Environmental and occupational causes of cancer: new evidence 2005-2007.

What do we currently know about the occupational and environmental causes of cancer? As of 2007, the International Agency for Research on Cancer (IARC) identified 415 known or suspected carcinogens. Cancer arises through an extremely complicated web of multiple causes, and we will likely never know the full range of agents or combinations of agents. We do know that preventing exposure to individual carcinogens prevents the disease. Declines in cancer rates-such as the drop in male lung cancer cases from the reduction in tobacco smoking or the drop in bladder cancer among cohorts of dye workers from the elimination of exposure to specific aromatic amines-provides evidence that preventing cancer is possible when we act on what we know. Although the overall age-adjusted cancer incidence rates in the United States among both men and women have declined in the last decade, the rates of several types of cancers are on the rise; some of which are linked to environmental and occupational exposures. This report chronicles the most recent epidemiologic evidence linking occupational and environmental exposures with cancer. Peer-reviewed scientific studies published from January 2005 to June 2007 were reviewed, supplementing our state-of-the-evidence report published in September 2005. Despite weaknesses in certain individual studies, we consider the evidence linking the increased risk of several types of cancer with specific exposures somewhat strengthened by recent publications, among them brain cancer from exposure to non-ionizing radiation, particularly radiofrequency fields emitted by mobile telephones; breast cancer from exposure to the pesticide dichlorodiphenyltrichloroethane (DDT) before puberty; leukemia from exposure to 1,3-butadiene; lung cancer from exposure to air pollution; non-Hodgkin's lymphoma (NHL) from exposure to pesticides and solvents; and prostate cancer from exposure to pesticides, polyaromatic hydrocarbons (PAHs), and metal working fluids or mineral oils. In addition to NHL and prostate cancer, early findings from the National Institutes of Health Agricultural Health Study suggest that several additional cancers may be linked to a variety of pesticides. Our report also briefly describes the toxicological evidence related to the carcinogenic effect of specific chemicals and mechanisms that are difficult to study in humans, namely exposures to bis-phenol A and epigenetic, trans-generational effects. To underscore the multi-factorial, multi-stage nature of cancer, we also present a technical description of cancer causation summarizing current knowledge in molecular biology. We argue for a new cancer prevention paradigm, one based on an understanding that cancer is ultimately caused by multiple interacting factors rather than a paradigm based on dubious attributable fractions. This new cancer prevention paradigm demands that we limit exposure to avoidable environmental and occupational carcinogens, in combination with additional important risk factors like diet and lifestyle. The research literature related to environmental and occupational causes of cancer is constantly growing, and future updates will be carried out in light of new biological understanding of the mechanisms and new methods for studying exposures in human populations. The current state of knowledge is sufficient to compel us to act on what we know. We repeat the call of ecologist Sandra Steingraber: "From the right to know and the duty to inquire flows the obligation to act."

Genetic effects of oxidative DNA damages: comparative mutagenesis of the imidazole ring-opened formamidopyrimidines (Fapy lesions) and 8-oxo-purines in simian kidney cells.

Fapy.dG and 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxo-dG) are formed in DNA by hydroxyl radical damage. In order to study replication past these lesions in cells, we constructed a single-stranded shuttle vector containing the lesion in 5'-TGT and 5'-TGA sequence contexts. Replication of the modified vector in simian kidney (COS-7) cells showed that Fapy.dG is mutagenic inducing primarily targeted Fapy.G-->T transversions. In the 5'-TGT sequence mutational frequency of Fapy.dG was approximately 30%, whereas in the 5'-TGA sequence it was approximately 8%. In parallel studies 8-oxo-dG was found to be slightly less mutagenic than Fapy.dG, though it also exhibited a similar context effect: 4-fold G-->T transversions (24% versus 6%) occurred in the 5'-TGT sequence relative to 5'-TGA. To investigate a possible structural basis for the higher G-->T mutations induced by both lesions when their 3' neighbor was T, we carried out a molecular modeling investigation in the active site of DNA polymerase beta, which is known to incorporate both dCTP (no mutation) and dATP (G-->T substitution) opposite 8-oxo-G. In pol beta, the syn-8-oxo-G:dATP pair showed greater stacking with the 3'-T:A base pair in the 5'-TGT sequence compared with the 3'-A:T in the 5'-TGA sequence, whereas stacking for the anti-8-oxo-G:dCTP pair was similar in both 5'-TGT and 5'-TGA sequences. Similarly, syn-Fapy.G:dATP pairing showed greater stacking in the 5'-TGT sequence compared with the 5'-TGA sequence, while stacking for anti-Fapy.G:dCTP pairs was similar in the two sequences. Thus, for both lesions less efficient base stacking between the lesion:dATP pair and the 3'-A:T base pair in the 5'-TGA sequence might cause lower G-->T mutational frequencies in the 5'-TGA sequence compared to 5'-TGT. The corresponding lesions derived from 2'-deoxyadenosine, Fapy.dA and 8-oxo-dA, were not detectably mutagenic in the 5'-TAT sequence, and were only weakly mutagenic (<1%) in the 5'-TAA sequence context, where both lesions induced targeted A-->C transversions. To our knowledge this is the first investigation using extrachromosomal probes containing a Fapy.dG or Fapy.dA site-specifically incorporated, which showed unequivocally that in simian kidney cells Fapy.G-->T substitutions occur at a higher frequency than 8-oxo-G-->T and that Fapy.dA is very weakly mutagenic, as is 8-oxo-dA.

Mirror image stereoisomers of the major benzo[a]pyrene N2-dG adduct are bypassed by different lesion-bypass DNA polymerases in E. coli.

The potent mutagen/carcinogen benzo[a]pyrene (B[a]P) is metabolically activated to (+)-anti-B[a]PDE, which induces a full spectrum of mutations (e.g., G-to-T, G-to-A, -1 frameshifts, etc.) via its major adduct [+ta]-B[a]P-N2-dG. We recently showed that the dominant G-to-T mutation depends on DNA polymerase V (DNAP V), but not DNAPs IV or II, when studied in a 5'-TG sequence in E. coli. Herein we investigate what DNAPs are responsible for non-mutagenic bypass with [+ta]-B[a]P-N2-dG, along with its mirror image adduct [-ta]-B[a]P-N2-dG. Each adduct is built into a 5'-TG sequence in a single stranded M13 phage vector, which is then transformed into eight different E. coli strains containing all combinations of proficiency and deficiency in the three lesion-bypass DNAPs II, IV and V. Based on M13 progeny output, non-mutagenic bypass with [-ta]-B[a]P-N2-dG depends on DNAP IV. In contrast, non-mutagenic bypass with [+ta]-B[a]P-N2-dG depends on both DNAPs IV and V, where arguments suggest that DNAP IV is involved in dCTP insertion, while DNAP V is involved in extension of the adduct-G:C base pair. Numerous findings indicate that DNAP II has a slight inhibitory effect on the bypass of [+ta]- and [-ta]-B[a]P-N2-dG in the case of both DNAPs IV and V. In conclusion, for efficient non-mutagenic bypass (dCTP insertion) in E. coli, [+ta]-B[a]P-N2-dG requires DNAPs IV and V, [-ta]-B[a]P-N2-dG requires only DNAP IV, while DNAP II is inhibitory to both, and experiments to investigate these differences should provide insights into the mechanism and purpose of these lesion-bypass DNAPs.

Molecular modeling benzo[a]pyrene N2-dG adducts in the two overlapping active sites of the Y-family DNA polymerase Dpo4.

The potent, ubiquitous environmental mutagen/carcinogen benzo[a]pyrene (B[a]P) induces a single major adduct [+ta]-B[a]P-N2-dG, whose bypass in most cases results in either no mutation (dCTP insertion) or a G-->T mutation (dATP insertion). Translesion synthesis (TLS) of [+ta]-B[a]P-N2-dG generally requires DNA polymerases (DNAPs) in the Y-family, which exist in cells to bypass DNA damage caused by chemicals and radiation. A molecular dynamics (MD) study is described with dCTP opposite [+ta]-B[a]P-N2-dG in Dpo4, which is the best studied Y-family DNAP from a structural point of view. Two orientations of B[a]P-N2-dG (BPmi5 and BPmi3) are considered, along with two orientations of the dCTP (AS1 and AS2), as outlined next. Based on NMR studies, the pyrene moiety of B[a]P-N2-dG is in the minor groove, when paired with dC, and can point toward either the base on the 5'-side (BPmi5) or the 3'-side (BPmi3). Based on published X-ray structures, Dpo4 appears to have two partially overlapping active sites. The architecture of active site 1 (AS1) is similar to all other families of DNAPs (e.g., the shape of the dNTP). Active site 2 (AS2), however, is non-canonical (e.g., the beta- and gamma-phosphates in AS2 are approximately where the alpha- and beta-phosphates are in AS1). In the Dpo4 models generated herein, using the BPmi3 orientation the pyrene moiety of [+ta]-B[a]P-N2-dG points toward the duplex region of the DNA, and is accommodated without distortions in AS1, but with distortions in AS2. Considering the BPmi5 orientation, the pyrene moiety points toward the ss-region of DNA in Dpo4, and sits in a hole defined by the fingers and little fingers domain ("chimney"); BPmi5 is accommodated in AS2 without significant distortions, but poorly in AS1. In summary, when dCTP is paired with [+ta]-B[a]P-N2-dG in the two overlapping active sites in Dpo4, the pyrene in the BPmi3 orientation is accommodated better in active site 1 (AS1), while the pyrene in the BPmi5 orientation is accommodated better in AS2. Finally, we discuss why Y-family DNAPs might have two catalytic active sites.

Homology modeling of four Y-family, lesion-bypass DNA polymerases: the case that E. coli Pol IV and human Pol kappa are orthologs, and E. coli Pol V and human Pol eta are orthologs.

Y-family DNA polymerases (DNAPs) are a superfamily of evolutionarily related proteins that exist in cells to bypass DNA damage caused by both radiation and chemicals. Cells have multiple Y-family DNAPs, presumably to conduct translesion synthesis (TLS) on DNA lesions of varying structure and conformation. The potent, ubiquitous environmental mutagen/carcinogen benzo[a]pyrene (B[a]P) induces all classes of mutations with G-->T base substitutions predominating. We recently showed that a G-->T mutagenesis pathway for the major adduct of B[a]P ([+ta]-B[a]P-N2-dG) in Escherichia coli depends on Y-family member DNAP V. Since no X-ray crystal study for DNAP V has been reported, no structure is available to help in understanding the structural basis for dATP insertion associated with G-->T mutations from [+ta]-B[a]P-N2-dG. Herein, we do homology modeling to construct a model for UmuC, which is the polymerase subunit of DNAP V. The sequences of eight Y-family DNAPs were aligned based on the positioning of conserved amino acids and an analysis of conserved predicted secondary structure, as well as insights gained from published X-ray structures of five Y-family members. Starting coordinates for UmuC were generated from the backbone coordinates for the Y-family polymerase Dpo4 for reasons discussed, and were refined using molecular dynamics with CHARMM 27. A survey of the literature revealed that E. coli DNAP V and human DNAP eta show a similar pattern of dNTP insertion opposite a variety of DNA lesions. Furthermore, E. coli DNAP IV and human DNAP kappa show a similar dNTP insertional pattern with these same DNA lesions, although the insertional pattern for DNAP IV/kappa differs from the pattern for DNAPs V/eta. These comparisons prompted us to construct and refine models for E. coli DNAP IV and human DNAPs eta and kappa as well. The dNTP/template binding pocket of all four DNAPs was inspected, focusing on the array of seven amino acids that contact the base of the incoming dNTP, as well as the template base. DNAPs V and eta show similarities in this array, and DNAPs IV and kappa also show similarities, although the arrays are different for the two pairs of DNAPs. Thus, there is a correlation between structural similarities and insertional similarities for the pairs DNAPs V/eta and DNAPs IV/kappa. Although the significance of this correlation remains to be elucidated, these observations point the way for future experimental studies.

A role for DNA polymerase V in G --> T mutations from the major benzo[a]pyrene N2-dG adduct when studied in a 5'-TGT sequence in E. coli.

Benzo[a]pyrene (B[a]P), a potent mutagen/carcinogen, is metabolically activated to (+)-anti-B[a]PDE, which induces a full spectrum of mutations (e.g. GC --> TA, GC --> AT, etc.) principally via its major adduct [+ta]-B[a]P-N2-dG. Recent findings suggest that different lesion bypass DNA polymerases may be involved in different mutagenic pathways, which is the subject of this report. [+ta]-B[a]P-N2-dG built into a plasmid in a 5'-TGT sequence gives approximately equal numbers of G --> T and G --> A mutations when host E. coli are UV irradiated prior to transformation, so this sequence context was chosen to investigate what DNA polymerases are involved in G --> T versus G --> A mutations. G --> T mutations decline (>10-fold) if E. coli either are not UV-irradiated or are deficient in DNA polymerase V ((delta)umuD/C), demonstrating a role for damage-inducible DNA Pol V in a G --> T pathway. G --> T mutations are not affected by transformation into E. coli deficient in either DNA polymerases II or IV. While the work herein was in progress, Lenne-Samuel et al. [Mol. Microbiol. 38 (2000) 299] built the same adduct into a plasmid in a 5'-GGA sequence, and showed that the frequency of G --> T mutations was similar in UV-irradiated and unirradiated host E. coli cells, suggesting no involvement by damage-inducible, lesion bypass DNA polymerases (i.e., not II, IV or V); furthermore, a role for DNA Pol V was explicitly ruled out. The easiest way to reconcile the findings of Lenne-Samuel et al. with the findings herein is if two G --> T mutagenic pathways exist for [+ta]-B[a]P-N2-dG, where sequence context dictates which pathway is followed. In contrast to the G --> T mutations, herein G --> A mutations from [+ta]-B[a]P-N2-dG in the 5'-TGT sequence context are shown not to be affected by UV-irradiation of host E. coli, and are not dependent on DNA Pol V, or Pol II, Pol IV, or the damage-inducible, but SOS-independent UVM system. Published studies, however, have shown that G --> A mutations are usually enhanced by UV-irradiation of host E. coli prior to the introduction of plasmids either site-specifically modified with [+ta]-B[a]P-N2-dG or randomly adducted with (+)-anti-B[a]PDE; both findings imply the involvement of a lesion-bypass DNA polymerase. These disparate results suggest the existence of two G --> A mutagenic pathways for [+ta]-B[a]P-N2-dG as well, although confirmation of this awaits further study. In conclusion, a comparison between the evidence presented herein and published findings suggests the existence of two distinct mutagenic pathways for both G --> T and G --> A mutations from [+ta]-B[a]P-N2-dG, where in each case one pathway is not damage-inducible and not dependent on a lesion-bypass DNA polymerase, while the second pathway is damage-inducible and dependent on a lesion-bypass DNA polymerase. Furthermore, DNA sequence context appears to dictate which pathway (as defined by the involvement of different DNA polymerases) is followed in each case.

A method to accurately quantitate intensities of (32)P-DNA bands when multiple bands appear in a single lane of a gel is used to study dNTP insertion opposite a benzo[a]pyrene-dG adduct by Sulfolobus DNA polymerases Dpo4 and Dbh.

Quantitating relative (32)P-band intensity in gels is desired, e.g., to study primer-extension kinetics of DNA polymerases (DNAPs). Following imaging, multiple (32)P-bands are often present in lanes. Though individual bands appear by eye to be simple and well-resolved, scanning reveals they are actually skewed-Gaussian in shape and neighboring bands are overlapping, which complicates quantitation, because slower migrating bands often have considerable contributions from the trailing edges of faster migrating bands. A method is described to accurately quantitate adjacent (32)P-bands, which relies on having a standard: a simple skewed-Gaussian curve from an analogous pure, single-component band (e.g., primer alone). This single-component scan/curve is superimposed on its corresponding band in an experimentally determined scan/curve containing multiple bands (e.g., generated in a primer-extension reaction); intensity exceeding the single-component scan/curve is attributed to other components (e.g., insertion products). Relative areas/intensities are determined via pixel analysis, from which relative molarity of components is computed. Common software is used. Commonly used alternative methods (e.g., drawing boxes around bands) are shown to be less accurate. Our method was used to study kinetics of dNTP primer-extension opposite a benzo[a]pyrene-N(2)-dG-adduct with four DNAPs, including Sulfolobus solfataricus Dpo4 and Sulfolobus acidocaldarius Dbh. Vmax/Km is similar for correct dCTP insertion with Dpo4 and Dbh. Compared to Dpo4, Dbh misinsertion is slower for dATP (∼20-fold), dGTP (∼110-fold) and dTTP (∼6-fold), due to decreases in Vmax. These findings provide support that Dbh is in the same Y-Family DNAP class as eukaryotic DNAP κ and bacterial DNAP IV, which accurately bypass N(2)-dG adducts, as well as establish the scan-method described herein as an accurate method to quantitate relative intensity of overlapping bands in a single lane, whether generated from (32)P-signals or by other means (e.g., staining).

How Y-Family DNA polymerase IV is more accurate than Dpo4 at dCTP insertion opposite an N2-dG adduct of benzo[a]pyrene.

To bypass DNA damage, cells have Y-Family DNA polymerases (DNAPs). One Y-Family-class includes DNAP κ and DNAP IV, which accurately insert dCTP opposite N(2)-dG adducts, including from the carcinogen benzo[a]pyrene (BP). Another class includes DNAP η and DNAP V, which insert accurately opposite UV-damage, but inaccurately opposite BP-N(2)-dG. To investigate structural differences between Y-Family-classes, regions are swapped between DNAP IV (a κ/IV-class-member) and Dpo4 (a η/V-class-member); the kinetic consequences are evaluated via primer-extension studies with a BP-N(2)-dG-containing template. Four key structural elements are revealed. (1) Y-Family DNAPs have discreet non-covalent contacts between their little finger-domain (LF-Domain) and their catalytic core-domain (CC-Domain), which we call "non-covalent bridges" (NCBs). Arg37 and Arg38 in DNAP IV's CC-Domain near the active site form a non-covalent bridge (AS-NCB) by interacting with Glu251 and Asp252, respectively, in DNAP IV's LF-Domain. Without these interactions dATP/dGTP/dTTP misinsertions increase. DNAP IV's AS-NCB suppresses misinsertions better than Dpo4's equivalent AS-NCB. (2) DNAP IV also suppresses dATP/dGTP/dTTP misinsertions via a second non-covalent bridge, which is ∼8Å from the active site (Distal-NCB). Dpo4 has no Distal-NCB, rendering it inferior at dATP/dGTP/dTTP suppression. (3) dCTP insertion is facilitated by the larger minor groove opening near the active site in DNAP IV versus Dpo4, which is sensible given that Watson/Crick-like [dCTP:BP-N(2)-dG] pairing requires the BP-moiety to be in the minor groove. (4) Compared to Dpo4, DNAP IV has a smaller major groove opening, which suppresses dGTP misinsertion, implying BP-N(2)-dG bulk in the major groove during Hoogsteen syn-adduct-dG:dGTP pairing. In summary, DNAP IV has a large minor groove opening to enhance dCTP insertion, a plugged major groove opening to suppress dGTP misinsertion, and two non-covalent bridges (near and distal to the active site) to suppress dATP/dGTP/dTTP misinsertions; collectively these four structural features enhance DNAP IV's dNTP insertion fidelity opposite a BP-N(2)-dG adduct compared to Dpo4.

Molecular modeling of the major benzo[a]pyrene N2-dG adduct in cases where mutagenesis results are known in double stranded DNA.

The potent mutagen/carcinogen benzo[a]pyrene (B[a]P) is metabolically activated to (+)-anti-B[a]PDE, which induces a full spectrum of mutations (e.g. GC-->TA, GC-->AT, etc.). One hypothesis for this complexity is that different mutations are induced by different conformations of its major adduct [+ta]-B[a]P-N2-dG when bypassed during DNA replication (probably by different DNA polymerases). Previous molecular modeling studies suggested that B[a]P-N2-dG adducts can in principle adopt at least 16 potential conformational classes in ds-DNA. Herein we report on molecular modeling studies with the eight conformations most likely to be relevant to base substitution mutagenesis in 10 cases where mutagenesis has been studied in ds-DNA plasmids in E. coli with B[a]P-N2-dG adducts of differing stereoisomers and DNA sequence contexts, as well as in five cases where the conformation is known by NMR. Of the approximately 11,000 structures generated in this study, the computed lowest energy structures are reported for 120 cases (i.e. eight conformations and 15 examples), and their conformations compared. Of the eight conformations, four are virtually always computed to be high in energy. The remaining four lower energy conformations include two with the BP moiety in the minor groove (designated: BPmi5 and BPmi3), and two base-displaced conformations, one with the dG moiety in the major groove (designated: Gma5) and one with the dG in the minor groove (designated: Gmi3). Interestingly, these four are the only conformations that have been observed for B[a]P-N2-dG adducts in NMR studies. Independent of sequence contexts and adduct stereochemistry, BPmi5 structures tend to look reasonably similar, as do BPmi3 structures, while the base-displaced structures Gma5 and BPmi3 tend to show greater variability in structure. A correlation was sought between modeling and mutagenesis results in the case of the low energy conformations BPmi5, BPmi3, Gma5 and Gma3. Plots of log[(G-->T)/(G-->A)] versus energy[(conformation X)-(conformation Y)] were constructed for all six pairwise combinations of these four conformations, and the only plot giving a straight line involved Gma5 and Gmi3. While this finding is striking, its significance is unclear (as discussed).

Structural model of the Y-Family DNA polymerase V/RecA mutasome.

To synthesize past DNA damaged by chemicals or radiation, cells have lesion bypass DNA polymerases (DNAPs), most of which are in the Y-Family. One class of Y-Family DNAPs includes DNAP η in eukaryotes and DNAP V in bacteria, which have low fidelity when replicating undamaged DNA. In Escherchia coli, DNAP V is carefully regulated to insure it is active for lesion bypass only, and one mode of regulation involves interaction of the polymerase subunit (UmuC) and two regulatory subunits (UmuD') with a RecA-filament bound to ss-DNA. Taking a docking approach, ∼150,000 unique orientations involving UmuC, UmuD' and RecA were evaluated to generate models, one of which was judged best able to rationalize the following published findings. (1) In the UmuD'(2)C/RecA-filament model, R64-UmuC interacts with S117-RecA, which is known to be at the UmuC/RecA interface. (2) At the model's UmuC/RecA interface, UmuC has three basic amino acids (K59/R63/R64) that anchor it to RecA. No other Y-Family DNAP has three basic amino acids clustered in this region, making it a plausible site for UmuC to form its unique interaction with RecA. (3) In the model, residues N32/N33/D34 of UmuC form a second interface with RecA, which is consistent with published findings. (4) Active UmuD' is generated when 24 amino acids in the N-terminal tail of UmuD are proteolyzed, which occurs when UmuD(2)C binds the RecA-filament. When UmuD is included in an UmuD(2)C/RecA-filament model, plausible UmuD/RecA contacts guide the UmuD cleavage site (C24/G25) into the UmuD proteolysis active site (S60/K97). One contact involves E11-UmuD interacting with R243-RecA, where the latter is known to be important for UmuD cleavage. (5) The UmuD(2)C/RecA-filament model rationalizes published findings that at least some UmuD-to-UmuD' cleavage occurs intermolecularly. (6) Active DNAP V is known to be the heterotetramer UmuD'(2)C/RecA, a model of which can be generated by a simple rearrangement of the RecA monomer at the 3'-end of the RecA-filament. The rearranged UmuD'(2)C/RecA model rationalizes published findings about UmuD' residues in proximity to RecA. In summary, docking and molecular simulations are used to develop an UmuD'(2)C/RecA model, whose structure rationalizes much of the known properties of the active form of DNA polymerase V.

Architecture of y-family DNA polymerases relevant to translesion DNA synthesis as revealed in structural and molecular modeling studies.

DNA adducts, which block replicative DNA polymerases (DNAPs), are often bypassed by lesion-bypass DNAPs, which are mostly in the Y-Family. Y-Family DNAPs can do non-mutagenic or mutagenic dNTP insertion, and understanding this difference is important, because mutations transform normal into tumorigenic cells. Y-Family DNAP architecture that dictates mechanism, as revealed in structural and modeling studies, is considered. Steps from adduct blockage of replicative DNAPs, to bypass by a lesion-bypass DNAP, to resumption of synthesis by a replicative DNAP are described. Catalytic steps and protein conformational changes are considered. One adduct is analyzed in greater detail: the major benzo[a]pyrene adduct (B[a]P-N(2)-dG), which is bypassed non-mutagenically (dCTP insertion) by Y-family DNAPs in the IV/κ-class and mutagenically (dATP insertion) by V/η-class Y-Family DNAPs. Important architectural differences between IV/κ-class versus V/η-class DNAPs are discussed, including insights gained by analyzing ~400 sequences each for bacterial DNAPs IV and V, along with sequences from eukaryotic DNAPs kappa, eta and iota. The little finger domains of Y-Family DNAPs do not show sequence conservation; however, their structures are remarkably similar due to the presence of a core of hydrophobic amino acids, whose exact identity is less important than the hydrophobic amino acid spacing.

Y-Family DNA polymerases may use two different dNTP shapes for insertion: a hypothesis and its implications.

Chemicals and radiation can damage DNA leading to the formation of adducts/lesions, which - if not removed by DNA repair pathways - usually block replicative DNA polymerases (DNAPs). To overcome such potentially lethal blockage, cells have lesion bypass DNAPs, which are often in the Y-Family and include several classes. One class includes human DNAP kappa and E. coli DNAP IV, and they insert dCTP in the non-mutagenic pathway opposite [+ta]-B[a]P-N(2)-dG, which is the major adduct formed by the environmental carcinogen benzo[a]pyrene. Another class includes hDNAP eta and ecDNAP V, and they insert dATP opposite [+ta]-B[a]P-N(2)-dG in the dominant G-->T mutagenic pathway. Herein we develop a hypothesis for why the IV/kappa-class preferentially does cellular dCTP insertion. On the minor groove side of the active site, Y-Family DNAPs have a cleft/hole that can be analyzed based on an analogy to a "chimney." Our models of DNAP IV show a large chimney opening from which the pyrene of [+ta]-B[a]P-N(2)-dG can protrude, which allows canonical adduct-dG:dCTP pairing. In contrast, our models of DNAP V have small chimney openings that forces adduct-dG downward in the active site such that canonical adduct-dG:dCTP pairing is not possible. Based on X-ray structures, sequence alignment and our modeled structures of Y-Family DNAPs, chimney opening size seems primarily controlled by one amino acid ("flue-handle"), which dictates whether nearby amino acids ("flue") plug the chimney or not. Based on this analysis, a correlation is apparent: the flue is closed in V/eta-class DNAPs giving small chimney openings, while the flue is open for the IV/kappa-class giving large chimney openings. Secondarily, a hypothesis is developed for why the V/eta-class might preferentially do cellular dATP insertion opposite [+ta]-B[a]P-N(2)-dG: the small chimney forces adduct-dG lower in the active site, possibly leading to catalysis using a non-canonical dNTP shape that permits syn-adenine:adduct-dG base pairing. In summary, a hypothesize is developed that the pyrene moiety of [+ta]-B[a]P-N(2)-dG protrudes from the large chimney opening of DNAP IV, thus permitting canonical dCTP:adduct-dG pairing, while the small chimney opening of DNAP V forces [+ta]-B[a]P-N(2)-dG lower down in the active site, in which syn-adenine can pair with adduct-dG via a non-canonical dNTP shape.

Amino acid architecture that influences dNTP insertion efficiency in Y-family DNA polymerase V of E. coli.

Y-family DNA polymerases (DNAPs) are often required in cells to synthesize past DNA-containing lesions, such as [+ta]-B[a]P-N(2)-dG, which is the major adduct of the potent mutagen/carcinogen benzo[a]pyrene. The current model for the non-mutagenic pathway in Escherichia coli involves DNAP IV inserting deoxycytidine triphosphate opposite [+ta]-B[a]P-N(2)-dG and DNAP V doing the next step(s), extension. We are investigating what structural differences in these related Y-family DNAPs dictate their functional differences. X-ray structures of Y-family DNAPs reveal a number of interesting features in the vicinity of the active site, including (1) the "roof-amino acid" (roof-aa), which is the amino acid that lies above the nucleobase of the deoxynucleotide triphosphate (dNTP) and is expected to play a role in dNTP insertion efficiency, and (2) a cluster of three amino acids, including the roof-aa, which anchors the base of a loop, whose detailed structure dictates several important mechanistic functions. Since no X-ray structures existed for UmuC (the polymerase subunit of DNAP V) or DNAP IV, we previously built molecular models. Herein, we test the accuracy of our UmuC(V) model by investigating how amino acid replacement mutants affect lesion bypass efficiency. A ssM13 vector containing a single [+ta]-B[a]P-N(2)-dG is transformed into E. coli carrying mutations at I38, which is the roof-aa in our UmuC(V) model, and output progeny vector yield is monitored as a measure of the relative efficiency of the non-mutagenic pathway. Findings show that (1) the roof-aa is almost certainly I38, whose beta-carbon branching R-group is key for optimal activity, and (2) I38/A39/V29 form a hydrophobic cluster that anchors an important mechanistic loop, aa29-39. In addition, bypass efficiency is significantly lower both for the I38A mutation of the roof-aa and for the adjacent A39T mutation; however, the I38A/A39T double mutant is almost as active as wild-type UmuC(V), which probably reflects the following. Y-family DNAPs fall into several classes with respect to the [roof-aa/next amino acid]: one class has [isoleucine/alanine] and includes UmuC(V) and DNAP eta (from many species), while the second class has [alanine (or serine)/threonine] and includes DNAP IV, DNAP kappa (from many species), and Dpo4. Thus, the high activity of the I38A/A39T double mutant probably arises because UmuC(V) was converted from the V/eta class to the IV/kappa class with respect to the [roof-aa/next amino acid]. Structural and mechanistic aspects of these two classes of Y-family DNAPs are discussed.

Mutagenesis studies of the major benzo[a]pyrene N2-dG adduct in a 5'-TG versus a 5'-UG sequence: removal of the methyl group causes a modest decrease in the [G->T/G->A] mutational ratio.

The potent mutagen/carcinogen benzo[a]pyrene (B[a]P) is metabolically activated to (+)-anti-B[a]PDE, which induces a full spectrum of mutations primarily at the G:C base pairs (e.g. GC-->TA, GC-->AT, etc.). Each of these mutations can be induced by its major adduct [+ta]-B[a]P-N(2)-dG, where DNA sequence context appears to influence both the quantitative and qualitative pattern of mutagenesis. We noted previously that 5'-TG sequences tend to have a higher fraction of G-->T mutations for both [+ta]-B[a]P-N(2)-dG and (+)-anti-B[a]PDE in comparison with 5'-CG, 5'-GG or 5'-AG sequences. To investigate a possible structural element for this trend, the role (if any) of the methyl group on the 5'-T is considered. Using adduct site-specific means, the [G-->T/G-->A] mutational ratio for [+ta]-B[a]P-N(2)-dG is determined to be approximately 1.08 in a 5'-TGT sequence, and approximately 0.60 in a 5'-UGT sequence. (G-->C mutations are minor.) Although this modest approximately 1.8-fold decrease in [G-->T/G-->A] ratio is statistically significant (P = 0.03), it suggests that the methyl group on the 5'-T is not the main reason why a 5'-T tends to enhance G-->T mutations. This study was prompted by an adduct conformational hypothesis, which predicted that the removal of the methyl group in a 5'-TG sequence would lower the fraction of G-->T mutations; however, the approximately 1.8-fold decrease is too small to do additional experiments to assess whether this conformational hypothesis, or other hypotheses, are the true cause of the decrease, which is discussed in this paper.

Free-energy perturbation methods to study structure and energetics of DNA adducts: results for the major N2-dG adduct of benzo[a]pyrene in two conformations and different sequence contexts.

The potent mutagen/carcinogen benzo[a]pyrene (B[a]P) is activated to (+)-anti-B[a]PDE, which induces a variety of mutations (e.g., G --> T, G --> A, etc.) via its major adduct [+ta]-B[a]P-N2-dG. One hypothesis is that adducts (such as [+ta]-B[a]P-N2-dG) induce different mutations via different conformations, probably when replicated by different lesion-bypass DNA polymerases (DNAPs). We showed that Escherichia coli DNAP V was responsible for G --> T mutations with [+ta]-B[a]P-N2-dG in a 5'-TGT sequence (Yin et al., (2004) DNA Repair 3, 323), so we wish to study conformations of this adduct/sequence context by molecular modeling. The development of a CHARMM-based molecular dynamics (MD) simulations protocol with free-energy calculations in the presence of solvent and counterions is described. A representative base-pairing and base-displaced conformation of [+ta]-B[a]P-N2-dG in the 5'-TGT sequence are used: (1) BPmi5, which has the B[a]P moiety in the minor groove pointing toward the base on the 5'-side of the adduct, and (2) Gma5, which has the B[a]P moiety stacked with the surrounding base pairs and the dG moiety displaced into the major groove. The MD output structures are reasonable when compared to known NMR structures. Changes in DNA sequence context dramatically affect the biological consequences (e.g., mutagenesis) of [+ta]-B[a]P-N2-dG. Consequently, we also developed a MD-based free-energy perturbation (FEP) protocol to study DNA sequence changes. FEP involves the gradual "fading-out" of atoms in a starting structure (A) and "fading-in" of atoms in a final structure (B), which allows a realistic assessment of the energetic and structural changes when two structures A and B are closely related. Two DNA sequence changes are described: (1) 5'-TGT --> 5'-TGG, which involves two steps [T:A --> T:C --> G:C], and (2) 5'-TGT --> 5'-TGC, which involves three steps [T:A --> T:2AP --> C:2AP --> C:G], where 2AP (2-aminopurine) is included, because T:2AP and C:2AP retain more-or-less normal pairing orientations between complementary bases. FEP is also used to evaluate the impact that a 5'-TGT to 5'-UGT sequence change might have on mutagenesis with [+ta]-B[a]P-N2-dG. In summary, we developed (1) a CHARMM-based molecular dynamics (MD) simulations protocol with free-energy calculations in the presence of solvent and counterions to study B[a]P-N2-dG adducts in DNA duplexes, and (2) a MD-based free-energy perturbation (FEP) protocol to study DNA sequence context changes around B[a]P-N2-dG adducts.

Mutagenesis studies with four stereoisomeric N2-dG benzo[a]pyrene adducts in the identical 5'-CGC sequence used in NMR studies: G->T mutations dominate in each case.

Benzo[a]pyrene (B[a]P) is a polycyclic aromatic hydrocarbon (PAH) and a potent mutagen/carcinogen found ubiquitously in the environment. B[a]P is primarily metabolized to diol epoxides, which react principally at N2-dG in DNA. B[a]P-N2-dG adducts have been shown to induce a variety of mutations, notably G-->T, G-->A, G-->C and -1 frameshifts. Four stereoisomers of B[a]P-N2-dG (designated: [+ta]-;, [+ca]-, [-ta] and [-ca]) were studied by NMR in duplex 11mers in a 5'-CGC sequence context, and each adopted a different adduct conformation (Geacintov, et al. (1997) Chem. Res. Toxicol., 10, 111). Herein these four identical B[a]P-containing 11mers are built into duplex plasmid genomes and mutagenesis studied in Escherichia coli following SOS-induction. In nucleotide excision repair (NER) proficient E.coli, no adduct-derived mutants are detected. In NER deficient E.coli, G-->T mutations dominate for all four stereoisomers [+ta]-, [+ca]-, [-ta] and [-ca]-B[a]P-N(2)-dG, and mutation frequency is similar. Thus, the mutagenic pattern for these four B[a]P-N2-dG stereoisomers is the same, in spite of the fact that they adopt dramatically different conformations in ds-oligonucleotides as determined by NMR. These findings suggest that adduct conformation must be fluid enough in the 5'-CGC sequence that the duplex DNA conformation can interconvert to mutagenic and non-mutagenic conformations during lesion-bypass. A comparison of all published studies with these four B[a]P-N2-dG stereoisomers in E.coli reveals that B[a]P-N2-dG adduct stereochemistry tends to have a lesser impact on mutagenic pattern (e.g. G-->T versus G-->A mutations) than does DNA sequence context, which is discussed.

Molecular modeling of four stereoisomers of the major B[a]PDE adduct (at N(2)-dG) in five cases where the structure is known from NMR studies: molecular modeling is consistent with NMR results.

The potent mutagen/carcinogen benzo[a]pyrene (B[a]P) is metabolically activated to (+)-anti-B[a]PDE, which is known to induce a variety of mutations (e.g., GC --> TA, GC --> AT, etc.). One hypothesis for this complexity is that different mutations are induced by different conformations of its major adduct [+ta]-B[a]P-N(2)-dG when bypassed during DNA replication (perhaps by different DNA polymerases). Our previous molecular modeling studies have suggested that conformational complexity might be extensive in that B[a]P-N(2)-dG adducts appeared capable of adopting at least sixteen potential conformational classes in ds-DNA [e.g., Kozack and Loechler (1999) Carcinogenesis 21, 1953], although only eight seemed likely to be relevant to base substitution mutagenesis. Such molecular modeling studies are only likely to be valuable for the interpretation of mutagenesis results if global minimum energy conformations for adducts are found and if the differences in the energies of these different conformations can be computed reasonably accurately. One approach to assessing the reliability of our molecular modeling techniques is considered herein. Using a five-step molecular modeling protocol, which importantly included a molecular dynamics version of simulated annealing, eight conformations are studied in each of five cases. (The five cases are listed below, and were chosen because in each case the preferred solution conformation is known from a NMR study.) Of the eight conformations studied, the one computed to be lowest in energy is the same conformation as the one observed by NMR in four of the five cases: 5'-CGC sequence with [+ta]-, [-ta]-, and [+ca]-B[a]P-N(2)-dG, and 5'-TGC sequence with [+ta]-B[a]P-N(2)-dG. In the fifth case (5'-CGC sequence with [-ca]-B[a]P-N(2)-dG), the known NMR conformation is computed to be second lowest in energy, but it is within approximately 1.7 kcal of the computed lowest energy conformation. These results suggest that molecular modeling is surprisingly accurate in computing lowest energy conformations and that it should be useful in assessing the relative energies of different conformations. This is especially important given that currently molecular modeling is the only means available to study the energetics of minor conformations of DNA adducts.