Fig. 1. The hypothesis of amylin-induced impairment of brain Aβ clearance.

Putative amylin function (green panel) and pathology (red panels) (a) along with work flow and methods (b). c Blood amylin concentrations in dementia (DEM; n = 19), mild cognitive impairment (MCI; n = 19), and cognitively unimpaired (CU; n = 42) individuals. d Pairwise correlation coefficient (r) between amylin and insulin concentrations in same blood samples as in (c). e Flow cytometry sorting of amylin positive (Q₂) or negative (Q₁) circulating CD14⁺ monocytes in blood with lower quartile vs. upper quartile amylin concentrations. f Confocal microscopic images showing amylin engulfed in CD14⁺ monocytes (n = 3). Pairwise correlation coefficient (r) between amylin and Aβ₄₂ concentrations in human brains, including persons with sAD (n = 42) and without AD (n = 18) (g), and between matched antemortem plasma amylin concentration and amylin concentration in autopsied brain tissue, including persons with sAD (n = 12) and without AD (n = 8) individuals (h) (potential outliers were removed from the analysis; shown in Supplemental Fig. S1e). i IHC analysis using anti-amylin (brown) and anti-Aβ (green) antibodies on serial sections from a sAD brain (n = 18). j Confocal microscopic analysis and amylin-Aβ proximity ligation assay (PLA) showing vascular amylin-Aβ deposits in a sAD brain. IHC analysis of fAD brains showing Aβ deposits in perivascular spaces and vessel walls with amylin accumulation within the lumen (k, l), and amylin deposits in the vessel wall (m) or vessel wall (n), and Aβ deposits in perivascular spaces (n = 32). Data are presented as box and whiskers or correlation analyses; Kruskal–Wallis one-way of variance, Data are means ± SEM.