Genome-wide DNA Methylation Variability in Adolescent Monozygotic Twins Followed since Birth

Lévesque ML1,2, Casey KF1, Szyf M3, Ly V3, Ismaylova E1,2, Verner M-P1,2, Provencal N3, Suderman M3, Brendgen M4, Dionne G5, Vitaro F1,6, Boivin M5, Tremblay RE1,7,8, Booij L1,2,9,10

1. Sainte-Justine Hospital research Center; 2. Department of Psychiatry, University of Montreal; 3. Department of Pharmacology and Therapeutics, McGill University; 4. Department of Psychology, UQAM; 5. Department of Psychology, University of Laval; 6. Department of Psycho-education, University of Montreal; 7. School of Public Health, Physiotherapy and Population Science, University College Dublin, Ireland; 8. Department of Psychology, University of Montreal; 9. Department of Psychiatry, McGill University; 10 Department of Psychology, Queen’s University.

Background: Although monozygotic (MZ) twins have identical genomes, DNA methylation patterns allow for divergent gene expression. Largely determined during gestation, methylation is highly influenced by the intrauterine environment. As children grow, divergence increases as a function of unique experiences. Filled with important physical and psychosocial changes, adolescence is an interesting but largely unexplored period of study. We thus aimed to assess within- pair variability in genome-wide DNA methylation patterns in adolescents.

Methods: Adolescent MZ twin pairs from the Quebec study of newborn twins who have been followed longitudinally since birth were epigenotyped and analysed using the 450K illumina assay. Raw illumina data were analysed using Bioconductor’s R package minfi and matlab.

Results: Preliminary analyses on the first 37 twin pairs showed that even unrelated individuals had highly similar methylation patterns across the epigenome (r2 ≥ 0.98), and twin pairs were even more similar (r2 ≥ 0.99). Data clustering confirmed that even the most discordant twins were better predictors of each others' methylation pattern than that of any unrelated individual. Importantly, within-pair differences were variable across twin pairs, and CPG sites. Pathway analysis revealed that high within-pair variability was found in genes associated state-like biological functions such as immune responses, and intriguingly with potentially more trait like networks regulating organismal growth and development.

Conclusion: Although overall similarity in DNA methylation was high, some variability was found in biological pathways related to human development. Such variability could plausibly underlie phenotypic differences in this monozygotic twin population. Functional, structural, and connectivity based neuroimaging is being concurrently conducted in this sample and we intend to examine epigenetic variability as a function of brain development.