Evidence suggests that repair of loopouts in E. coli may involve excision repair proteins UvrA, UvrB, and SbcC (Parniewski et al. 1999; Oussatcheva et al. 2001). d(CTG) n and d(CAG)n loopouts in plasmids were repaired (removed) when introduced into E. coli, and repair was less effective, but not prevented, in cells lacking certain excision repair proteins (UvrA, SbcC) (Oussatcheva et al. 2001). The binding of UvrA to d(CAG)n loopouts in vitro supports the hypothesis that loopout structures can be repaired in E. coli by excision repair functions. Panigrahi et al. (2005) have carefully characterized the ability of plasmid DNA containing a slipped intermediate DNA with a d(CAG)n loopout or a d(CTG)n hairpin in a continuous template or nicked nascent strand to be repaired in mammalian cell extracts. These templates mimic products of replication slippage or strand exchange during replication restart or during double-strand break repair. The stability of the repeats was analyzed in situations where the nicks were 3' or 5' to the loopouts. Different substrates were repaired, or not repaired, with remarkably different efficiencies (Panigrahi et al. 2005). First, repair required a nick. Second, a substrate containing d(CAG)5o on the continuous strand opposite d(CTG)30 on the nicked strand was repaired to d(CAG)50 ■ d(CTG)50. Third, when d(CTG)50 was opposite d(CAG)30 in the nicked strand, no repair occurred in cell extracts. Fourth, when the excess d(CAG) or d(CTG) slipped-out repeats were present on the nicked strand, variable-sized products corresponding to all possible lengths from 30 to 50 repeats were observed. These events did not require mismatch or excision repair proteins or DNA polymerase p. This work clearly shows that different DNA repeat structures at various positions relative to the direction, and strand, of replication can have very different consequences.
Was this article helpful?