Identification of differential DNA methylation associated with multiple sclerosis: a family-based study

Multiple Sclerosis (MS) is known to be caused by a genetic predisposition triggered by environmental factors. Despite the knowledge of genetic factors and environmental agents involved, it is largely unclear how they interact. Epigenetics, particularly DNA methylation, represents a model for environmental factors to influence the genome. In this paper, we studied 26 affected and 26 unaffected relatives from 8 MS multiplex families (≥ 3 affected relatives) as part of a multicentric Italian cohort study using Methylated DNA immunoprecipitation sequencing (MeDIP-Seq) run on an Illumina® HiSeq 2500 (2x100 bp). Technical validation in 6/8 of the original families and biological replication in 2 additional families (FDR < 0.1 and a concordant fold change, FC) were performed in suggestive differentially methylated regions (DMRs) between affected and unaffected relatives using SeqCap Epi Choice Enrichment (Roche®). Evidence of association from MeDIP-Seq across 8 families was combined with aggregation statistics, separately for hypo- and hyper-methylated regions with concordant FC in ≥ 6/8 families, yielding 162 DMRs at FDR ≤ 0.1. Technical validation and biological replication led to 2 hypomethylated regions, which point to NTM and BAI3 genes, and 2 hypermethylated regions in PIK3R1 and CAPN13 (mean size: 3.1 kb). Multiplex families represent a privileged setting for the study of regions of differential methylation as they reduce the impact of potential confounders like shared genetics and environmental factors. We identified 4 novel regions which contain genes of potential interest that need to be tested in larger cohorts of patients. Whole Blood samples from both MS affected individuals and healthy realatives were collected from 8 families and were subjected to MeDIP Sequencing. After quality filtering reads and identifying peaks of methylated regions, differentially methylated regions were called (DMRs). An aggregation strategy was used to select for thos regions with hyper or hypomethylation in MS compared to Healthy individuals based on metaanalyzed p.values across the 8 families. After increasing those DMRs by 1kb down and upstream, a bisulfite panel of methylation sequencing (by Roche) was used to validate DMRs using 6 families of the discovery phase and replicate them on two unrelated MS families