The SOS system of Escherichia coli plays a central role in mutagenesis in this organism. The system is not normally present in the cell but becomes induced upon blockage of DNA replication by DNA damage. Its induction entails the expression of a large number of gene products, several of which are postulated to interact with the process of DNA replication, rendering it error prone and producing mutations on both damaged and undamaged DNA. The evidence for the existence of these components rests largely on genetic experiments. However, the elucidation of the nature of these components and their mechanisms of action requires a more direct biochemical approach. We have designed an in vitro DNA replication system in which the existence of the error-prone replication components may be tested. The system uses the conversion of single-stranded bacteriophage M13 DNA to its double-stranded form (ss -> RF conversion) by cell-free extracts derived from either normal or SOS-induced cells. After replication, the product DNA is transfected to produce intact bacteriophage. The accuracy of the in vitro replication step is determined from the frequency of mutant phage before and after replication. Since insights into SOS-modified DNA replication requires knowledge of the factors involved in maintaining normal accuracy, the latter is investigated as well, using, amongst others, E. coli mutator strains with known (or presumed) DNA replication defects. We have found DNA replication in E. coli extracts to be extremely accurate, with error rates approaching (or identical to) estimated in vivo rates. Extracts of two different E. coli mutator strains, mutD and mutT, display enhanced error rates in this assay. In case of mutT, this observation was exploited to uncover the role of the mutT gene product, namely to specifically prevent the otherwise frequent misinsertion of dGTP opposite template adenines. The system is currently used to investigate error rates during DNA synthesis by extracts from cells induced for the SOS response.
