Accueil  >  Séminaires  >  DNA double strand break repair in single E. coli cells
DNA double strand break repair in single E. coli cells
Par Meriem El Karoui (The University of Edimbourgh)
Le 19 Juin 2018 à 11h00 - Salle de séminaires 5ème étage, Tour 32-33


It has been shown in recent years that even genetically identical cells behave differently because many central processes involve molecules present in small numbers. The inherent randomness of chemical reactions then generates spontaneous fluctuations that can enslave all dependent processes. Double strand breaks are one of the most deleterious types of DNA damage because they lead death if not repaired. In Escherichia coli, the main repair pathway involves the multifunctional RecBCD enzyme, which salvages broken chromosomes by catalyzing the first step of homologous recombination. RecBCD is a heterotrimeric complex that is reportedly present in very low numbers in bacterial cells. This should lead to spontaneous fluctuations in RecBCD levels and non-genetic heterogeneity in the population. Qualitative studies based on population averages show that bacterial cells that do not express RecBCD are barely viable, while over-expression of the RecBCD protein leads to less efficient DNA repair. This suggests that the level of RecBCD expression needs to be tightly controlled, and raises the question of how bacterial cells cope with potentially large cell-to-cell fluctuations in this complex.

We have quantified cell-to-cell variability of RecBCD transcription using chromosomal transcriptional fusions and observed significant fluctuations that are consistent with very low levels of transcription. We developed a microfluidic device, combined with highly inclined illumination using laser excitation, which allows us to detect single molecules of GFP in vivo.  Using translational fusions of RecBCD to GFP we have shown that RecBCD is present in less than 10 molecules per cell. However, the level of fluctuations of each protein of the complex is surprisingly much lower than predicted by stochastic modeling. This suggests that a previously unknown regulatory network is controlling RecBCD expression to avoid potentially harmful fluctuations in RecBCD copy number.