L’ubiquitination est une modification post-traductionnelle qui joue un rôle majeur chez les organismes vivants. Chez Arabidopsis thaliana, le transporteur de fer racinaire IRT1 est endocyté à la suite de la monoubiquitination de deux résidus lysine situés au niveau de sa grande boucle cytosolique. Cependant, les mécanismes régissant l’endocytose médiée par l’ubiquitine ainsi que son rôle biologique restent flous. Au cours de ma Thèse, j’ai mis en évidence que la dynamique d’IRT1 était contrôlée par les métaux substrats secondaires du transporteur (à savoir le zinc, le manganèse et le cobalt). En l’absence de ces métaux, IRT1 est localisé à la membrane plasmique avec une polarité latérale le positionnant sur la face externe des cellules de l’épiderme racinaire. La présence de ces mêmes métaux à un niveau physiologique entraîne la monoubiquitination d’IRT1 et son internalisation vers les endosomes précoces. J’ai démontré que lorsque les métaux substrats secondaires d’IRT1 sont présents en excès, les modifications monoubiquitine sont alors allongées en chaînes de polyubiquitines liées par le résidu lysine-63, entrainant ainsi son adressage vers la vacuole et sa dégradation. Mes travaux ont par ailleurs permis d’élucider les mécanismes moléculaires impliquées dans la réponse des plantes à l’excès de métaux substrats d’IRT1. J’ai notamment montré que l’endocytose d’IRT1 était dépendante i) d’un motif riche en résidus histidine dans la séquence d’IRT1 qui est capable de fixer ces métaux autres que le fer, ii) de la phosphorylation d’IRT1 au niveau d’un résidu thréonine par une protéine kinase en cours d’investigation, et iii) de l’E3 ligase à domaine RING IDF1. D’un point de vue physiologique, l’endocytose d’IRT1 médiée par l’ubiquitine et dépendante des métaux protège la plante d’une suraccumulation de ces métaux autres que le fer qui sont hautement réactifs. / Ubiquitination is a post-translational modification playing a major role in living organisms. In Arabidopsis thaliana, the root iron transporter IRT1 is endocytosed following the monoubiquitination of two lysine residues located in its large cytosolic loop. However, the mechanisms driving IRT1 ubiquitin-mediated endocytosis and its biological relevance remains unclear. During my PhD, I uncovered that IRT1 dynamics is controlled by its secondary metal substrates (i.e. zinc, manganese and cobalt). In the absence of these non-iron metals, IRT1 is found at the cell-surface of root epidermal cells with an outer lateral polarity, while their presence at physiological levels triggers IRT1 monoubiquitination, internalization and accumulation in early endosomes. However, upon non-iron metal excess, monoubiquitin modifications are extended into K63 polyubiquitin chains to promote the vacuolar targeting of IRT1 and its degradation. I investigated further the molecular mechanisms driving plant responses to non-iron metal excess. I notably showed that this regulation by non-iron metals is dependent on i) a histidine-rich stretch in IRT1 that is able to directly bind to non-iron metals, ii) the subsequent recruitment of a kinase currently under investigation which phosphorylates IRT1 at a threonine residue, and iii) the RING E3 ligase IDF1. Altogether, the metal-dependent ubiquitin-mediated endocytosis of IRT1 protects the plant from overaccumulation of highly reactive non-iron metals.

Characterization of phosphoproteins-including detection, identification of phosphoproteins and identification of phosphorylation sites-is mostly done with radiolabeling and proteomic techniques. Three main topics related to phosphoprotein characterization are included in this dissertation. First, large-scale characterization of the CHO (Chinese hamster ovary) cell phosphoproteome was done using two dimensional gel electrophoresis (2DE) separation, visualization of phosphoproteins by radiolabeling or a phosphoprotein specific dye, followed by MALDI-TOF identification. Because radiolabeling of phosphoproteins is very sensitive and straightforward to quantify, such analysis can give a clear picture of the relative phosphosphorylation of proteins present in a sample. But there are also limitations to this approach, such as the inability of 2DE to separate hydrophobic, acidic and large proteins and the poor detection limits of common protein stains such as Coomassie stain. Additionally, it is difficulty to excise the right spots for identification because of the low abundance of phosphoproteins which have been visualized by radiolabeling. Furthermore, there are problems associated with metabolic radiolabeling. A second topic of the dissertation is the development of a novel strong cation exchange monolithic column for MudPIT (multidimensional protein identification technology) and phosphopeptide isolation. This column, a poly(AMPS-co-PEGDA) monolith containing as high as 40% AMPS, has several favorable features, such as high binding capacity, extraordinarily high resolution, and high peak capacity, making it ideal for resolving complex peptide samples. Application of this novel column to isolate model phosphopeptides was shown. More general use of this column in MudPIT (strong cation exchange column followed by reverse-phased MS/MS) is probably somewhat limited, due to the hydrophobicity of the AMPS monomer. A better monolith could be obtained if a more hydrophilic monomer was used. In the third area of the dissertation, several individual protein phosphorylation sites were analyzed, employing different strategies. Phosphorylation sites of one multiply phosphorylated tryptic peptide from CK2-phosphorylated phosducin-like protein (PhLP) was well characterized using enrichment with a MonoTip® TiO Pipette Tip. Analysis of syntaxin 1a phosphorylation by AMPK (AMP-activated protein kinase) was done by peptide level mapping for potential phosphopeptides after its unsuccessful trial with enrichment using the MonoTip® TiO Pipette Tip. Several criteria such as existence of non-phosphorylated forms of potential phosphopeptides, controls and reasonable retention times were used to rule out false positives. Phosphorylation of syntaxin 1a by AMPK was narrowed down to tryptic peptide T32 with evidence from different sources. Three phosphorylation sites of syntaxin 4 by AMPK were identified within the same peptide (Q65QVTILATPLPEESMK80). Further pinpointing of phosphorylation site(s) for syntaxin 1a by AMPK and further confirmation of these phosphorylation sites in syntaxin 4 by AMPK are required in vivo. The role of phosphorylation in syntaxin 4 by AMPK is the next step toward elucidation of AMPK activation and regulation of the glucose uptake mechanism.