Archives
Electron paramagnetic resonance studies of mixtures
Electron paramagnetic resonance studies of mixtures of Aβ1–40 and Aβ1–42 suggest that interlaced fibrils might also form [40] where Aβ1–40 and Aβ1–42 is found within the same fibril. Considering the Aβ fold rather than the sequence, these observations imply two mechanisms that might lower the total in vivo load of Aβ assemblies that have the properties of Aβ1–42 fibrils – (i) the impaired ability of monomeric Aβ1–40 to be incorporated into the ends of Aβ1–42 fibrils and (ii) the ability of Aβ1–42 monomers to adopt the properties of the less pathogenic Aβ1–40 when incorporated into Aβ1–40 fibrils via cross-templating. Although speculative, this might be of importance, because Aβ1–40 monomers are found in vivo at a 7-fold molar excess over Aβ1–42 monomers [41]. Fibrillar Aβ1–40 and Aβ1–42 structures have been extensively studied using solid-state nuclear magnetic resonance and have been shown to have significant structural differences. Several studies of Aβ1–40 are consistent with an U-shaped strand-loop-strand arrangement in which two β-strands, spanning residues 12–19 and 30–40 are separated by a loop region that forms an in register parallel β-sheet assembly [11], [42], [43]. Within this form both a two-fold [11], [43] and a three-fold symmetry [44] has been shown. In both models the first β-strand, residues 12–19, faces the solvent while the β-strand intervening residues 30–40 is buried within the structure. The fibrillar architecture of Aβ1–42 has also been extensively investigated [10], [45], [46], and these Zoniporide hydrochloride mg also fold into a parallel β-strand arrangements, but the fibrillar assembly of Aβ1–42 differs significantly in several respects. Recent studies reported that the fibril core consists of four β-strands in an S-shaped fold. The C-terminus of Aβ1–42 forms an additional coil-β-sheet structure that spans residues Met35-Ala42, and here a salt bridge forms between Lys28 and the carboxyl end of the peptide [8]. The template incorporation of Aβ1–42 monomers into the ends of Aβ1–40 fibrils (as per an Aβ1–42 fibrillar model description) is not compatible with any of the reported Aβ1–40 fibril structures. This therefore supports that Aβ1–42 monomers, upon being cross-templated by the fibrillar ends of Aβ1–40 fibrils, instead adopt the properties (and likely also a structure) similar to an Aβ1–40 fibril. Currently, we cannot provide a structural explanation for why Aβ1–40 is poorly incorporated into the fibrillar form of Aβ1–42. Moreover, an open question remains whether the observed small signal from probing Aβ1–40 monomers on Aβ1–42 fibrils actually represents a monomer to fibril-end interaction. An alternative explanation is provided by Saric, A. et al., [47] suggesting that the initial fast decay seen in the SPR sensograms instead is caused by a lateral interaction between the monomer and the immobilized fibril and that the interaction is a part of the SCSN [47]. Such interaction would hence provide an explanation to why a small signal can be seen when probing the fibrillar form of Aβ1–42 with Aβ1–40 monomers and imply that the barrier preventing Aβ1–40 to incorporate into Aβ1–42 fibrils is even stronger and more stringent than indicated by the SPR results. The ability of our system to discriminate between the fibrillar properties of Aβ1–40 and Aβ1–42 using SPR interestingly also allowed us to determine whether the properties of the parental fibril are also transferred when fibrils are cross-nucleated via SCSN. Our results clearly showed that fibrils derived through cross-nucleation via SCSN in this system do not adopt the properties of the parental fibril, but instead retain the same, or a very similar, fibrillar form as they would adopt through primary nucleation in the absence of pre-existing fibrils. This suggests that nuclei formation via SCSN is fundamentally different from the template-dependent elongation process. Fig. 5 summarizes the different pathways of fibril formation. In a previous study based on mixing Aβ1–40 and Aβ1–42 in solution, it was suggested that Aβ1–40 and Aβ1–42 might only interact at the primary nucleation level and that only the smallest aggregates might exist as mixed species [37]. These authors also suggested that fibril elongation and fibril-catalyzed secondary nucleation are highly specific events resulting in the formation of distinct fibrils composed of either Aβ1–40 or Aβ1–42 and that cross-seeding does not occur. Differences in experimental conditions probably explain this discrepancy; in this work, we used physiological ion strength, while the previous investigation [37] was performed under low ionic strength. We also used a fully native Aβ sequence, while the previous study was based on peptides that contained a non-native methionine prior to the first residue in the Aβ sequence.