Abstract
Background Although monozygotic (MZ) twins share the majority of their genetic makeup, they can be phenotypically discordant on several traits and diseases. DNA methylation is an epigenetic mechanism that can be influenced by genetic, environmental and stochastic events and may have an important impact on individual variability. Methods In this study we explored epigenetic differences in peripheral blood samples in three MZ twin studies on major depressive disorder (MDD). Epigenetic data for twin pairs were collected as part of a previous study using 8.1-K-CpG microarrays tagging DNA modification in white blood cells from MZ twins discordant for MDD. Data originated from three geographical regions: UK, Australia and the Netherlands. Ninety-seven MZ pairs (194 individuals) discordant for MDD were included. Different methods to address non independently-and-identically distributed (non-i.i.d.) data were evaluated. Machine-learning methods with feature selection centered on support vector machine and random forest were used to build a classifier to predict cases and controls based on epivariations. The most informative variants were mapped to genes and carried forward for network analysis. A mixture approach using principal component analysis (PCA) and Bayes methods allowed to combine the three studies and to leverage the increased predictive power provided by the larger sample. Results A machine-learning algorithm with feature reduction classified affected from non-affected twins above chance levels in an independent training-testing design. Network analysis revealed gene networks centered on the PPAR − γ (NR1C3) and C-MYC gene hubs interacting through the AP-1 (c-Jun) transcription factor. PPAR − γ (NR1C3) is a drug target for pioglitazone, which has been shown to reduce depression symptoms in patients with MDD. Discussion Using a data-driven approach we were able to overcome challenges of non-i.i.d. data when combining epigenetic studies from MZ twins discordant for MDD. Individually, the studies yielded negative results but when combined classification of the disease state from blood epigenome alone was possible. Network analysis revealed genes and gene networks that support the inflammation hypothesis of MDD.
Highlights
Major depressive disorder (MDD) is a pervasive psychiatric disorder characterized by a number of clinical symptoms including: persistent low mood, anhedonia, insomnia, low energy, feelings of guilt and ideation of death or suicide.[1,2]
The results from this study support previous human and animal studies on MDD that have uncovered genes centered around a stress-response cascade involving the activator protein 1 (AP-1) and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κβ).[30]
AP-1 downregulation has been implicated as part of the mechanism by which administration of IFN-alpha therapy induces depression symptoms.[31]
Summary
Major depressive disorder (MDD) is a pervasive psychiatric disorder characterized by a number of clinical symptoms including: persistent low mood, anhedonia, insomnia, low energy, feelings of guilt and ideation of death or suicide.[1,2] MDD is associated with a range of social impairments, including educational and occupational problems and with an increased risk of developing systemic disease, such as cardiovascular disease and Type 2 diabetes.[3]. Behavioral genetic research into the etiology of depression reports heritability estimates between 31 and 42% (refs 6–8) but uncovering common sequence variants associated with the pathology has been challenging. There are no common genetic variants of sufficiently high penetrance to account for the pathology that have clinical significance, variants associated with the disease at the genome-wide level have been recently announced.[9] Environmental factors, such as early- and late-life stressors, are thought to increase the risk of developing depression; the interaction between genetic and environmental factors remains poorly understood.[10]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
