Genomic imprinting refers to the process of epigenetic change that occurs during germ cell development that results in either maternal- or paternal- specific gene expression. Imprinting appears to be involved in the development of several inherited disorders in humans as well as in the ontogeny of several forms of cancer. Because only a few imprinted genes have been discovered, identification of other imprinted genes is of primary importance to the understanding of imprinting as well as to the isolation of candidate disease genes. Most regions of the genome replicate synchronously with respect to their chromosomal homologue. However, it has been established that chromosomal regions currently known to contain imprinted genes replicate asynchronously. This proposal describes a general method for scanning the human genome far imprinted genes based on the replication asynchrony of their chromosomal alleles. The method involved the separation of dividing cells into different fractions of S- phase by flow cytometry and isolating newly replicated DNA from each of these fractions. The late-replicating DNA is predicted to be enriched in asynchronously-replicating inter-Alu sequences and will be used as a template for Alu PCR amplification; a plasmid library will be prepared from these products and replication profiles of inter-Alu products from different fractions of S-phase will be used to analyze the clones; the inter-Alu clones will also be screened for their presence on single human chromosomes by hybridization to a monochromosomal hybrid panel; sequences that exhibit biphasic or broad replication profiles and that are localized to a single chromosome will be sequenced; locus-specific PCR primers will be developed to verify replication asynchrony, determine allele-specific replication timing and test for gene expression; positive inter-Alu clones will be used for the isolation of cosmid clones; cosmids will be examined for chromosomal location and replication asynchrony by in situ hybridization, for methylation imprinting in genomic DNA, and for transcriptional expression. Monoallelic expression of candidate genes will be examined after obtaining expressed polymorphisms. Preliminary data indicate that the scanning method is likely to identify several replication-imprinted regions. Initial screening resulted in 3/123 inter- Alu clones that replicated asynchronously and localized to single chromosomes. A chromosome 15 clone was further characterized because of its possible localization to a known region of replication-imprinting, the Prader-Willi/Angelman locus. This sequence was localized outside the region usually deleted in Prader-Willi patients and was expressed in human cell lines. Other candidates were localized to chromosomes 17 and 19, each of which may have imprinted regions based on synteny with mouse chromosomes. Further characterization of these, as well as future clones, may help to identify imprinted genes that have major roles in human disease.
