A subset of mammalian genes are expressed exclusively from a single parental allele. These genes must obtain a mark, or imprint, such that a developing organism can distinguish between the maternal and paternal alleles. It has been suggested that cytosine methylation in CpG dinucleotides serves as the imprinting mark for several imprinted genes, including H19, which is expressed solely from the hypomethylated on the silent paternal allele in mice and humans. Understanding the mechanism of imprinting may lead to an understanding of some human diseases that are likely encoded by imprinted genes, such as Beckwith-Weidemann Syndrome. It has also been suggested that imprinting may play a role in cancer, based on the observation that some tumors have abnormal expression of imprinted genes. The goal of this proposal is to understand how and when the imprint is established during development. The first specific aim will determine the time at which parental-specific methylation patterns are established in the germline through an analysis of methylation patterns in staged populations of germ cells. Such information will provide a foundation for the identification of trans-acting factors required for imprinting.
The second aim of this proposal is to determine if the differentially methylated region of H19 is sufficient to confer imprinting on a unrelated gene using transgenic mice. This information will help to delineate the minimal requirements for establishing and maintaining imprinting.
