Insights into the interactions between replication protein A and the ubiquitin ligase Rad18 from Saccharomyces cerevisiae.
Doctoral thesis, UCL (University College London).
DNA damage may lead to mutations and loss of genome integrity. Lesions encountered during replication cause the replication machinery to stall and, unless repaired or bypased, can result in lethality of the cell. The DNA polymerase processivity clamp, PCNA (proliferating cell nuclear antigen), mediates either mutagenic damage bypass or error-fee damage avoidance through its post-translational modification states. Mono-ubiquitylated PCNA stimulates the activity of translesion DNA polymerases, while poly-ubiquitylation of PCNA is a pre-requisite for error-free damage avoidance by a yet unknown mechanism. Recent findings in the laboratory suggested that Replication Protein A (RPA), an essential single-stranded (ss) DNA-binding protein, is required for induction of PCNA ubiquitylation upon DNA damage. Consequently, the aim of my thesis was to gain further insight into the mechanism by which RPA is involved in the up-stream signals that activate PCNA modification. The Rad18 protein from Saccharomyces cerevisiae (S. cerevisiae) is the ubiquitin ligase (E3) responsible for PCNA mono-ubiquitylation. The interactions of Rad18 with DNA and RPA, and the effects of this interaction on Rad18 binding to ssDNA, were studied in detail. Recombinant Rad18 was purified as a complex with its ubiquilin-conjugating enzyme, Rad6. Their stable association and ubiquitin conjugation activity was verified. Furthermore, basal levels of PCNA ubiquitylation were reconstituted in vitro. Yeast Rad18 was reported by others to bind preferentially to ssDNA over dsDNA. The intrinsic ssDNA-binding activity of the recombinant Rad18 protein was confirmed by pull-down assays using biotinylated oligonucleotides. Importantly, Rad18 is able to bind to ssDNA and to other ssDNA-containing structures, but also to forked-DNA consisting entirely of double-stranded (ds) DNA regions. Rad18 binding to DNA was demonstrated to be dependent on the ionic strength of the buffer. At low salt concentrations Rad18 was found to stably associate with ssDNA. At moderate to high salt concentrations, including in ionic strength conditions that could be considered physiological, Rad18 did not bind to ssDNA. Interestingly, binding to ssDNA at low ionic strength confers a stable association of Rad18 with the DNA for subsequent high ionic strength conditions. Taken together, these findings suggest that Rad18 binds to ssDNA at low salt concentrations with low affinity. Thereafter, a slow conformational change leads to an increased binding affinity that renders the ssDNA-bound Rad18 stable association with the ssDNA in high salt concentrations. Furthermore, these findings argue against the speculation that Rad18 can bind to sites of DNA damage in vivo by itself. In vitro experiments further demonstrated that the yeast Rad18-Rad6 complex interacts physically with RPA. The findings in this work support a mechanism through which both proteins bind directly to RPA. Further analysis of recombinant RPA subunits revealed that both Rfa2 and the DNA-binding domain of Rfa1 contribute independently to the specific interactions with the complex. Interestingly, the association between the Rad18-Rad6 complex and the DNA-binding domain of Rfa1 is stimulated by the presence of ssDNA. Furthermore, at physiological ionic strength, RPA recruits the Rad18-Rad6 complex to ssDNA. These findings support a model by RPA-coated ssDNA recruits Rad18 to sites of DNA damage. Thus, ssDNA-bound RPA may provide the up-stream signal for the activation of the DNA damage tolerance pathway and for PCNA ubiquitylation. Although this has yet to be clarified, most likely the interactions of Rad18 with both RPA and DNA contribute to its localisation to stalled replication forks in vivo. Although the SAP domain of human Rad18 was reported to be both necessary and sufficient for its interactions with DNA, the results obtained in this work suggest that this function may not be conserved in yeast. Nevertheless, this domain is essential for the in vivo function of yeast Rad18. Although its effect may be indirect, the SAP domain appears to contribute to the correct conformation of the Rad18 protein and to facilitate the interaction of the E3 with PCNA thereby allowing the ubiquitylation of the clamp.
|Title:||Insights into the interactions between replication protein A and the ubiquitin ligase Rad18 from Saccharomyces cerevisiae|
|Additional information:||Authorisation for digitisation not received|
Archive Staff Only