d-Amino Acid Pseudopeptides as Potential Amyloid-Beta Aggregation Inhibitors

Abstract

A causative factor for neurotoxicity associated with Alzheimer's disease is the aggregation of the amyloid-β (Aβ) peptide into soluble oligomers. Two all d-amino acid pseudo-peptides, SGB1 and SGD1, were designed to stop the aggregation. Molecular dynamics (MD) simulations have been carried out to study the interaction of the pseudo-peptides with both Aβ_13⁻23 (the core recognition site of Aβ) and full-length Aβ_1⁻42. Umbrella sampling MD calculations have been used to estimate the free energy of binding, ∆G, of these peptides to Aβ_13⁻23. The highest ∆G_binding is found for SGB1. Each of the pseudo-peptides was also docked to Aβ_1⁻42 and subjected up to seven microseconds of all atom molecular dynamics simulations. The resulting structures lend insight into how the dynamics of Aβ_1⁻42 are altered by complexation with the pseudo-peptides and confirmed that SGB1 may be a better candidate for developing into a drug to prevent Alzheimer's disease.

Keywords: Alzheimer’s; amyloid-beta; d-amino acids; inhibitors; molecular dynamics; umbrella sampling.

Figure 1

The most stable structures of SGB1 (a) and SGD1 (b) obtained from the MD simulation. The numbering is from the N-terminal amino-acid. (c) Reference geometry of Aβ₄₂. (d) Extended geometry of Aβ₄₂ for docking. The R region of Aβ is highlighted as a red cartoon with side chains.

Figure 2

Binding of SGB1 to R at (a) R^T-SGB1, on the top edge, (b) R^B-SGB1, on the bottom edge. The right-hand axis is for PMF curve (purple line) in kJ/mol, the error bars are included (±1σ). The left-hand axis has two functions: a numerical count of the average intermolecular H-bonds (blue line) and a distance measure in nm for the separations of the polar charged residues (salt-bridges), Lys16-glu8′ (green line) and minimum distances for Glu22-daba1′, Glu22-orn2′, Asp23-daba1′, Asp23-orn2′ (red line). The horizontal axis is the separation of the centers of mass in nm.

Figure 3

Binding of SGD1 to R at (a) R^T-SGD1, on the top, (b) R^B-SGD1, in the bottom. The right-hand axis is for PMF curve (purple line) in kJ/mol, the error bars are included (±1σ). The left-hand axis is the average intermolecular H-bonds (blue line) and salt-bridge distances at the polar charged residues in nm; Lys16-glu8′ (green line) and minimum distances for Glu22-daba1′, Glu22-orn2′, Asp23-daba1′, Asp23-orn2′ (red line). The horizontal axis is the separation of the centers of mass in nm.

Figure 4

The homodimers; the structures and the related PMF curve: (a) SGB1-self (b) SGD1-self.

Figure 5

Aβ₄₂^T-SGB1-a; The cluster Ba1 structure is shown, which persist through almost 1 μs. The vertical axis is the cluster number count based on population, where cluster 1 has the highest population. The horizontal axis is simulation time in nanoseconds.

Figure 6

Aβ₄₂^B-SGB1-b; Three major clusters are depicted; Cluster Bb1, Bb2 and Bb3. The vertical axis is the cluster number count based on population, where cluster 1 has the highest population. The horizontal axis is simulation time in nanoseconds.

Figure 7

Aβ₄₂^B-SGB1-c; cluster Bc1 with the highest population of similar conformations and Bc4, the last significant cluster in the trajectory are shown. The vertical axis is the cluster number count based on population, where cluster 1 has the highest population. The horizontal axis is simulation time in nanoseconds.

Figure 8

Aβ₄₂^T-SGD1-a: represented the two most populated clusters; cluster Da1 and cluster Da2 in the trajectory.

Figure 9

Aβ₄₂^B-SGD1-b: The two ensembles with the highest population are shown. Both Cluster Db1 and cluster Db2 disappear in the middle of trajectory but return for the last 300 and 200 ns of simulation, respectively.

Figure 10

Aβ₄₂-SGD1-c: The first cluster, Dc1, exists almost through all of trajectory, with no β-sheet with R region. The total number of clusters through 1 μs was only 40.

Figure 11

Aβ^B-SGD1-d. Clusters Dd1, Dd2, Dd3 and Dd5 are shown.