During the last two decades of genome-wide sequencing, millions of single-nucleotide polymorphisms (SNPs) or variants (SNVs), copy-number variants (CNVs), and other types of sequence variations have been identified. These human genome maps have now served as catalysts for thousands of genome-wide association studies (GWAS), and have provided insights into diverse processes — such as mechanisms of disease, mutation, evolution, human migrations, selection/adaptation due to environmental stressors, and chromosomal recombination.
As provided in another GEITP email earlier this week, alterations in DNA sequence are not the only type of genomic variations; remember that epigenetic effects reflect DNA-methylation, RNA-interference, histone modifications, and chromatin remodeling. There are now well-documented examples of epigenetic marks — such as DNA methylation and histone modifications — that show significant interindividual variation, and increasing numbers of epigenome-wide association studies (EWAS) of these marks are now being performed.
Familial and twin studies in human and mouse have shown that a substantial fraction of sites showing variable DNA-methylation levels are highly heritable (i.e. passed on to next generation); for some loci, this epigenetic polymorphism has been linked with nearby genetic variation. However, these same studies have also demonstrated that a subset of methylation variation exhibits low heritability. Whereas stochastic (i.e. random events) variation or technical variability could explain decreased heritability levels — differing environmental exposures (e.g. smoking, diet, in-utero environment, stress, and even aging and X-chromosome inactivation) have all been shown to modify the epigenome. Whatever the cause of epigenetic polymorphisms, it seems clear that a subset of these variations is functionally significant and associated with expression levels of nearby genes.
Authors [see attached article] identified sites containing clusters of CpG sites (cytosine-guanine dinucleotides are where DNA-methylation occurs) with high interindividual epigenetic variation [termed Variably Methylated Regions (VMRs)] in five purified cell types. They found the VMRs occur preferentially at enhancers and 3′-untranslated regions (UTRs; part of the gene that doesn’t code for the protein). While the majority of VMRs have high heritability, a subset of VMRs within the genome shows highly correlated variation in trans (distant from the gene), forming co-regulated networks that
have low heritability, [b] differ between cell types, and [c] are enriched for specific transcription-factor-binding sites and biological pathways of functional relevance to each tissue.
For example, in T cells, authors found 95 co-regulated VMRs enriched for genes having roles in T-cell activation; in fibroblasts a network of 34 co-regulated VMRs comprising all four HOX (homeobox, important in early embryogenesis) gene clusters enriched for control of tissue growth; and in neurons a network of 18 VMRs enriched for roles in synapse-signaling. By culturing genetically-identical fibroblasts under varying environmental conditions, authors showed that some VMR networks are responsive to the environment, with methylation levels at these loci changing in a coordinated fashion in trans— and dependent on cellular growth. Intriguingly, these environmentally-responsive VMRs showed a strong enrichment for imprinted loci (epigenetic effects that cause genes to be expressed in a parent-of-origin-specific manner), suggesting that these are particularly sensitive to environmental conditions. This study provides a detailed map of common epigenetic variation in the human genome — showing that both genetic and environmental causes underlie these variations in a complex trait.
PLoS Genet Oct 2o18; 14: e1007707