A possible mechanism for the enhanced toxicity of beta-amyloid protofibrils in Alzheimer's disease

Abstract

The amyloid-beta peptide (Aβ) is a driver of Alzheimer's disease (AD). Aβ monomers can aggregate and form larger soluble (oligomers/protofibrils) and insoluble (fibrils) forms. There is evidence that Aβ protofibrils are the most toxic form, but the reasons are not known. Consistent with a critical role for this form of Aβ in AD, a recently FDA-approved therapeutic antibody targeted against protofibrils, lecanemab, slows the progression of AD in patients. The plasma contact system, which can promote coagulation and inflammation, has been implicated in AD pathogenesis. This system is activated by Aβ which could lead to vascular and inflammatory pathologies associated with AD. We show here that the contact system is preferentially activated by protofibrils of Aβ. Aβ protofibrils bind to coagulation factor XII and high molecular weight kininogen and accelerate the activation of the system. Furthermore, lecanemab blocks Aβ protofibril activation of the contact system. This work provides a possible mechanism for Aβ protofibril toxicity in AD and why lecanemab is therapeutically effective.

Keywords: Alzheimer’s; beta-amyloid; coagulation.

Fig. 1.

Aβ protofibrils are the most effective activators of the contact system in human plasma compared to other forms of Aβ. (A) Electron micrographs of various Aβ42 species. (Scale bar, 200 nm.) (B) Aβ42 was incubated with human plasma, and Western blots were run under reducing conditions. FXII activation was determined using an antibody that detects the activated form, FXIIa. PK activation was determined by the presence of PKa/inhibitor complexes that only form with activated PK. Aβ protofibrils induced complete HK cleavage and activation of FXII and PK. Transferrin (TRF) was a loading control. (C–E) Quantitation of Western blots. Aβ fibrils induced some contact system activation but significantly less than that of Aβ protofibrils. Aβ monomers and small oligomers did not induce contact system activation. AU, arbitrary units. (F) FXIIa and (G) PKa activities were measured by chromogenic assays. (H) Bradykinin levels were measured by ELISA. (I) Clotting time was determined in normal human plasma. Results are representative of three independent experiments. Data were analyzed by one-way ANOVA with Tukey’s multiple comparisons analysis and are denoted as mean ± SEM. *P ≤ 0.05, **P ≤ 0.01, ****P ≤ 0.0001. (J) Binding of biotinylated human HK (B-HK) to Aβ was assessed by ELISA. Binding constants were as follows: Aβ protofibrils/HK, K_D = 1.3 nM; Aβ fibrils/HK, K_D ~ 268 nM. Other binding interactions were too weak to be measured. (K) Aβ42 protofibrils pulled down FXII and HK in human plasma, but minimal to no FXII or HK was pulled down with other Aβ species. Transferrin (TRF) was used as a negative control.

Fig. 2.

Lecanemab prevents Aβ protofibrils from activating the contact system in normal human plasma. Plasma was incubated with or without Aβ42 protofibrils in the presence of vehicle, lecanemab, human IgG control (Hu-IgG), or G2-11, another anti-Aβ antibody. Analyses as in Fig. 1. (A) Contact system activation was examined by Western blotting. (B–D) Quantitation of Western blots. AU, arbitrary units. (E and F) Activity assays for FXIIa and PKa. (G) Bradykinin levels. (H) Analysis of proteins bound to biotinylated Aβ42. Both FXII and HK bound to Aβ42, which was blocked by lecanemab but not by human control IgG. TRF was used as a negative control. Results are representative of three independent experiments in A–G and four independent experiments in H. (I) Intrinsic clotting. n = 8/group. Lecanemab blocked Aβ-induced accelerated clotting. Data were analyzed by one-way ANOVA with Tukey’s multiple comparisons test and are denoted as mean ± SEM. **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001.