Fig. 5.

Common de novo regions of cell and proteopathy composition are identified across different individuals. a Computational workflow used to isolate common regions of pathology (de novo regions) across CN, CIND, and ADD hippocampi independent of patient diagnosis. b De novo regions displayed across the entire tissue of each sample, CN, CIND, and ADD. N = Neuronal dominant region, G = glial (astrocyte/oligodendrocyte) dominant region, M = Mixed proteopathy region, AP = Aβ Plaque dominant region, TT = Tau NFT-NTs dominant region. c Relative composition of each de novo region across each individual hippocampus. d Relative composition of each cell and proteopathy subtype within each de novo region of each individual. e UMAP projection of all cells and proteopathies, calculated using much of the full panel (list in methods). De novo regions (left), cell and proteopathy classification (middle) and individual (right). Black ellipse denotes area of clustered cell lineages associated with TT region. f Ratio of proteopathy-associated cells to proteopathy-free cells in each de novo region of each individual. Ratio of 1 indicates an equal number of proteopathy cells to proteopathy-free cells. Tau tangle cell associations (Top) and Aβ plaque cell associations (Bottom). Top callout indicates high levels of tau tangle overlap in TT region; bottom callout indicates high Aβ plaque overlap in AP region. g Synaptic proteopathy presence. For each de novo region, all positive synaptic pixels with positive proteopathy signal (Top: tau, Bottom: amyloid) relative to all positive synaptic pixels. Excitatory synaptic channels (SYP(+) PSD95(+), and either VGLUT1(+) or VGLUT2(+)) and inhibitory synaptic channels (SYP(+), VGAT(+), GAD(+)) considered. Abbreviations: CN, cognitively normal; CIND, cognitive impairment no dementia; ADD, AD dementia; N, Neuronal dominant; G, Glial dominant; M, Mixed disease; AP, Aβ plaque dominant; TT, PHF1-TAU NFT-NTs dominant