The number of gene products available for structural and functional study is increasing at an unprecedented rate as a result of the successful whole genome sequencing projects. Systematic structure determination of proteins on a genomic scale, called structural genomics, can significantly contribute to the field of protein science and to functional annotation of newly identified genes. This thesis covers different aspects of protein production in Eschericiha coli for structural studies in the context of structural genomics. Protocols have been downscaled and standardized to allow for a rapid assessment of the production characteristics for multiple proteins in parallel under a number of different conditions. Foremost, the ability of different proteins and peptide tags to affect the solubility of the recombinant protein when produced as fusion proteins has been systematically studied. Large differences in the success-rate for production of soluble protein in E. coli were found depending on the fusion partner used, with a more than two-fold increase in the number of proteins produced as soluble when comparing the best and the poorest fusion tags. For different constructs with a histidine tag, commonly used to facilitate protein purification, large differences in yield depending on the design of the expression vector were found. When comparing different fusion proteins produced from identical expression vectors, fusions to the GB1 domain were found to result in the highest yield of purified target protein, on average 25 % higher than any of the other fusions. The suitability for further structural studies was tested at an intermediate scale for proteins that were identified as soluble in the expression screening. For this purpose, protocols for rapid purification and biophysical characterization using nuclear magnetic resonance and circular dichroism spectroscopy were developed and tested on 19 proteins, of which four were structured. / QC 20100826

Les cassures double brin de l'ADN altèrent l'intégrité physique du chromosome et constituent l'un des types les plus sévères de dommages à l'ADN. Pour préserver l'intégrité du génome contre les effets potentiellement néfastes des cassures double brin de l'ADN, les cellules humaines ont développé plusieurs mécanismes de réparation, dont la réparation par recombinaison de l'ADN et la jonction d'extrémités non-homologues (NHEJ), catalysés par des enzymes spécifiques. Pendant ma thèse, nous avons caractérisé la dynamique de certaines des interactions protéines/ADN impliquées dans ces mécanismes au niveau de la molécule unique. Dans ce but, nous avons combiné des pinces optiques et de la micro-fluidique avec de la microscopie de fluorescence à champ large afin de manipuler une ou deux molécules d'ADN individuelles et d'observer directement les protéines de la réparation marquées par fluorescence agissant sur l'ADN. Nous avons concentré notre analyse sur trois protéines/complexes essentiels impliqués dans la réparation de l'ADN: (i) la protéine humaine d’appariement de brin RAD52, (ii) les protéines humaines XRCC4, XLF et le complexe XRCC4/Ligase IV de la NHEJ et (iii) le complexe humain MRE11/RAD50/NBS1. / DNA double-strand breaks disrupt the physical continuity of the chromosome and are one of the most severe types of DNA damage. To preserve genome integrity against the potentially deleterious effects of DNA double-strand breaks, human cells have evolved several repair mechanisms including DNA recombinational repair and Non-Homologous End Joining (NHEJ), each catalyzed by specific enzymes. In this thesis we aimed at unraveling the dynamics of protein/DNA transactions involved in DNA double-strand break repair mechanisms at single molecule level. To do this, we combined optical tweezers and microfluidics with wide-field fluorescence microscopy, which allowed us to manipulate individual DNA molecules while directly visualize fluorescently-labeled DNA repair proteins acting on them. We focused the study on three crucial proteins/complexes involved in DNA repair: (i) the human DNA annealing protein RAD52, (ii) the non-homologous end joining human proteins XRCC4 and XLF and the complex XRCC4/Ligase IV, and (iii) the human MRE11/RAD50/NBS1 complex.