A reference protein atlas of the adult mouse brain vasculature

Abstract

Current methods for assessing low-abundance proteins in individual cells are limited. As a result, cell functions are often inferred from single-cell RNA sequencing (scRNA-seq) data, which can be misleading due to the poor cross-gene correlation (different genes in the same cells) between messenger RNA (mRNA) and protein levels. To address this issue, we used isolated brain vascular fragments to study the blood–brain barrier. We applied a combination of deep bulk proteomic analysis, a proteomic ruler approach, and scRNA-seq, assuming a high within-gene correlation (same gene in different cells) between mRNA and protein. This approach allowed us to estimate protein copy numbers per cell for 9,940 proteins across eight cell types, including endothelium, smooth muscle, pericytes, fibroblasts, and microglia. We also evaluated protein abundance in astrocyte end-feet attached to the vessel fragments. Our data are available through an Online Database, providing a searchable resource and reference protein atlas for future studies of neurovascular proteomics in health and disease.