Brain Vascular Multi-Omics Atlas Maps Genetic Risk for Alzheimer's and Stroke

Cerebrovascular dysfunction is central to many neurological disorders, yet the specific cell types and molecular mechanisms through which noncoding genetic variants drive disease have remained largely unresolved. This study addresses that gap by introducing MultiVINE-seq, a multimodal single-cell sequencing method designed to simultaneously capture RNA expression and chromatin accessibility (ATAC-seq) in vascular, perivascular, and immune cells isolated from human brain tissue.

The team applied MultiVINE-seq to postmortem brain samples from 30 individuals, constructing a comprehensive multi-omic atlas of the human brain vasculature. This atlas encompasses endothelial cells, smooth muscle cells, pericytes, fibroblasts, and various brain-resident and infiltrating immune cell populations. The simultaneous profiling of transcriptomes and open chromatin regions allowed the researchers to infer gene regulatory networks and identify active enhancer elements in each cell type.

By integrating this atlas with GWAS summary statistics for cerebrovascular diseases and Alzheimer's disease, the researchers performed cell-type-specific enrichment and fine-mapping analyses. They linked thousands of noncoding risk variants to putative target genes and regulatory elements, including 2,605 variants that had not previously been mapped to any cell type or gene. This represents a substantial expansion of the functional annotation of neurological disease GWAS loci.

A key discovery is the mechanistic divergence between cerebrovascular and neurodegenerative disease variants. Cerebrovascular disease risk variants are concentrated in open chromatin regions active in endothelial cells, mural cells (pericytes and smooth muscle cells), and fibroblasts, and they disproportionately regulate extracellular matrix genes essential for vessel wall integrity. In contrast, Alzheimer's disease risk variants are enriched in endothelial and immune cell regulatory elements and converge on inflammatory adaptor proteins and signaling pathways. Most strikingly, a lead AD GWAS variant was found to enhance expression of PTK2B—a gene encoding a focal adhesion kinase involved in immune signaling—specifically in brain CD8 T cells, providing the first direct genetic evidence linking adaptive immunity to AD pathogenesis through a specific brain vascular immune cell type.

The atlas and associated analytical framework have been made available through a searchable web interface, enabling the broader research community to query variant-to-cell-type-to-gene relationships. The findings reframe how the field should think about vascular contributions to brain disease: not as a single unified mechanism, but as cell-type-specific and disease-specific regulatory programs driven by distinct noncoding genetic architectures.