Aurèle PIAZZA Department of Genomes and Genetics Institut Pasteur, Paris

DNA double-strand breaks are genotoxic lesions whose repair can be templated off an intact DNA
duplex through the conserved Homologous Recombination (HR) pathway. HR entails a genome-wide
homology search step by the broken molecule. The successful encounter of homology leads the
formation of a joint molecule called a “displacement loop” (D-loop). The accuracy of these early
steps likely underlie the high fidelity of DNA break repair by HR. Yet, the mechanism of homology
search and the metabolism of D-loops in cells remain elusive.
To address this fundamental “needle-in-a-haystack” search problem I developed physical assays for
the study of D-loop metabolism in S. cerevisiae. It revealed the existence of “multi-invasions”
in cells, a byproduct of the homology search process we had postulated. This byproduct is at the
basis of a genomic instability and DNA break amplification mechanism that we named “multi-invasionsinduced
rearrangements” (MIR), with implications for the formation of complex chromosomal
rearrangements observed in human. Several proteins inhibit MIR, including conserved HR-related
helicases such as Mph1 (human FANCM), Sgs1-Top3-Rmi1 (BLM-TOPO3a-RMI1/2), and Srs2. Physical
examination of perfectly homologous D-loop levels in cells revealed that the majority are reversed
by these proteins. In addition to guard against MIR, this constitutive D-loop reversal may enabled
exhaustive and stringent sampling of the genomic space, and thus be an integral component of
homology search.
Building upon these methodologies and findings, the goal of my future lab will be to investigate
the interplay between events occurring at the chromosomal scale (structure and dynamics) and at
the molecular scale (homology search and joint molecules metabolism) by combining experiments and
computational modelling. A first research axis, in the continuity of my previous work, will be
directed at the development of a quantitative framework of homology search in vegetative cells. A
second research axis will aim at defining the elusive HR-driven phenomenon of homolog bias and
crossover patterning underlying homologs pairing during meiosis.

Contact: D. Auboeuf