A hypothetical model of skewed DNA methylation balance in the enhancer regions containing differentially methylated cytosines associated with non-malignant complex diseases

Epigenome-wide association studies have been widely conducted over the past decade to identify the epigenetic determinants of human complex diseases. Numerous differentially methylated cytosines (DMCs) associated with diseases or environmental exposure have been identified till date. An important hallmark of DMCs in non-malignant complex diseases is that they are strongly enriched in CpG sites showing small changes to intermediate levels of methylation. This paradigm of subtle change in DMC methylation hampers our understanding of epigenetic regulation in non-malignant complex diseases and the application of DMCs as biomarkers in clinical practice. Herein, we propose a hypothetical model of skewed DNA methylation balance in the enhancer regions containing differentially methylated cytosines associated with non-malignant complex diseases. The DMCs are strongly enriched in enhancers, which are characterized by a highly dynamic balance of DNA methylation. The DNA methylation balance plays an important role in cellular adaptation to adverse environment; however, in the case of certain diseases, the balance becomes skewed toward one end of dynamic range. Binding stress-inducible transcription factors determines DNA methylation at the enhancer region, resulting in epigenetic heterogeneity. The dynamics of DNA methylation balance can be evaluated by analyzing the DNA co-methylation status of DMCs and their neighboring CpG sites. Targeted bisulfite resequencing method is suggested to analyze the change in DNA co-methylation in both case-control and longitudinal cohort studies. Cell models are necessary to study the molecular mechanisms of DNA methylation changes. We also suggest that complex obstetrical syndromes associated with placental disorders may be excellent models for investigating common features of DNA methylation in non-malignant complex diseases.