Integrating spatial transcriptomics and snRNA-seq data enhances differential gene expression analysis results of AD-related phenotypes.

Abstract

Spatial transcriptomics (ST) data provide spatially informed gene expression profiles. However, power is limited for spatially informed differential gene expression (DGE) of complex diseases such as Alzheimer disease (AD), due to small sample sizes of ST data. Conversely, single-nucleus RNA sequencing (snRNA-seq) data offer larger sample sizes for cell-type-specific (CTS) analyses but lack spatial information. Here, we integrated ST and snRNA-seq data to enhance the power of spatially informed CTS DGE analysis of AD-related phenotypes. We first utilized the CeLEry tool to infer six cortical layers of ∼1.5 million cells in the snRNA-seq data that were profiled from the dorsolateral prefrontal cortex (DLPFC) tissue of 436 postmortem brains. Then, we conducted cortical layer- and cell-type-specific (LCS) and CTS DGE analyses based on the linear mixed model, for β-amyloid, tangle density, and cognitive decline. We identified 138 LCS significant genes with false discovery rate (FDR) q <0.05, including 103 for β-amyloid, 24 for tangle density, and 25 for cognitive decline. The majority of these LCS significant genes, including known AD risk genes such as APOE, KCNIP3, and CTSD, cannot be detected by CTS analyses. We also identified 2 genes shared across all 3 phenotypes and 10 shared between 2 phenotypes. Gene set enrichment analyses with the LCS DGE results of microglia in cortical layer 6 of β-amyloid identified 12 significant AD-related pathways. In conclusion, incorporating spatial information with snRNA-seq data enhanced the power of spatially informed DGE analyses. These identified LCS significant genes not only help illustrate the pathogenesis of AD but they also provide potential targets for developing therapeutics of AD.