Investigation of the multi-domain aggregation mechanism of ataxin-3

Spinocerebellar Ataxia Type 3 (SCA3) is one of nine polyglutamine (polyQ) diseases which are all characterized by progressive neuronal dysfunction. A hallmark of these diseases is the presence of neuronal inclusions which contain aggregated polyQ protein, implicating protein misfolding as a key part of this disease. The only sequence homology which exists between the polyQ proteins is the polyQ tract, thus suggesting that protein context is responsible for dictating the specific characteristics of each disease. Increasingly, flanking domains to the polyQ tract have been demonstrated to modulate the aggregation of polyQ proteins. Fibrillogenesis of ataxin-3, the causative protein of SCA3, occurs via a multi-stage pathway when studied in vitro. The isolated Josephin domain of ataxin-3 has the intrinsic ability to aggregate under native conditions and thus the multistage model of ataxin-3 aggregation has proposed the first stage of aggregation is mediated by interactions of the Josephin domain. This thesis has further characterized the multistage aggregation pathway of ataxin-3, with a focus on the role of the Josephin domain during misfolding and aggregation. This thesis initially determined that changing the stability of the flanking Josephin domain dramatically impacts upon ataxin-3 fibrillogenesis in vitro. Mutations within the Josephin domain that increased stability led to decreased aggregation rates and vice versa. This provided support to the hypothesis that Josephin domain interactions are involved in the first stage of ataxin-3 aggregation. The stability mutants also distinguished between two previously proposed models of misfolding to demonstrate that ataxin-3 aggregates from an intermediate which is on the folding pathway. Defining the conformational landscape of ataxin-3 is one mechanism which may provide insights into the initial misfolding events involved in aggregation. As Josephin domain interactions were demonstrated to mediate the first stage of aggregation, hydrogen exchange mass spectrometry was used investigate the dynamics of this folded domain. The Josephin domain was found to be highly dynamic, and under native conditions exists in rapid equilibrium with several partially unfolded conformations. From these data it is proposed the unfolded conformations are not involved in the fibrillogenic pathway, as mutants with either increased or decreased aggregation rates do not change the proportions of the partially unfolded conformations. In addition, the proportion of the partially unfolded conformations was unaffected by several protein interaction partners of ataxin-3. Therefore native ataxin-3 exists within a more complex conformational landscape than previously thought, which is potentially linked to the normal function of the protein. In a further exploration of the conformational landscape of ataxin-3, addition of the membrane mimetic sodium dodecyl sulfate resulted in a number of alternative aggregation pathways. Ataxin-3 was found to bind several phospholipids only when in a fibrillar conformation, which may be linked to the mechanism of toxicity, as has been reported for other amyloidogenic proteins. Together, these studies have demonstrated that native ataxin-3 exists within a complex conformational landscape, and provided evidence for a multi-domain aggregation mechanism for ataxin-3. This thesis has added to the growing body of evidence that suggests flanking domains influence the misfolding and aggregation of polyQ disease proteins.