The role of water in promoting the formation of protofilaments (the basic building blocks of amyloid fibrils) is investigated using fully atomic molecular dynamics simulations. Our model protofilament consists of two parallel β-sheets of Alzheimer Amyloid-β 16-22 peptides (Ac-K 16 -L 17 -V 18 -F 19 -F 20 -A 21 -E 22 -NH 2 ). Each sheet presents a distinct hydrophobic and hydrophilic face and together self-assemble to a stable protofilament with a core consisting of purely hydrophobic residues (L 17 ,F 19 ,A 21 ), with the two charged residues (K 16 , E 22 ) pointing to the solvent. Our simulations reveal a subtle interplay between a water mediated assembly and one driven by favorable energetic interactions between specific residues forming the interior of the protofilament. A dewetting transition, in which water expulsion precedes hydrophobic collapse, is observed for some, but not all molecular dynamics trajectories. In the trajectories in which no dewetting is observed, water expulsion and hydrophobic collapse occur simultaneously, with protofilament assembly driven by direct interactions between the hydrophobic side chains of the peptides (particularly between F-F residues). For those same trajectories, a small increase in the temperature of the simulation (on the order of 20 K) or a modest reduction in the peptide-water van der Waals attraction (on the order of 10%) is sufficient to induce a dewetting transition, suggesting that the existence of a dewetting transition in simulation might be sensitive to the details of the force field parametrization.

Alzheimer's disease is a debilitating neurodegenerative disorder associated with the abnormal self-assembly of amyloid-beta (Abeta) peptides into fibrillar species. N-methylated peptides homologous to the central hydrophobic core of the Abeta peptide are potent inhibitors of this aggregation process. In this work, we use fully atomistic molecular dynamics simulations to study the interactions of the N-methylated peptide inhibitor Abeta16-20m (Ac-Lys(16)-(Me)Leu(17)-Val(18)-(Me)Phe(19)-Phe(20)-NH(2)) with a model protofilament consisting of Alzheimer Abeta16-22 peptides. Our simulations indicate that the inhibitor peptide can bind to the protofilament at four different sites: 1), at the edge of the protofilament; 2), on the exposed face of a protofilament layer; 3), between the protofilament layers; and 4), between the protofilament strands. The different binding scenarios suggest several mechanisms of fibrillogenesis inhibition: 1), fibril inhibition of longitudinal growth (in the direction of monomer deposition); 2), fibril inhibition of lateral growth (in the direction of protofilament assembly); and 3), fibril disassembly by strand removal and perturbation of the periodicity of the protofilament (disruption of fibril morphology). Our simulations suggest that the Abeta16-20m inhibitor can act on both prefibrillar species and mature fibers and that the specific mechanism of inhibition may depend on the structural nature of the Abeta aggregate. Disassembly of the fibril can be explained by a mechanism through which the inhibitor peptides bind to disaggregated or otherwise free Abeta16-22 peptides in solution, leading to a shift in the equilibrium from a fibrillar state to one dominated by inhibitor-bound Abeta16-22 peptides.

No abstract