Spatial Transcriptomics of Human Postmortem Brain Tissue Uncovers Molecular Pathways Associated with Amyloid‐Beta Plaques

Denis Avey,Bernard Ng,Artemis Iatrou,Ricardo A Vialle,Katia de Paiva Lopes,Shinya Tasaki,Jishu Xu,Sashini L De Tissera,Daniel Flood,David A. A Bennett,Yanling Wang
DOI: https://doi.org/10.1002/alz.077668
2023-12-01
Abstract:Abstract Background Alzheimer’s disease (AD) pathologies include extracellular amyloid‐beta (A‐beta) plaques, intracellular neurofibrillary tangles, neuronal cell loss, and chronic glial inflammation. Recent studies profiling gene expression in human postmortem brain tissue have uncovered cell‐type‐specific changes in AD. However, these studies did not provide simultaneous, direct measurements of AD pathologies in the native microenvironment. Method To this end, we applied the 10X Genomics Visium platform to post‐mortem dorsolateral prefrontal cortex (DLPFC) from 20 ROSMAP participants with or without AD dementia. In total, we obtained spatial transcriptomic (ST) data for ∼260,000 tissue microdomains (“spots” comprised of ∼1‐10 cells). Result We clustered spots and annotated clusters according to expression of genes with known enrichment in anatomical layers (i.e. cortical layers 1‐6, meninges, and white matter). We deconvoluted the cell type(s) comprising each spot using cell2location, and found significant changes in the abundance of specific cell types in AD. Using a two‐stage pseudobulk regression approach, we identified hundreds of genes whose spatial expression is significantly associated with nearby A‐beta plaques and/or astrocyte reactivity (GFAP). A‐beta‐associated genes are enriched for diverse biological processes, including translation, metabolism, inflammation, gliosis, and immune cell infiltration. We also detected genes that distinguish A‐beta plaques surrounded by reactive glial nets. We validated the spatial expression pattern of a subset of plaque‐associated genes by RNAScope. Conclusion In summary, our study reveals genes and pathways that shape molecular states in the spatial context of AD pathology. Our data represent a comprehensive application of ST technology in AD brains, providing a valuable resource for the research community.
clinical neurology
What problem does this paper attempt to address?