Fig. 3: Gene-set analysis uncovers a link between impaired meningeal lymphatics and microglial activation in AD.
From: Meningeal lymphatics affect microglia responses and anti-Aβ immunotherapy
a, SNPs in the 1-Mb region of genes that are highly expressed in meningeal LECs are in cis-expression quantitative expression loci (cis-eQTL) in microglia from the parietal cortex of human brains (controls and individuals with sporadic AD). Violin plots show the distribution of gene expression and the first, median and third quantiles for each genotype. b, tSNE plots demonstrating segregation of human or 5xFAD microglia into five clusters upon cross-species RNA-seq data integration (Extended Data Fig. 10). c, d, Violin plots showing expression of homeostatic genes (c) and activation genes (d) in different human microglial clusters. e, Graph showing cluster proportions in microglia from humans without AD or with presymptomatic, familial or sporadic AD, and in microglia from Vis + mIgG and Vis/photo + mIgG 5xFAD mice. Proportions of microglia in clusters 1 (P = 0.00026), 2 (P = 0.00191) and 3 (P = 0.03535) were statistically different between Vis + mIgG and Vis/photo + mIgG groups. f, tSNE representation of the scaled average expression (range in scale bar) of the module of 54 human gene orthologues of the 5xFAD microglial gene signature of meningeal lymphatic dysfunction, with aggregated expression of randomly chosen control feature gene-sets subtracted. g, Violin plot showing each microglial cluster signature score obtained by partial residuals from linear mixed models. Data in b–g were obtained from the integrated analysis of the following numbers of microglia from each group: 5,462 (non-AD), 618 (presymptomatic AD), 4,548 (familial AD) and 6,461 (sporadic AD) (single-nucleus RNA-seq data) from human parietal lobes; 781 (Vis + mIgG) and 770 (Vis/photo + mIgG) (single-cell RNA-seq data) from 5xFAD mice. Results involving human microglia were corrected for age of death, sex and disease status; statistically significant differences were calculated by two-proportion Z-test (e) and by linear mixed models (g; individual comparisons between the score in cluster 1 and the other scores).
