Human epidemiological and animal experimental data indicate that the risk of developing adult-onset diseases is influenced by persistent adaptations to prenatal and early postnatal exposure to environmental conditions such as nutritional privation [Jirtle and Skinner, 2007]. Moreover, the link between what we are exposed to in utero and disease formation in adulthood appears to involve epigenetic modifications, like DNA methylation, at metastable epiallele and imprinted gene loci. The overall objective of our laboratory is to determine the role of epigenetics in the pathogenesis of human diseases and neurological disorders.

Figure 1: Despite their appearance, these mice are genetically identical. Their mothers ate different diets during pregnancy, which led to epigenetic changes in the Agouti gene that influenced coat color and body weight.

Metastable Epialleles. Genes with metastable epialleles have highly variable expressions because of stochastic allelic changes in the epigenome rather than mutations in the genome. The viable yellow agouti (Avy) mouse (Figure 1) harbors a metastable Agouti gene because of an upstream insertion of a transposable element. We showed with the use of this mouse that dietary supplementation during pregnancy, with either methyl donors (i.e. folic acid, vitamin B12, choline and betaine) [Waterland and Jirtle, 2003] or genistein [Dolinoy et al., 2006], decreased the adult disease incidence in the offspring by increasing DNA methylation at the Avy locus. These supplements also guard the epigenome from the negative hypomethylating effects of bisphenol A, a common environmental toxicant and endocrine disruptor used in the production of plastic [Dolinoy et al., 2007]. These findings suggest novel approaches for protecting against the harmful effects of environmental pollutants.

Imprinted Genes. Genomic imprinting is a non-Mendelian, germline-inherited epigenetic form of gene silencing that results in parent-of-origin dependent monoallelic expression [Murphy and Jirtle, 2003]. Imprinting first evolved in an ancestor to Therian mammals (marsupials and eutherians) following divergence from the egg-laying Prototherian mammals (monotremes) during the late Triassic / early Jurassic period (i.e. 180 to 200 million years ago) [Killian et al., 2000]. Our comparative phylogenetic studies also showed that imprinting at specific loci could be lost, as well as gained, during evolution. For example, the IGF2R (Insulin-like growth factor 2 receptor), which was imprinted about 180 million years ago, lost its imprinted status in an early primate ancestor about 75 million years ago [Killian et al., 2001 ]. Consequently, people have two functional copies of this tumor suppressor gene; whereas, mice have only one.

Because imprinted genes are functionally haploid, their inactivation requires only a single genetic or epigenetic event. The high vulnerability of imprinted genes to becoming dysfunctional led us to develop machine learning algorithms to identify them throughout the human genome. We uncovered 156 novel candidate imprinted genes in the human [Luedi et al., 2007] and 600 in the mouse [Luedi et al. , 2005]. Humans are predicted to have not only fewer imprinted genes than mice, but also a markedly different repertoire. Thus, mice may not be a suitable choice for studying diseases resulting principally from the epigenetic deregulation of imprinted genes, or for assessing human risk from environmental factors that alter the epigenome. As simply stated by the English poet Alexander Pope in the early eighteenth century, “The proper study of Mankind is Man.”

By mapping these newly identified human imprinted gene candidates onto the disease landscape defined by linkage analysis, we are now determining the role of imprinting in the etiology of some of the most dreaded human diseases and neurological disorders, such as autism, bipolar disorder, cancer, and schizophrenia.