Nucleotide incorporation by human DNA polymerase gamma opposite benzo[a]pyrene and benzo[c]phenanthrene diol epoxide adducts of deoxyguanosine and deoxyadenosine

Mitochondria are major cellular targets of benzo[a]pyrene (BaP), a known carcinogen that also inhibits mitochondrial proliferation. Here, we report for the first time the effect of site-specific N²-deoxyguanosine (dG) and N⁶-deoxyadenosine (dA) adducts derived from BaP 7,8-diol 9,10-epoxide (BaP DE) and dA adducts from benzo[c]phenanthrene 3,4-diol 1,2-epoxide (BcPh DE) on DNA replication by exonuclease-deficient human mitochondrial DNA polymerase (pol γ) with and without the p55 processivity subunit. The catalytic subunit alone primarily misincorporated dAMP and dGMP opposite the BaP DE–dG adducts, and incorporated the correct dTMP as well as the incorrect dAMP opposite the DE–dA adducts derived from both BaP and BcPh. In the presence of p55 the polymerase incorporated all four nucleotides and catalyzed limited translesion synthesis past BaP DE–dG adducts but not past BaP or BcPh DE–dA adducts. Thus, all these adducts cause erroneous purine incorporation and significant blockage of further primer elongation. Purine misincorporation by pol γ opposite the BaP DE–dG adducts resembles that observed with the Y family pol η. Blockage of translesion synthesis by these DE adducts is consistent with known BaP inhibition of mitochondrial (mt)DNA synthesis and suggests that continued exposure to BaP reduces mtDNA copy number, increasing the opportunity for repopulation with pre-existing mutant mtDNA and a resultant risk of mitochondrial genetic diseases.     

Translesion replication of benzo[a]pyrene and benzo[c]phenanthrene diol epoxide adducts of deoxyadenosine and deoxyguanosine by human DNA polymerase iota

Human DNA polymerase ι (polι) is a Y-family polymerase whose cellular function is presently unknown. Here, we report on the ability of polι to bypass various stereoisomers of benzo[a]pyrene (BaP) diol epoxide (DE) and benzo[c]phenanthrene (BcPh) DE adducts at deoxyadenosine (dA) or deoxyguanosine (dG) bases in four different template sequence contexts in vitro. We find that the BaP DE dG adducts pose a strong block to polι-dependent replication and result in a high frequency of base misincorporations. In contrast, misincorporations opposite BaP DE and BcPh DE dA adducts generally occurred with a frequency ranging between 2 × 10^–3 and 6 × 10^–4. Although dTMP was inserted efficiently opposite all dA adducts, further extension was relatively poor, with one exception (a cis opened adduct derived from BcPh DE) where up to 58% extension past the lesion was observed. Interestingly, another human Y-family polymerase, polκ, was able to extend dTMP inserted opposite a BaP DE dA adduct. We suggest that polι might therefore participate in the error-free bypass of DE-adducted dA in vivo by predominantly incorporating dTMP opposite the damaged base. In many cases, elongation would, however, require the participation of another polymerase more specialized in extension, such as polκ.     

Inhibition of Werner syndrome helicase activity by benzo[c]phenanthrene diol epoxide dA adducts in DNA is both strand-and stereoisomer-dependent

Helicases are among the first enzymes to encounter DNA damage during DNA processing within the cell and thus are likely to be targets for the adverse effects of DNA lesions induced by environmental chemicals. Here we examined the effect of cis- and trans-opened 3,4-diol 1,2-epoxide (DE) DNA adducts of benzo[c]phenanthrene (BcPh) at N6 of adenine on helicase activity. These adducts are derived from the highly tumorigenic (-)-(1R,2S,3S,4R)-DE as well as its less carcinogenic (+)-(1S,2R,3R,4S)-DE enantiomer in both of which the benzylic 4-hydroxyl group and epoxide oxygen are trans. The hydrocarbon portions of these adducts intercalate into DNA on the 3' or the 5' side of the adducted deoxyadenosine for the 1S- and 1R-adducts, respectively. These adducts inhibited the human Werner (WRN) syndrome helicase activity in a strand-specific and stereospecific manner. In the strand along which WRN translocates, cis-opened adducts were significantly more effective inhibitors than trans-opened isomers, indicating that WRN unwinding is sensitive to adduct stereochemistry. WRN helicase activity was also inhibited but to a lesser extent by cis-opened BcPh DE adducts in the displaced strand independent of their direction of intercalation, whereas inhibition by the trans-opened stereoisomers in the displaced strand depended on their orientation, such that only adducts oriented toward the advancing helicase inhibited WRN activity. A BcPh DE adduct positioned in the helicase-translocating strand did not sequester WRN, nor affect the rate of ATP hydrolysis relative to an unadducted control. Although the Bloom (BLM) syndrome helicase was also inhibited by a cis-opened adduct in a strand-specific manner, this helicase was not as severely affected as WRN. Because BcPh DEs form substantial amounts of deoxyadenosine adducts at dA, their adverse effects on helicases could contribute to genetic damage and cell transformation induced by these DEs. Thus, the unwinding activity of RecQ helicases is sensitive to the strand, orientation, and stereochemistry of intercalated polycyclic aromatic hydrocarbon adducts.     

Factors determining mutagenic potential for individual cis and trans opened benzo[c]phenanthrene diol epoxide-deoxyadenosine adducts

Four adducts that would result from trans opening at C-1 of benzo[c]phenanthrene 3,4-diol 1,2-epoxide (B[c]PhDE) isomers (i.e., DE-1 enantiomers, where the epoxide oxygen and benzylic hydroxyl group are cis, and DE-2 enantiomers, where they are trans) by the N(6)-amino group of dAdo, together with the two cis opened N(6)-dAdo adducts of B[c]PhDE-1, were incorporated into two oligonucleotides at the underlined site in 5'-TTTAGAGTCTGCTCCC [context I(A)] and 5'-CAGATTTAGAGTCTGC [context II(A)]. After ligation of these, and the corresponding unsubstituted oligonucleotides, into single-stranded M13mp7L2 bacteriophage and transfection into SOS-induced Escherichia coli SMH77, base substitution mutations induced by the different B[c]PhDE-dAdo adducts were determined. These findings were compared with data [Pontén et al. (1999) Biochemistry 38, 1144-1152] for cis opened B[c]PhDE-2-dAdo adducts in the same sequence contexts. In most cases, adducts with S absolute configuration at the site of attachment of the nucleoside to the hydrocarbon had higher mutation frequencies (1.9-56.5%) than the corresponding adducts with R configuration (0.05-5.6%). For adducts derived from B[c]PhDE-1, the predominant mutations were A-->T transversions in context I(A) and A-->G transitions for most of these adducts in context II(A). For adducts derived from B[c]PhDE-2, A-->T base substitutions predominated for most of the trans adducts, but A-->G mutations were favored by the cis adduct with S configuration in either context. Thus, the structural feature that most dramatically affected mutagenic activity was the configuration of the carbon at the attachment point, with S configuration mostly being associated with greater mutagenicity than the R configuration. However, other structural variations and sequence context also affected mutagenicity, indicating that a combination of structure and context effects define mutagenicity.     

Benzo[c]phenanthrene adducts and nogalamycin inhibit DNA transesterification by vaccinia topoisomerase

Vaccinia DNA topoisomerase forms a covalent DNA-(3'-phosphotyrosyl)-enzyme intermediate at a specific target site 5'-C(+5)C(+4)C(+3)T(+2)T(+1)p downward arrow N(-1) in duplex DNA. Here we study the effects of position-specific DNA intercalators on the rate and extent of single-turnover DNA transesterification. Chiral C-1 R and S trans-opened 3,4-diol 1,2-epoxide adducts of benzo[c]phenanthrene (BcPh) were introduced at single N2-deoxyguanosine and N6-deoxyadenosine positions within the 3'-G(+5)G(+4)G(+3)A(+2)A(+1)T(-1)A(-2) sequence of the nonscissile DNA strand. Transesterification was unaffected by BcPh intercalation between the +6 and +5 base pairs, slowed 4-fold by intercalation between the +5 and +4 base pairs, and virtually abolished by BcPh intercalation between the +4 and +3 base pairs and the +3 and +2 base pairs. Intercalation between the +2 and +1 base pairs by the +2R BcPh dA adduct abolished transesterification, whereas the overlapping +1S BcPh dA adduct slowed the rate of transesterification by a factor of 2700, with little effect upon the extent of the reaction. Intercalation at the scissile phosphodiester (between the +1 and -1 base pairs) slowed transesterification by a factor of 450. BcPh intercalation between the -1 and -2 base pairs slowed cleavage by two orders of magnitude, but intercalation between the -2 and -3 base pairs had little effect. The anthracycline drug nogalamycin, a non-covalent intercalator with preference for 5'-TG dinucleotides, inhibited the single-turnover DNA cleavage reaction of vaccinia topoisomerase with an IC50 of 0.7 microM. Nogalamycin was most effective when the drug was pre-incubated with DNA and when the cleavage target site was 5'-CCCTT/G instead of 5'-CCCTT/A. These findings demarcate upstream and downstream boundaries of the functional interface of vaccinia topoisomerase with its DNA target site.     

32P-Postlabeling high-performance liquid chromatographic improvements to characterize DNA adduct stereoisomers from benzo[a]pyrene and benzo[c]phenanthrene,and to separate DNA adducts from 7,12-dimethylbenz[a]anthracene

Single compounds can generate complex DNA adduct patterns by reactions through different pathways, with different target nucleotides and through different configurations of the products. DNA adduct analysis by ³²P-HPLC was improved by adding an isocratic plateau in an otherwise linear gradient, thereby enhancing resolution of predictable retention time intervals. This enhanced ³²P-HPLC technique was used to analyze and at least partly resolve 14 out of 16 available benzo[c]phenanthrene deoxyadenosine and deoxyguanosine adduct standards, 8 out of 8 available benzo[a]pyrene deoxyadenosine and deoxyguanosine adduct standards, and 51 peaks from 7,12-dimethylbenz[a]anthracene-calf thymus DNA reaction products. The same type of gradient modifications could be used to enhance resolution in analyses of other complex DNA adduct mixtures, e.g., in vivo in humans.     

DNA polymerase eta participates in the mutagenic bypass of adducts induced by benzo[a]pyrene diol epoxide in mammalian cells

Y-family DNA-polymerases have larger active sites that can accommodate bulky DNA adducts allowing them to bypass these lesions during replication. One member, polymerase eta (pol eta), is specialized for the bypass of UV-induced thymidine-thymidine dimers, correctly inserting two adenines. Loss of pol eta function is the molecular basis for xeroderma pigmentosum (XP) variant where the accumulation of mutations results in a dramatic increase in UV-induced skin cancers. Less is known about the role of pol eta in the bypass of other DNA adducts. A commonly encountered DNA adduct is that caused by benzo[a]pyrene diol epoxide (BPDE), the ultimate carcinogenic metabolite of the environmental chemical benzo[a]pyrene. Here, treatment of pol eta-deficient fibroblasts from humans and mice with BPDE resulted in a significant decrease in Hprt gene mutations. These studies in mammalian cells support a number of in vitro reports that purified pol eta has error-prone activity on plasmids with site-directed BPDE adducts. Sequencing the Hprt gene from this work shows that the majority of mutations are G>T transversions. These data suggest that pol eta has error-prone activity when bypassing BPDE-adducts. Understanding the basis of environmental carcinogen-derived mutations may enable prevention strategies to reduce such mutations with the intent to reduce the number of environmentally relevant cancers.     

Excision of benzo[a]pyrene diol epoxide I adducts from nucleosomal DNA of confluent normal human fibroblasts

The formation and removal of covalent adducts of racemic 7 beta, 8 alpha-dihydroxy-9 alpha,10 alpha-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (BPDE I) was studied in nucleosomal DNA of confluent cultures of normal human fibroblasts (NF). For this purpose NF were prelabeled in their DNA with [14C]-thymidine and treated with [3H]BPDE I. The adducts were composed of 77% (7R)-N2-(7 beta, 8 alpha, 9 alpha-trihydroxy-7,8,9,10-tetrahydrobenzo[a]pyrene-10-yl)deoxyguanosine, 12% of the corresponding 7S-enantiomer and of minor amounts of adducts to cytosine and adenine. The adduct composition did not change significantly in 24-h post treatment incubation. Bulk mononucleosomes were prepared from micrococcal nuclease digested nuclei and their DNA analyzed by gel electrophoresis. The adduct concentrations were determined in 145 base pair (b.p.) nucleosomal core-DNA, 165 b.p. chromatosomal DNA and in total nuclear DNA. From these data the concentration in nucleosomal linker-DNA was calculated. The initial adduct distribution was non-random and 6.3 times higher in 47 b.p. linker-DNA relative to 145 b.p. core-DNA and 9.2 times higher in 27 b.p. linker-DNA relative to 165 b.p. chromatosomal DNA. Adduct removal was very rapid during the first 8 h and more efficient from linker-DNA than from core-DNA. After this early phase the adducts located in 145 b.p. core-DNA became refractory to further excision and represent a major fraction of the adducts persisting in DNA of NF over a prolonged period. In contrast, further adduct removal was observed from nucleosomal linker-DNA.     

Inhibition of SV40 DNA replication by benzo[a]pyrene diol epoxide adducts: two recovery modes

Anti-benzo[a]pyrene diol epoxide (BPDE) adducts produced in vitro in SV40 initially inhibit SV40 DNA replication in vivo, in cells unexposed to BPDE. A single adduct in a replicon is probably sufficient to block DNA replication. The recovery process appears to begin immediately after infection. The rate of recovery of replicative capacity is inversely related to the initial adduct number. Holding the infected cells temporarily under conditions that prevent viral DNA replication results subsequently in increased recovery, proportional to the holding time. The mechanism of recovery appears to be constitutive and prereplicative. In addition, there is a second mode of recovery which is induced by pretreatment of the host cells with BPDE before infection. The effect of pretreatment is similar to that of extending the holding time before replication: the first molecules begin to replicate earlier but the subsequent rate of recovery is unchanged. The induced mechanism may be either a limited stoichiometric repair process or a slow replicative bypass.     

Benzo[a]pyrene diol epoxide-deoxyguanosine adducts are accurately bypassed by yeast DNA polymerase zeta in vitro

The possible role of bypass DNA polymerase zeta in mutagenic translesion synthesis past benzo[a]pyrene (BP) 7,8-diol-9,10-epoxide (DE) N(2)-deoxyguanosine (dG) adducts has been examined. We prepared 59-mer DNA templates containing dG adducts derived from trans opening of enantiomers of BP DE-2, in which the 7-hydroxyl group and epoxide oxygen are trans. The 10S-BP DE-dG and 10R-BP DE-dG adducts derive from the (+)- and (-)-DE-2 enantiomers, respectively. The adducted dG is located at a site identified as a G-->T mutational hotspot in random mutagenesis studies of (+)-BP DE-2 in Chinese hamster V-79 cells. Yeast pol zeta (complex of Gst-Rev3p and Rev7p) formed extension products (total of all lengths) of 71, 74 and 88% of a primer annealed to the 10S-BP DE-dG, 10R-BP DE-dG and non-adducted 59-mer templates, respectively. However, only 18 and 19% of the primer was extended to the full-length product on 10S-BP DE-dG and 10R-BP DE-dG adducted templates compared to 55% of the primer on the non-adducted template. A major 34-mer product corresponding to primer elongation up to and including the base before the adduct indicated that nucleotide incorporation opposite both adducts was strongly blocked. Full-length products were isolated from gels and subjected to PCR amplification and cloning. Sequence analysis of more than 300 clones of these full-length products on each template showed that only the correct dCMP was incorporated opposite both the adducted and non-adducted G-hotspot in the template. This corresponds to a probability of mutation lower than 0.3%, the limit of detection, and demonstrates the remarkable fidelity of yeast pol zeta in translesion synthesis past these BP DB-dG lesions in vitro.     

H2AX phosphorylation in A549 cells induced by the bulky and stable DNA adducts of benzo[a]pyrene and dibenzo[a,l]pyrene diol epoxides

Early events in the cellular response to DNA damage, such as double strand breaks, rely on lesion recognition and activation of proteins involved in maintenance of genomic stability. One important component of this process is the phosphorylation of the histone variant H2AX. To investigate factors explaining the variation in carcinogenic potency between different categories of polycyclic aromatic hydrocarbons (PAHs), we have studied the phosphorylation of H2AX (H2AXγ). A549 cells were exposed to benzo[a]pyrene diol epoxide [(+)-anti-BPDE] (a bay-region PAH) and dibenzo[a,l]pyrene diol epoxide [(−)-anti-DBPDE] (a fjord-region PAH) and H2AXγ was studied using immunocytochemistry and Western blot. Hydrogen peroxide (H₂O₂) was used to induce oxidative DNA damage and strand breaks. As showed with single cell gel electrophoresis, neither of the diol epoxides resulted in DNA strand breaks relative to H₂O₂. Visualisation of H2AXγ formation demonstrated that the proportion of cells exhibiting H2AXγ staining at 1 h differed between BPDE, 40% followed by a decline, and DBPDE, <10% followed by an increase. With H₂O₂ treatment, almost all cells demonstrated H2AXγ at 1 h. Western blot analysis of the H2AXγ formation also showed concentration and time-dependent response patterns. The kinetics of H2AXγ formation correlated with the previously observed kinetics of elimination of BPDE and DBPDE adducts. Thus, the extent of H2AXγ formation and persistence was related to both the number of adducts and their structural features.     

Phenylalanine 171 is a molecular brake for translesion synthesis across benzo[a]pyrene-guanine adducts by human DNA polymerase kappa

Human cells possess multiple specialized DNA polymerases (Pols) that bypass a variety of DNA lesions which otherwise would block chromosome replication. Human polymerase kappa (Pol κ) bypasses benzo[a]pyrene diolepoxide-N(2)-deoxyguanine (BPDE-N(2)-dG) DNA adducts in an almost error-free manner. To better understand the relationship between the structural features in the active site and lesion bypass by Pol κ, we mutated codons corresponding to amino acids appearing close to the adducts in the active site, and compared bypass efficiencies. Remarkably, the substitution of alanine for phenylalanine 171 (F171), an amino acid conserved between Pol κ and its bacterial counterpart Escherichia coli DinB, enhanced the efficiencies of dCMP incorporation opposite (-)- and (+)-trans-anti-BPDE-N(2)-dG 18-fold. This substitution affected neither the fidelity of TLS nor the efficiency of dCMP incorporation opposite normal guanine. This amino acid change also enhanced the binding affinity of Pol κ to template/primer DNA containing (-)-trans-anti-BPDE-N(2)-dG. These results suggest that F171 functions as a molecular brake for TLS across BPDE-N(2)-dG by Pol κ and that the F171A derivative of Pol κ bypasses these DNA lesions more actively than does the wild-type enzyme.     

In vitro and in vivo modulations of benzo[c]phenanthrene-DNA adducts by DNA mismatch repair system

Benzo[c]phenanthrene dihydrodiol epoxide (B[c] PhDE) is well known as an important environmental chemical carcinogen that preferentially modifies DNA in adenine residues. However, the molecular mechanism by which B[c]PhDE induces tumorigenesis is not fully understood. In this report, we demonstrate that DNA mismatch repair (MMR), a genome maintenance system, plays an important role in B[c]PhDE-induced carcinogensis by promoting apoptosis in cells treated with B[c]PhDE. We show that purified human MMR recognition proteins, MutSα and MutSβ, specifically recognized B[c]PhDE-DNA adducts. Cell lines proficient in MMR exhibited several-fold more sensitivity to killing than cell lines defective in either MutSα or MutLα by B[c]PhDE; the nature of this sensitivity was shown to be due to increased apoptosis. Additionally, wild-type mice exposed to B[c]PhDE had intestinal crypt cells that underwent apoptosis significantly more often than intestinal crypt cells found in B[c]PhDE-treated Msh2^–/– or Mlh1^–/– mice. These findings, combined with previous studies, suggest that the MMR system may serve as a general sensor for chemical-caused DNA damage to prevent damaged cells from mutagenesis and carcinogenesis by promoting apoptosis.     

Sequence preferences of covalent DNA binding by anti-(+)- and anti-(-)-benzo[a]pyrene diol epoxides

The sequence preferences of formation of piperidine-labile adducts of guanine by individual (+)- and (-)-isomers of trans-7,8-dihydroxy-anti-9,10-epoxy-7,8,9,10- tetrahydrobenzo[a]pyrene [anti-(+)- and anti-(-)-BPDE] were examined by techniques analogous to chemical DNA sequencing. Data were obtained on over 1200 bases with anti-(-)-BPDE and 1000 bases with anti-(+)-BPDE. Guanines on average yielded more labile adducts than other bases, and the reactivities of guanines with both anti-(+)- and anti-(-)-BPDE isomers were found to be distinctly nonrandom with respect to DNA sequence. The most and least reactive guanines, defined in terms of the upper and lower 10 percentiles of reactivity, differed on average by a factor of 17. This range of guanine reactivities was correlated with distinct sequence preferences, which differed in part for the two isomers. The strongest determinant for preferred reaction of anti-(-)-BPDE to form a labile adduct at a guanine was the presence of a 3'-flanking guanine, but a thymine 5'-flanking a guanine also generally enhanced reactivity. The triplets containing central guanines most preferred by anti-(-)-BPDE were AGG, CGG, and TG(G greater than T greater than C,A). anti-(+)-BPDE also formed labile adducts preferentially at AGG and CGG triplets, but not at TGN triplets. Significant effects of next-nearest-neighbor bases on guanine reactivities were also noted.     

P-Postlabeling high-performance liquid chromatographic improvements to characterize DNA adduct stereoisomers from benzo[a]pyrene and benzo[c]phenanthrene, and to separate DNA adducts from 7,12-dimethylbenz[a]anthracene

Single compounds can generate complex DNA adduct patterns by reactions through different pathways, with different target nucleotides and through different configurations of the products. DNA adduct analysis by ³²P-HPLC was improved by adding an isocratic plateau in an otherwise linear gradient, thereby enhancing resolution of predictable retention time intervals. This enhanced ³²P-HPLC technique was used to analyze and at least partly resolve 14 out of 16 available benzo[c]phenanthrene deoxyadenosine and deoxyguanosine adduct standards, 8 out of 8 available benzo[a]pyrene deoxyadenosine and deoxyguanosine adduct standards, and 51 peaks from 7,12-dimethylbenz[a]anthracene-calf thymus DNA reaction products. The same type of gradient modifications could be used to enhance resolution in analyses of other complex DNA adduct mixtures, e.g., in vivo in humans.     

Preferential misincorporation of purine nucleotides by human DNA polymerase eta opposite benzo[a]pyrene 7,8-diol 9,10-epoxide deoxyguanosine adducts

Human DNA polymerase eta was used to copy four stereoisomeric deoxyguanosine (dG) adducts derived from benzo[a]pyrene 7,8-diol 9,10-epoxide (diastereomer with the 7-hydroxyl group and epoxide oxygen trans (BaP DE-2)). The adducts, formed by either cis or trans epoxide ring opening of each enantiomer of BaP DE-2 by N(2) of dG, were placed at the fourth nucleotide from the 5'-end in two 16-mer sequence contexts, 5' approximately CG*A approximately and 5' approximately GG*T. poleta was remarkably error prone at all four diol epoxide adducts, preferring to misincorporate G and A at frequencies 3- to more than 50-fold greater than the frequencies for T or the correct C, although the highest rates were 60-fold below the rate of incorporation of C opposite a non-adducted G. Anti to syn rotation of the adducted base, consistent with previous NMR data for a BaP DE-2 dG adduct placed just beyond a primer terminus, provides a rationale for preferring purine misincorporation. Extension of purine misincorporations occurred preferentially, but extension beyond the adduct site was weak with V(max)/K(m) values generally 10-fold less than for misincorporation. Mostly A was incorporated opposite (+)-BaP DE-2 dG adducts, which correlates with published observations that G --> T is the most common type of mutation that (+)-BaP DE-2 induces in mammalian cells.     

In vivo evidence that phenylalanine 171 acts as a molecular brake for translesion DNA synthesis across benzo[a]pyrene DNA adducts by human DNA polymerase κ

Humans possess multiple specialized DNA polymerases that continue DNA replication beyond a variety of DNA lesions. DNA polymerase kappa (Pol κ) bypasses benzo[a]pyrene diolepoxide-N²-deoxyguanine (BPDE-N²-dG) DNA adducts in an almost error-free manner. In the previous work, we changed the amino acids close to the adducts in the active site and examined the bypass efficiency. The substitution of alanine for phenylalanine 171 (F171A) enhanced by 18-fold in vitro, the efficiencies of dCMP incorporation opposite (−)- and (+)-trans-anti-BPDE-N²-dG. In the present study, we established human cell lines that express wild-type Pol κ (POLK+/−), F171A (POLK F171A/−) or lack expression of Pol κ (POLK−/−) to examine the in vivo significance. These cell lines were generated with Nalm-6, a human pre-B acute lymphoblastic leukemia cell line, which has high efficiency for gene targeting. Mutations were analyzed with shuttle vectors having (−)- or (+)-trans-anti-BPDE-N²-dG in the supF gene. The frequencies of mutations were in the order of POLK−/− > POLK+/− > POLK F171A/− both in (−)- and (+)-trans-anti-BPDE-N²-dG. These results suggest that F171 may function as a molecular brake for bypass across BPDE-N²-dG by Pol κ and raise the possibility that the cognate substrates for Pol κ are not BP adducts in DNA but may be lesions in DNA induced by endogenous mutagens.