•
•
•
•
Members of the steroid receptor superfamily, particularly the progesterone receptor (PR) and estrogen receptor (ER), play critical roles in the etiology of human breast cancer. Human mammary cancers can be categorized as estrogen dependent or independent, and treatment of early-diagnosed tumors with tamoxifen, an ER antagonist, can be effective. It is clear that a comprehensive grasp of the genomic regulatory actions of these receptors is critical to the ultimate solution of the breast cancer problem.
Our current understanding of the molecular mechanisms by which these receptors effect regulatory transitions is based on experimentation in a very limited number of model systems. Recent technical developments now permit the examination of regulatory factor/genome interactions on the genome-wide scale. In particular, high density tiling of oligonucleotide probes on microarray chips allows one to characterize genome-receptor interactions and a number of resulting regulatory processes in great detail throughout the mammalian genome.
We have successfully adapted this exciting new technology to the study of nuclear receptor interactions with the genome. Using standard expression microarrays, we have identified approximately 100 genes that are inducible by the glucocorticoid receptor in a murine mammary carcinoma cell line. Furthermore, in a congenic cell line expressing an inducible dominant negative Brg1 component of the Swi/Snf chromatin remodeling complex, we show that chromatin remodeling is critically important for the activation of most GR regulated genes. With these identified targets of GR action, we have employed high density tiled arrays from NimbleGen to characterize GR binding sites (by ChIP-Chip analysis) throughout the genomic regions associated with these genes. The data obtained from these experiments are of surprisingly high quality, allowing the unambiguous identification of receptor interaction sites for most of the regulated genes. Indeed, clear binding sites were found for genes that were not identified in the expression array experiments. When GR regulation of these promoters was subsequently examined by RT-PCR analysis, they were in fact observed to be responsive to GR action, confirming the high reliability of the ChIP-Chip data.
We now intend to adapt these approaches to a complete screening for PR, ER and GR interaction sites in the genome, and to characterize the epigenetic modifications associated with these events. In collaboration with John Stamatoyannopoulos, Regulome, Seattle, we have shown that we can map global DNaseI hypersensitive sites in the genome associated with nuclear receptor action. We have also initiated collaborative experiments with Oliver Rando, Harvard, to examine global nucleosome positioning in genome domains associated with nuclear receptor function. We intend to develop a detailed characterization of histone modifications induced in these domains by hormone stimulation.
Thus, we propose the complete, genome wide analysis of ER, PR and GR regulatory elements, and of the epigenetic transitions associated with hormone action at the promoters associated with these sites. This analysis will provide an unprecedented description of hormone receptor interaction with the genome, and will serve as the eventual basis for a comprehensive understanding of mechanisms utilized to effect hormone regulation of mammalian genes.
