Reactive oxygen species (ROS) are generated during normal metabolic processes and their levels increase in response to environmental stress. Exposure of DNA to ROS results in oxidative DNA damage and the generation of multiple types of DNA lesions. Left unrepaired these lesions contribute to mutagenesis, cancer and human disease (1
). One of the most abundant lesions caused by oxidative DNA damage in both the DNA and nucleotide pools is 8-oxo-7,8-dihydro-2′-deoxyguanosine (8-oxoG). At physiological pH, 8-oxoG has a carbonyl at C8 and is protonated at N7 (2
). These modifications to guanine lead to additional hydrogen bonding interactions at the Hoogsteen edge and promote rotation about its glycosidic bond thereby encouraging both anti-
and syn-conformations. This rotation about the glycosidic bond to the syn-conformation is further facilitated by a backbone clash between the adducted oxygen (O8) and the sugar–phosphate backbone in the anti-8-oxoG conformation (3
The coding potential of 8-oxoG is dictated by the anti- or syn-conformation of the modified base. Similar to unmodified deoxyguanine, anti-8-oxoG base pairs with cytosine through traditional Watson–Crick hydrogen bonding interactions forming a non-mutagenic DNA lesion. In contrast, the mutagenic syn-8-oxoG conformation is able to base pair with adenine through its Hoogsteen edge. Kinetic studies have shown that DNA polymerases insert adenine opposite 8-oxoG frequently and with enhanced catalytic efficiency when compared with guanine (4
). When compared with guanine, the catalytic efficiency of adenine insertion opposite 8-oxoG increases almost three orders of magnitude. If left unrepaired, the pro-mutagenic syn-8oxoG-dA base pair results in a G–C to T–A transversion mutation following DNA replication (5
). This highlights the importance of the dual-coding properties of 8-oxoG being either error free or mutagenic when it is in the anti- or syn-conformations, respectively.
The prevalence of oxidative stress and subsequent mutagenic properties of 8-oxoG has resulted in an elegant cellular defense mechanism against 8-oxoG (4
). This lesion is removed from DNA by a glycosylase that generates a gap in the DNA following removal of the damaged base and subsequent initiation of base excision repair (BER). DNA polymerase (pol) β is required to fill the single-nucleotide-gapped DNA intermediate. In the repair of the 8oxoG-C base pair, OGG1 glycosylase removes 8-oxoG resulting in a gapped DNA substrate with a templating cytosine that will be filled with high fidelity by pol β. When 8-oxoG escapes repair, there is a high probability that replication will result in adenine insertion. Thus, the cell also codes for a glycosylase, MYH, which removes adenine paired with 8-oxoG and initiates BER. In this situation, pol β encounters 8-oxoG as the templating nucleotide and will insert dCMP or dAMP opposite 8-oxoG resulting in error-free or mutagenic gap filling, respectively. Kinetic studies with a templating 8-oxoG have shown that pol β only incorporates dCTP over dATP by ~2-fold (6
). This poor 8-oxoG fidelity results in futile mutagenic ‘repair’ following insertion of dATP opposite 8-oxoG. Accordingly, mutagenic repair would be a cellular burden during times of elevated metabolic or environmental stress. These examples emphasize the importance of BER and pol β in modulating the repair of oxidative DNA damage and lesion bypass.
Currently, pol β is the only wild-type DNA polymerase that has been structurally characterized with a templating 8-oxoG base paired with either an incoming dCTP or dATP (7
). Overall these structures indicate that the 8-oxoG lesion paired with dCTP or dATP is well tolerated in the polymerase active site. The structure with an incoming dCTP paired with anti-8-oxoG shows the greatest structural change at the phosphate backbone of the anti-8-oxoG, shifting 2.9 Å and rotating 200° to accommodate the carbonyl oxygen at C8 position (A). This movement is a result of a steric clash between O8 and the phosphate backbone. Recently, the structure of pol β with a templating syn-8-oxoG Hoogsteen base pairing with dATP was determined and indicates that it is similar to that with an incoming dCTP (7
). With the syn-8-oxoG structure, however, Arg283 is observed to stabilize the syn-conformation through a hydrogen bond to O8 in the DNA minor groove (B). With guanine or anti-8-oxoG as the templating base, Arg283 interacts with the templating nucleotide upstream of the coding base. Importantly, Arg283 only interacts with the templating strand after nucleotide binding that induces subsequent polymerase subdomain repositioning.
Figure 1. Structural overview of wild-type pol β with templating 8-oxoG in the active site. (A) Overlay of the templating anti-G:dCMPCPP (PDB ID 2FMP), anti-8-oxoG:dCMP(CF2)PP (PDB ID 3RJI) and syn-8-oxoG:dAMPCPP (PDB ID 3RJF) in the pol β active (more ...)
Combining rational site-directed mutagenesis with kinetic and structural approaches, we are able to identify DNA minor groove interactions that modulate the coding potential of 8-oxoG. These results provide mechanistic insights into substrate specificity, lesion bypass and the role that the polymerase and incoming nucleotide have on the anti-/syn-equilibrium of 8-oxoG. In addition to the mechanistic information obtained, we have also gained insight concerning the impact the incoming nucleotide has on the conformation of the templating 8-oxoG and the point at which the 8-oxoG switches between the syn- or anti-conformations.