Although replication through many of the DNA lesions can be accomplished by the action of TLS polymerases, template switching provides for a more general and error-free means of lesion bypass. The biochemical data presented here give strong support for a role of Rad5 in mediating error-free lesion bypass by template switching wherein its DNA helicase activity promotes replication fork regression, a conclusion that is in keeping with genetic observations that have been made with the various rad5
mutations (Gangavarapu et al., 2006; Johnson et al., 1992; Liefshitz et al., 1998; Torres-Ramos et al., 2002; Zhang and Lawrence, 2005
Recently, RecQ family members the Werner, Bloom, and RecQ5beta proteins, with well-established role in recombinational repair, have been suggested to perform fork regression in higher-order eukaryotes (Kanagaraj et al., 2006; Machwe et al., 2006; Ralf et al., 2006
). However, these RecQ helicases also yield single-stranded DNAs from fork substrates, and importantly, the presence of ssDNA-binding proteins was shown to bias these reactions toward processing forks to structures other than double-stranded products (Kanagaraj et al., 2006
). By contrast, Rad5 does not generate single-stranded DNA from fork DNA and is not affected by ssDNA-binding protein; this together with all the other evidence we present here indicates that Rad5 concertedly unwinds and anneals the nascent and the parental strands without exposing extended single-stranded regions. This unique property of Rad5 will ensure that the stalled replication forks are not processed to structures, which are abortive for template switching and the consequent lesion bypass. To our knowledge, Rad5 is the only eukaryotic protein shown to have such an activity, and certainly the only known DNA helicase for which genetic studies have implicated a role in postreplication repair.
In , we present a model for Rad6-Rad18-dependent lesion bypass in eukaryotes. If replication stalls upon encountering an unrepaired DNA lesion, PCNA monoubiquitination by Rad6-Rad18 could enable replication through the lesion site by the TLS polymerases. Alternatively, Rad5-mediated template switching could promote lesion bypass. In this case, if the lesion is located on the leading strand template, the leading and the lagging strand polymerases can become uncoupled, and synthesis on the lagging strand can continue way past the blocked nascent leading strand (Cordeiro-Stone et al., 1997, 1999; Pages and Fuchs, 2003; Svoboda and Vos, 1995
). Next, the Rad5 helicase action unwinds the nascent strands from their respective templates and then anneals them with one another as well as reanneals the parental strands. The overall outcome of these reactions is the regression of the replication fork to form a four-way junction, similar to a Holliday junction. Following that, the sequences complementary to the damaged region are synthesized on the nascent leading strand using the nascent lagging strand as the template. The reversed fork is then regressed, and synthesis resumes beyond the point of the lesion that completes the error-free damage bypass. We should point out that the proposed model for lesion bypass by Rad5-catalyzed fork reversal is only effective for stalling on the leading strand. Because DNA damage on the leading strand template presents a serious impediment to the progression of the replication fork, whereas lagging strand damage could be bypassed from replication from an adjacent Okazaki fragment and could utilize other means for its bypass, Rad5 would play an indispensable role in rescuing the stalled fork when the lesion is on the leading strand.
Model for the Role of Rad5 in Lesion Bypass by Template Switching
The replication fork reversal mediated by Rad5 must be tightly regulated, for otherwise the DNA intermediates generated from fork reversal could provide potential substrates for dangerous recombination events leading to chromosomal rearrangements. One such regulatory circuit is based on PCNA polyubiquitination. Although PRR depends on the polyubiquitination of PCNA by Mms2-Ubc13-Rad5 complex, the precise mechanistic role of PCNA polyubiquitination has remained elusive. We envisage that PCNA polyubiquitination could be important for disrupting the association of some PCNA-bound protein such as the replicative polymerase or another component of the replication ensemble that is otherwise inhibitory to the uncoupling of leading and lagging strand synthesis and for fork reversal.
The replication intermediates that accumulate in yeast cells in response to hydroxyurea (HU) treatment and ultraviolet (UV) exposure have been examined using electron microscopy (EM) and two-dimensional (2D) gel electrophoresis (Sogo et al., 2002; Cotta-Ramusino et al., 2005; Lopes et al., 2006
). Although single-stranded gaps accumulate in response to replication blockage in wild-type cells, reversed forks are seen only in rad53
mutants, where the replication ensemble becomes destabilized. These observations have suggested that the reversed forks that are seen in rad53
cells represent pathological structures that arise at collapsed replication forks in the absence of Rad53-dependent checkpoint control. We suggest that the fork reversal mediated by Rad5 is a highly regulated and tightly controlled event, and in addition to PCNA polyubiquitination, it likely requires the Rad53-dependent stabilization of the replication fork. We envisage the DNA intermediates of Rad5-mediated fork reversal in vivo to be rather transient, and not subject to processing by nucleases such as the Exo1 processing of reversed forks that occurs in rad53
cells (Cotta-Ramusino et al., 2005
Given that the many elements of PRR are conserved from yeast to humans, it is highly probable that the replication fork regression activity of Rad5 has also been conserved during eukaryotic evolution. In this regard, we note that, similar to Rad5, the human SHPRH protein functions as a ubiquitin ligase for Mms2-Ubc13-dependent PCNA polyubiquitination, and similar to Rad5, SHPRH harbors an SWI/SNF2-type helicase domain (Motegi et al., 2006; Unk et al., 2006
). Thus, SHPRH too could have the fork regression activity and promote thereby the error-free bypass of DNA lesions by template switching. Interestingly, the SHPRH gene is mutated in a number of cancer cell lines, including those from melanomas and ovarian cancers, which indicates that SHPRH function is an important deterrent to mutagenesis and carcinogenesis in human cells (Sood et al., 2003