Mechanochemistry of the Rad5 Double-Stranded DNA Translocase

Abstract

Replication forks may stall as a result of unrepaired DNA lesions or premature dissociation of the components of the replication machinery. Stalled forks are unstable and their collapse is extremely dangerous, as it may result in double stranded DNA breaks. Rad5 is a yeast protein with human homologs having a proposed replication fork reversal activity, which may facilitate the restart of stalled replication forks. Rad5 is a SWI/SNF ATPase having no canonical helicase activity. It does not unwind unbranched dsDNA substrates, but it can process four-armed forks with homologous arms. Our study is aimed at deciphering Rad5 mechanochemical activities. Beside full-length Rad5, we also used an N-terminal truncated construct lacking the HIRAN domain. The relatively low basal (DNA-free) ATPase activity of Rad5 is accelerated by both ssDNA and dsDNA. Poly-dT ssDNA and dsDNA enhance this activity 5 and 50 times, respectively. This difference suggests that the enzyme acts differently on the two types of substrate. We observed limited ssDNA length dependence of the ATPase activity. Our results indicate a large Rad5 effective binding site of 60 nt. We are now investigating dsDNA length dependence to decipher dsDNA translocation mechanochemistry proposed to support fork dissolution. We are in progress of determining the mechanochemical coupling between ATP consumption and fork regression by applying single-turnover kinetic methods.