Under optimal conditions, the fidelity of DNA replication is extremely high; on average, only one error occurs for every 1010 bases replicated. Unfortunately, optimal conditions rarely occur in vivo as most living organisms are continually subjected to a variety of chemicals, both synthetic and natural, that damage their DNA. Although many organisms have evolved elaborate repair processes to deal with this damage, under certain conditions not all of the damage can be processed by error-free repair mechanisms. As a result, the damaged DNA is replicated, but with much lower fidelity. This process is commonly referred to as translesion DNA synthesis (TLS) or translesion replication (TR).

Until recently, it was believed that the proteins involved in translesion replication were merely accessory proteins, somehow coercing each organism's main replicase to bypass lesions that might otherwise pose as blocks to continued genome duplication. However, in the past couple of years, our understanding of this process has changed dramatically, and it is now known that the key participants in this process are bona fide DNA polymerases that facilitate direct lesion bypass rather than somehow modifying the fidelity of an existing polymerase.

Phylogenetic analysis of these polymerases suggests that they can be broadly subdivided into four groups typified by Escherichia coli UmuC, E. coli DinB, Saccharomyces cerevisiae Rev1, and the S. cerevisiae Rad30 protein, all of which have been collectively classified as the Y-family DNA polymerases.

The general aim of this project is to understand the mechanisms by which mutations are introduced into damaged DNA. Historically, most of our efforts have focused on Escherichia coli, primarily because its cellular, molecular, and genetic characterization was the most advanced of any organism. However, within the past few years, there has been considerable progress in understanding this process in higher organisms, including Saccharomyces cerevisiae and human cells. As a consequence, we have investigated this fascinating process in all three kingdoms of life: bacteria, archea, and eukaryotic cells.

Translesion Replication in Escherichia coli
Woodgate, in collaboration with Myron F. Goodman (University of Southern California)
In E. coli, efficient translesion replication occurs only when the UmuC protein physically interacts with a dimer of UmuD' to form a heterotrimeric complex of UmuD'2C, that is known as E. coli pol V. Because pol V is a low-fidelity enzyme, its activities within the cell are strictly controlled at the transcriptional level as well as at multiple posttranslational steps. For example, recent experiments demonstrated that the in vitro catalytic activity of pol V is greatly stimulated through a physical interaction with the RecA protein. RecA normally binds to regions of single-stranded DNA that are generated at the site of DNA damage and, in doing so, blocks genome duplication by the cell's main replicase, pol III. However, studies within the past year revealed that unlike pol III, pol V efficiently catalyzes RecA filament disassembly in the 3' to 5' direction with an activity that has been likened to a "locomotive cow-catcher." Concurrent ATP hydrolysis-driven filament disassembly occurs in the opposite direction. The bidirectional collapse of the RecA filament and the concomitant decrease in pol V's enzymatic activity therefore provide a mechanism whereby the cell can restrict the generation of pol V-dependent untargeted mutations in undamaged DNA.

Y-Family Polymerases in Archea
Boudsocq, Woodgate in collaboration with Hong Ling and Wei Yang (NIDDK, NIH)
Scientists in the laboratory have recently identified and cloned a DinB homolog from the archaeon Sulfolobus solfataricus P2, called DNA polymerase IV (Dpo4). Characterization of the enzyme reveals that the protein possesses many biochemical properties similar to other DinB polymerases. However, in contrast to DinB polymerases, which are unable to bypass a thymine-thymine cyclobutane dimer, Dpo4 bypasses the lesion moderately efficiently. In this regard, the enzyme is more akin to the distantly related eukaryotic DNA polymerase eta (Rad30 protein). S. solfataricus Dpo4 has been overproduced, purified, and its structure recently solved by x-ray crystallography. Like all DNA polymerases characterized to date, the enzyme possesses a topology similar to a right hand with domains that resemble "fingers," a "palm," and a "thumb." Dpo4 also possesses a unique domain called the "little finger," which helps the enzyme bind to DNA. Interestingly, the active site of the enzyme is large enough to accommodate two bases at one time, thus potentially explaining its ability to bypass thymine-thymine dimers and other bulky DNA adducts.

Enzymatic Characterization of DNA Polymerase Iota
Chumakov, Vaisman, Woodgate in collaboration with Patricia J. Gearhart (NIA, NIH) and Thomas A. Kunkel (NIEHS, NIH)
Scientists in the section recently discovered a novel human DNA polymerase called pol-iota. This enzyme is unique in its template-specific nucleotide misincorporation pattern. At T, on a recessed template, pol-iota prefers to misincorporate the "wobble base," G, three- to 11-fold more frequently than the correct "Watson and Crick" base, A. In contrast, at the end of a template, pol-iota misincorporates C and A eight- and three-fold, respectively, over the correct base, G. Such fidelities are 100- to-1,000-fold lower than most DNA polymerases. In addition to misincorporating bases with high frequency, pol-iota can extend the mispairs relatively efficiently, thereby fixing the misincorporated base as a mutation. It seems unlikely that a human cell would need such an error-prone DNA polymerase unless its activities might be of some evolutionary advantage. For example, pol-iota might be used during the hypermutation of rearranged immunoglobulin genes so as to increase antibody diversity. Another possibility might be at a uracil moiety or its derivatives. In living cells, uracil frequently arises from the spontaneous deamination of cytosine residues. The result is an increase in spontaneous mutagenesis, as the U normally base pairs with T, not G, as it would if the base were C. However, recent studies suggest that pol-iota not only misinserts G opposite U with high frequency but also extends the mispair efficiently. Thus, the unique ability of pol-iota to misinsert guanosine opposite uracils (which were once cytosines) provides a potential mechanism for human cells to reduce the extent of spontaneous mutagenesis caused by deamination of cytosine.

In addition to exhibiting a remarkable template-dependent misincorporation spectrum in vitro, pol-iota has been shown to possess deoxyribose phosphate lyase activity, and it has been hypothesized that under certain conditions in vivo, pol-iota may substitute for the better-characterized pol-beta during base-excision repair of damaged DNA.