Epigenome erosion in Alzheimer’s disease brain cells and induced neurons

Abstract

ABSTRACTLate-onset Alzheimer’s disease (LOAD) is typically sporadic, correlated only to advanced age, and has no clear genetic risk factors. The sporadic nature of LOAD presents a challenge to understanding its pathogenesis and mechanisms. Here, we comprehensively investigated the epigenome of LOAD primary entorhinal cortex brain tissues via single-cell multi-omics technologies, simultaneously capturing DNA methylation and 3D chromatin conformation. We identified AD-specific DNA methylation signatures and found they interact with bivalent promoters of AD differentially expressed genes. In addition, we discovered global chromosomal epigenome erosion of 3D genome structure within and across brain cell types. Furthermore, to evaluate whether these age- and disease-dependent molecular signatures could be detected in thein vitrocellular models, we derived induced neurons (iNs) converted directly from AD patients’ fibroblasts and found a set of conserved methylation signatures and shared molecular processes. We developed a machine-learning algorithm to identify robust and consistent methylation signatures of LOADin vivoprimary brain tissues andin vitrofibroblast-derived iNs. The results recapitulate the age- and disease-related epigenetic features in iNs and highlight the power of epigenome and chromatin conformation for identifying molecular mechanisms of neuronal aging and generating biomarkers for LOAD.HIGHLIGHTAD-specific DNA methylation signatures are identified in entorhinal cortex brain cell typesThe AD differentially expressed genes linked with differentially methylated regions via loop interactions are enriched in a bivalent chromatin stateChromosomal epigenome erosion of 3D genome structures occurs in LOAD brain cell types.Shared and reliable methylation signatures are observed in bothin vitrocellular iN models and primary brain tissues.Machine learning models identify robust and reliable methylation loci as AD biomarkers across cell types.