Overcoming Obstacles in Epigenetic Analysis of Human Twins

Cardiovascular disease (CVD) is the leading cause of death in the US, emphasizing an incomplete understanding of its pathology and translation into effective interventions. The fetal origins hypothesis posits that the risk of CVD is partially set in utero. Individual differences in epigenetic DNA methylation are a leading candidate mechanism to explain the association between CVD risk and the fetal environment. Indeed, differences in DNA methylation have been implicated in the pathogenesis of CVD and its risk factors. Previous studies in rats have been characterizing a specific class of epigenetic loci known as metastable epialleles (MEs). Methylation of DNA in utero can be in response to non-genetic factors and the degree of hypo- or hyper- methylation remains stable throughout the lifespan. Studies in mice further have shown that methylation has dramatic and permanent phenotypic effects. In recent studies over 100 human MEs have been identified. Several of these loci have been associated with phenotypes related to CVD risk.

Investigators now need to determine the phenotypic consequences of interindividual variation in DNA methylation at human MEs. Studies in inbred mice have been instrumental in showing that there are non-genetic influences on the establishment of methylation at MEs. Twin studies, typically using saliva or peripheral blood cells (PBCs), have been used to assess whether individual epigenetic variation in humans is influenced by non-genetic factors. Indeed, in translating epigenetic findings from mice to humans, monozygotic (MZ; genetically-identical) twins are commonly viewed as the human equivalent of isogenic mice. However, there is an important difference between MZ twins and isogenic mice. Whereas each inbred mouse littermate has its own placenta, about 75% of MZ twins share a single placenta (monochorionicity) and therefore have intermingled circulation during fetal development. It is therefore essential to determine whether the resulting "cross pollination" of hematopoietic stem cells (HSCs) results in peripheral blood (and salivary) DNA of two monochorionic MZ twins being more epigenetically similar than that of other somatic tissues. Accurately quantifying epigenetic effects of periconceptional nutrition in twin studies, and thereby providing future intervention targets to reduce CVD risk in offspring, will likely require analyses of DNA methylation in tissues other than saliva or peripheral blood cells.

Investigators need to examine whether Monozygotic:Dizygotic twin comparisons, without regard to placental sharing in utero (chorionicity), are a suitable model for future work which will characterize ME loci and establish their associations with phenotypic consequences. Infancy is the ideal time to examine the association between MEs and CVD factors, since it provides the earliest time for risk prediction and intervention. Therefore, for this study investigators plan to recruit 40 twins pairs under 4 months of age to identify more easily available tissues for epigenetic analyses than hair follicles (which has been previously used as an alternative to PBCs or saliva). The plucking of hair follicles from infants is too burdensome on participants to accrue the large samples intended for future analyses.

Summary: Studies in rats provide evidence that maternal nutrition around the time of conception can alter epigenetic mechanisms affecting obesity, and obesity itself is a strong CVD risk factor. This relationship is mediated by the early embryonic methylation at metastable alleles. Following the identification of over 100 MEs in humans many of which are associated with obesity or other CVD risk factors, investigators are now ready to implement studies to examine whether periconceptional nutrition influences CVD risk via methylation. Investigators seek to translate this work on MEs into human populations since it suggests an important CVD risk. Further, investigators want to establish the importance of chorionicity as a factor in epigenetic analyses of MZ twins, and to validate a new tissue for such studies.

Aim 1: Validate nail clippings as an acceptable DNA source for the study for human MEs Hypothesis 1: At MEs, DNA methylation in infant nail clippings (which can be easily collected) is highly correlated with that in hair follicles, urine and buccal swabs.

Aim 2: Compare epigenetic discordance at MEs in monochorionic vs. dichorionic MZ twins Hypothesis 2: At MEs, dichorionic MZ twin pairs (with separate placentae) will show greater epigenetic dissimilarity in PBCs than monochorionic MZ twin pairs (which share a placenta), but not in other tissues.