Solid-State-NMR-Structure-Based Inhibitor Design to Achieve Selective Inhibition of the Parallel-in-Register β-Sheet versus Antiparallel Iowa Mutant β-Amyloid Fibrils.

Abstract

Solid-state nuclear magnetic resonance (ssNMR) spectroscopy has been widely applied to characterize the high-resolution structures of β-amyloid (Aβ) fibrils. While these structures provide crucial molecular insights on the deposition of amyloid plaques in Alzheimer's diseases (AD), ssNMR structures have been rarely used so far as the basis for designing inhibitors. It remains a challenge because the ssNMR-based Aβ fibril structures were usually obtained with sparsely isotope-labeled peptides with limited experimental constraints, where the structural models, especially the side-chain coordinates, showed restricted precision. However, these structural models often possess a higher accuracy within the hydrophobic core regions with more well-defined experimental data, which provide potential targets for the molecular design. This work presents an ssNMR-based molecular design to achieve selective inhibition of a particular type of Aβ fibrillar structure, which was formed with the Iowa mutant of Aβ with parallel-in-register β-sheet hydrophobic core. The results show that short peptides that mimic the C-terminal β-strands of the fibril may have a preference in binding to the parallel Aβ fibrils rather than the antiparallel fibrils, mainly due to the differences in the high-resolution structures in the fibril elongation interfaces. The Iowa mutant Aβ fibrils are utilized in this work mainly as a model to demonstrate the feasibility of the strategy because it is relatively straightforward to distinguish the parallel and antiparallel fibril structures using ssNMR. Our results suggest that it is potentially feasible to design structure-selective inhibitors and/or diagnostic agents to Aβ fibrils using ssNMR-based structural models.