Kinetic diversity of amyloid oligomers

Abstract

The spontaneous assembly of proteins into amyloid fibrils is a phenomenon central to many increasingly common and currently incurable human disorders, including Alzheimer's and Parkinson's diseases. Oligomeric species form transiently during this process and not only act as essential intermediates in the assembly of new filaments but also represent major pathogenic agents in these diseases. While amyloid fibrils possess a common, defining set of physicochemical features, oligomers, by contrast, appear much more diverse, and their commonalities and differences have hitherto remained largely unexplored. Here, we use the framework of chemical kinetics to investigate their dynamical properties. By fitting experimental data for several unrelated amyloidogenic systems to newly derived mechanistic models, we find that oligomers present with a remarkably wide range of kinetic and thermodynamic stabilities but that they possess two properties that are generic: they are overwhelmingly nonfibrillar, and they predominantly dissociate back to monomers rather than maturing into fibrillar species. These discoveries change our understanding of the relationship between amyloid oligomers and amyloid fibrils and have important implications for the nature of their cellular toxicity.

Keywords: Alzheimer’s; amyloid; kinetics; modeling; oligomers.

Fig. 1.

(A) The different possible reactions that can produce and deplete nonfibrillar amyloid oligomers. Oligomers (concentration $S$) may be generated by free association of monomeric protein (concentration $m$) with rate constant $k_{\text{o}1}$. Once formed, they may dissociate back to monomers (rate constant $k_{\text{d}1}$) or undergo a structural conversion process to generate fibrillar species (concentration $P$) with rate constant $k_c$. Reverse conversion of fibrils to oligomers is neglected as it is found experimentally that fibrils are far more thermodynamically stable than nonfibrillar oligomers. (B) The coarse-grained reaction network describing oligomer-mediated fibril formation, represented in Petri net form (34), with reactions represented by boxes and chemical species of interest as circles. Oligomers, and reactions involving them, are highlighted in red. Both oligomer formation through monomer association and oligomer dissociation may in some cases be catalyzed by the surfaces of existing fibrils (fibril mass concentration $M$) with rate constants $k_{\text{o}2}$ and $k_{\text{d}2}$, respectively. Once formed, fibrillar species undergo rapid elongation by monomer addition (rate constant $k_{+}$), increasing the fibril mass concentration. They may also fragment to generate new fibrils (rate constant $k_{-}$), increasing $P$.

Fig. 2.

Global fits of experimental kinetic data on aggregating protein systems to the analytical models derived in SI Appendix, using the reaction steps identified in Table 1. Soluble oligomer concentration (measured using single molecule techniques) and fibril mass concentration (usually measured using thioflavin T [ThT] dye fluorescence) are fitted simultaneously, with fitting parameters summarized in SI Appendix, Table S1. See Materials and Methods for summaries of experimental techniques and fitting methodologies. (A) Tau data taken from ref. ; type-A oligomers (pale green) are on-pathway and type-B oligomers (dark green) off-pathway. (B) Ure2 data from ref. . (C and D) Aβ40 and Aβ42 oligomer data from ref. . Fibril data are from earlier papers (26, 38); for visual clarity, only one initial monomer concentration of the many available is shown for each. (E) αS data from the first quantitative FRET-based study (ref. 31); experimental accuracy thus lower than more recent studies with optimized protocols, resulting in lower-quality fits. Both oligomer and fibril concentrations measured via FRET. Initial monomer concentrations are 10 (lightest), 35, 70, and 140 $\mathrm{\mu }$M (darkest). (F) Data on αS aggregated in the presence of nanobodies NbSyn2 (darker) and NbSyn87 (lighter). Data are taken from ref. . Note that for most proteins, the reaction processes involved in aggregation depend sensitively on the reaction conditions. See SI Appendix, Table S4 for a summary of the conditions under which the data shown here were collected. $*$Oligomer concentrations in E and F were measured in micromoles per liter.

Fig. 3.

(A) Categorizing oligomers from Fig. 2 by their key properties. “Persistence” is the kinetic stability of the oligomers as indicated by their half-life $t_h=\ln \left(2\right)/k_e$ (with $k_e=k_c+k_d$). “Abundance” is the maximal rate of formation $\alpha$ divided by the maximal rate of depletion $k_e$ and indicates the maximal steady-state oligomer concentration. “Productivity” indicates the relative contributions of conversion and dissociation to overall oligomer depletion, defined as $k_c/k_e$. (B) Of the systems hitherto studied, in every case, the oligomers dissociate more rapidly than they convert, and only αS oligomers, Aβ42 oligomers, and type-B (off-pathway) tau oligomers persist longer than the corresponding monomeric protein. These oligomers are also relatively abundant, as might be expected. $*$Prion protein PrP data were taken from ref. (proteinase K-sensitive species); abundance was not measured.