Models of α-Syn with fatty acids.
Fatty acids facilitate formation of α-Syn oligomers [57] but not β-Syn oligomers [58]. van Diggelen et al. [18] have isolated two conformationally distinct α-Syn oligomers: one complexed with the naturally occurring polyunsaturated fatty acid docosahexaenoic acid (DHA) and the other with the lipid peroxidation product 4-hydroxynonenal (HNE). The sequence-independent polyclonal A11 antibody recognizes numerous soluble amyloids, possibly by recognizing antiparallel β-sheets. It recognizes HNE-αSOs (α-Syn oligomers), but not DHA-αSOs. They report that both types of oligomers have a relatively high content of β-structure; DHA-αSOs have 62.3% random coil, 12.8% α-helical and 12.7% β-sheet and HNE-αSOs have 45% random coil, 14.6% α-helical and 28.4% β-sheet. Circular Dichroism and FTIR spectra analyses indicate the presence of antiparallel β-structure in both types of oligomers. Transmission Electron Microscopy images (Fig. 8 g & h) suggest that both types of oligomers are spherical with average diameters of 20.0±4.6 nm for DHA-αSOs and 19.5±6.3 nm for HNE-αSOs. However, these images also have much smaller bright spots that appear to self-associate. The circles in Fig. 8 indicate that the spacings between these spots are consistent with models of interactions between the Nt-NAC beta barrels with diameters of 6.0 nm for Type 2A hexamers (red) for HNE-αSOs and 7.1 nm for Type 2A octamers (yellow) for DHA-αSOs. Like EM images of numerous other oligomers (see Fig. 4), strings or clusters of apparently dense particles are surrounded by other material; possibly disordered Ct domains and entrapped monomers and/or fatty acids that may compose the remaining portions of the 20 nm diameter spheres.
Unlike models described above, the concentric β-barrels models that follow have central pores. The ratio of strands/monomer of the inner barrel to the surrounding barrel decreases as the number of strands increases: for the tetramer described above the ratio was 2/5, for the hexamers described below it is 3/5, and for octamers to dodecamers it is 3/4. These pores may act as lipid and fatty-acid binding sites; they are sufficiently large to surround alkyl chains and their entrances contain positively charged lysines that should interact favorably with negatively charged head-groups. It is difficult to predict based on the data whether the smaller HNE-oligomers are more consistent with Type 2A (Fig. 8 a-c) hexamers or Type 2P hexamers described later in the annular protofibril section because the Nt-NAC barrel region of both models have the same dimensions. The Type 2A model has more antiparallel β-strands but the Type 2P models are more consistent with the absence of A11 binding. The hexamer model differs from our other models in that Sy7 forms a continuous β-strand that is part of a β-barrel and Sy7 and Sy8 form a β-hairpin. Perhaps the absence of Sy7 in β-Syn explains why its oligomerization is unaffected by fatty acids [58]. The data for the DHA oligomer are more consistent with a Type 2A octamer (Fig. 8 d-f).