DESCRIPTION: Sequence analysis of deletion mutations in both E. coli and humans has shown that deletions occur most frequently between short repeated sequences. Models to explain this observation have been of two main types: recombinational (unequal crossing over) and replicational (strand slippage or misalignment). There has been a tendency to discount recombinational mechanisms because deletion mutations in E. coli lack two traditional hallmarks of homologous recombination: the repeated sequences are very often much shorter than necessary to serve as a substrate for RecA protein, and the deletions occur quite readily in RecA deletion strains, in which conventional recombination is eliminated. Recent work by Dr. Lovett using a plasmid-based system for monitoring deletions has shown, unexpectedly, that RecA-independent deletion formation can have unmistakable recombinational features, most notably, deletion-associated plasmid dimerization. On the basis of this and related observations, Dr. Lovett has proposed a mechanism for deletion formation initiated by RecA-independent pairing between nascent strands in the vicinity of an arrested replication fork to form a Holiday junction, followed by processing by other enzymes associated with homologous recombination reactions. Dr. Lovett furthermore proposes that the same mechanism may apply to recombinational "postreplication" repair observed on damaged DNA templates, whose mechanism has been obscure. In this application, Dr. Lovett proposes to continue her study on the mechanism of deletion formation and its relationship to recombinational DNA repair and sister strand exchange (plasmid dimerization) in E. coli. The proposed work has two broad components. In one component, Dr. Lovett will analyze the effect of single DNA lesions (thymine dimers) on the occurrence of deletions and plasmid dimerization.Using variations of this basic idea, she will examine if a thymine dimer has different effects when placed in the leading vs. lagging strand vs. both together; when placed in different locations relative to the tandemly repeated sequences that are recombining; and when processed in strains with mutations in various components of recombination or replication. In the other part of the investigation, Dr. Lovett will define the genetics of recA-independent deletion formation by examining the effects of known components of replication and recombination on deletion formation, and by searching for novel genes that affect the frequency of deletion formation.