Fig. 8

APOE expression in hiNS is driven by hiMG. A APOE immunofluorescence in hiNS slices reveals large quantities of extracellular APOE in 5w Aβ hiNS(+). B (Upper part) Quantification of APOE immunofluorescence (ctrl hiNS(−/+): N = 4 each; 3w Aβ hiNS(−/+): N = 5 each; 5w Aβ hiNS(−/+): N = 5 each). (Lower part) Intracellular APOE identified in hiMG (stained for IBA1, cell labeled ‘mg’) and astrocytes in 5w Aβ hiNS(+). C: Soluble APOE and Aβ levels in hiNS supernatant at DIV 60 measured by APOE western blot and Aβ ELISA (4G8) (N = 3; pooled supernatant from six individual hiNS each). APOE was only detectable in supernatant of hiNS containing hiMG. At 5w Aβ hiNS(+) soluble Aβ compared to 5w Aβ hiNS(−)) and APOE levels compared to 3w Aβ hiNS(+)) were significantly reduced. Gel loading was normalized to protein concentrations and the raw intensities were compared with each other. D Co-staining of APOE with 4G8 indicates that extracellular APOE is in close proximity to Aβ aggregates close to the surface of hiNS (N = 3 hiNS each). E snRNA-seq clusters of neurons, microglia and astrocytes from ctrl, 3w Aβ and 5w Aβ hiNS(+). hiMG express APOE independent of Aβ exposure, while astrocytic and neuronal APOE expression depends on the presence of hiMG after 5w Aβ exposure. Note that hiMG expression of APOE remains higher than that of neurons/astrocytes. F: Immunofluorescence staining in cleared and expanded human AD post-mortem brain sections confirms the close proximity of APOE with Aβ deposits and intracellular APOE in astrocytes and microglia in AD (N = 3, 1 sporadic AD, 2 familial AD). All statistical testing performed using two-way ANOVA followed by Holm-Sidak post-hoc tests on logarithmically transformed data