Alternative Structures of α-Synuclein

Abstract

The object of our analysis is the structure of alpha-synuclein (ASyn), which, under in vivo conditions, associates with presynaptic vesicles. Misfolding of ASyn is known to be implicated in Parkinson's disease. The availability of structural information for both the micelle-bound and amyloid form of ASyn enables us to speculate on the specific mechanism of amyloid transformation. This analysis is all the more interesting given the fact that-Unlike in Aβ(1-42) amyloids-only the central fragment (30-100) of ASyn has a fibrillar structure, whereas, its N- and C-terminal fragments (1-30 and 100-140, respectively) are described as random coils. Our work addresses the following question: Can the ASyn chain-as well as the aforementioned individual fragments-adopt globular conformations? In order to provide an answer, we subjected the corresponding sequences to simulations carried out using Robetta and I-Tasser, both of which are regarded as accurate protein structure predictors. In addition, we also applied the fuzzy oil drop (FOD) model, which, in addition to optimizing the protein's internal free energy, acknowledges the presence of an external force field contributed by the aqueous solvent. This field directs hydrophobic residues to congregate near the center of the protein body while exposing hydrophilic residues on its surface. Comparative analysis of the obtained models suggests that fragments which do not participate in forming the amyloid fibril (i.e., 1-30 and 100-140) can indeed attain globular conformations. We also explain the influence of mutations observed in vivo upon the susceptibility of ASyn to undergo amyloid transformation. In particular, the 30-100 fragment (which adopts a fibrillar structure in PDB) is not predicted to produce a centralized hydrophobic core by any of the applied toolkits (Robetta, I-Tasser, and FOD). This means that in order to minimize the entropically disadvantageous contact between hydrophobic residues and the polar solvent, ASyn adopts the form of a ribbonlike micelle (rather than a spherical one). In other words, the ribbonlike micelle represents a synergy between the conformational preferences of the protein chain and the influence of its environment.

Keywords: A-synuclein; amyloid; fibril; hydrophobicity; misfolding; protein folding.

Figure 1

Structure of ASyn in its micelle-bound form (1XQ8); (A)—Theoretical (T, blue) and observed (O, red) hydrophobicity density distribution profiles calculated for the complete chain (1–140). Random coil C-terminal fragment (96–140) is highlighted in green, (B)—Theoretical (T, blue) and observed (O, red) hydrophobicity density distribution profiles calculated for the 1–95 fragment treated as an individual unit. Blue markers indicate residues No. 36–38, 47–49, 51–53, 67–69, 85–94 which exhibit local excess of hydrophobicity, while red markers correspond to local hydrophobicity deficiency (residues No. 20–22); (C)—3D presentation with color-coding corresponding to figures A and B.

Figure 2

3D structure of the amyloid form of ASyn, with a clearly distinguished amyloid-like section (30–100). Chain A has been marked red, while random coil fragments (1–29 and 101–140) are shown in green.

Figure 3

Theoretical (T, blue) and observed (O, red) hydrophobicity density distribution profiles for the ASyn amyloid (2N0A). Each chart shows two profiles for every chain from the complex (10 in total); (A)—Calculations performed for the entire complex (1–140). Random coil fragments (1–29, 101–140) are highlighted in green; (B)—Calculations performed for the amyloid fragment (30–100) treated as part of the complex; (C)—Calculations performed for the amyloid fragment (30–100) treated as an individual molecule.

Figure 4

Theoretical (T, blue) and observed (O, red) hydrophobicity density distribution profiles for chain E (central) from the ASyn amyloid (2N0A); (A)—Treated as part of the complex; (B)—Treated as a standalone structure.

Figure 5

3D visualizations along with theoretical (T, blue) and observed (O, red) hydrophobicity density distribution profiles for models of the 1–140 fragment produced by each software package and characterized by the lowest RD. The 95–102 fragment is highlighted in green; (A,B)—FOD_1_(202); (C,D)—ITASSER_1_(1); (E,F)—ROBETTA_1_(2).

Figure 6

3D visualizations along with theoretical (T, blue) and observed (O, red) hydrophobicity density distribution profiles for models involving the 1–30 fragment, produced by each software package and characterized by the lowest RD. (A,B)—FOD_1_(053); (C,D)—ITASSER_1_(3); (E,F)—ROBETTA_1_(4).

Figure 7

3D visualizations along with theoretical (T, blue) and observed (O, red) hydrophobicity density distribution profiles for models of the 30–100 fragment, produced by each software package and characterized by the lowest RD. (A,B)—FOD_1_(289); (C,D)—ITASSER_1_(1); (E,F)—ROBETTA_1_(2).

Figure 8

3D visualizations, theoretical (T, blue) and observed (O, red) hydrophobicity density distribution profiles for models of 100–140 fragment, produced by each software package and characterized by the lowest RD. (A,B)—FOD_1_(002); (C,D)—ITASSER_1_(5); (E,F)—ROBETTA_1_(1).

Figure 9

Presentation of the fibrillar part (30–100) of ASyn (2N0A); (A)—Theoretical (T, blue) and observed (O, red) hydrophobicity density distribution profiles for chain E. The teal chart represents the observed profile for the mutated sequence; (B)—3D visualization; Orange stars in Figure 9A, and orange/green spheres in Figure 9B correspond to loci of mutations which lower intrinsic hydrophobicity: A30P, E46K, H50Q and A53T. Green fragments correspond to residues No. 43–46 and 80, which–when eliminated from FOD computations—Produce a distribution consistent with the theoretical model (RD lowered from 0.506 to 0.490).

Figure 10

Visualization of the presented model, reduced to a single dimension for the sake of clarity: (A)—Gaussian distribution superimposed onto the protein molecule (T); (B)—Observed distribution in the molecule under consideration (O) (C)—Uniform distribution (R) (D)—Part of the Gaussian distribution plotted for the selected fragment (T_i–f) (E)—Observed distribution plotted for the selected fragment (O_i–f) (F)—Intrinsic distribution plotted for the selected fragment (H_i–f) (G)— RD scale for the example illustrated in Figures (A–C,H)—RD scale for the example illustrated in Figures (D–F) (“f” indicates “fragment”).