Post-translational modification sites are present in hydrophilic cavities of alpha-synuclein, tau, FUS, and TDP-43 fibrils: A molecular dynamics study.

Hydration plays a crucial role in the refolding of intrinsically disordered proteins into amyloid fibrils; however, the specific interactions between water and protein that may contribute to this process are still unknown. In our previous studies of alpha-synuclein (aSyn), we have shown that waters confined in fibril cavities are stabilizing features of this pathological fold; and that amino acids that hydrogen bond with these confined waters modulate primary and seeded aggregation. Here, we extend our aSyn molecular dynamics (MD) simulations with three new polymorphs and correlate MD trajectory information with known post-translational modifications (PTMs) and experimental data. We show that cavity residues are more evolutionarily conserved than non-cavity residues and are enriched with PTM sites. As expected, the confinement within hydrophilic cavities results in more stably hydrated amino acids. Interestingly, cavity PTM sites display the longest protein-water hydrogen bond lifetimes, three-fold greater than non-PTM cavity sites. Utilizing the deep mutational screen dataset by Newberry et al. and the Thioflavin T aggregation review by Pancoe et al. parsed using a fibril cavity/non-cavity definition, we show that hydrophobic changes to amino acids in cavities have a larger effect on fitness and aggregation rate than residues outside cavities, supporting our hypothesis that these sites are involved in the inhibition of aSyn toxic fibrillization. Finally, we expand our study to include analysis of fibril structures of tau, FUS, TDP-43, prion, and hnRNPA1; all of which contained hydrated cavities, with tau, FUS, and TDP-43 recapitulating our PTM results in aSyn fibril cavities.

Threonine Cavities Are Targetable Motifs That Control Alpha-Synuclein Fibril Growth.

Recent high-resolution structures of alpha-synuclein (aSyn) fibrils offer promise for rational approaches to drug discovery for Parkinson's disease and Lewy body dementia. Harnessing the first such structures, we previously used molecular dynamics and free energy calculations to suggest that threonines 72 and 75─which line water-filled cavities within the fibril stacks─may be of central importance in stabilizing fibrils. Here, we used experimental mutagenesis of both wild-type and A53T aSyn to show that both threonine residues play important but surprisingly disparate roles in fibril nucleation and elongation. The T72A mutant, but not T75A, resulted in a large increase in the extent of fibrillization during primary nucleation, leading us to posit that T72 acts as a "brake" on run-away aggregation. An expanded set of simulations of five recent high-resolution fibril structures suggests that confinement of cavity waters around T72 correlates with this finding. In contrast, the T75A mutation led to a modest decrease in the extent of fibrillization. Furthermore, both T72A and T75A completely blocked the initial fibril elongation in seeded fibrillization. To test whether these threonine-lined cavities are druggable targets, we used computational docking to identify potential small-molecule binders. We show that the top-scoring hit, aprepitant, strongly promotes fibril growth while specifically interacting with aSyn fibrils and not monomer, and we offer speculation as to how such compounds could be used therapeutically.