Thesis title: Role of WRN interactors and backup factors during replicative stress in normal and cancer cells
DNA Replication stress is one of the characteristics of cancer cells and is associated to proto-oncogene activation. Cells evolved a sophisticated network of proteins to deal with DNA replication stress and prevent accumulation of genome instability, which is another feature of cancer cells. Among the several proteins responding to DNA replication stress, there is the Werner syndrome protein (WRN). WRN is an exo/helicase belonging to the RecQ family of helicases mutated in the Werner syndrome (WS). One of the main WRN interactors is RPA, a single-strand DNA binding protein crucial in many DNA metabolisms. RPA represents one of the most abundant interactors of WRN and stimulates WRN helicase in vitro. However, the functional role of WRN-RPA interaction in the cell is poorly understood. Hence, my PhD work focused on the characterization of the functional implications of WRN-RPA association for the roles of WRN in response to replication stress.
Post-translational modifications, like phosphorylation, are often involved in regulating protein-protein interactions. WRN undergoes multiple phosphorylation events by different kinases each of them regulates specific functions in the cell. I focused on some clustered WRN residues phosphorylated by Casein Kinase 2 (CK2). These putative CK2 sites are embedded into the acidic domain localised at the N-terminal region of WRN and partially overlaps with the RPA70 binding region. Thus, I exploited mutants of WRN abrogating phosphorylation at CK2 sites to dissect the defect associated to loss or reduction of WRN-RPA interaction.
Using an immortalized WS-patient derived cell line, which is a natural knockout, complemented with the wild-type or the CK2 unphosphorylable form of WRN as experimental cell model, I performed multiple biochemical and cell biology experiments to analyze different aspects of the replication stress response in which WRN is implicated.
From pull-down, coimmunoprecipitation and proximity-ligation assays, I confirmed that CK2 targets the acidic domain of WRN and regulates the binding with RPA in vitro and in the cell. Using phospho-specific antibodies recognising two of the CK2 sites, I observed that the pool of WRN associating with RPA in response to replication stress is almost completely phosphorylated at CK2-dependent sites and that loss of CK2-dependent phosphorylation substantially abrogates RPA association of WRN. Moreover, to understand if differences in the binding with RPA of WRN could confer defects in the response to perturbed replication, I analyzed the formation of two replication intermediates, nascent or parental ssDNA, upon transient or persistent replication fork arrest. From these experiments, I found that WRN-RPA interaction does not influence the function of WRN in fork processing although it has mild effects on WRN presence on DNA. Most importantly, through DNA fiber assays I defined that WRN interaction with RPA affects the ability of the replication fork to recover after prolonged replication arrest leading to accumulation of template gaps and increased load of DNA damage processed by MRE11. Accumulation of parental gaps observed in the unphosphorylable WRN mutant apparently correlates with an impaired helicase activity consistent with an RPA-dependent stimulation of WRN helicase.
Thus, my data define that the predominant role of RPA in the function of WRN is stimulation of helicase activity to favour fruitful replication fork restart. In contrast, WRN-RPA association is not involved in the role of WRN to process and protect reversed forks.
The absence of WRN function has been recently shown to be synthetic lethal in MSI cancer cells, identifying WRN as a therapeutic target. In a distinct part of my PhD study, I started the dissection of the WRN interactome using a proximity-labelling strategy coupled with mass spectrometry to identify differences of proteome in different cancer cells. I focused on developing WRN stable constructs associated with TurboID biotinylating technology, which can be used to catch also subtle and transient interactions over time. As model conditions, I focused on two colorectal cancer cell lines characterized by the microsatellite instability (MSI) phenotype. I established models cells in which WRN depletion can be induced by doxycycline and I performed proof of principle experiments to test the TurboID-WRN fusion protein in the cell. The aim of this dynamic interactome study will be to understand the specific role and interactions of WRN in MSI cell lines compared with a MSS condition and eventually characterize the MSI-specific WRN-dependent pathway.