Relation structure-état nucléotidique-capacité à s’assembler : la tubuline dans tous ses états. – Mec-Tub

Microtubules are dynamic structures of the eukaryotic cytoskeleton. They are involved in functions as different as the intracellular traffic or the formation of the mitotic spindle. They alternate between growth and shrinking phases in a process termed dynamic instability. The ''-tubulin heterodimer is the microtubule building block. In physiological conditions, only tubulin in the GTP state (i.e. with GTP bound to the ' subunit) assembles in microtubules. This is accompanied by GTP hydrolysis giving rise to the paradox that microtubules are mainly constituted of GDP-tubulin which does not assemble on its own. Upon disassembly, nucleotide exchange should occur for tubulin to be able to assemble again. Therefore, dynamic instability is linked to the tubulin nucleotide cycle. If the assembly-disassembly of tubulin in microtubules linked to the nucleotide cycle has been unravelled, the underlying structural cycle is only partially deciphered. Two types of GDP-tubulin structures have been determined so far, both at about 3.5 Å resolution. They correspond to tubulin in the core of the microtubule and after disassembly. By contrast, the GTP-tubulin structure is unknown. This proposal aims at identifying the structural changes linked to GTP binding that render tubulin competent for assembly. This involves both the determination of the GTP-tubulin structure and the improvement of the current diffracting power of the tubulin crystals so as to ascertain the structural variations with confidence. Tubulin has long resisted crystallization attempts, both because of its instability and its propensity to self-associate. Stabilizing it is by all current criteria a prerequisite. Therefore, we propose to develop new stabilizing tools. As soon as new stable tubulin entities will be obtained, we will set up crystallization experiments with GDP, with GTP, and with its far more stable analogue GMPCPP. Several approaches based on known tubulin sequesterers will be investigated. First, we will introduce mutations in stathmin domains to enhance their interaction with tubulin. We will also design stathmin peptides that react covalently with tubulin in a way that prevents its self-assembly. Second, we will evaluate PN2-3 based constructs. PN2-3 is a domain of the centrosomal CPAP protein that sequesters tubulin. We have already used it in crystallization experiments, without success so far. We propose to design new CPAP constructs that encompass PN2-3 and the region immediately downstream, which has recently been shown to interact with microtubules. We will also precise the mapping of the PN2-3:tubulin interaction on tubulin. If applicable PN2 3:stathmin domain chimeras that sum up the properties of both parent molecules will be produced. Finally, we will use our recent results to guide the design of modified vinca domain molecules in order to convert them from inducers of tubulin self-assembly into sequesterers. The approaches described above will provide us with several very different tubulin entities. As additional routes to diversify further the species to be tried in crystallization experiments, we propose to select ankyrin repeat proteins (DARPins) for tubulin binding so as to modify its surface and hence its crystallization properties. Finally, we have noticed a strong influence of some tubulin ligands upon the crystallization of the tubulin:stathmin domain complex. We therefore propose to screen chemical libraries for tubulin ligands, starting with the 4400 compounds library gathered at ICSN. The main objective will be the identification of two types of molecules: those targeting the vinca domain but with a simpler structure than the ones currently available and those that bind to an as yet unidentified tubulin binding site. Beyond its main objective, which is structural, this proposal will have additional benefits of two kinds. Firstly, it will provide us with new tools to study the microtubule cytoskeleton, useful in particular for biochemical and structural approaches. Secondly, the identification of new inhibitors will also potentially provide leads to pharmacologists for the development of new microtubule targeted drugs, in particular in oncology.