Year: 2019
Type: article
Abstract: It is hypothesised that Alzheimer's disease (AD) has an epigenetic component. In recent years a number of epigenome-wide association studies (EWAS) have been undertaken in AD brain. We have recently performed a cross-brain region meta-analysis of all available DNA methylomic studies of AD to identify numerous robust disease-associated changes. We have collected matched genome-wide association study (GWAS) data on the same samples to allow us to also identify methylation quantitative trait loci (mQTLs) and look at the relationship between genotype and methylation. Raw Illumina Methylation 450K BeadChip data was collated from seven EWAS of AD, which equated to 1,576 individuals, with in some cases multiple brain regions available. Within R data was re-normalized and quality checked before looking for an association with disease separately in each cohort, controlling for age, sex and cell proportions. After, meta-analyses were performed across cortical tissues for combined p-value, effect size and correlation. The R package Impute was used to impute additional genotypes form the 1000 genome project and then the R package matrix eQTL was used to identify mQTLs associated with disease. Within R we have used a Fisher's meta-analysis to identify robust DNA methylation associated with Braak stage using multiple cortical brain regions from >1,500 unique donors and identified >300 experiment-wide significant loci. We have employed a leave-one-out approach to demonstrate consistency in changes across cohorts. We have used the matrix eQTL package to identify mQTLs in our integrated genetic-epigenetic data. Finally, we have used the weighted effect size of the epigenetic data to generate polyepigenic risk scores (PeRS) for AD cortex. Our approach represents the largest meta-analysis of AD EWAS to date, including data from >1,500 individuals. Our study has highlighted a number of loci previously nominated from AD EWAS (e.g. ANK1, RHBDF2) as well as many novel loci, which were not identified in the smaller individual cohort studies. Our mQTL analysis has identified loci where genetic variation alters DNA methylation, most notably within the APOE gene. Finally, by generating PeRS, we have been able to investigate the proportion of disease risk accounted for by changes in methylation in AD cortex.
Source: Alzheimer s & Dementia
Institutions: University of Exeter, Maastricht University, Icahn School of Medicine at Mount Sinai, Banner Sun Health Research Institute, Arizona State University, University of Essex, Rush University Medical Center, University of Würzburg, Columbia University Irving Medical Center, Columbia University
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: bronze
OpenAlex ID: W2980613795