Structures of the (+)- and (-)-trans-7,8-dihydroxy-anti-9,10-epoxy-7,8,9,10-tetrahydrobenzo(a)pyre ne adducts to guanine-N2 in a duplex dodecamer.

The structures of the mirror image (+)- and (-)-trans-anti-adducts of 7,8-dihydroxy-9,10-epoxy-7,8,9,10-tetrahydrobenzo(a)pyrene to guanine N2 have been of great interest because the high biological activity of 7,8-dihydroxy-9,10-epoxy-7,8,9,10-tetrahydrobenzo(a)pyrene in mammalian mutagenesis and tumorigenesis has been attributed to the predominant (+)-trans-anti-adduct. We have carried out new potential energy minimization studies, involving wide-scale conformational searches on small modified DNA subunits, followed by energy-minimized build-up techniques, to generate atomic resolution views of these adducts. These energy-minimized duplex dodecamers were then subjected to 100-ps molecular dynamic simulations with solvent and salt to yield animated molecular structures. The most favored computed structure for the (+)-adduct places the pyrenyl moiety in the B-DNA minor groove, with its long axis directed toward the 5' end of the modified strand, and with a pronounced bend in the helix axis. In the (-)-adduct, there are 2 favored structures. One places the pyrenyl moiety in the minor groove, whereas the other positions it in the major groove; in both cases, the pyrenyl long axis is directed more toward the 3' end of the modified strand, and with much less helix axis bend. Structures with intercalation character computed for these adducts are less preferred. The favored computed structures agree with spectroscopic data on the (+)- and (-)-trans-anti-adducts, whereas recent experimental evidence suggests that cis-adducts assume intercalation-type structures. Perhaps the conformational distinctions elucidated for the (+)- and (-)-trans anti-adducts play a role in their differential tumorigenic properties in mammalian systems.

Evading the proofreading machinery of a replicative DNA polymerase: induction of a mutation by an environmental carcinogen.

DNA replication fidelity is dictated by DNA polymerase enzymes and associated proteins. When the template DNA is damaged by a carcinogen, the fidelity of DNA replication is sometimes compromized, allowing mispaired bases to persist and be incorporated into the DNA, resulting in a mutation. A key question in chemical carcinogenesis by metabolically activated polycyclic aromatic hydrocarbons (PAHs) is the nature of the interactions between the carcinogen-damaged DNA and the replicating polymerase protein that permits the mutagenic misincorporation to occur. PAHs are environmental carcinogens that, upon metabolic activation, can react with DNA to form bulky covalently linked combination molecules known as carcinogen-DNA adducts. Benzo[a]pyrene (BP) is a common PAH found in a wide range of material ingested by humans, including cigarette smoke, car exhaust, broiled meats and fish, and as a contaminant in other foods. BP is metabolically activated into several highly reactive intermediates, including the highly tumorigenic (+)-anti-benzo[a]pyrene diol epoxide (BPDE). The primary product of the reaction of (+)-anti-BPDE with DNA, the (+)-trans-anti-benzo[a]pyrene diol epoxide-N(2)-dG ((+)-ta-[BP]G) adduct, is the most mutagenic BP adduct in mammalian systems and primarily causes G-to-T transversion mutations, resulting from the mismatch of adenine with BP-damaged guanine during replication. In order to elucidate the structural characteristics and interactions between the DNA polymerase and carcinogen-damaged DNA that allow a misincorporation opposite a DNA lesion, we have modeled a (+)-ta-[BP]G adduct at a primer-template junction within the replicative phage T7 DNA polymerase containing an incoming dATP, the nucleotide most commonly mismatched with the (+)-ta-[BP]G adduct during replication. A one nanosecond molecular dynamics simulation, using AMBER 5.0, has been carried out, and the resultant trajectory analyzed. The modeling and simulation have revealed that a (+)-ta-[BP]G:A mismatch can be accommodated stably in the active site so that the fidelity mechanisms of the polymerase are evaded and the polymerase accepts the incoming mutagenic base. In this structure, the modified guanine base is in the syn conformation, with the BP moiety positioned in the major groove, without interfering with the normal protein-DNA interactions required for faithful polymerase function. This structure is stabilized by a hydrogen bond between the modified guanine base and dATP partner, hydrophobic interactions between the BP moiety and the polymerase, a hydrogen bond between the modified guanine base and the polymerase, and several hydrogen bonds between the BP moiety and polymerase side-chains. Moreover, the G:A mismatch in this system closely resembles the size and shape of a normal Watson-Crick pair. These features reveal how the polymerase proofreading machinery may be evaded in the presence of a mutagenic carcinogen-damaged DNA, so that a mismatch can be accommodated readily, allowing bypass of the adduct by the replicative T7 DNA polymerase.

Accomodation of S-cis-tamoxifen-N(2)-guanine adduct within a bent and widened DNA minor groove.

The non-steroidal anti-estrogen tamoxifen [TAM] has been in clinical use over the last two decades as a potent adjunct chemotherapeutic agent for treatment of breast cancer. It has also been given prophylactically to women with a strong family history of breast cancer. However, tamoxifen treatment has also been associated with increased endometrial cancer, possibly resulting from the reaction of metabolically activated tamoxifen derivatives with cellular DNA. Such DNA adducts can be mutagenic and the activities of isomeric adducts may be conformation-dependent. We therefore investigated the high resolution NMR solution conformation of one covalent adduct (cis-isomer, S-epimer of [TAM]G) formed from the reaction of tamoxifen [TAM] to N(2)-of guanine in the d(C-[TAM]G-C).d(G-C-G) sequence context at the 11-mer oligonucleotide duplex level. Our NMR results establish that the S-cis [TAM]G lesion is accomodated within a widened minor groove without disruption of the Watson-Crick [TAM]G. C and flanking Watson-Crick G.C base-pairs. The helix axis of the bound DNA oligomer is bent by about 30 degrees and is directed away from the minor groove adduct site. The presence of such a bulky [TAM]G adduct with components of the TAM residue on both the 5'- and the 3'-side of the modified base could compromise the fidelity of the minor groove polymerase scanning machinery.

Molecular topology of polycyclic aromatic carcinogens determines DNA adduct conformation: a link to tumorigenic activity.

We report below on the solution structures of stereoisomeric "fjord" region trans-anti-benzo[c]phenanthrene-N2-guanine (designated (BPh)G) adducts positioned opposite cytosine within the (C-(BPh)G-C).(G-C-G) sequence context. We observe intercalation of the phenanthrenyl ring with stereoisomer-dependent directionality, without disruption of the modified (BPh)G.C base-pair. Intercalation occurs to the 5' side of the modified strand for the 1S stereoisomeric adduct and to the 3' side for the 1R stereoisomeric adduct, with the S and R-trans-isomers related to one another by inversion in a mirror plane at all four chiral carbon atoms on the benzylic ring. Intercalation of the fjord region BPh ring into the helix without disruption of the modified base-pair is achieved through buckling of the (BPh)G.C base-pair, displacement of the linkage bond from the plane of the (BPh)G base, adaptation of a chair pucker by the BPh benzylic ring and the propeller-like deviation from planarity of the BPh phenanthrenyl ring. It is noteworthy that intercalation without base-pair disruption occurs from the minor groove side for S and R-trans-anti BPh-N2-guanine adducts opposite C, in contrast to our previous demonstration of intercalation without modified base-pair disruption from the major groove side for S and R-trans-anti BPh-N6-adenine adducts opposite T. Further, these results on fjord region 1S and 1R-trans-anti (BPh)G adducts positioned opposite C are in striking contrast to earlier research with "bay" region benzo[a]pyrene-N2-guanine (designated (BP)G) adducts positioned opposite cytosine, where 10S and 10R-trans-anti stereoisomers were positioned with opposite directionality in the minor groove without modified base-pair disruption. They also are in contrast to the 10S and 10R-cis-anti stereoisomers of (BP)G adducts opposite C, where the pyrenyl ring is intercalated into the helix with directionality, but the modified base and its partner on the opposite strand are displaced out of the helix. These results are especially significant given the known greater tumorigenic potential of fjord region compared to bay region polycyclic aromatic hydrocarbons. The tumorigenic potential has been linked to repair efficiency such that bay region adducts can be readily repaired while their fjord region counterparts are refractory to repair. Our structural results propose a link between DNA adduct conformation and repair-dependent mutagenic activity, which could ultimately translate into structure-dependent differences in tumorigenic activities. We propose that the fjord region minor groove-linked BPh-N2-guanine and major groove-linked BPh-N6-adenine adducts are refractory to repair based on our observations that the phenanthrenyl ring intercalates into the helix without modified base-pair disruption. The helix is therefore minimally perturbed and the phenanthrenyl ring is not available for recognition by the repair machinery. By contrast, the bay region BP-N2-G adducts are susceptible to repair, since the repair machinery can recognize either the pyrenyl ring positioned in the minor groove for the trans-anti groove-aligned stereoisomers, or the disrupted modified base-pair for the cis-anti base-displaced intercalated stereoisomers.

A conformational analysis of the (+) anti BPDE adduct to the guanine amino group of dCpdG.

The (+) anti isomer of benz[a]pyrene diol epoxide (BPDE), 7 beta, 8 alpha-dihydroxy-9 alpha, 10 alpha-epoxy-7,8,9,10-tetrahydrobenz [a] pyrene has been identified as the probable tumorigenic lesion in mammalian systems. It forms a predominant adduct with DNA at N2 of guanine. In order to elucidate its conformation in atomic resolution detail, minimized conformational potential energy calculations were performed for the adduct with dCpdG. A global conformation search involving about 1000 trials was made. The lowest energy conformation had stacking between the hydrocarbon and the adjacent cytidine, in agreement with CD studies on modified GpU and UpG. This conformer differed from the B form most notably in the guanine glycosidic torsion, which is high anti. The next lowest energy form had torsion angles like the B form, with guanine-cytidine stacking. These two conformers differ in energy by only 2.1 kcal./mole, suggesting that their relative stability could easily be reversed in larger polymers, or under specific environmental conditions. Other conformations, with base-hydrocarbon or base-base stacking are also found, at somewhat higher energies. The Z form is at 7.8 kcal./mole. Thus, this adduct stabilizes the B form, in contrast with the N2 linked AAF adduct, which stabilizes the Z conformation.

Energy minimized structures of carcinogen-DNA adducts: 2-acetylaminofluorene and 2-aminofluorene.

Energy minimized structures of DNA modified by the aromatic amines 2-acetylaminofluorene (AAF) and 2-aminofluorene (AF), for which no experimental atomic resolution data exist, are presented. These have been computed with a new molecular mechanics program specifically designed to define distortions imposed by such adducts, and employing a rational strategy for searching the conformation space of a DNA molecule with covalently linked carcinogen. In alternating G-C sequences, the AAF adduct prefers to reside at the exterior of an undeformed Z-helix. It can also induce base displacement with attendant denaturation and helix bending in sequences that disfavor the Z form, but undeformed B helices are excluded. The AF adduct, by contrast, prefers the major groove of an unperturbed B-helix, but can also induce carcinogen-base stacking in single stranded regions of the DNA, such as at the replication fork. The different biological properties of these two adducts may be related to their distinct

Minor-groove binding models for acetylaminofluorene modified DNA.

Minimized potential energy calculations have been employed to locate and evaluate energetically a number of different models for DNA modified at carbon-8 of guanine by acetylaminofluorene (AAF). Three different duplex nonamer sequences were investigated. In addition to syn guanine models which have some denaturation and a Z-DNA model, we have found two new types of structures in which guanine remains syn and the AAF is placed in the minor groove of a B-DNA helix. One type features Hoogsteen base pairing between the modified guanine and protonated cytosine, with a sharply bent helix. The other (here termed the "wedge" model because the aromatic amine is wedged into the minor groove) maintains a single hydrogen bond between O6 of the modified guanine and N3 of protonated cytosine, with much less deformation of the helix, and close Van der Waals contacts between the AAF and the walls of the minor groove. Both types of structures (as well as the related forms produced by deprotonation of cytosine) are energetically important in all three sequences examined. The wedge-type model, which is most favored except in alternating G-C sequences, has been previously observed in a combined NMR and computational characterization of an aminofluorene (AF) modified guanine opposite adenine in a DNA duplex undecamer (D. Norman, P. Abuaf, B.E. Hingerty, D. Live, D. Grunberger, S. Broyde and D.J. Patel, Biochemistry 28, 7462 (1989)).

An internal fluorene model for iodo-N-2-acetylaminofluorene modified DNA.

Minimized conformational potential energy calculations have been performed for the 7-iodo (AAIF) and 7-fluoro (AAFF) derivatives of N-2-acetylaminofluorene (AAF), linked covalently to guanine C-8 in dCpdG. Both the iodo and the fluoro derivatives are carcinogenic and mutagenic. The lowest energy forms on the dinucleoside monophosphate level have syn guanine and fluorene-cytidine stacking. However, the iodo adduct cannot adopt this conformation in larger polymers, according to earlier experimental studies (Fuchs et al., Biochemistry, 15 (1976) 3347) and model building, because of iodine's large Van der Waal's radius. Therefore, a model consistent with all the experimental evidence, incorporating the second lowest energy conformation in B form duplex (dCdG)3 was constructed. In this model the modified guanine is syn, yet still stacked with the adjacent cytidine in one direction, the fluorene is located primarily at the helix interior between the base pairing sites, rupturing two base pairs, and the iodine atom and its adjoining ring protrude to the helix exterior.

AAF linked to the guanine amino group: a B-Z junction.

Minimized conformational potential energy calculations have been performed for AAF linked to dCpdG at the guanine amino group. This is a model for the minor AAF adduct observed in DNA, whose conformational influence has been difficult to ascertain. A global minimum energy conformation was computed with torsion angles like those of the dCpdG residue of Z-DNA. This conformation was incorporated into a larger polymer model at a B-Z junction, with the carcinogen residing in the groove in the Z direction. Local minimum energy conformations of the B type were also computed. In addition, two minima were found with fluorenecytidine stacking. These results suggest that existing B-Z junctions may be vulnerable to modification by AAF at the guanine amino group, or that such junctions may be induced by the carcinogen if the base sequence is appropriate. Otherwise the carcinogen can be located in the minor groove of the B helix (5, 10, 11) or covalently intercalated (13-15).

Helix geometry of single stranded DNA 'A' and 'B' forms from minimum energy conformations of dimeric subunits.

Low energy conformations with dihedral angles similar to those occurring in fibers of the 'A' and 'B' forms of DNAs have been calculated for the deoxydinucleoside phosphates dApdA, dCpdC, dTpdT, dGpdG and dGpdC (1-3). These conformers have been used as building blocks for generating larger single stranded polymers, whose helical parameters we have calculated. We find that single stranded 'A' and 'B' form helices tend to be narrower and more tightly wound than the duplexes obtained in fibers (4,5). This is consistent with experimental observations on single stranded fibers of poly (rC) (6). We also find that the different sequences have different helix geometries. In addition, it is observed that large variations in helix geometry for a given sequence are achievable at little energetic cost.

The influence of terminal 3', 5' phosphates on conformations of dApdA.

Addition of 3' and 5' terminal phosphates to dApdA causes a decrease in conformational flexibility. pdApdAp has much fewer conformers with energies below 2.5 kcal./mole than dApdA. THE A, B and Watson-Crick (34) helices are the most preferred forms. Other important conformations are in the trans domain of psi. Thus, flexibility in psi as well as in omega and omega, and in the sugar pucker is indicated. The transformation from the B helix to the Watson-Crick helix follows a low energy path. This is significant since Watson-Crick conformations may be important for intercalation into nucleic acid polymers (40-42) above the dimer level. The B helix is preferred over the A form in these large DNA subunits.

Conformation of the deoxydinucleoside monophosphate dCpdG modified at carbon 8 of guanine with 2-(acetylamino)fluorene.

Minimized conformational potential energy calculations were performed for dCpdG modified with the carcinogen 2-(acetylamino)fluorene (AAF). The major adduct, linked via a covalent bond between guanine C-8 and N-2 of AAF, was investigated. The 12 variable torsion angles and both deoxyribose puckers were independent flexible parameters in the energy minimizations. Three categories of low-energy conformers were calculated in which the guanine was syn and nearly perpendicular to the plane of the fluorene: (1) forms in which fluorene is stacked with cytidine (included among these is the global minimum energy conformation); (2) conformers which preserve guanine-cytidine stacking while placing the fluorene in a base-pair obstructing position; (3) conformers which maintain guanine-cytidine stacking and place the fluorene at the helix exterior, without interfering with base pairing. The Z form is important in this group. In addition, a low-energy conformation with guanine anti, but still nearly perpendicular to fluorene, was computed. Molecular models were constructed for the most important conformations incorporated into larger polymers. These indicated that the fluorene-cytidine stacked forms induce a severe kink in the B helix. Conformers with guanine-cytidine stacking and AAF in a base-pair obstructing position place the AAF at the B-type helix interior with little distortion in the helix direction. Conformers with the guanine-cytidine stack in which AAF does not affect base pairing place the fluorene at the Z or alternate helix exterior. It is suggested that base sequence, extent of modification, and external conditions such as salt concentration determine which of a number of possible conformational effects is actually induced by AAF. The variety of observed experimental results with AAF-modified DNA may reflect there various conformational possibilities.

A model for the single stranded random coil form of polydeoxyadenylic acid from minimum energy conformations of the dimeric subunit.

The minimum energy conformations of dApdA have been examined for their suitability as buildings blocks of the single stranded coil form of polynucleotides. Calculations of the characteristic ratio C difference = less than ro greater than 2/n liter2 were made for a polymer generated from all the low energy conformers, as well as for selected combinations. A polymer composed of a conformer with omega', omega = t*,g+,(skewed) psi = t, C-(2)-endo type pucker, in combination with the 'B' form, has a C difference equal to that observed in coils of apurinic acid (6) when the fraction of 'B' form conformers is approximately 25% and approximately 91%. The t*,g+ conformer is the second lowest energy form in the C-(2)-endo puckering domain, following the 'B' form.

'A' forms of RNAs in single strands, duplexes and RNA-DNA hybrids.

Helical parameters have been calculated for the 'A' form minimum energy conformations of ApA, CpC, GpG, UpU, GpC and UpA. The helix geometries are base sequence dependent. The single strands are narrower and more tightly wound than that duplex RNA-11 form. 9-12 kcal./mole are needed to convert these single strands to the RNA-11 conformation. However, in some sequences other 'A' type conformers capable of complementary base pairing may be formed at lower energetic cost. There is substantially more base stacking in the calculated single strands than in the RNA-11 conformation. Calculated intrastrand base stacking energies reflect these differences, and also are sequence dependent. The 'A' form RNA subunits differ from the analogous DNAs in possessing a larger rise per residue, needed to accomodate the 2'-OH. RNA-DNA hybrids are consequently more likely to be in the 'A-RNA than in the 'A'-DNA conformation, although the base sequence determines the extent of the preference.

Base displacement in AAF-modified Z-DNA.

A new model is presented for acetylaminofluorene (AAF)-modified Z-DNA, in which the carcinogen is in a base-displaced position, inserted into the helix interior and parallel with an adjacent base.

Visualization of an AAF induced frameshift mutation: molecular views of base displacement in B-DNA from minimized potential energy calculations.

Energy minimized structures of base displacement in an AAF modified B-DNA dodecamer are presented. A rational search strategy, beginning with a global search of the conformation space of the modified deoxydinucleoside monophosphate, together with model building by computer graphics, has been employed. A number of different minimum energy conformations have been located which reveal base displaced structures. These show fluorene interstrand stacking, fluorene inter- and intrastrand stacking, and non-stacked fluorene situated in the denatured bulge. The local helix axis is bent to various extents in the different forms, and one or two base pairs are fully denatured. One structure of special interest offers a molecular view that suggests how AAF can induce the -2 deletion mutation observed in AAF modified E. coli.