Epigenome-wide association study of alcohol use disorder in five brain regions

Lea Zillich,Rainer Spanagel,Anita C Hansson,Franziska Degenhardt,Per Hoffmann,Stephanie H Witt,Stefanie Heilmann-Heimbach,Marion M Friske,Markus M Nöthen,Jerome C Foo,Josef Frank,Fabian Streit,Helene Dukal,Lea Sirignano,Marcella Rietschel

doi:10.1038/s41386-021-01228-7

Abstract

Alcohol use disorder (AUD) is closely linked to the brain regions forming the neurocircuitry of addiction. Postmortem human brain tissue enables the direct study of the molecular pathomechanisms of AUD. This study aims to identify these mechanisms by examining differential DNA-methylation between cases with severe AUD (n = 53) and controls (n = 58) using a brain-region-specific approach, in which sample sizes ranged between 46 and 94. Samples of the anterior cingulate cortex (ACC), Brodmann Area 9 (BA9), caudate nucleus (CN), ventral striatum (VS), and putamen (PUT) were investigated. DNA-methylation levels were determined using the Illumina HumanMethylationEPIC Beadchip. Epigenome-wide association analyses were carried out to identify differentially methylated CpG-sites and regions between cases and controls in each brain region. Weighted correlation network analysis (WGCNA), gene-set, and GWAS-enrichment analyses were performed. Two differentially methylated CpG-sites were associated with AUD in the CN, and 18 in VS (q < 0.05). No epigenome-wide significant CpG-sites were found in BA9, ACC, or PUT. Differentially methylated regions associated with AUD case-/control status (q < 0.05) were found in the CN (n = 6), VS (n = 18), and ACC (n = 1). In the VS, the WGCNA-module showing the strongest association with AUD was enriched for immune-related pathways. This study is the first to analyze methylation differences between AUD cases and controls in multiple brain regions and consists of the largest sample to date. Several novel CpG-sites and regions implicated in AUD were identified, providing a first basis to explore epigenetic correlates of AUD.

Highlights

Every year, ~5.3% of all deaths worldwide are a result of the harmful use of alcohol and ~230 diseases are associated with alcohol use [1]
Multi-marker Analysis of GenoMic Annotation [30] was used to Epigenome-wide association analysis In the caudate nucleus (CN), two CpG-sites were epigenome-wide significantly hypomethylated in alcohol use disorder (AUD) cases compared to controls
In the ventral striatum (VS), 18 CpG-sites were epigenome-wide significantly associated with AUD

Summary

Introduction

Every year, ~5.3% of all deaths worldwide are a result of the harmful use of alcohol and ~230 diseases are associated with alcohol use [1]. It has been proposed that drug-induced alterations in gene expression in the neurocircuitry of the brain contribute to addiction [4]. Recent evidence suggests that alterations in DNAmethylation, an epigenetic mechanism affecting gene expression, play an important role in addiction (for reviews see: [5, 6]). Differential DNA-methylation is associated with alcohol consumption and AUD both in peripheral blood and postmortem brain tissue (for an overview see: Wedemeyer et al [7]). Examining alterations in DNA-methylation in epigenome-wide association studies (EWAS) allows for the investigation of inter-individual differences which are attributable to a phenotype [8]. A recent EWAS of AUD in peripheral blood suggests that networks in glucocorticoid signaling and inflammation-related genes are associated with AUD [9]

Objectives

Results

Conclusion