Skip Navigation


Mutagenesis Advance Access originally published online on November 25, 2005
Mutagenesis 2005 20(6):441-448; doi:10.1093/mutage/gei061
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
20/6/441    most recent
gei061v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (10)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Seo, K.-Y.
Right arrow Articles by Loechler, E. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Seo, K.-Y.
Right arrow Articles by Loechler, E. L.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?


© The Author 2005. Published by Oxford University Press on behalf of the UK Environmental Mutagen Society. All rights reserved. For permissions, please email: journals.permissions@oxfordjournals.org

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

Kwang-Young Seo, Arumugam Nagalingam, Matthew Tiffany and Edward L. Loechler*

Biology Department, Boston University, 24 Cummington Street, Boston, MA 02215, USA


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 References
 
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–N2-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.


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 References
 
Benzo[a]pyrene (B[a]P) is an important example of a polycyclic aromatic hydrocarbon (PAH), which are ubiquitous environmental pollutants found in internal combustion engine exhaust, power plant emissions, cigarette smoke and charred foods (1GoGoGoGoGoGo–7Go). B[a]P and its derivatives are known mutagens in cell types from bacteria to humans (8GoGoGoGoGoGoGoGoGoGoGo–19Go), are teratogenic (see Ref. 20Go and references therein), can cause mutations relevant to cancer causation (21GoGoGoGo–25Go), and are thought to be important in human cancer (consider Refs 26GoGoGoGo–30Go).

B[a]P is metabolically activated to a mutagenic and carcinogenic diol epoxide (+)-anti-B[a]PDE, which reacts with DNA via trans- and cis-addition to give [+ta]- and [+ca]-B[a]P–N2-dG (Figure 1), the former being the dominant adduct (31Go,32Go). Analogously, the minor metabolite (–)-anti-B[a]PDE gives [–ta]- and [–ca]-B[a]P–N2-dG (1GoGo–3Go). B[a]P can also react with DNA following other types of metabolic activation (representative Refs include 33GoGoGoGo–37Go).



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 1.. Structures of the four stereoisomers of N2-dG adducts studied herein: [+ta]-, [+ca]-, [– ta]- and [– ca]-B[a]P–N2-dG.

 
Mutational spectra with racemic (+)- and (–)-anti-B[a]PDE (10GoGoGoGo–14Go), and with pure (+)-anti-B[a]PDE have been reported (8Go,9Go,15GoGoGoGo–19Go). In Escherichia coli, (+)-anti-B[a]PDE gave deletions, insertions, frameshift and base substitutions (9, 24, 23 and 44%, respectively), where the latter primarily occurred at G:C base pairs with GC->TA (57%), GC->AT (23%) and GC->CG (20%) each being significant (8Go,9Go).

The type of mutation induced by (+)-anti-B[a]PDE depends on DNA sequence context, which has also been observed with the major adduct [+ta]-B[a]P–N2-dG (38GoGoGoGoGoGoGoGoGoGoGoGo–50Go), where the most dramatic examples include: G->T mutations dominated in a 5'-TGC sequence [>95% (38Go)], and G->A mutations dominated in a 5'-AGA sequence (~95% (41Go)), whereas –1 frameshift mutations dominated in a 5'-GGG run (~80% (51GoGo–53Go)).

These results raise the question: by what mechanism(s) can a single adduct induce multiple kinds of mutations? This question is best addressed with examples from studies in E.coli with AAF-C8-dG, which is the major adduct of N-2-acetylaminofluorene (AAF). AAF-C8-dG in a 5'-CGCG sequence either can cause a –2 frameshift mutation in a process dependent on DNA polymerase II (DNAP II), or can result in no mutation in a process dependent on DNAP V (52GoGoGo–55Go). The current model is that AAF-C8-dG at a replication fork can exist in two different conformations: one conformation in which the adducted dG moiety forms a –2 slipped intermediate that includes a primer terminus that DNAP II can elongate, or a second conformation in which the adducted dG is properly positioned to form a base pair with dC, as catalyzed by DNAP V. The simplest interpretation of these findings is that the pathway of lesion bypass (be it error-free or mutagenic) for a single adduct depends on its conformation, which in turn dictates which DNAP is involved in the trans lesion synthesis (TLS) event (56GoGoGoGoGoGoGoGoGoGoGoGoGo–69Go). Thus, for a given adduct, the lesion bypass event is most likely defined by both adduct conformation and the DNAP doing the bypass, and both factors must be understood in order to understand mutagenic mechanism.

Herein, we consider whether different stereoisomers of B[a]P–N2-dG adducts, which are known to be in dramatically different conformations in ds-DNA (70Go), give rise to different patterns of mutagenesis. In a 5'-CGC sequence, the adducts [+ta]-B[a]P–N2-dG (71Go) and [–ta]-B[a]P–N2-dG (72Go) were shown by NMR to have their B[a]P moiety in the minor groove and pointing toward the base on the 5' side or the 3' side of the adduct, respectively, but with more-or-less normal adduct-G:C base pairing. These conformations are designated BPmi5 and BPmi3, respectively. The dominant conformations for [+ca]-B[a]P–N2-dG (73Go) and [–ca]-B[a]P–N2-dG (74Go) are base displaced, in that the B[a]P moiety is stacked with the surrounding base pairs, while the dG moiety is displaced into the minor groove or the major groove, respectively; these conformations are designated Gmi3 and Gma5, respectively. If adduct conformation does indeed influence adduct mutagenesis, then we felt that a mutagenesis study with these four B[a]P–N2-dG stereoisomers in the identical NMR 5'-CGC sequence context might be revealing about the relationship between conformation and mutation.

We show that [+ta]-, [–ta]-, [+ca] and [–ca]-B[a]P–N2-dG in the NMR 5'-CGC sequence context each induce predominantly G->T mutations at approximately the same frequency using SOS-induced E.coli deficient in nucleotide excision repair (NER). Given the dramatically different conformations for these adducts, it seems unlikely that the dominant adduct conformation as determined in ds-oligonucleotides could be responsible for both the non-mutagenic and mutagenic outcomes, which is discussed.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 References
 
The B[a]P–N2-dG-containing 11mers (5'-CCATCGCTACC) were synthesized, purified by HPLC and characterized in the laboratory of Dr Nicholas Geacintov as described previously (75GoGoGo–78Go). Thereafter, each B[a]P-containing oligonucleotide ([+ta]-, [–ta]-, [+ca]- and [–ca]-B[a]P–CGC) and the unadducted control (C-CGC, which was taken through the same synthesis procedures) was 5' end, 32P-radiolabeled and purified successively by both denaturing and native polyacrylamide (20%) gel electrophoresis as described previously (39GoGoGoGoGoGo–45Go). No contaminants (~<0.5%) were detected following gel electrophoresis and visualization of the purified products. E.coli strains were described previously (8Go,9Go,38GoGoGoGoGoGoGo–45Go), including PFB50 (79Go). All other materials were as described previously (38GoGoGoGoGoGoGo–45Go).

Each oligonucleotide was built into the plasmid pRB1 using procedures identical to those we have used previously [Figure 2; (38GoGoGoGoGoGoGo–45Go)]. pRB1 itself was constructed from pTZ19R (ColE1 origin) in three steps. First, a wobble base (Position 2010) was changed (using ‘overlap extension’ PCR (80Go,81Go)) to eliminate the original BcgI site in the ampicillin gene. Second, a 10mer (5'-GTACGCAGTC) was inserted into the HincII site of the polycloning region ((82Go)) to give pRB0, which has a +1 frameshift mutation in its lacZ' fragment. Third, an 11mer (5'-CCATCGCTACC) was inserted into the reconstituted HincII site to the right of the 10mer to give pRB1, whose complete 21 bp insert (5'-GTACGCAGTCCCATCGCTACC) contains the targeted G (bold/underlined) in a unique BcgI site (underlined). The sequence of the 11mer is identical to the sequence studied by NMR (70GoGoGoGo–74Go).



View larger version (16K):
[in this window]
[in a new window]
  		Fig. 2.. The strategy to construct [+ta]-B[a]P–CGC-pRB1, which has [+ta]-B[a]P–N2-dG in a 5'-CGC sequence (5'-CCATCGCTACC) in the plasmid pRB1. (In parallel, [–ta]-, [+ca]- and [–ca]-B[a]P–CGC-pRB1 were constructed similarly, as well as C-CGC-RB1, which has no adduct.) Steps a and b: [+ta]-B[a]P–5'-CGC was formed by reaction of the starting oligonucleotide C-CGC with (+)-anti-B[a]PDE, and purified by HPLC, as well as native and denaturing gel electrophoresis. Steps 1 and 2: PRB0 itself was constructed by inserting a duplex 10-mer [5'-GTAGCGAGTC/5'-GACTCGCTAC] into the HincII site of pTZ19R. Thereafter the duplex 11mer [5'-CCATCGCTACC/5'-GGTAGCGATGG] was inserted into the re-formed HincII site in pRB0. Steps 3 and 4: ss-pRB1 was isolated and UV irradiated to eliminate progeny from this strand (Materials and methods). Steps 5–8: ds-pRB0 was digested with HincII and mixed with UV irradiated ss-pRB1, and denatured/renatured to give gapped heteroduplex DNA (GHD), which was isolated and purified. Step 9: [+ta]-B[a]P–CGC was covalently incorporated into the GHD via ligation to give [+ta]-B[a]P–CGC-pRB1. The adducted G (bold) is in a BcgI recognition site (5'-GCA[N]6TCG, underlined), although only the right half of the BcgI site is shown in the figure itself.

 
pRB1 with an 11 bp gap in one strand was constructed using techniques described previously (38GoGoGoGoGoGoGo–45Go), and the 11mers [+ta]-, [–ta]-, [+ca]- and [–ca]-B[a]P–CGC, as well as unadducted C-CGC, were each ligated into this gap to give products designated: [+ta]-, [–ta]-, [+ca]-, [–ca]- and C-B[a]P–CGC-pRB1. Ligation efficiencies were similar for each construct (e.g. 56, 65, 50, 54 and 63%, respectively, in one experiment). Following ligation, each plasmid product has a single adduct in a unique BcgI site (underlined in 5'-GCA[N]6TCGC, where the bolded ‘G’ is adducted) in the polycloning region of pRB1. Using methods we have employed repeatedly, a large number of UV-lesions (~15) were incorporated into the strand not containing the B[a]P–N2-dG adduct, which virtually eliminates ’strand bias‘ (83GoGo–85Go); i.e. output progeny plasmids derived from the strand not containing B[a]P–N2-dG are minimized.

In parallel, pRB1 containing [+ta]-, [–ta]-, [+ca]-, and [–ca]-B[a]P–N2-dG, as well as the unadducted control, were each transformed into ES87 cells that were UV-irradiated using a procedure that we have used consistently [UV irradiation at 254 nm; total dose, 12.6 J/m2; giving a cell survival of 30–60%; (see Refs 38GoGoGoGoGoGoGo–45Go)]. ES87 cells were used so as to be consistent with many of our previous experiments (8Go,9Go,39GoGoGoGoGo–44Go). The BcgI site in pRB1 is in the lacZ' gene, which permits {alpha}-complementation to give ß-galactosidase activity in appropriate cells (e.g. DH5{alpha}). ß-galactosidase activity can be monitored with X-gal, which gives colonies a blue color when cleaved. Blue colonies have plasmids with lacZ' in-frame, which includes wild type or base substitution mutations at the BcgI site, as well as in-frame deletion/insertion mutants.

Mutations in the unique BcgI site in pRB1 render progeny plasmids resistant to cleavage by BcgI. This property was used to eliminate wild-type BcgI sensitive plasmids, by successive rounds of plasmid cleavage with BcgI, transformation and plasmid re-isolation, as described previously (38GoGoGoGoGoGoGo–45Go). After mutant enrichment, progeny plasmids from blue colonies were isolated and sequenced. From ES87 cells, few progeny plasmids from [–ca]-B[a]P–CGC-pRB1 and C-CGC-pRB1 and none from [+ta]-, [–ta]-, [+ca]-CGC-pRB1 contained targeted, base substitution mutations. Most blue mutants were either genetic engineering-derived deletions or contained an unrecognizable DNA sequence. We have observed such genetic engineering-derived mutants in previous work (38GoGoGoGoGoGoGo–45Go).

Thereafter, [+ta]-, [–ta]-, [+ca]-, [–ca]- and C-B[a]P–CGC-pRB1 were each transformed into NER deficient PFB50 cells (uvrA6), following SOS-induction by UV-irradiation at a lower UV dose (3.0 J/m2), which also gave a cell survival of ~40%. Progeny yield for [+ta]-, [+ca]-, [–ta]-, and [–ca]-B[a]P–CGC-pRB1, and C-CGC-pRB1 was 2.8, 5.0, 6.2, 2.0 and 8.0 x 10+5, respectively, in one transformation. From blue colonies, progeny plasmids were isolated and sequenced, and these gave a higher fraction of targeted base substitution mutations (see below). In this phase of our study, we developed a method to screen out the largest class of contaminating mutants, which have a –1 frameshift mutation in the 10mer insert and no 11mer insert. Since the 10mer/11mer junction has a unique BsmFI site [underlined: 5'-GTACGCAGTC|CCATCGCTACC], only progeny plasmids that were cleavable at this BsmFI were sequenced.

Base substitution mutation frequency (MF) at the original genome location of the position occupied by the adducts was determined as described previously (38GoGoGoGoGoGoGo–45Go), and as outlined here. The BcgI site in pRB1 is in the lacZ' gene, which permits {alpha}-complementation to give ß-galactosidase activity in appropriate cells (e.g. DH5{alpha}). ß-galactosidase activity can be monitored with X-gal, which gives colonies a blue color when cleaved. Blue colonies have plasmids with lacZ' in-frame, which includes wild type or base substitution mutations at the BcgI site, as well as in-frame deletion/insertion mutants. Approximately 10% of the colonies were white from the original transformations with [+ta]-, [–ta]-, [+ca]- and [–ca]-B[a]P–CGC-pRB1, and were shown to contain plasmids of sequence corresponding to pRB0, which is a starting plasmid (Figure 2), is an 11 bp deletion of pRB1 and is also a BcgI resistant mutant. (Progeny pRB0 are genetic engineering-derived, side products of plasmid construction, as discussed previously (38Go,42Go).) We define: B/WR0 = [(WTb + Mb)/ (Mw)]. B/WR0 is the ratio of (blue colonies/white colonies) derived from the original transformation with e.g. [+ta]-B[a]P–CGC-pRB1, where ‘R0’ means ‘round zero’ (i.e. before mutant enrichment). WTb represents the number of blue wild-type colonies, and Mb and Mw represent the number of blue and white mutant colonies, respectively. After three rounds (R3) of mutant enrichment, wild-type BcgI sensitive plasmids (i.e., WTb) were eliminated and the ratio [blue colonies/white colonies] corresponds to: B/WR3 = [(Mb)/(Mw)]. The apparent mutation frequency is calculated according to: MFapp = [(B/WR3)/(B/WR0)], because [(B/WR3)/(B/WR0)] = [(Mb)/(Mw)]/[(WTb + Mb)/ (Mw)] = [(Mb)/(WTb + Mb)], the latter term being MF by definition. Plasmids from individual blue colonies were isolated after R3 and sequenced. In addition to base substitution mutations at the original genome location of the adduct, some plasmids had deletion mutations that restored the reading frame, where such mutants are derived from genetic engineering side-reactions as noted in the past (38Go,42Go). Thus, true base substitution mutation frequency was calculated as: MFtrue = [true mutants/total mutants] [(B/WR3)/(B/WR0)].


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 References
 
Oligonucleotides of sequence 5'-CCATCGCTACC containing four different stereoisomers of B[a]P–N2-dG (i.e., [+ta], [+ca], [–ta] and [–ca]; Figure 1) were synthesized and purified by HPLC using procedures identical to those used in their preparation for NMR studies (75GoGoGo–78Go). The oligonucleotides were purified further by both denaturing and native polyacrylamide gel electrophoresis (38GoGoGoGoGoGoGo–45Go), were characterized as described previously (39GoGoGo–42Go), and were shown to contain no contaminants at the limit of detection (<0.5%) using methods described previously (39GoGoGo–42Go).

These purified oligonucleotides were incorporated into the double stranded plasmid pRB1 (Materials and methods) to give four adduct-containing products designated: [+ta]-, [+ca]-, [–ta]- and [–ca]-B[a]P–CGC-pRB1, as well as an unadducted control C-CGC-pRB1 (Figure 2). Each plasmid was transformed in parallel into UV-irradiated (SOS-induced) ES87 cells, which are wild type for NER. No mutant progeny plasmids were detected, except for several G->A and G->C mutants with [–ca]-B[a]P–CGC-pRB1, although these were at or below the level observed with the control plasmid C-CGC-pRB1 that contained no adduct (Table I).


View this table:
[in this window]
[in a new window]
  		Table I.. Base substitution mutations at the original genome location of B[a]P–N2-dG adducts, when studied in a ds-plasmid in NER-proficient and -deficient E.coli cells that are SOS-induceda

 
The adduct-containing plasmids [+ta]-, [+ca]-, [–ta]- and [–ca]-B[a]P–CGC-pRB1, as well as C-CGC-pRB1, were then each transformed into NER deficient PFB50 cells (uvrA6). Mutant progeny plasmids were isolated and sequenced, and mutations targeted to the original genome location of the adducts are listed in Table I, along with MF.


   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
References
 
The low MF in NER proficient cells for B[a]P–N2-dG adducts in a 5'-CGC sequence
In our previous work in NER proficient cells using ds-plasmids, MF was 0.05–1.0% for [+ta]-B[a]P–N2-dG in five different sequence contexts (38GoGoGoGo–42Go,45Go), 0.6 and 1.8% for [+ca]-B[a]P–N2-dG in two sequence contexts (40Go,41Go), and 0.24% for [-ta]-B[a]P–N2-dG in one context (40Go). Herein, the four stereoisomers of B[a]P–N2-dG yielded no mutations above background in NER proficient cells (~<0.03% for C-CGC-pRB1, Table I). Thus, in our previous work MF was on average higher in NER proficient cells than observed herein for the 5'-CGC sequence.

The following estimate suggests that this low MF in NER proficient cells is not unreasonable. When (+)-anti-B[a]PDE mutagenesis was studied in SOS-induced NER proficient E.coli, no G114->T mutants were detected out of 578 mutants sequenced (9Go), where G114 is the only 5'-CGC sequence in the supF mutational target gene. (G114->T mutations are phenotypically detectable (8Go).) The (+)-anti-B[a]PDE mutational hotspot G115 had 48 out of the 578 mutants (9Go), making the ratio of mutants (G114/G115) = (0/48), which is better expressed as = (<1/48). In a sequence virtually identical to G115, the major adduct [+ta]-B[a]P–N2-dG gave MF ~1% in SOS-induced NER proficient cells (38Go). If we assume that [+ta]-B[a]P–N2-dG in the 5'-CGC sequence studied herein is like G114, and it is proportionally less mutagenic than G115, then we estimate MF ~<0.02% [= ~1% x (<1/48)], which is in the range for [+ta]-B[a]P–N2-dG in the 5'-CGC sequence in NER proficient cells (Table I). Although this estimate is not rigorous and is only for one stereoisomer, it suggests that our inability to detect mutants above background in NER proficient cells is not unreasonable.

In NER deficient cells G->T mutations dominate for all four stereoisomers
G->T mutations predominated for [+ta]-, [+ca]-, [–ta]- and [–ca]-B[a]P–CGC-pRB1 (Table I). These G->T mutations are likely to be adduct-derived for three reasons. (i) G->T MF was ~20-to-100-fold above MF for the unadducted, control C-CGC-pRB1 (Table I). (ii) [+ta]-, [+ca]-, [-ta]- and [-ca]-B[a]P–CGC-pRB1 gave a different mutagenic pattern (G->T mutations predominated) than did the control C-CGC-pRB1 (G->A mutations predominated). (iii) G->T mutations increased in going from NER proficient cells (ES87) to NER deficient cells (PFB50) for [+ta]-, [+ca]-, [–ta]- and [–ca]-B[a]P–CGC-pRB1 (~11–240-fold, Table I), which argues that they are adduct-derived, especially given that MF for G->A mutations did not change significantly for the control C-CGC-pRB1 in going from NER proficient to NER deficient cells. Furthermore, each adduct had a similar MF, with the possible exception of [–ta]-B[a]P–N2-dG, whose MF was slightly lower (~4-fold, Table I).

MF was low (<0.05%) for G->A mutations in NER deficient cells for all of the stereoisomers (Table I). G->A MF was slightly above control MF (C-CGC-pRB1) for two of the adduct-containing plasmids (i.e., [+ta]- and [+ca]-B[a]P–CGC-pRB1), while it was below the control for the other two adduct-containing plasmids (i.e., [-ta]- and [–ca]-B[a]P–CGC-pRB1). A case can be made that the G->A mutations obtained with [+ta]- and [+ca]-B[a]P–CGC-pRB1 are probably adduct-derived. MF for G->A mutations increased in going from NER proficient cells (ES87) to NER deficient cells (PFB50) by at least ~5-fold with [+ta]-B[a]P–CGC-pRB1 and ~10-fold with [+ca]-B[a]P–CGC-pRB1 (Table I), which suggests that G->A mutations with these samples are adduct-derived. This notion is reinforced by the observation that MF for G->A mutations did not increase in the control C-CGC-pRB1 in going from NER proficient cells to NER deficient cells (Table I). In contradiction, however, G->A MF is about the same for the control C-CGC-pRB1 (0.028%) as for [+ta]-B[a]P–CGC-pRB1 (0.035%) and [+ca]-B[a]P–CGC-pRB1 (0.044%) in NER deficient cells, which suggests that these G->A mutations are ‘not’ adduct-derived. This apparent contradiction can be reconciled if the G->A mutations from the control C-CGC-pRB1 arise from the presence of a low level contaminant in the control C-CGC oligonucleotide, which contains an ‘A-like’ base (i.e. a 5'-CCATC(A-like)CTACC 11mer), where this ‘A-like’ base could not react with either (+)- or (–)-anti-B[a]PDE, and, thus, could not provide a source of G->A mutations in the B[a]P-containing vectors following ligation. Two results are consistent with the hypothesis that G->A mutations in the control C-CGC-pRB1 sample came from a 5'-CCATC(A-like)CTACC contaminant that was not present in the B[a]P-containing oligonucleotides. (i) If this were true, then MF should be lower if the C-CGC oligonucleotide (which contains the putative A-like contaminant) were left out of the ligation reaction that generated the non-adduct-containing control plasmid. Call the control with no oligonucleotide ‘No/oligo-CGC’. This expectation was tested in a separate experiment and shown to be true: G->A MF was at least 6-fold lower for No/oligo-CGC (MF < ~0.02%; no mutants isolated) compared with C-CGC (MF ~0.12%, 12 G->A mutants isolated) (ii) G->A MF was higher with [+ta]- and [+ca]-B[a]P–CGC-pRB1 than was G>A MF for [ta] - and [-ca]- where the latter two are actually more revealing controls for G->A mutations than C-CGC-pRB1 itself. Thus, G->A mutations observed with [+ta]- and [+ca]-B[a]P–CGC-pRB1 are most likely due to the adducts, [+ta]- and [+ca]-B[a]P–N2-dG, respectively. Using similar reasoning, G->C mutations in the [+ca]-B[a]P–CGC-pRB1 sample are most likely due to the adduct. We will draw no conclusions about G->A mutations nor about G->C mutations, however, because they both were low in abundance and, thus, are not definitive.

Adduct conformation and mutagenic outcome
G->T mutations dominated with all four stereoisomers [+ta]-, [–ta]-, [+ca]- and [–ca]-B[a]P–N2-dG in the 5'-CGC sequence context (Table I), which is of interest because this is the identical 5'-CGC sequence context used to study adduct conformation by NMR, where different conformations dominate. BPmi5 and BPmi3, which were observed with [+ta]- and [–ta]-B[a]P–N2-dG, respectively (71Go,72Go), form more-or-less normal G:C base pairs with the B[a]P moiety in the minor groove pointing toward the base on the 5' side and the 3' side, respectively. Gmi3 and Gma5, which were observed with [+ca]- and [–ca]-B[a]P–N2-dG, respectively (73Go,74Go), are base displaced with the B[a]P moiety stacked with the surrounding base pairs and the dG moiety displaced into the minor groove and the major groove, respectively. Thus, the dominant conformation differs dramatically with adduct stereochemistry, while the dominant mutation does not. Furthermore, MF is similar for each of these adducts, which implies that non-mutagenic bypass occurs at a reasonably similar frequency (i.e. ~99.8% based on the data in Table I).

Little is known about the detailed relationship between adduct conformation and dNTP insertional preference during adduct bypass. We naively thought that adduct mutagenesis pattern might vary with adduct stereochemistry, given that adduct stereochemistry has such a significant effect on adduct conformation in the 5'-CGC sequence, and we consider why this is not the case.

The dominant bypass event is no mutation (~99.8%; Table I), which seems intuitively reasonable for [+ta]- and [–ta]-B[a]P–N2-dG, given that they adopt the BPmi5 and BPmi3 conformations that form more-or-less normal adduct-G:C base pair in ds-oligonucleotides (71Go,72Go), and might also be expected to do so during lesion bypass. However, a low MF makes less sense intuitively for [+ca]- and [–ca]-B[a]P–N2-dG, which have their B[a]P moiety co-planar with adjacent base pairs in the Gmi3 and Gma5 conformations, respectively (73Go,74Go), making it impossible to accurately pair with a dCTP. The most obvious way to rationalize this is if the adducts are sufficiently flexible in the 5'-CGC sequence, so that they rearrange to non-mutagenic conformations before bypass. For example, we (86Go), and others (70Go,87Go), have noted e.g. that Gma5 could interconvert to BPmi5 by a relatively simple one-step mechanism. The notion that B[a]P–N2-dG adducts are flexible in a 5'-CGC sequence contrasts with the situation for [+ta]-B[a]P–N2-dG in a 5'-CGG sequence, where several lines of investigation suggested that conformations do not readily interconvert (43Go). If these interpretations are true, then B[a]P–N2-dG adducts may behave differently in 5'-CGC versus 5'-CGG sequences in terms of conformational flexibility. The fact that G->T mutations predominate for all B[a]P–N2-dG stereoisomers in the 5'-CGC sequence (Table I) also suggests that interconversion occurs between the dominant conformation and the mutagenic conformation, since it seems unlikely that the mutagenic conformation would be the same as the dominant conformation in ds-DNA in each of these cases.

The effect of stereochemistry and sequence context on mutagenic pattern for B[a]P–N2-dG adducts
A significant literature exists concerning the effect of adduct stereochemistry, as well as DNA sequence context, on the pattern of B[a]P–N2-dG mutagenesis (Table II). In this section, we compare the results obtained herein with published findings, and analyze the similarities and differences. In each case when a comparison is made, it involves data collected by a single laboratory (Refs in Table II). We begin by comparing G->T and G->A mutations, which are quantitatively most important with (+)-anti-B[a]PDE (8Go,9Go) and with B[a]P–N2-dG adducts in general (e.g. G->T, G->A and G->C mutations represent 58, 24 and 19% of mutations listed in Table II).


View this table:
[in this window]
[in a new window]
  		Table II.. Base substitution mutations from different stereoisomers of B[a]P–N2-dG in different DNA sequence contexts in E.coli, as reported herein and in the literaturea

 
In four sequence contexts (i.e. 5'-CGA, 5'-CGC, 5'-GGT and 5'-TGG), G->T mutations predominate for all four stereoisomers [+ta]-, [– ta]-, [+ca]- and [– ca]-B[a]P–N2-dG (Table II). In another sequence context (i.e. 5'-CGT), G->A mutations are preferred for all but one stereoisomer [+ca]-B[a]P–N2-dG, where G->T predominates (Table II). These findings show that changes in B[a]P–N2-dG stereochemistry do not usually affect the dominant base substitution mutation.

There are several examples, however, where the [G->A/G->T] mutational ratio changes with B[a]P–N2-dG stereochemistry. The largest effect is in a 5'-CGT sequence, where the [G->A/G->T] ratio was ~17-fold higher for [-ca]-B[a]P–N2-dG versus [+ca]-B[a]P–N2-dG (Table II). In both a 5'-CGA and a 5'-GGT sequence, the [G->A/G->T] ratio was ~10-fold higher for [–ca]-B[a]P–N2-dG versus [–ta]-B[a]P–N2-dG (Table II).

In contrast, sequence context can have a significant effect on the dominant mutation. The most striking example is that G->T mutations overwhelmingly dominate in a 5'-TGC sequence, whereas G->A mutations overwhelmingly dominate in a 5'-AGA sequence with [+ta]-B[a]P–N2-dG (Table II). The [G->A/G->T] ratio is ~1300-fold greater for 5'-AGA, although this value is imprecise due to the small number of G->A and G->T mutations obtained in the 5'-TGC and 5'-AGA sequences, respectively. Another dramatic example is with [–ca]-B[a]P–N2-dG, where G->A is preferred with 5'-CGT, while G->T is preferred with 5'-CGC sequences, and the [G->A]/[G->T] ratios differ by at least ~600-fold (Table II).

G->C mutations vary most in comparison to G->A mutations, and the [G->A/G->C] ratio is affected more similarly by stereochemistry and sequence context. The biggest impact of stereochemistry is in a 5'-CGT sequence, where the [G->A/G->C] ratio was ~57-fold higher for [+ta]-B[a]P–N2-dG compared to [–ta]-B[a]P–N2-dG. In a 5'-CGA sequence, the effect is almost as great: the [G->A/G->C] ratio was ~44-fold higher for [–ta]-B[a]P–N2-dG versus [+ca]-B[a]P–N2-dG. In terms of sequence context, the [G->A/G->C] ratio was ~>108-fold higher for [+ca]-B[a]P–N2-dG in a 5'-AGA compared to a 5'-CGT sequence, while it is ~>39-fold for [+ta]-B[a]P–N2-dG in a 5'-AGA compared to a 5'-CGG sequence.

Thus, DNA sequence context seems to affect B[a]P–N2-dG base substitution mutagenic outcome more significantly than does adduct stereochemistry e.g. with G->T versus G->A mutations. One well-defined mechanism by which sequence context can affect mutagenic outcome is via a slippage-type process called ‘dislocation’ or ‘templated’ mutagenesis (88GoGoGoGoGoGoGoGo–96Go). A DNA polymerase is envisioned to stall at an adduct, to slip to the next 5' base along the template, and to use this base to direct incorporation (e.g. dATP insertion opposite the 5'-T in a 5'-TG sequence context), whereupon the newly incorporated dA slips back to form an adduct-G:A mispair, from which extension yields the intermediate that eventually gives the G->T mutation. The dislocation/templated mechanism has been suggested as a possible rationale for why G->T mutations dominate with [+ta]-B[a]P–N2-dG in a 5'-TGC sequence (38Go), and why G->A mutations dominate in a 5'-AGA sequence (41Go), perhaps being the reason why the [G->A]/[G->T] ratio varies so dramatically between these sequences (i.e., ~1300-fold, Table II).

Although logically attractive, a variety of observations raise doubts about the significance of the dislocation/templated mechanism to explain the mutagenic patterns observed with B[a]P–N2-dG adducts (discussed in Ref. 45Go), while findings herein add new objections. The dislocation/templated mechanism cannot explain the ~600-fold difference in the [G->A]/[G->T] ratio for the 5'-CGT versus 5'-CGC sequence with [–ca]-B[a]P–N2-dG (Table II), since neither dATP nor dTTP insertion is predicted in a 5'-CG sequence. Furthermore, the fact that G->T mutations are overwhelmingly favored for [–ca]-B[a]P–N2-dG in the 5'-CGC sequence (Table I) is the clearest example to date that the dislocation/templated mechanism is not obligate for the G->T mutational pathway.

Summary and conclusions
In NER proficient E.coli, MF is low for all of the stereoisomers [+ta]-, [+ca]-, [–ta]- and [–ca]-B[a]P–N2-dG (Table I). In NER deficient cells, G->T mutations predominated for all of the stereoisomers (Table I), in spite of the fact that they each adopt significantly different conformations (i.e. BPmi5, BPmi3, Gmi3 and Gma5, respectively) in the identical sequence context as determined by NMR. This suggests that adduct conformational interconversion occurs in the 5'-CGC sequence, which contrasts with conclusions from several published studies in a 5'-CGG sequence, where evidence suggested that conformational interconversion was relatively slow (43Go). Analysis of the findings herein and published findings reveal that changes in DNA sequence context have a bigger effect on B[a]P–N2-dG mutagenic patterns than do changes in adduct stereochemistry (Table II). While no satisfactory explanation for this observation can currently be offered, the work herein adds new arguments to suggest that a slippage-type, dislocation/templated mechanism is unlikely to be the explanation for how sequence context affects B[a]P–N2-dG mutagenic pattern.


   Acknowledgments
 
This work was supported by United States Public Health Services Grant R01ES03775. The phosphorimager used in this work was purchased from National Institutes of Health Shared Instrumentation Grant RR11397.


   Notes
 
* To whom correspondence should be addressed. Tel: +1 617 353 9259; Fax: +1 617 353 6340; Email: loechler{at}bu.edu


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 

    1. Harvey,R.G. (1997) Polycyclic Aromatic Hydrocarbons: Chemistry and Cancer, Wiley-VCH, Inc., New York.

    2. Singer,B. and Grunberger,D. (1983) Molecular Biology of Mutagens and Carcinogens, Plenum Press, New York.

    3. Conney,A.H. (1982) Induction of microsomal enzymes by foreign chemicals and carcinogens by polycyclic aromatic hydrocarbons. Cancer Res., 42, 4875–4917.[Free Full Text]

    4. Phillips,D.H. (1983) Fifty years of benzo[a]pyrene. Nature, 303, 468–472.[CrossRef][Medline]

    5. Dipple,A. (1985) Polycyclic aromatic hydrocarbon carcinogens. In: Harvey,R.G. (ed.), Polycyclic Aromatic Hydrocarbons and Carcinogenesis, American Chemical Society Press, Washington, DC, pp 1–17.

    6. Jones,P.W. (1982) Polynuclear aromatic hydrocarbons. In: Bowman,M.G. (ed.), Handbook of Carcinogens and Hazardous Substances, Marcel Dekker, Inc., New York, pp. 573–639.

    7. Grasso,P. (1984) Carcinogens in food. In: Searle,C.E. (ed.), Chemical Carcinogens, 2nd edn. ACS Monograph 182, American Chemical Society Press, Washington, DC, pp. 1205–1239.

    8. Rodriguez,H. and Loechler,E.L. (1993) Mutational spectra of the (+)-anti-diol epoxide of benzo[a]pyrene in a supF gene of an Escherichia coli plasmid: DNA sequence context influences hotspots, mutational specificity and the extent of SOS enhancement of mutagenesis. Carcinogenesis, 14, 373–383.[Abstract/Free Full Text]

    9. Rodriguez,H. and Loechler,E.L. (1993) Mutagenesis by the (+)-anti-diol epoxide of benzo[a]pyrene: what controls mutagenic specificity? Biochemistry, 32, 373–383.

    10. Eisenstadt,E., Warren,A.J., Porter,J., Atkins,D. and Miller,J.H. (1982) Carcinogenic epoxides of benzo[a]pyrene and cyclopenta[cd]pyrene induce base substitutions via specific transversions. Proc. Natl Acad. Sci. USA, 82, 1945–1949.

    11. Bernelot-Moens,C., Glickman,B.W. and Gordon,A.J.E. (1990) Induction of specific frameshift and base substitution events by benzo[a]pyrene diol epoxide in excision-repair-deficient. Escherichia coli Carcinogenesis, 11, 781–785.

    12. Yang,J.L., Maher,V.M. and McCormick,J.J. (1987) Kinds of mutations formed when a human shuttle vector containing adducts of (+/-)-7ß,8{alpha}-dihydroxy-9{alpha},10{alpha}-epoxy-7,8,9,10- dihydrobenzo[a]pyrene replicates in human cells. Proc. Natl Acad. Sci. USA, 84, 3787–3791.[Abstract/Free Full Text]

    13. Yang,J.-L., Chen,R.-H., Maher,V.M. and McCormick,J.J. (1990) Kinds and locations of mutations induced by (+/-)- 7ß,8{alpha} -dihydroxy-9{alpha},10{alpha} -epoxy-7,8,9,10-dihydrobenzo[a]pyrene in the coding region of the hypoxanthine (guanine) phosphoribosyltransferase gene in diploid human fibroblasts. Carcinogenesis, 12, 71–75.

    14. Carothers,A.M. and Grunberger,D. (1990) DNA base changes in benzo[a]pyrene diol epoxide-induced dihydrofolate reductase mutants of Chinese hampster ovary cells. Carcinogenesis, 11, 189–192.[Abstract/Free Full Text]

    15. Wei,S.-J.C., Chang,R.L., Wong,C.-Q. et al. (1991) Dose-dependent differences in the profile of mutations induced by an ultimate carcinogen from benzo[a]pyrene Proc. Natl Acad. Sci. USA, 88, 11227–11230.[Abstract/Free Full Text]

    16. Wei,S.-J.C., Chang,R.L., Bhachech,N., Cui,X.X., Merkler,K.A., Wong,C.Q., Hennig,E., Yagi,H., Jerina,D.M. and Conney,A.H. (1993) Dose-dependent differences in the profile of mutations induced (+)-7R,8S-dihydroxy-9S,10R-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene in the coding region of the hypoxanthine (guanine) phosphoribosyltransferase gene in Chinese hampster V79 cells. Cancer Res., 53, 3294–3301.[Abstract/Free Full Text]

    17. Wei,S.J., Chang,R.L., Hennig,E., Cui,X.X., Merkler,K.A., Wong,C.Q., Yagi,H., Jerina,D.M. and Conney,A.H. (1995) Mutagenic selectivity at the HPRT locus in V-79 cells: comparison of mutations caused by bay-region benzo[a]pyrene 7,8-diol-9,10-opxide enantiomers with high and low carcinogenic activity. Carcinogenesis, 15, 1729–1735.

    18. Wei,S.J., Chang,R.L., Wong,C.Q., Cui,X.X., Dandamudi,N., Lu,Y.P., Merkler,K.A., Sayer,J.M., Conney,A.H. and Jerina,D.M. (1999) The ratio of deoxyadenosine to deoxyguanosine adducts formed by (+)-(7R, 8S, 9S, 10R)-7,8-dihydroxy-9,10-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene in purified calf thymus DNA and DNA in V-79 cells is independent of dose. Int. J. Oncol., 14, 509–513[ISI][Medline]

    19. Schiltz,M., Cui,X.X., Lu,Y.P., Yagi,H., Jerina,D.M., Zdzienicka,M.Z., Chang,R.L., Conney,A.H. and Wei,S.J. (1999) Characterization of the mutational profile of (+)-7R, 8S-dihydroxy-9S, 10R-epoxy-7, 8, 9, 10-tetrahydrobenzo[a]pyrene at the hypoxanthine (guanine) phosphoribosyltransferase gene in repair-deficient Chinese hamster V-H1 cells. Carcinogenesis, 20, 2279–2286.[Abstract/Free Full Text]

    20. Nicol,C.J., Harrison,M.L., Laposa,R.R., Gimelshtein,I.L. and Wells,P.G. (1995) A teratologic suppressor role for p53 in benzo[a]pyrene-treated transgenic-deficient mice. Nat. Genet., 10, 181–187.[CrossRef][ISI][Medline]

    21. Chakravarti,D., Pelling,J., Cavalieri,E.L. and Rogan,E.G. (1995) Relating aromatic hydrocarbon-induced DNA adducts and the c-H-ras mutations in mouse skin papillomas: the role of apurinic sites. Proc. Natl Acad. Sci. USA, 92, 10422–10426.[Abstract/Free Full Text]

    22. Chen,L., Devanesan,R., Higginbotham,S., Ariese,F., Jankowiak,R., Small,G.J., Rogan,E.G. and Cavalieri,E. (1996) Expanded analysis of benzo[a]pyrene-DNA adducts formed in vitro and in mouse skin: their significance in tumor initiation. Chem. Res. Toxicol., 9, 897–903.[CrossRef][ISI][Medline]

    23. Balmain,A. and Brown,K. (1988) Oncogene activation in chemical carcinogenesis. Adv. Cancer Res., 51, 147–182.[ISI][Medline]

    24. Ruggeri,B.A., Bauer,B., Zhang,S.-Y. and Klein-Szanto,A.J.P. (1994) Murine squamous cell carcinoma cell lines produced by a complete carcinogenesis protocol with benzo[a]pyrene exhibit characteristic p53 mutations and the absence of H-ras and cyl1/cyclin D1 abnormalities. Carcinogenesis, 15, 6113–1619.

    25. Hall,M. and Grover,P.L. (1990)Chemical carcinogenesis and mutagenesis. In: Cooper,G.S. and Grover,P.L. (eds.), Handbook of Experimental Pharmacology, Vol. 94/1, Springer-Verlag, Heidelberg, pp 327–372.

    26. Denissenko,M.F., Pao,A., Tang,M. and Pfeifer,G.P. (1996) Preferential formation of benzo[a]pyrene adductrs at lung cancer mutational hotspots in p53. Science, 274, 430–432.[Abstract/Free Full Text]

    27. Denissenko,M.F., Pao,A., Pfeifer,G.P. and Tang,M.-S. (1998) Slow repair of bulky DNA adducts along the nontranscribed strand of the human p53 gene may explain the strand bias of transversion mutations in cancers. Oncogene, 16, 1241–1247.[CrossRef][ISI][Medline]

    28. Pfeifer,G.P. and Denissenko,M.F. (1998) Formation and repair of DNA lesions in the p53 gene: relation to cancer mutations? Environ. Mol. Mutagen., 31, 197–205.[CrossRef][ISI][Medline]

    29. Pfeifer,G.P., Denissenko,M.F., Olivier,M., Tretyakova,N., Hecht,S.S. and Hainaut,P. (2002) Tobacco smoke carcinogens, DNA damage and p53 mutations in smoking-associated cancers. Oncogene, 21, 7435–7451.[CrossRef][ISI][Medline]

    30. Smith,L.E., Denissenko,M.F., Bennett,W.P., Li,H., Amin,S., Tang,M. and Pfeifer GP. (2000) Targeting of lung cancer mutational hotspots by polycyclic aromatic hydrocarbons. J. Natl Cancer Inst., 92, 803–811.[Abstract/Free Full Text]

    31. Cheng,S.C., Hilton,B.D., Roman,J.M. and Dipple,A. (1989) DNA adducts from carcinogenic and noncarcinogenic enantiomers of benzo[a]pyrene dihydrodiol epoxide. Chem. Res. Toxicol., 2, 334–340.[CrossRef][ISI][Medline]

    32. Sayer,J.M., Chadha,A., Agarwal,H.S.K., Yeh,H.J.C., Yagi,H. and Jerina,D.M. (1991) Covalent nucleoside adducts of benzo[a]pyrene 7,8-diol 9,10-epoxides: Structural reinvestigation and characterization of a novel adenosine adduct on the ribose moiety. J. Org. Chem., 56, 20–29.[CrossRef]

    33. Phillips,D.H., Hewer,A. and Grover,P.L. (1985) Aberrant activation of benzo[a]pyrene in cultured rat mammary cells in vitro and following direct application to rat mammary glands in vivo. Cancer Res., 45, 4167–4174.[Abstract/Free Full Text]

    34. Marnett,L.J. (1987) Peroxyl free radicals: potential mediators of tumor initiation and promotion. Carcinogenesis, 8, 1365–1373.[Free Full Text]

    35. Devanesan,P.D., RamaKrishna,N.V.S., Todorovic,R., Rogan,E.G., Cavalieri,E.L., Jeong,H., Jankowiak,R. and Small,G.J. (1992) Identification and quantitation of benzo[a]pyrene–DNA adducts formed by rat liver microsomes. Chem. Res. Toxicol., 5, 302–309.[CrossRef][ISI][Medline]

    36. Chakravarti,D., Mailander,P.C., Cavalieri,E.L. and Rogan,E.G. (2000) Evidence that error-prone DNA repair converts dibenzo[a,l]pyrene-induced depurinating lesions into mutations: formation, clonal proliferation and regression of initiated cells carrying H-ras oncogene mutations in early preneoplasia. Mutat. Res., 456, 17–32.[ISI][Medline]

    37. Chakravarti,D., Pelling,J., Cavalieri,E.L. and Rogan,E.G. (1995) Relating aromatic hydrocarbon-induced DNA adducts and the c-H-ras mutations in mouse skin papillomas: the role of apurinic sites. Proc. Natl Acad. Sci. USA, 92, 10422–10426.[Abstract/Free Full Text]

    38. Mackay,W., Benasutti,M., Drouin,E. and Loechler,E.L. (1992) Mutagenesis by the major adduct of activated Benzo[a]pyrene, (+)-anti-BP–N2-Gua, when studied in an Escherichia coli plasmid using site-directed methods. Carcinogenesis, 13, 1415–1425.[Abstract/Free Full Text]

    39. Shukla,R., Liu,Y., Geacintov,N. and Loechler,E.L (1997) The major, N2-dG adduct of (+)-anti-B[a]PDE shows a dramatically different mutagenic specificity (predominantly, G->A) in a 5'-CGT-3' sequence context. Biochemistry, 36, 10256–10261.[CrossRef][Medline]

    40. Shukla,R, Jelinsky,S., Liu,T., Geacintov,N.E. and Loechler,E.L (1997) How stereochemistry affects mutagenesis by N2-dG adducts of B[a]PDE: configuration of the adduct bond is more important than of the hydroxyl groups. Biochemistry, 36, 13263–13269.[CrossRef][Medline]

    41. Shukla,R., Geacintov,N. and Loechler,E.L. (1999) The major, N2-dG adduct of (+)-anti-B[a]PDE induces G->A mutations in a 5'-AGA-3' sequence context. Carcinogenesis, 20, 261–268.[Abstract/Free Full Text]

    42. Jelinsky,S.A., Mao,B., Geacintov,N.E. and Loechler,E.L. (1995) The major, N2-Gua adduct of the (+)-anti-benzo[a]pyrene diol epoxide is capable of inducing G->A and G->C, in addition to G->T, mutations. Biochemistry, 34, 13545–13553.[CrossRef][Medline]

    43. Seo,K.-Y., Jelinsky,S.A. and Loechler,E.L. (2000) Adduct conformational complexity can cause adduct mutational complexity: evidence from mutagenic studies of the potent environmental carcinogen benzo[a]pyrene. Mutat. Res., 463, 215–245.[CrossRef][ISI][Medline]

    44. Loechler,E.L. (1995) How are potent bulky carcinogens able to induce such a diverse array of mutations? Mol. Carcinog., 13, 213–219[ISI][Medline]

    45. Yin,J., Seo,K.-Y. and Loechler,E.L. (2004) A role for DNA polymerase V in G->T mutagenesis from the major benzo[a]pyrene N2-dG adduct when studied in a 5'-TGT sequence in Escherichia coli. DNA Repair, 3, 323–334.[Medline]

    46. Page,J.E., Zajc,B., Oh-hara,T., Lakshman,M.K., Sayer,J.M., Jerina,D.M. and Dipple,A. (1998) Sequence context profoundly influences the mutagenic potency of trans-opened benzo[a]pyrene 7,8-diol 9,10-epoxide-purine nucleoside adducts in site-specific mutation studies. Biochemistry, 37, 9127–9137.[CrossRef][Medline]

    47. Chary,P., Latham,G.J.