ACCURACY OF DNA REPLICATION IN VITRO

点击翻译按钮获取中文摘要

We are interested in determining the mechanisms by which human cells control spontaneous and induced mutation rates. While DNA synthesis by purified DNA polymerases in vitro is not accurate enough to account for low spontaneous mutation rates in vivo, actual DNA replication involves the concerted action of a number of proteins. We have therefore been examining the fidelity of semiconservative bidirectional DNA replication by proteins present in extracts of human HeLa cells. The data obtained using mutagenesis vectors that monitor base substitution and frameshift fidelity indicate that replication is highly accurate. The major accomplishments for this year include completion of a detailed quantitative determination of replication error rates for a variety of errors. This analysis has shown that frameshift fidelity is dramatically greater than that of purified polymerases, providing the basis for complementation assays for identifying fidelity components. We have also used specialized vectors that place the replication origin on either side of the mutational target to determine the fidelity of leading versus lagging strand replication for minus-one base frameshifts and two transition mispairs. We have seen small differences (2- to 4-fold), but nothing dramatic yet. This observation is relevant the current model for a eukaryotic replication fork, which would predict that a difference should exist. Not unexpectedly, the current model is too simple. We have also demonstrated that efficient repair of mismatched base pairs occurs in the extract. Repair is strand-specific and directed by a nick. It requires the presence of the mismatch and is efficient for some mispairs and inefficient for others. ATP hydrolysis is required, repair can proceed in either direction, and the resynthesis tract size can be in excess of 1000 nucleotides. We intend to continue these studies, with emphasis on defining the details of mismatch repair, identifying fidelity factors, possibly including exonucelolytic proofreading activity, and examining molecular models for base-substitution and frameshift fidelity.