Error-free lesion bypass and error-prone lesion bypass are important cellular responses to DNA damage during replication, both of which require a DNA polymerase (Pol). To identify lesion bypass DNA polymerases, we have purified human Polkappa encoded by the DINB1 gene and examined its response to damaged DNA templates. Here, we show that human Polkappa is a novel lesion bypass polymerase in vitro. Purified human Polkappa efficiently bypassed a template 8-oxoguanine, incorporating mainly A and less frequently C opposite the lesion. Human Polkappa most frequently incorporated A opposite a template abasic site. Efficient further extension required T as the next template base, and was mediated mainly by a one-nucleotide deletion mechanism. Human Polkappa was able to bypass an acetylaminofluorene-modified G in DNA, incorporating either C or T, and less efficiently A opposite the lesion. Furthermore, human Polkappa effectively bypassed a template (-)-trans-anti-benzo[a]pyrene-N:(2)-dG lesion in an error-free manner by incorporating a C opposite the bulky adduct. In contrast, human Polkappa was unable to bypass a template TT dimer or a TT (6-4) photoproduct, two of the major UV lesions. These results suggest that Polkappa plays an important role in both error-free and error-prone lesion bypass in humans.

The effectiveness of in vitro primer elongation reactions catalyzed by human bypass DNA polymerases κ (hDinB1), pol η (hRad30A), pol ι (hRad30B), and yeast pol ζ (Rev3 and Rev7) in site-specifically modified template oligonucleotide strands were studied in vitro. The templates contained single bulky lesions derived from the trans-addition of the mutagenic (+)- or (−)-enantiomers ofr7,t8-dihydroxy-t9,10-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (a metabolite of the environmental carcinogen benzo[a]pyrene), to the exocyclic amino groups of guanine or adenine in oligonucleotide templates 33, or more, bases long. In “running start” primer extension reactions, pol κ effectively bypassed both the stereoisomeric (+)- and (−)-trans-guanine adducts but not the analogous adenine adducts. In sharp contrast, pol η, which exhibits considerable sequence homology with pol κ (both belong to the group of Y family polymerases), is partially blocked by the guanine adducts and the (−)-trans-adenine adduct, although the stereoisomeric (+)-trans-adenine adduct is more successfully bypassed. Neither pol ι nor pol ζ, either alone or in combination, were effective in trans-lesion synthesis past the same adducts. In all cases, the fidelity of insertion is dependent on adduct stereochemistry and structure. Generally, error-free nucleotide insertion opposite the lesions tends to depend more on adduct stereochemistry than error-prone insertion. None of the polymerases tested are a universal bypass polymerase for the stereoisomeric bulky polycyclic aromatic hydrocarbon-DNA adducts derived from anti-BPDE.

DNA polymerase activity is essential for replication, recombination, repair, and mutagenesis. All DNA polymerases studied so far from any biological source synthesize DNA by the Watson-Crick base-pairing rule, incorporating A, G, C, and T opposite the templates T, C, G, and A, respectively. Non-Watson-Crick base pairs would lead to mutations. In this report, we describe the ninth human DNA polymerase, Pol(iota), encoded by the RAD30B gene. We show that human Pol(iota) violates the Watson-Crick base-pairing rule opposite template T. During base selection, human Pol(iota) preferred T-G base pairing, leading to G incorporation opposite template T. The resulting T-G base pair was less efficiently extended by human Pol(iota) compared to the Watson-Crick base pairs. Consequently, DNA synthesis frequently aborted opposite template T, a property we designated the T stop. This T stop restricted human Pol(iota) to a very short stretch of DNA synthesis. Furthermore, kinetic analyses show that human Pol(iota) copies template C with extraordinarily low fidelity, misincorporating T, A, and C with unprecedented frequencies of 1/9, 1/10, and 1/11, respectively. Human Pol(iota) incorporated one nucleotide opposite a template abasic site more efficiently than opposite a template T, suggesting a role for human Pol(iota) in DNA lesion bypass. The unique features of preferential G incorporation opposite template T and T stop suggest that DNA Pol(iota) may additionally play a specialized function in human biology.

DNA is frequently damaged by various physical and chemical agents. DNA damage can lead to mutations during replication. In the yeast Saccharomyces cerevisiae, the damage-induced mutagenesis pathway requires the Rev1 protein. We have isolated a human cDNA homologous to the yeast REV1 gene. The human REV1 cDNA consists of 4255 bp and codes for a protein of 1251 amino acid residues with a calculated molecular weight of 138 248 Da. The human REV1 gene is localized between 2q11.1 and 2q11.2. We show that the human REV1 protein is a dCMP transferase that specifically inserts a dCMP residue opposite a DNA template G. In addition, the human REV1 transferase is able to efficiently and specifically insert a dCMP opposite a DNA template apurinic/apyrimidinic (AP) site or a uracil residue. These results suggest that the REV1 transferase may play a critical role during mutagenic translesion DNA synthesis bypassing a template AP site in human cells. Consistent with its role as a fundamental mutagenic protein, the REV1 gene is ubiquitously expressed in various human tissues.

DNA lesion bypass is an important cellular response to genomic damage during replication. Human DNA polymerase eta (Pol(eta)), encoded by the Xeroderma pigmentosum variant (XPV) gene, is known for its activity of error-free translesion synthesis opposite a TT cis-syn cyclobutane dimer. Using purified human Pol(eta), we have examined bypass activities of this polymerase opposite several other DNA lesions. Human Pol(eta) efficiently bypassed a template 8-oxoguanine, incorporating an A or a C opposite the lesion with similar efficiencies. Human Pol(eta) effectively bypassed a template abasic site, incorporating an A and less frequently a G opposite the lesion. Significant -1 deletion was also observed when the template base 5' to the abasic site is a T. Human Pol(eta) partially bypassed a template (+)-trans-anti-benzo[a]pyrene-N:(2)-dG and predominantly incorporated an A, less frequently a T, and least frequently a G or a C opposite the lesion. This specificity of nucleotide incorporation correlates well with the known mutation spectrum of (+)-trans-anti-benzo[a]pyrene-N:(2)-dG lesion in mammalian cells. These results show that human Pol(eta) is capable of error-prone translesion DNA syntheses in vitro and suggest that Pol(eta) may bypass certain lesions with a mutagenic consequence in humans.