Potential multiple disease progression pathways in female patients with Alzheimer’s disease inferred from transcriptome and epigenome data of the dorsolateral prefrontal cortex

Abstract

Late-onset Alzheimer’s disease (AD) is a typical type of dementia for which therapeutic strategies have not yet been established. The database of the Rush Alzheimer’s Disease study by the ENCODE consortium contains transcriptome and various epigenome data. Although the Rush AD database may contain a satisfactory amount of data for women, the amount of data for men remains insufficient. Here, based on an analysis of publicly available data from female patients, this study found that AD pathology appears to be nonuniform; AD patients were divided into several groups with differential gene expression patterns, including those related to cognitive function. First, cluster analysis was performed on individuals diagnosed with “No Cognitive Impairment (NCI),” “Mild Cognitive Impairment (MCI),” and “Alzheimer’s Disease (AD)” stages in clinical trials using gene expression, and multiple substages were identified across AD progression. The epigenome data, in particular genome-wide H3k4me3 distribution data, also supported the existence of multiple AD substages. However, APOE gene polymorphisms of individuals seemed to not correlate with disease stage. An inference of adjacency networks among substages, evaluated via partition-based graph abstraction using the gene expression profiles of individuals, suggested the possibility of multiple typical disease progression pathways from NCI to different AD substages through various MCI substages. These findings could refine biomarker discovery or inform personalized therapeutic approaches.