Project Summary/Abstract
Dementia afflicts over 55 million people worldwide. With an aging population and no effective treatments
available, this number is projected to nearly triple by 2050. Cerebral small vessel disease (CSVD) causes up to
45% of all dementia, accounts for ~20% of strokes, and appears in most Alzheimer’s disease patients. CSVD
arises from pathologies of the brain’s ~400 miles of oxygen and nutrient-delivering small arteries, arterioles,
venules, and capillaries; and is recognized as the most important vascular contributor to cognitive impairment
and dementia (VCID). CSVD pathologies, such as arteriolosclerosis and microinfarcts, develop insidiously to
strangulate cerebral blood flow, compromise the blood-brain barrier, and trigger neuroinflammation. Recent
developments in single-nucleus sequencing technologies have begun elucidating fundamental molecular and
cellular processes underlying human neurological disease. They have revealed selectively vulnerable cell
populations, transcriptional dysfunction across disease stages, and the mechanisms of genetic risk variants to
inform new research areas and therapeutic targets. Yet, single-nucleus studies have mostly lost human brain
vascular cells unknown reasons, leaving our understanding of CSVD lagging. To address this challenge, we
invented Vessel Isolation and Nuclei Extraction for Sequencing (VINE-seq) to efficiently capture human brain
vascular cell types from frozen postmortem brains for single-nucleus profiling. Here, we will combine VINE-seq
with multimodal single-nucleus RNA and ATAC sequencing to generate a first ~1.9 million vascular atlas of
early, mid-, and late-stage CSVD alongside age-matched cognitively normal controls. Samples come from a
specially organized CSVD set of 191 postmortem midfrontal gyrus watershed tissue, prioritized via MRI and
nerve tractography, from the Religious Order Study and Memory and Aging Project (ROSMAP). Samples
exhibit no confounding non-vascular neurodegenerative pathology and are richly annotated with demographic,
genomic, pathologic, and longitudinal cognitive data to link molecular readouts to phenotypes. Thus, we will
combine VINE-seq with multimodal single-nucleus RNA- and ATAC- sequencing (snRNA/ATAC-seq) to
generate a first vascular atlas of CSVD (A1); reveal the molecular signatures of CSVD progression (A2); and
determine the causal mechanisms of CSVD genetic risk (A3). As vascular risk factors are modifiable, insights
revealed here may be critical to understanding and ultimately treating VCID and mixed pathology dementias.
