Fig. 1: Establishment of methods to identify astrocyte endfoot proteins.
From: Molecular profiling of brain endothelial cell to astrocyte endfoot communication in mouse and human
a Schematic of study aims. b Dot plot illustrating the endfoot-enriched dystrophin-associated protein complex (left) and other known endfoot proteins (gene names, right) in published endfoot proteome attempts36,37,38,39. Colored dots indicate presence of the protein, gray dots indicate absence (Stokum et al,. whole brain, black; Kameyama et al., whole brain, orange; Alonso-Gardon et al., whole brain, teal; Soto et al., striatum, pink). Horizontal bars indicate total proteins per dataset. c Top: AAV constructs generated and TurboID mechanism. GfaABC1D prom: minimal GFAP promoter, HA: Human Influenza Hemagglutinin tag peptide, NES: nuclear export signal, WPRE: Woodchuck Hepatitis Virus Posttranscriptional Regulatory Element. Bottom: Schematic of AAV expression protocol. d–f Line analysis demonstrating biotinylated proteins (streptavidin) in endfeet (AQP4) vs brain endothelial cells (BECs; PECAM1) in Astrocyte-TurboID injected mice. Example 10 µm dotted line (d) and its intensity profiles (e) shown. f Quantified distance between PECAM1-streptavidin, PECAM1-AQP4, and AQP4-streptavidin. A linear mixed-effects model (LMM) was used to account for repeated measures per mouse: Measurement~Distance + (1|Animal), N = 4 mice, 3 images/mouse; F(2,30) = 23.638, Tukey post hoc. Dots are average per mouse. Dotted lines connect measurements by mouse. Scale bar = 10 µm. g Vessel isolation schematic (top) and representative magnified image of isolated blood vessels with endfeet (AQP4, magenta) around the vasculature (lectin, green). Scale bar = 10 µm. h Representative tile-scan image of isolated microvessels stained with lectin (vessels) and Hoechst for nuclei (cyan). Scale bars = 250 µm (left) and 20 µm (right). i Normalized Hoechst+ nuclei count. N = 3 images from one vessel isolation: Unpaired two-tailed t-test, (t(4) = 12.59). Data are mean ± SEM. j Proportional composition of vessel sizes and types (grouped by diameter) in isolated microvessels calculated as area fraction of the total vessel area in (h). k Percent AQP4 coverage (fraction of lectin+ vessel area colocalizing with AQP4) by vessel diameter. Dots left of 0 indicate arteries/arterioles; right of 0, venules/veins. Data points are shaded by mouse. N = 108 vessels from 5 mice, 19–23 vessels/mouse. Diagrams below represent continuum of vessel diameter/type. l Representative images of vessels in (k) stained with AQP4, lectin, and alpha-smooth muscle actin (α-SMA; striated smooth muscle lining arterioles/arteries). Percent AQP4 coverage indicated in the first column. Scale bars = 20 µm. m Percent AQP4 coverage by vessel type. Horizontal line indicates median, and data points are shaded by mouse. LMM was used to account for repeated measures per mouse: Measurement~Vessel_type + (1|Animal), N = 108 vessels from 5 mice, 19–22 vessels per mouse with 1–8 vessels/category/mouse; F(4,100.69) = 41.664, Tukey post hoc. n Proportion of endfeet isolated from each vessel type. Exact sample sizes per animal and source data are provided in a Source Data file.
