Premium
O4‐10‐06: INTEGRATED GENETIC‐EPIGENETIC ANALYSES OF ALZHEIMER'S DISEASE
Author(s) -
Smith Rebecca G.,
Pishva Ehsan,
Shireby Gemma,
Smith Adam R.,
Han Eilis,
Sharp Andrew J.,
Mastroeni Diego,
Schalkwyk Leonard C.,
Haroutunian Vahram,
Coleman Paul D.,
Bennett David A.,
Hove Daniel L.A.,
De Jager Philip L.,
Mill Jonathan,
Lun Katie
Publication year - 2019
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2019.06.4799
Subject(s) - genome wide association study , dna methylation , epigenetics , expression quantitative trait loci , biology , quantitative trait locus , genetic association , genetics , dnam , epigenome , disease , genotype , methylation , computational biology , single nucleotide polymorphism , bioinformatics , gene , medicine , gene expression
population of mice harboring familial AD mutations (AD-BXDs) and their non-transgenic littermates (Ntg-BXDs) (Neuner et al, 2018). Gene co-expression modules from hippocampal RNAsequencing were built using a consensus clustering algorithm (Gaiteri et al, 2015). To identify transcriptional networks and driver genes that 1) explain significant variation in cognitive outcomes, and 2) have relevance to humans, we calculated Pearson correlations between module expression and cognitive outcomes, and performed hypergeometric overlap tests between mouse and previously characterized human modules (Mostafavi et al, 2018). We then focused on mouse modules that were significantly correlated to cognitive outcomes and highly conserved in humans, and performed genetic mapping of these modules to identify loci and putative regulators of gene co-expression in the selected modules (mQTLs). Results: We identified a subset of modules that significantly associated with variation in cognitive decline across our panel. Additionally, the majority of these modules significantly overlapped with human AD modules. We also identified a number of modules modified by AD mutations, including two immune-enriched modules that significantly correlated with individual differences in shortand long-term memory. The top module significantly correlated to working memory was enriched for pathways associated with protection from oxidation. Genetic mapping of this module identified a significant mQTL on chromosome 6 that highlighted a novel putative driver of module expression and memory outcomes. Conclusions: We identified transcriptional networks that provide new mechanistic insight into processes regulating individual differences in cognitive function across aging and AD. In addition, we identified a gene within an mQTL in a module conserved in humans that may be targeted to promote resilience to memory decline during aging. Finally, the high degree of overlap between mouse and human modules reflects the translatability of our model to transcriptional networks in human AD.