Community Research and Development Information Service - CORDIS

Periodic Report Summary 1 - MULTIPLE SEARCH (Mechanism of homology search during homologous recombination)

DNA double strand breaks (DSBs) are toxic lesions that endanger cell survival and threaten the stability of the genome. Homologous recombination is a cellular mechanism that faithfully repairs DSBs by using an intact, identical DNA copy as a template. To achieve identification of this identical (homologous) sequence, the Rad51 recombinase coats the DNA flanking the break site and harnesses its sequence information to sample the genome by a process called homology search. How this daunting genome-wide search operates is unknown. My ongoing research project aims at characterizing the mechanism of homology search in cells, the proteins involved, and identify putative defects with advert effects for genomic stability.
First, I developed a technique to detect Displacement-loops (D-loop; the first DNA strand exchange produced by the successful encounter of a homologous sequence by the searching Rad51-DNA molecule) in cells upon controlled site-specific induction of DSB. This innovation closes a long-lasting technical gap in the field and allows studying the early stages of homologous recombination in cells. In particular, the D-loop Capture (DLC) assay enabled the study of the kinetics of homology search and the factors involved in the process. I addressed the role in regulating D-loop stability of several proteins known from end-point genetic assays to be involved in homologous recombination fidelity. I defined two D-loop disruption pathways. The first is supported by the Mph1 (human FANCM) helicase and the Sgs1-Top3-Rmi1 (human BLM-TOPO3alpha-RMI1/RMI2) complex. The second relies on the Srs2 helicase. These results reveal unanticipated granularities in D-loops regulation, with two D-loop subtypes that are substrate for independent unwinding activities. Apical to this two pathways is the translocase Rdh54, which models D-loops for the Mph1 and Sgs1-Top3-Rmi1 axis. Strangely, these multiple DNA strand-exchange turnover activities operate on a perfectly homologous sequence, questioning their homology search fidelity-enforcing role. We are currently developing a computational homology search framework to understand how these D-loop disruption activities are productive in homology search. My future goal in the return laboratory will be to develop a high-throughput version of the DLC assay to monitor DNA strand exchange genome-wide.
Second, I conducted a series of experiment to address our postulated multiplexed homology search model. This model entails that not only the extremity, but the whole length of the DNA flanking the break is used by Rad51 to sample the genome. A straightforward prediction of this model is that homology can be simultaneously encountered at different positions along the searching DNA molecule, forming a “multi-invasion” intermediate (Figure 1A). In collaboration with William Wright in the laboratory, I demonstrated the existence of this intermediate in vitro, which is formed both with the yeast or the human recombination proteins. Furthermore, a major effort has been to improve and adapt the DLC assay to provide physical evidence for multi-invasion in cells and study their in vivo regulation.
Third, I postulated that the “multi-invasion” byproduct of homology search might be at risk for genomic stability. Indeed, the multiple DNA branch points constitutive of the intermediate are known substrates for enzymes cleaving DNA (called structure-selective endonucleases) that may jeopardize the integrity of the intact donors engaged. I developed an experimental system in yeast to address this possibility, and demonstrated that indeed a break formed on one chromosome can causes translocation of the two intact donors on other chromosomes. I showed that this multi-invasion-induced rearrangement (MIR) depends on structure-selective endonucleases Mus81-Mms4 (human MUS81-EME1), Yen1 (human GEN1) and Slx1-Slx4. On the contrary MIR is inhibited by the D-loop disruption activities characterized earlier (Srs2, Mph1, Sgs1-Top3-Rmi1) with the DLC assay, and the Rad1-Rad10 flapase (human XPF-ERCC1). By modifying the length and position and performing physical analysis of the translocants, I proposed a detailed mechanism of MIR (Figure 1B). Importantly, integral to MIR is the propagation of the DSBs onto both donors, for which I could provide physical evidence. This amplification of the original damage results in frequent additional chromosomal rearrangements. This unanticipated genomic instability mechanism has major implications for the cancer fields. Indeed, several features of MIR provide a parsimonious explanation for the formation of complex “chromothriptic” chromosomal rearrangements observed in cancer genomes.
In conclusion, my work so far resulted in (i) the development of an assay to study homology search and regulation of DNA strand exchange stability in cells, closing a decades-long limitation in the field, (ii) the characterization of several D-loop disruption activities and their interactions, (iii) the development of a computational framework for homology search, (iv) the demonstration of the existence of multi-invasion, a new homologous recombination intermediate that directly results from our postulated homology search mechanism, and (v) the discovery and characterization of MIR, a new genomic instability mechanism. These findings major implications for our basic understanding of homologous recombination and its defects underlying simple and complex genomic rearrangements in human.

Related information

Documents and Publications

Reported by



Life Sciences
Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top