The centrosome is the cell’s main microtubule organizing center. In order to assemble a proper bipolar spindle in mitosis, the cell has to duplicate its inherited single centrosome. During centrosome duplication, a single procentriole assembles approximately perpendicular to the base of each of the two parental centrioles. Members of the evolutionary conserved SAS-6 protein family are essential for procentriole formation, and have recently been shown to be at the root of the universal nine-fold symmetry of centrioles. However, the features to HsSAS-6 function in human cells remain incompletely understood. In this thesis, we further addressed the function of HsSAS-6 by performing a structure/function analysis of HsSAS-6 . We have generated a refined map of HsSAS-6 domain organization, and have attributed new roles to each domain. We have identified the coiled-coil domain as being sufficient for centriolar localization, and the N- and C-terminal domains for maintenance of HsSAS-6 at centrioles until mitosis and function in procentriole formation. Through collaborative work, we have been able to reveal the homodimerization of HsSAS-6 molecules driven by the coiled-coil, and the role of the N-terminal domain in mediating oligomerization of HsSAS-6 dimers. In particular, we have highlighted the importance of a conserved residue F131 involved in that interaction, whose mutation abolishes HsSAS-6 function and prevents centriole formation. We aimed at investigating the nanoscale organization of HsSAS-6 , and thus developed a protocol for super-resolution 3D-STORM enabling us to achieve near-isotropic resolution on a fairly simple optical setup. Using this protocol, we have revealed HsSAS-6 nine-fold symmetrical ring-like organization at the onset of procentriole formation in S phase. We then analyzed the dynamics and mobility of HsSAS-6 using several means. FRAP experiments revealed that HsSAS-6 exhibited a higher mobility at centrosomes in S phase as comparedtotheG2phaseofthecellcycle. Byperturbingthemicrotubulenetwork,wealso showed that centriolar HsSAS-6 levels were decreased, thus indicating that HsSAS-6 recruitment or its maintenance at centrosome was microtubule-dependent. Finally, using FCS we have started to address the cytoplasmic dynamics of HsSAS-6 and revealed a complex-mobility mode that needs further analysis. Taken together, our findings have helped put HsSAS-6 at the root of the universal nine-fold symmetry of centrioles, and have led to the development of new tools to further investigate this process.