Date et lieu

Lundi 31 janvier à 10h30 au CRI - Seminar room n°2006 au 2ème étage
Faculté de médecine Cochin
24 rue du faubourg St Jaques
75014 Paris

Composition du jury

• Lionel MOISAN, directeur de thèse
• Ariel LINDNER, co-directeur de thèse
• Hugues BERRY, LIRI, Lyon, rapporteur
• Rafael GIRALDO, CIB, Madrid , rapporteur
• Pr Reiner VEITIA, Institut Jacques Monod, Paris, examinateur
• Pr Ivo SBALZARINI, ETH, Zürich, examinateur
• Miroslav RADMAN, U1001, Paris, examinateur

Abstract

An increase in the probability of death and a decrease in the reproduction rate (both accounting for decrease of the fitness) are signatures of ageing in living organisms. Using a morphological criteria, allowed to demonstrate that Escherichia coli a symmetrically dividing micro-organism is subject to ageing.

Ageing is studied by using time-lapse movie of growth of microcolonies emanating from a single cells. This result in a huge amount of images to analyse. The duration of semi- or non-automated analyses of those images brings a serious limit to the rate of data available for statistical and biological analysis. Hence, the processing of images had to be automated to speed up the process and make possible the studies of large data set. To address this key issue, I developed a new approach based on one main idea : considering segmentation and tracking at the same time, whereby the spatio-temporal segmentation uses the advantage of the large time redundancy of data, contrary to existing methods relying on successive spatial segmentation and tracking.

Specifically, I applied the image analysis tools to address the role of protein aggregation in bacterial ageing. Aggregation of misfolded proteins is known to be an hallmark of many age-related disease e.g. Alzheimer disease, Parkinson disease, cataract, and known to have cell toxicity action. Thanks to a translational fusion between a small heat shock protein (IbpA) and a fluorescent protein, effect of natural protein aggregation on growth rate of bacteria were observed. We were able to show, among other things, that protein aggregation is associated with the decrease of growth rate associated with ageing in E. coli .

In conclusion in this work I developed new image analysis methodologies that improved speed, accuracy and reliability of the results on one hand, and shed light on the dynamics and effects of natural aggregates in bacterial ageing on the other hand.