​​Homologous recombination (HR) is an error-free mechanism for repairing DNA double-strand breaks (DSBs) that occur spontaneously through endogenous DNA damage or through exposure to radiation and other DNA-damaging agents. Abnormalities of HR are associated with a number of genetic diseases, including ataxia-telangiectasia, Nijmegen break syndrome, Fanconi anemia, and Bloom’s syndrome (reviewed in: {Thompson and Schild, 2002, #76614). A number of these involve increased cancer susceptibility, highlighting the critical role of HR in maintaining genome integrity.

​HR involves the use of an intact template homologous DNA molecule and a key step in homology search and template invasion is the polymerization of Rad51 protein at 3’ single-stranded DNA (ssDNA) ends that are exposed after DSB processing, resulting in the formation of a highly ordered, right-handed helical protein filament (reviewed in: {Holthausen et al., 2010, #35878}). The polymerization process involves Rad51 nucleation on ssDNA followed by cooperative filament elongation. In HR, RPA first binds to the initiating ssDNA, preventing Rad51 fixation. Recombination Mediator Proteins (RMPs) are capable of overcoming the inhibitory effect of single strand binding (SSB) proteins (such as RPA) on nucleofilament assembly by promoting the recombinase nucleation on ssDNA. The subsequent powerful filament elongation leads to RPA displacement (reviewed in: {Beernink and Morrical, 1999, #78610}). Up to now, the RMP family includes the bacterial RecO and DprA proteins, the yeast Rad52 protein and the heterodimeric Rad55–Rad57 complex and the human breast and ovarian tumor suppressor protein BRCA2. In yeast, the Rad55–Rad57 complex Rad51 paralogs was also shown in vitro to protect the filament from dismantling by helicase activity {Liu et al., 2011, #82047}.

The yeast helicase Srs2 dismantles Rad51 filaments through its translocase activity on ssDNA {Veaute et al., 2003, #16321}, thus removing inappropriate Rad51 filaments that can potentially lead to cell death {Esta et al., 2013, #41591}. In higher eukaryotes, FBH1, PARI, RTEL1, FANCJ or BLM have been identified as potential Srs2 functional homologues (for a review: {Heyer et al., 2010, #98538}) but their function remains to be defined. 

Study of inappropriate Rad51 filament formation following irradiation of yeast cells lacking Srs2 led us to isolate a Rad52 mutant (rad52-L264P) with attenuated mediator activity that does not affect HR but allows a complete bypass of Srs2 activity in eliminating toxic Rad51 filaments {Esta et al., 2013, #41591}. We found that this mutant mimics the regulation of Rad51 filament formation through Rad52 sumoylation, revealing a complex regulation of this early step of HR.

We are now conducting genetic and biochemical approaches to precise our view of Rad51 filaments formation and regulation. We are exploring the specific role of each mediator/protector activities (mainly Rad52 and Rad55/Rad57) in Rad51 filament formation. We are also investigating the role of Rad52 sumoylation in regulating Rad51 filament formation. Finally, we are developing genetic systems coupled with whole genomic analyses to detect in vivo toxic Rad51 filaments. In parallel, we are also studying the involvement of other proteins in the generation of toxic Rad51 filament formation.