol in the 1R (+) B[c]Ph‐dA adduct and –23.5 kcal/mol in the 1S (–) B[c]Ph‐dA adduct. Watson–Crick hydrogen bonding is maintained in both the 1R (+) and 1S (–) B[c]Ph‐dA adducts Hydrogen bond occupancy for the two Watson–Crick hydrogen bonds N3–H3···N1 and N6–H6···O4 is, respectively, 93.8 and 97.1% for the 1R (+) B[c]Ph‐dA adduct and 95.5 and 94.9% for the 1S (−) B[c]Ph‐dA adduct. The quality of Watson–Crick hydrogen bonding at the lesion site A*6 in the 1R (+) and 1S (–) adducts was analyzed using our hydrogen bond quality index IH (see Materials and Methods) for each of the dynamics trajectories over the 500–2000 ps time frame, a total of 3000 structures (Fig. S7). The summed value of IH for the 3000 structures is 206 for the unmodified control, 252 for the 1R (+) adduct and 398 for the 1S (–) adduct. The lower the value of IH, the better the hydrogen bond quality, so deviation from ideal Watson–Crick hydrogen bonding is exhibited in both the 1R (+) and 1S (–) B[c]Ph‐dA adducts. However, the Watson–Crick hydrogen bonds are still maintained for both isomers throughout the simulation, although perturbed. Interestingly, our hydrogen bond quality index, IH, shows that the S isomer is somewhat more distorted than the R isomer, since IH is higher. Also, propeller, buckle and opening are higher (Table 1). This difference is caused by the shifting of the glycosidic torsional angle χ toward the high‐anti domain (47) only in the 1S (–) adduct. The distribution of the glycosidic torsion angles χ at the lesion site A*6 for the 1R (+) B[c]Ph‐dA adduct, the 1S (–) B[c]Ph‐dA adduct and the unmodified control is displayed in Figure S8, which shows the number of structures from the 500–2000 ps dynamic trajectories as a function of χ. For the 1R (+) adduct and the unmodified control, χ is found only in the normal anti domain; however, for the 1S (–) adduct, χ is shifted toward the high‐anti domain (47). In the 1R (+) adduct, the 1S (–) adduct and the unmodified control, the χ values are –175° to –50° (–103° ± 17°), –108° to –15° (–61° ± 14°) and –171° to –51° (–100° ± 18°), respectively. By comparison, the anti domain of B‐DNA has χ values in the range of ∼(–120° to –100°) (51). The right‐handed helical twist of B‐DNA introduces different steric effects stemming from the OH‐containing benzylic ring in R and S adducts (29,33). The R adduct is inserted on the 5′ side of the modified base (A*6), and has the benzylic ring positioned toward the 3′ side of A*6, which is not crowded. However, in the S adduct, the B[c]Ph is inserted on the 3′ side of the modified base (A*6), and the benzylic ring is positioned toward the 5′ side of A*6, in a sterically crowded location (Fig. 3). The shifting of χ helps to alleviate this steric crowding caused by the benzylic ring in the S isomer (28,33), and accounts for the greater deviation from ideal Watson– Crick base pairing in this isomer. Distortion free energies are similar in 1R (+) and 1S (–) B[c]Ph‐dA adducts and are greater in the bay region B[a]P‐dA adducts Using the MD trajectories generated in our previous study for the B[a]P‐dA adducts (28) and the present work, we calculated distortion free energies for the 1R (+) and 1S (–) B[c]Ph‐dA, and also for the 10S (+) and 10R (–) B[a]P‐dA adducts, as described in Materials and Methods (Table S5). These distortion free energies evaluate the cost of distorting an unmodified B‐DNA duplex to the conformation of the adduct. We found that the 1S (–) B[c]Ph‐dA adduct has the lowest distortion free energy (ΔΔGdistort = 0), and it is similar in the 1R (+) B[c]Ph‐dA adduct (ΔΔGdistort = 0.2 kcal/mol); however, the 10S (+) B[a]P‐dA adduct distorts the normal B‐DNA most severely with ΔΔGdistort of 8.3 kcal/mol, while the 10R (–) B[a]P‐dA adduct only moderately distorts the B‐DNA (ΔΔGdistort = 3.8 kcal/mol). These thermodynamic stability data indicate that both the 10S (+) and 10R (–) B[a]P‐dA adducts are more destabilized than the B[c]Ph‐dA adducts compared with the unmodified duplex. Similar enhanced van der Waals interaction energies in the intercalation pocket in 1R (+) and 1S (–) B[c]Ph‐dA adducts compared with unmodified B‐DNA We calculated the van der Waals interaction energy between the entire B[c]Ph modified adenine residue (A*6) and the adjacent bases in the intercalation pocket for both 1R (+) and 1S (–) B[c]Ph‐dA isomers and compared these values with the van der Waals interaction energies between these bases in the unmodified B‐DNA duplex. The results show that intercalation of the B[c]Ph enhances the van der Waals interaction energy in the intercalation pocket by nearly the same amount, 10.5 and 9.9 kcal/mol, for the 1R (+) and 1S (–) isomers respectively, over the unmodified B‐DNA control (Table S6). DISCUSSION Origin of similar stabilities in 1R (+) and 1S (–) isomeric B[c]Ph‐dA adducts Our computations show that the 1R (+) and 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA adducts have similar free energies, enthalpies and entropies (Table 2). This is consistent with experimental thermal melting studies, which have shown that duplexes containing the 1R (+) or 1S (–) B[c]Ph‐dA adduct on A* in a 13mer duplex containing the 11mer of the present study are thermally as stable as the unmodified DNA duplex, since all three duplexes exhibit essentially the same melting points (Tm): 54.7 ± 0.5°C for the unmodified duplex, 55.2 ± 0.5°C for the 1R (+) B[c]Ph‐dA adduct and 56.5 ± 0.5°C for the 1S (–) B[c]Ph‐dA adducts. Similar thermal melting temperatures for this pair of adducts have also been reported in the case of another sequence context (52). The experimental results indicate that the secondary structure of the DNA is not destabilized even when the modified bases are present. It should be noted, however, that the free energy computations cannot compare DNA adducts with unmodified DNA since they have different numbers of atoms. Hence, we cannot compare experimental Tm values of modified and unmodified DNA with computed thermodynamic parameters. The following structural features can account for the similar stabilities of the 1R (+) and 1S (–) B[c]Ph‐dA adducts: the degree of helical backbone distortion in the two adducts is similar (Figs S4 and S6), the van der Waals interactions between the B[c]Ph aromatic moiety with the adjacent base pairs in the intercalation pocket are similar (Fig. 2 and Table S4) and the helix unwinding and stretching, induced by the intercalation, is about the same in the two isomers (Table 1). The components analyses of the distortion free energies (Table S5) reveal that unfavorable van der Waals energy changes () are in part compensated by the favorable electrostatic energy changes ( ); therefore, the differences in the distortion free energies are mainly caused by the internal distortion energy which is derived form helix perturbations, such as unwinding and stretching, due to the intercalation. Furthermore, both the intercalation‐induced distortions and the compensating van der Waals interactions are quite similar in the case of these two isomeric B[c]Ph‐dA adducts (Tables S4 and S5). In addition, the ∼95% occupancies of the Watson–Crick base pair hydrogen bonds show that, though non‐ideal, the modified adenines remain paired with their partner thymines during the simulations. Origin of different stabilities in fjord region B[c]Ph‐dA and bay region B[a]P‐dA adducts Our previous computations showed a significant free energy difference between the 10R (–)‐ and 10S (+)‐trans‐anti‐B[a]P‐N6‐dA adducts [ΔGtot = 13 kcal/mol, of which ∼10 kcal/mol is enthalpic contribution, favoring the 10R (–) B[a]P‐dA adduct] (28). This is in line with differences in thermal melting temperatures for this pair of adducts. Specifically, the thermal melting temperature (Tm) for the sequence studied in our computations, an 11mer DNA duplex d(CGGACA*AGAAG)· d(CTTCTTGTCCG), containing a 10R (–)‐trans‐anti‐B[a]P‐N6‐dA adduct is 33°C, while that of the DNA duplex containing a 10S (+)‐trans‐anti‐B[a]P‐N6‐dA adduct is 25°C (53). Both are much lower than that of the unmodified DNA duplex control, which is 46°C (53). These experimental thermal stability data indicate that both the 10R (–)‐ and 10S (+)‐trans‐anti‐B[a]P‐N6‐dA adducts are destabilized relative to the unmodified duplex. Moreover, the 10S (+) B[a]P‐dA isomer is more destabilized than the 10R (−) B[a]P‐dA isomer. However, the 1R (+)‐ and 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA adducts, as discussed above, have the same stabilities. Calculated distortion free energies of the B[a]P‐dA adducts are also much higher than those of the B[c]Ph‐dA adducts (Table S5). Taken together, these findings indicate that the bay region B[a]P modified DNA duplexes are structurally more distorted than their fjord region counterparts. However, as discussed in our earlier work, comparisons of experimental Tm values with computed enthalpies in R and S isomers of the adducts is only a rough approximation, with the assumption that the two modified isomeric single‐stranded coils would have the same enthalpies (28,54). Origins of differences in B[c]Ph‐ and B[a]P‐dA adducts are attributable to a combination of factors Greater backbone torsional distortions of the B[a]P‐dA adducts. The backbone torsional parameters in the bay region B[a]P‐dA adducts are more distorted, especially in the intercalation pockets (28), and this is also manifested by the higher distortion free energies for these adducts (Table S5). Greater rise values at the intercalation pockets of the B[a]P‐dA adducts compared with the B[c]Ph‐dA adducts. The bay region B[a]P‐dA adducts cause greater stretch than the fjord region B[c]Ph‐dA adducts; this arises from the necessity of accommodating this flat and extended aromatic ring system within the double helix (Table 3) (28). Greater extent of unwinding of the B[a]P‐dA adducts. Intercalation of an aromatic ring system into the DNA double helix requires stretching and unwinding (47). Because of the intercalation of the B[a]P residue, the 10R (−) B[a]P‐dA isomer causes an unwinding of ∼29° at the intercalation pocket. However, the 10S (+) B[a]P‐dA isomer is unwound by a total of ∼41°. In the latter case, an unwinding of ∼26° at the intercalation pocket is directly due to the insertion of the aromatic residue. However, an additional ∼15° of unwinding occurs at an adjacent site, to alleviate steric crowding between the S isomer benzylic ring and the right‐handed twisted B‐DNA helix (28,29). Interestingly, unwinding to alleviate steric crowding was not needed in the S isomer of the B[c]Ph‐dA adduct (Table S3), as explained below. Less optimal stacking interactions between the B[a]P residue and the neighboring base pairs. There is better overlap between the polycyclic aromatic portion of the B[c]Ph residue and neighboring bases than there is in the case of B[a]P adducts since the former are distorted out of the plane to optimize stacking in each isomer (26,27) (see Fig. 3 and figure 3 in ref. 28). More disturbed Watson–Crick hydrogen bonding in analogous B[a]P‐dA adducts, especially in the case of the 10S (+) isomer with its sterically crowded benzylic ring (28,29,33). This is reflected by the hydrogen bond quality index IH at the lesion site, which is higher when the quality of hydrogen bonding is worse (Table 3). The topological differences between the fjord and bay region adducts can account for the greater destabilization of the bay region adducts. The more distal aromatic ring system is oriented closer to the linkage site in the fjord than in the bay region adducts (Scheme S1). The B[c]Ph residue experiences steric hindrance between the H1 and H12 atoms (Fig. 1), causing the aromatic ring system to twist to alleviate it. This twist adjusts its direction to optimize stacking with neighboring base pairs; it is in the 5′ direction for the 1R (+) B[c]Ph‐dA adduct and in the 3′ direction for the 1S (–) B[c]Ph‐dA adduct (Fig. 2) (26,27,34). On the other hand, the aromatic ring system of the B[a]P is planar and rigid. Moreover, B[a]P has a larger aromatic moiety, with one more aromatic ring than B[c]Ph, and the arrangement of the aromatic ring system is extended in the B[a]P and curved in the B[c]Ph (Scheme S1). These topological differences account for the differences in rise, in degree of carcinogen–base stacking, in unwinding and in quality of Watson–Crick hydrogen bonding. The greater distortion due to the rise difference and consequent less favorable stacking in B[a]P‐dA adducts results from the rigid nature of the B[a]P residue compared with the adaptable, twisted B[c]Ph residue. When inserted into the DNA duplexes, the planar B[a]P ring system is not parallel to the adjacent base pairs, because its orientation is completely governed by the R or S linkage site, with no adaptable twist. Furthermore, the topology compels this ring system to orient itself away from the C10 linkage site (Scheme S1 and Fig. 3) (29–31,33). Thus, the DNA duplex has to stretch more in order to accommodate the rigid B[a]P ring system. The more unwinding of the B[a]P‐dA adducts stems from the extended nature of the B[a]P ring system. In the bay region B[a]P residue, the aromatic ring system points away from the linkage site (solid black dots in Scheme S1), while in the fjord region B[c]Ph residue, the twisted aromatic ring is oriented towards the linkage site. Therefore, the flat B[a]P aromatic moiety is more extended than the B[c]Ph moiety and causes more unwinding of the DNA in order to accommodate itself within the duplex (29–31,33). Conversely, the B[c]Ph aromatic ring system has a more compact shape and fits better into a less distorted intercalation pocket, and the adaptability in direction of its twist is employed to achieve optimal stacking with neighboring base pairs (26,27,34). The worse quality of Watson–Crick hydrogen bonding at the lesion site in the 10S (+) B[a]P‐dA isomeric adduct is due to the more severe steric crowding between the benzylic ring and the C5 residue adduct than in the B[c]Ph adenine adduct (Fig. 3) (28). In the S isomer of the B[c]Ph‐dA adduct, this steric crowding is alleviated in part by a movement of the benzylic ring slightly toward the 3′ side of the modified base, and rotation of the glycosidic torsional angle χ toward the high‐anti region. However, in the S isomer of the B[a]P‐dA adduct, the more severe steric crowding causes a syn/anti equilibrium, which produces diminished quality of hydrogen bonding. Different susceptibilities to nucleotide excision repair of the bay region B[a]P‐dA and fjord region B[c]Ph‐dA adducts: structural and thermodynamic factors In mammals NER is the only repair mechanism to remove PAH‐modified DNA adducts (1). Both modifications to the DNA chemistry and distortions to DNA helical structure are necessary for NER (1,6,10). A ‘thermodynamic probing’ mechanism is suggested to be involved in the NER damage recognition process (12), which probes the DNA helix for destabilization, stemming from such factors as diminished Watson–Crick base pairing, DNA bending, and unwinding of the DNA strands (6,7,10,11). The fjord region B[c]Ph‐dA adducts and the bay region B[a]P‐dA adducts, located in the identical sequence context, the human N‐ras mutational hotspot codon 61 sequence, and with exactly the same stereochemical properties, show very different repair susceptibilities in the cell‐free human NER system: both the 10R (–)‐ and 10S (+)‐trans‐anti‐B[a]P‐N6‐dA adducts are excised, while neither the 1R (+)‐ nor the 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA adducts are removed (19). Specifically, the relative NER excision activity for the 1R (+) and 1S (–) B[c]Ph‐dA and 10S (+) and 10R (–) B[a]P‐dA adducts are about 1, 1, 18 and 8%, respectively (relative to 100% for the 2‐acetylaminofluorene adduct, e.g. [AAF]‐C8‐dG adduct) (19). We have found minimal distortions and thermodynamic destabilization needed to accommodate intercalation in the case of the B[c]Ph‐dA adducts, and we suggest these are insufficient for recognition and excision of these lesions by the human NER apparatus. In contrast, we suggest that the significantly greater thermodynamic destabilization and structural distortion induced by the B[a]P‐dA adducts are recognized by the DNA repair enzymes, leading to the excision of these bulky bay region adducts. Our computed distortion free energies correlate with the experimentally determined human NER susceptibilities for these adducts (Fig. 4), and support the hypothesis that DNA helix destabilization facilitates human NER (6,7,10–12). It remains to be determined whether such a correlation might provide a way to estimate the NER excision activity for the intercalated family of DNA adducts using MD and free energy calculations, through studies of other such systems. CONCLUSIONS The limited extent of destabilization and distortion of the B[c]Ph‐modified DNA duplexes provides a possible explanation for the observed lack of excision of these fjord region adducts by NER enzymes. Comparisons between the fjord region 1R (+)‐ and 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA adducts and the bay region 10R (–)‐ and 10S (+)‐trans‐anti‐B[a]P‐N6‐dA adducts suggests a plausible structural basis for the differences in stabilities and susceptibilities to NER of these pairs of fjord and bay region PAH‐N6‐dA adducts. In turn, the resistance to repair of the fjord PAH‐N6‐dA adducts may contribute to the high tumorigenic potency of the B[c]PhDE metabolites. SUPPLEMENTARY MATERIAL Supplementary Material is available at NAR Online. ACKNOWLEDGEMENTS This research is supported by NIH Grant CA‐28038 to S.B., NIH Grants CA‐76660 and CA‐20851 to N.E.G. and NIH Grant CA‐46533 to D.J.P. Computations were carried out at the National Science Foundation (NSF) San Diego Supercomputer Center, the NSF Advanced Computing Center for Engineering and Science at the University of Texas at Austin, the Department of Energy (DOE) National Energy Research Super Computing Center and on our own SGI workstations. View largeDownload slide Figure 1. Structures of (A) B[c]Ph, (+)‐ and (–)‐anti‐B[c]PhDE, and 1R (+)‐ and 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA adducts. Torsion angles α′, β′, γ′ and δ′ are defined as follows: α′, N1–C6–N6–C1(B[c]Ph); β′, C6–N6–C1(B[c]Ph)–C2(B[c]Ph); γ′, C1–C2–C3–C4 (defines benzylic ring conformation); δ′, C12–C8B–C6B–C4B (defines fjord region twist). (B) 10R (–)‐ and 10S (+)‐trans‐anti‐B[a]P‐N6‐dA adducts. Torsion angles α′, β′ are defined as follows: α′, N1–C6–N6–C10(B[a]P); β′, C6–N6– C10(B[a]P)–C9(B[a]P). (C) Human N‐ras codon 61 sequence; A* is the modified adenine. View largeDownload slide Figure 1. Structures of (A) B[c]Ph, (+)‐ and (–)‐anti‐B[c]PhDE, and 1R (+)‐ and 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA adducts. Torsion angles α′, β′, γ′ and δ′ are defined as follows: α′, N1–C6–N6–C1(B[c]Ph); β′, C6–N6–C1(B[c]Ph)–C2(B[c]Ph); γ′, C1–C2–C3–C4 (defines benzylic ring conformation); δ′, C12–C8B–C6B–C4B (defines fjord region twist). (B) 10R (–)‐ and 10S (+)‐trans‐anti‐B[a]P‐N6‐dA adducts. Torsion angles α′, β′ are defined as follows: α′, N1–C6–N6–C10(B[a]P); β′, C6–N6– C10(B[a]P)–C9(B[a]P). (C) Human N‐ras codon 61 sequence; A* is the modified adenine. View largeDownload slide Figure 2. Stereo views of the MD average structures over 500–2000 ps. (A) Central 5mer of the 1R (+) B[c]Ph‐dA adduct with B[c]Ph in red, A*6–T17 in green and C5–G18 in blue. (B) Intercalation pocket of the 1R (+) B[c]Ph‐dA adduct [d(C5·A*6)·d(T17·G18)]. B[c]Ph is in red stick, A*6–T17 is in green stick and C5–G18 is in blue ball‐and‐stick. (C) Central 5mer of the 1S (–) B[c]Ph‐dA adduct with B[c]Ph in red, A*6–T17 in green and A7–T16 in blue. (D) Intercalation pocket of the 1S (–) B[c]Ph‐dA adduct [d(A*6·A7)·d(T16·T17)]. B[c]Ph is in red stick, A*6–T17 is in green stick and A7–T16 is in blue ball‐and‐stick. (E) Central 5mer of the unmodified control duplex with C5–G18 and A7–T16 in blue and A6–T17 in green. The backbone, sugar atoms and other residues are in gray. Structures in (A), (C) and (E) are aligned to have 5′‐OH at the top right. The view in (B) and (D) is along the helix axis from 5′ (top) to 3′ (bottom) of the modified strand. All stereo images are constructed for viewing with a stereoviewer. View largeDownload slide Figure 2. Stereo views of the MD average structures over 500–2000 ps. (A) Central 5mer of the 1R (+) B[c]Ph‐dA adduct with B[c]Ph in red, A*6–T17 in green and C5–G18 in blue. (B) Intercalation pocket of the 1R (+) B[c]Ph‐dA adduct [d(C5·A*6)·d(T17·G18)]. B[c]Ph is in red stick, A*6–T17 is in green stick and C5–G18 is in blue ball‐and‐stick. (C) Central 5mer of the 1S (–) B[c]Ph‐dA adduct with B[c]Ph in red, A*6–T17 in green and A7–T16 in blue. (D) Intercalation pocket of the 1S (–) B[c]Ph‐dA adduct [d(A*6·A7)·d(T16·T17)]. B[c]Ph is in red stick, A*6–T17 is in green stick and A7–T16 is in blue ball‐and‐stick. (E) Central 5mer of the unmodified control duplex with C5–G18 and A7–T16 in blue and A6–T17 in green. The backbone, sugar atoms and other residues are in gray. Structures in (A), (C) and (E) are aligned to have 5′‐OH at the top right. The view in (B) and (D) is along the helix axis from 5′ (top) to 3′ (bottom) of the modified strand. All stereo images are constructed for viewing with a stereoviewer. View largeDownload slide Figure 3. Stereo views of central tetramers. d(CA*AG)·d(CTTG) region of the (A) 1S (–) B[c]Ph‐dA and (B) 10S (+) B[a]P‐dA adducts. d(ACA*A)·d(TTGT) region of the (C) 10R (–) B[a]P‐dA adduct. B[c]Ph and B[a]P are in red and A*6–T17 is in green. A7–T16 is in blue in (A) and (B), C5–G18 is in blue in (C). Backbone, sugar atoms and other residues are in gray. The structures are aligned to have the 5′ residue of the modified strand at top left. View largeDownload slide Figure 3. Stereo views of central tetramers. d(CA*AG)·d(CTTG) region of the (A) 1S (–) B[c]Ph‐dA and (B) 10S (+) B[a]P‐dA adducts. d(ACA*A)·d(TTGT) region of the (C) 10R (–) B[a]P‐dA adduct. B[c]Ph and B[a]P are in red and A*6–T17 is in green. A7–T16 is in blue in (A) and (B), C5–G18 is in blue in (C). Backbone, sugar atoms and other residues are in gray. The structures are aligned to have the 5′ residue of the modified strand at top left. View largeDownload slide Figure 4. Correlation between distortion free energy and relative excision activity for the 1S (–) and 1R (+) B[c]Ph‐dA and 10S (+) and 10R (–) B[a]P‐dA adducts. View largeDownload slide Figure 4. Correlation between distortion free energy and relative excision activity for the 1S (–) and 1R (+) B[c]Ph‐dA and 10S (+) and 10R (–) B[a]P‐dA adducts. View largeDownload slide Scheme 1. Topological difference of bay region B[a]P (left) and fjord region B[c]Ph (right). View largeDownload slide Scheme 1. Topological difference of bay region B[a]P (left) and fjord region B[c]Ph (right). Table 1. Trajectory averaged structural parameters at lesion site   1R (+) B[c]Ph‐dA  1S (–) B[c]Ph‐dA  Unmodified control  Buckle  –28.3 (8.3)  36.0 (7.1)  7.1 (11.4)  Propeller  20.2 (7.5)  –44.4 (12.0)  –6.8 (9.7)  Opening  6.8 (6.5)  14.9 (7.7)  1.8 (5.1)  Rise  7.6 (0.5)  7.0 (0.4)  3.4 (0.4) 3.2 (0.3)  Roll  4.0 (6.7)  23.2 (6.8)  8.6 (7.9) 0.4 (5.4)  Tilt  –8.3 (5.2)  –17.7 (3.9)  1.6 (4.9) –4.4 (4.5)  α′  161 (8) [154]  –163 (10) [–131]    β′  103 (8) [104]  –88 (11) [–111]    γ′  –63 (5) [–66]  –62 (5) [–53]    δ′  19 (5) [18]  –12 (7) [–16]      1R (+) B[c]Ph‐dA  1S (–) B[c]Ph‐dA  Unmodified control  Buckle  –28.3 (8.3)  36.0 (7.1)  7.1 (11.4)  Propeller  20.2 (7.5)  –44.4 (12.0)  –6.8 (9.7)  Opening  6.8 (6.5)  14.9 (7.7)  1.8 (5.1)  Rise  7.6 (0.5)  7.0 (0.4)  3.4 (0.4) 3.2 (0.3)  Roll  4.0 (6.7)  23.2 (6.8)  8.6 (7.9) 0.4 (5.4)  Tilt  –8.3 (5.2)  –17.7 (3.9)  1.6 (4.9) –4.4 (4.5)  α′  161 (8) [154]  –163 (10) [–131]    β′  103 (8) [104]  –88 (11) [–111]    γ′  –63 (5) [–66]  –62 (5) [–53]    δ′  19 (5) [18]  –12 (7) [–16]    All values are in degrees except for the rise values, which are in Å. Standard deviations are listed in parentheses for the calculated average values. Experimental values for torsion angles α′, β′, γ′ and δ′ taken from the NMR solution structures are in brackets (26,27). The base pair parameters buckle, propeller and opening are for the A*6–T17 base pair in 1R (+) and 1S (–) B[c]Ph‐dA adducts, and the unmodified control. The base pair step parameters rise, roll and tilt are for the C5–G18 to A*6–T17 step in the 1R (+) B[c]Ph‐dA adduct, the A*6–T17 to A7–T16 base pair step in the 1S (–) B[c]Ph‐dA adduct, and the C5–G18 to A*6–T17 (left column) and A*6–T17 to A7–T16 (right column) base pair steps in the unmodified control. Since it has been shown that the rise values calculated by the CURVES algorithm (49) in Dials and Windows (50) are systematically larger than those determined by other methods (55), we also employed another nucleic acid structure analysis method, RNA (Run Nucleic Acid) (56), to calculate the rise values. They are 6.2 and 5.6 Å for the 1R (+) and 1S (−) B[c]Ph‐dA adducts, respectively, while for the 10S (+) and 10R (−) B[a]P‐dA adducts these rise values calculated by RNA (56) are 7.0 and 6.4 Å, respectively, preserving the trend of greater rise in the B[a]P‐dA adducts (see text). View Large Table 2. Free energy component analysis of 1R (+) and 1S (–)‐trans‐anti‐B[c]Ph‐N6‐dA DNA adducts   1R (+) B[c]Ph‐dA   1S (–) B[c]Ph‐dA  1R (+) – 1S (–)  Eelectrostatic  304.0 (45.0)  356.2 (38.2)  –52.2  EvdW  –186.2 (9.7)  –187.7 (10.4)  1.5  Eint  1015.6 (17.8)  1012.3 (17.3)  3.3  EMM  1133.5 (50.5)  1180.8 (38.4)  –47.3  Gnonpolar  25.2 (0.2)  25.1 (0.2)  0.1  GPB  –5596.8 (44.8)  –5645.1 (35.0)  48.3  Gsolvation  –5571.7 (44.8)  –5620.0 (35.0)  48.3  Eelectrostatic + GPB  –5292.8 (12.2)  –5288.9 (11.7)  –3.9  EMM + GPB  –4438.2 (17.0)  –4439.3 (15.1)  1.1  –TΔS  –593.9  –592.6  –1.3  Gtot  –5032.1  –5031.9  –0.2    1R (+) B[c]Ph‐dA   1S (–) B[c]Ph‐dA  1R (+) – 1S (–)  Eelectrostatic  304.0 (45.0)  356.2 (38.2)  –52.2  EvdW  –186.2 (9.7)  –187.7 (10.4)  1.5  Eint  1015.6 (17.8)  1012.3 (17.3)  3.3  EMM  1133.5 (50.5)  1180.8 (38.4)  –47.3  Gnonpolar  25.2 (0.2)  25.1 (0.2)  0.1  GPB  –5596.8 (44.8)  –5645.1 (35.0)  48.3  Gsolvation  –5571.7 (44.8)  –5620.0 (35.0)  48.3  Eelectrostatic + GPB  –5292.8 (12.2)  –5288.9 (11.7)  –3.9  EMM + GPB  –4438.2 (17.0)  –4439.3 (15.1)  1.1  –TΔS  –593.9  –592.6  –1.3  Gtot  –5032.1  –5031.9  –0.2  All energies are in kcal/mol. Standard deviations are listed in parentheses. View Large Table 3. Comparison of lesion site structural and thermodynamic parameters for the fjord region B[c]Ph‐dA versus the bay region B[a]P‐dA adducts   B[c]Ph‐dA adduct  B[a]P‐dA adduct    1R (+)  1S (–)  10R (–)  10S (+)  Rise (Å)  7.6  7.0  8.6  7.6  Unwinding (°)  18  19  29  41  Relative distortion free energy (kcal/mol)  0.2  0  3.8  8.3  Quality of Watson–Crick hydrogen bonding (IH)  252  398  363  599    B[c]Ph‐dA adduct  B[a]P‐dA adduct    1R (+)  1S (–)  10R (–)  10S (+)  Rise (Å)  7.6  7.0  8.6  7.6  Unwinding (°)  18  19  29  41  Relative distortion free energy (kcal/mol)  0.2  0  3.8  8.3  Quality of Watson–Crick hydrogen bonding (IH)  252  398  363  599  View Large References 1. Wood,R.D. ( 1999) DNA damage recognition during nucleotide excision repair in mammalian cells. Biochimie , 81, 39–44. Google Scholar 2. Sancar,A. ( 1996) DNA excision repair. Annu. Rev. Biochem. , 65, 43–81. Google Scholar 3. Benhamou,S. and Sarasin,A. ( 2000) Variability in nucleotide excision repair and cancer risk: a review. Mutat. Res. , 462, 149–158. Google Scholar 4. Agarwal,S.K., Sayer,J.M., Yeh,H.J.C., Pannell,L.K., Hilton,B.D. and Pigott,M.A. ( 1987) Chemical characterization of DNA adducts derived from the configurationally isomeric benzo[c]phenanthrene‐3,4‐diol‐1,2‐epoxides. J. Am. Chem. Soc. , 109, 2497–2504. Google Scholar 5. Topinka,J., Schwarz,L.R., Kiefer,F., Wiebel,F.J., Gajdos,O., Vidova,P., Dobias,L., Fried,M., Sram,R.J. and Wolff,T. ( 1998) DNA adduct formation in mammalian cell cultures by polycyclic aromatic hydrocarbons (PAH) and nitro‐PAH in coke even emission extract. Mutat. Res. , 419, 91–105. Google Scholar 6. Batty,D.P. and Wood,R.D. ( 2000) Damage recognition in nucleotide excision repair of DNA. Gene , 241, 193–204. Google Scholar 7. Sugasawa,K., Okamoto,T., Shimizu,Y., Masutani,C., Iwai,S. and Hanaoka,F. ( 2001) A multistep damage recognition mechanism for global genomic nucleotide excision repair. Genes Dev. , 15, 507–521. Google Scholar 8. Volker,M., Mone,M.J., Karmakar,P., van Hoffen,A., Schul,W., Vermeulen,W., Hoeijmakers,J.H.J., van Driel,R., van Zeeland,A.A. and Mullenders,L.H.F. ( 2001) Sequential assembly of the nucleotide excision repair factors in vivo. Mol. Cell , 8, 213–224. Google Scholar 9. Hecht,S.S., el‐Bayoumy,K., Rivenson,A. and Amin,S. ( 1994) Potent mammary carcinogenicity in female CD rats of a fjord region diol‐epoxide of benzo[c]phenanthrene compared to a bay region diol‐epoxide of benzo[a]pyrene. Cancer Res. , 54, 21–24. Google Scholar 10. Hess,M.T., Schwitter,U., Petretta,M., Giese,B. and Naegeli,H. ( 1997) Bipartite substrate discrimination by human nucleotide excision repair. Proc. Natl Acad. Sci. USA ., 94, 6664–6669. Google Scholar 11. Zou,Y., Luo,C. and Geacintov,N.E. ( 2001) Hierarchy of DNA damage recognition in Escherichia coli nucleotide excision repair. Biochemistry , 40, 2923–2931. Google Scholar 12. Gunz,D., Hess,M.T. and Naegeli,H. ( 1996) Recognition of DNA adducts by human nucleotide excision repair: evidence for a thermodynamic probing mechanism. J. Biol. Chem. , 271, 25089–25098. Google Scholar 13. Hirshfeld,F.L. ( 1963) The structure of overcrowded aromatic compounds. Part VII. Out‐of‐plane deformation in benzo[c]‐phenanthrene and 1,12‐dimethyl‐benzo[c]phenanthrene. J. Chem. Soc. , 2126–2135. Google Scholar 14. Levin,W., Chang,R.L., Wood,A.W., Thakker,D.R., Yagi,H., Jerina,D.M. and Conney,A.H. ( 1986) Tumorigenicity of optical isomers of the diastereomeric bay‐region 3,4‐diol‐1,2‐epoxides of benzo[c]phenanthrene in murine tumor models. Cancer Res. , 46, 2257–2261. Google Scholar 15. Cavalieri,E.L., Higginbotham,S., RamaKrishna,N.V., Devanesan,P.D., Todorovic,R., Rogan,E.G. and Salmasi,S. ( 1991) Comparative dose‐response tumorigenicity studies of dibenzo[α,l]pyrene versus 7,12‐dimethylbenz[α]anthracene, benzo[α]pyrene and two dibenzo[α,l]pyrene dihydrodiols in mouse skin and rat mammary gland. Carcinogenesis , 12, 1939–1944. Google Scholar 16. Levin,W., Wood,A.W., Chang,R.L., Ryan,D.E., Thomas,P.E., Yagi,H., Thakker,D.R., Vyas,K., Boyd,C., Chu,S.‐Y. et al. ( 1982) Oxidative metabolism of polycyclic aromatic hydrocarbons to ultimate carcinogens. Drug Metab. Rev. , 13, 555–580. Google Scholar 17. 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. Google Scholar 18. McKay,I.A., Paterson,H., Brown,R., Toksoz,D., Marshall,C.J. and Hall,A. ( 1986) N‐ras and Human Cancer. Anticancer Res. , 6, 483–490. Google Scholar 19. Buterin,T., Hess,M.T., Luneva,N., Geacintov,N.E., Amin,S., Kroth,H., Seidel,A. and Naegeli,H. ( 2000) Unrepaired fjord region polycyclic aromatic hydrocarbon–DNA adducts in ras codon 61 mutational hot spots. Cancer Res. , 60, 1849–1856. Google Scholar 20. Jayaram,B., Sprous,D., Young,M.A. and Beveridge,D.L. ( 1998) Free energy analysis of the conformational preferences of A and B forms of DNA in solution. J. Am. Chem. Soc. , 120, 10629–10633. Google Scholar 21. Srinivasan,J., Cheatham,T.E., Cieplak,P., Kollman,P.A. and Case,D.A. ( 1998) Continuum solvent studies of the stability of DNA, RNA, and phosphoramidate–DNA helices. J. Am. Chem. Soc. , 120, 9401–9409. Google Scholar 22. Cheatham,T.E.,III, Srinivasan,J., Case,D.A. and Kollman,P.A. ( 1998) Molecular dynamics and continuum solvent studies of the stability of polyG‐polyC and polyA‐polyT DNA duplexes in solution. J. Biomol. Struct. Dyn. , 16, 265–280. Google Scholar 23. Kollman,P.A., Massova,I., Reyes,C., Kuhn,B., Huo,S., Chong,L., Lee,M., Lee,T., Duan,Y., Wang,W. et al. ( 2000) Calculating structures and free energies of complex molecules: combining molecular mechanics and continuum models. Acc. Chem. Res. , 33, 889–897. Google Scholar 24. Jayaram,B., McConnell,K., Dixit,S.B., Das,A. and Beveridge,D.L. ( 2002) Free‐energy component analysis of 40 protein–DNA complexes: a consensus view on the thermodynamics of binding at the molecular level. J. Comput. Chem. , 23, 1–14. Google Scholar 25. Cheatham,T.E. and Young,M.A. ( 2001) Molecular dynamics simulation of nucleic acids: successes, limitations, and promise. Biopolymers , 56, 232–256. Google Scholar 26. Cosman,M., Fiala,R., Hingerty,B.E., Laryea,A., Lee,H., Harvey,R.G., Amin,S., Geacintov,N.E., Broyde,S. and Patel,D. ( 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. Google Scholar 27. 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. Google Scholar 28. Yan,S., Shapiro,R., Geacintov,N.E. and Broyde,S. ( 2001) Stereochemical, structural and thermodynamic origins of stability differences between stereoisomeric benzo[a]pyrene diol epoxide deoxyadenosine adducts in a DNA mutational hot spot sequence. J. Am. Chem. Soc. , 123, 7054–7066. Google Scholar 29. Schwartz,J.L., Rice,J.S., Luxon,B.A., Sayer,J.M., Xie,G., Yeh,H.J., Liu,X., Jerina,D.M. and Gorenstein,D.G. ( 1997) Solution structure of the minor conformer of a DNA duplex containing a dG mismatch opposite a benzo[a]pyrene diol epoxide/dA adduct: glycosidic rotation from syn to anti at the modified deoxyadenosine. Biochemistry , 36, 11069–11076. Google Scholar 30. Yeh,H.J., Sayer,J.M., Liu,X., Altieri,A.S., Byrd,R.A., Lakshman,M.K., Yagi,H., Schurter,E.J., Gorenstein,D.G. and Jerina,D.M. ( 1995) NMR solution structure of a nonanucleotide duplex with a dG mismatch opposite a 10S adduct derived from trans addition of a deoxyadenosine N6‐amino group to (+)‐(7R,8S,9S,10R)‐7,8‐dihydroxy‐9,10‐epoxy‐7,8,9,10‐tetrahydrobenzo[a]pyrene: an unusual syn glycosidic torsion angle at the modified dA. Biochemistry , 34, 13570–13581. Google Scholar 31. 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‐dihydroxy‐9,10‐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. Google Scholar 32. Schurter,E.J., Yeh,H.J., Sayer,J.M., Lakshman,M.K., Yagi,H., Jerina,D.M. and Gorenstein,D.G. ( 1995) NMR solution structure of a nonanucleotide duplex with a dG mismatch opposite a 10R adduct derived from trans addition of a deoxyadenosine N6‐amino group to (‐)‐(7S,8R,9R,10S)‐7,8‐dihydroxy‐9,10‐epoxy‐7,8,9,10‐tetrahydrobenzo[a]pyrene. Biochemistry , 34, 1364–1375. Google Scholar 33. Volk,D.E., Rice,J.S., Luxon,B.A., Yeh,H.J.C., Liang,C., Xie,G., Sayer,J.M., Jerina,D.M. and Gorenstein,D.G. ( 2000) NMR evidence for syn‐anti interconversion of a trans opened (10R)‐dA adduct of benzo[a]pyrene (7S,8R)‐diol (9R,10S)‐epoxide in a DNA duplex. Biochemistry , 39, 14040–14053. Google Scholar 34. Wu,M., Yan,S., Patel,D.J., Geacintov,N.E. and Broyde,S. ( 2001) Cyclohexene ring and fjord region twist inversion in stereoisomeric DNA adducts of enantiomeric benzo[c]phenanthrene diol epoxides. Chem. Res. Toxicol. , 14, 1629–1642. Google Scholar 35. Hingerty,B.E., Figueroa,S., Hayden,T.L. and Broyde,S. ( 1989) Prediction of DNA structure from sequence: a build‐up technique. Biopolymers , 28, 1195–1222. Google Scholar 36. Arnott,S., Smith,P.J.C. and Chandrasekaran,R. ( 1976) Handbook of Biochemistry and Molecular Biology, 3rd Edn. CRC Press, Cleveland, OH. Google Scholar 37. Case,D., Pearlman,D., Caldwell,J., Cheatham,T., Ross,W., Simmerling,C., Darden,T., Merz,K., Stanton,R., Cheng,A. et al. ( 1997) AMBER 5.0. University of California, San Francisco, San Francisco, CA. Google Scholar 38. Cornell,W.D., Cieplak,P., Bayly,C.I., Gould,I.R., Merz,K.M.J., Ferguson,D.M., Spellmeyer,D.C., Fox,T., Caldwell,J.W. and Kollman,P.A. ( 1995) A second generation force field for the simulation of proteins and nucleic acids. J. Am. Chem. Soc. , 117, 5179–5197. Google Scholar 39. Cheatham,T.E., Cieplak,P. and Kollman,P.A. ( 1999) A modified version of the Cornell et al. force field with improved sugar pucker phases and helical repeat. J. Biomol. Struct. Dyn. , 16, 845–862. Google Scholar 40. Essmann,U., Perera,L., Berkowitz,M.L., Darden,T., Lee,H. and Pederson,L.G. ( 1995) A smooth particle mesh Ewald method. J. Chem. Phys. , 103, 8577–8593. Google Scholar 41. Ryckaert,J.P., Ciccotti,G. and Berendsen,H.J.C. ( 1977) Numerical integration of cartesian equations of motion of a system with constraints: molecular dynamics of n‐alkanes. J. Comp. Phys. , 23, 327–341. Google Scholar 42. Jorgensen,W.L., Chandreskhar,J., Madura,J.D., Imprey,R.W. and Klein,M.L. ( 1983) Comparison of simple potential functions for simulating liquid water. J. Chem. Phys. , 79, 926–935. Google Scholar 43. Berendsen,H.J.C., Postma,J.P.M., van Gunsteren,W.F., DiNola,A. and Haak,J.R. ( 1984) Molecular dynamics with coupling to an external bath. J. Chem. Phys. , 81, 3684–3690. Google Scholar 44. Honig,B. and Nicholls,A. ( 1995) Classical electrostatics in biology and chemistry. Science , 268, 1144–1149. Google Scholar 45. Nicholls,A., Sharp,K.A. and Honig,B. ( 1990) DelPhi. Department of Biochemistry and Molecular Biophysics, Columbia University, New York, NY. Google Scholar 46. Sanner,M.F., Olson,A.J. and Spehner,J.C. ( 1996) Reduced surface: an efficient way to compute molecular surfaces. Biopolymers , 38, 305–320. Google Scholar 47. Saenger,W. ( 1984) Principles of Nucleic Acid Structure. Springer‐Verlag New York, Inc., New York, NY. Google Scholar 48. Dickerson,R.E. ( 1989) Definitions and nomenclature of nucleic acid structure parameters. J. Biomol. Struct. Dyn. , 6, 627–634. Google Scholar 49. Lavery,R. and Sklenar,H. ( 1988) The definition of generalized helicoidal parameters and of axis curvature for irregular nucleic acids. J. Biomol. Struct. Dyn. , 6, 63–91. Google Scholar 50. Ravishanker,G., Swaminathan,S., Beveridge,D.L., Lavery,R. and Sklenar,H. ( 1989) Conformational and helicoidal analysis of 30 ps of molecular dynamics on the d(CGCGAATTCGCG) double helix: “curves”, dials and windows. J. Biomol. Struct. Dyn. , 6, 669–699. Google Scholar 51. Berman,H.M. ( 1997) Crystal studies of B‐DNA: the answers and the questions. Biopolymers , 44, 23–44. Google Scholar 52. 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 the binding of a 3,4‐dihydroxy‐1,2‐epoxybenzo[c]phenanthrene stereoisomer to an 11‐mer oligodeoxyribonucleotide. Chem. Res. Toxicol. , 8, 444–454. Google Scholar 53. Krzeminski,J., Ni,J., Zhuang,P., Luneva,N., Amin,S. and Geacintov,N.E. ( 1999) Total synthesis, mass spectrometric sequencing, and stabilities of oligonucleotide duplexes with single trans‐anti‐BPDE‐N6‐dA lesions in the N‐ras codon 61 and other sequence contexts. Polycyclic Aromatic Compounds , 17, 1–10. Google Scholar 54. Chalikian,V.T., Völker,J., Plum,G.E. and Breslauer,K.J. ( 1999) A more unified picture for the thermodynamics of nucleic acid duplex melting: A characterization by calorimetric and volumetric techniques. Proc. Natl Acad. Sci. USA , 96, 7853–7858. Google Scholar 55. Lu,X.J. and Olson,W.K. ( 1999) Resolving the discrepancies among nucleic acid conformational analyses. J. Mol. Biol. , 285, 1563–1575. Google Scholar 56. Babcock,M.S., Pednault,E.P.D. and Olson,W.K. ( 1993) Nucleic acid structure analysis: a users guide to a collection of new analysis programs. J. Biomol. Struct. Dyn. , 11, 597–628. Google Scholar
TI - Relating repair susceptibility of carcinogen‐damaged DNA with structural distortion and thermodynamic stability
JF - Nucleic Acids Research
DO - 10.1093/nar/gkf427
DA - 2002-08-01
UR - https://www.deepdyve.com/lp/oxford-university-press/relating-repair-susceptibility-of-carcinogen-damaged-dna-with-usI0j5WfYQ
SP - 3422
EP - 3432
VL - 30
IS - 15
DP - DeepDyve
ER -