Single‐nucleus open chromatin accessibility landscape of Pick’s and Alzheimer’s disease

Abstract Background Pick’s disease (PiD), a behavioral variant of frontotemporal dementia, is one of the common neurodegenerative dementia that is characterized by tau lesions. Despite the distinct tau aggregation observed in Pick’s disease compared with the tauopathy of Alzheimer’s disease (AD), the similarity in cognitive and behavioral impairments during the progression of the disease make their diagnosis challenging. What determines the differences and similarities between PiD and AD progression, especially at the epigenetic level, are largely unknown. We believe that recent advancement in single‐cell epigenomic profiling methods, Assay for Transposase‐Accessible Chromatin (ATAC) combined with high throughput sequencing, will enable us to map the chromatin‐regulatory landscapes of disease brains at a single‐cell resolution and enable us to understand underlying molecular changes in these diseases. Method In the present study, we have isolated single nuclei from the frontal cortex region and performed single‐nucleus ATAC sequencing (snATAC‐seq) of 198,722 nuclei to generate cell‐type specific chromatin accessibility profiles of postmortem brains of PiD and AD patients and to uncover their cellular heterogeneity and similarity. We used cellranger to process the paired‐end FastQ reads of each sample to obtain its chromatin accessibility count matrix. Next, we utilized ArchR and Signac pipelines to generate 501‐bp fixed‐width peak regions and performed downstream differential accessible regions analysis. Quality control was also performed to remove the contribution of low‐quality nuclei. Result We constructed chromatin accessibility profiles from 19,690 nuclei from PiD patients, 59,148 nuclei from AD patients, and 119,884 nuclei from healthy controls. We applied UMAP dimensionality reduction and clustered nuclei through open chromatin region to the batch‐corrected epigenomic datasets. We identified seven distinct cell types, including astrocyte, excitatory neurons, inhibitory neurons, microglia, oligodendrocytes, OPC, and pericyte/endothelial cells, and we applied pseudotime trajectory analysis to characterize disease‐associated cell state in both AD and PiD at the epigenomics level. Conclusion Our data identified the cell‐type‐specific open chromatin accessible regions in AD and PiD brains. Although the causative molecular mechanisms of AD and PiD remain unknown, our work helps to identify shared epigenomic changes in AD and PiD, especially in regards to cell‐type‐specific genomic loci with disease risk.