NOVEL IMAGING METHODS FOR GENE DISCOVERY IN CANCER

DESCRIPTION: (Applicant's Description) We propose to develop clinically relevant mouse models that resemble human breast tumors in etiology and histology. We propose to use these mouse models to procure pure populations of motile and invasive tumor cells from primary tumors in live intact mice under direct visualization using novel imaging technology. We propose to develop high sensitivity DNA microarrays for use with these cells so as to examine gene expression patterns unique to the invasive and metastatic population of cells in the primary tumor compared to other populations of tumor and normal cells. The coupling of cell behavior to gene expression will allow the rational interpretation of gene expression patterns in metastatic tumors. The specific aims for the R21 phase of the project are: 1. Prepare mouse models that develop metastatic breast tumors as seen in patients and that can be used for imaging of tumor cell behavior in vivo by multi-photon microscopy. 2. Develop methods for multi-photon imaging of normal and tumor bearing mouse mammary glands. 3. Refine methods for collecting chemotactic cells from the primary tumor that have been characterized by direct imaging and that represent the population of cells capable of chemotaxis to vessels and surrounding tissue in response to serum and growth factors. The specific aims for the R33 phase of the project are: 1. Develop sensitive DNA microarray techniques for comparing small numbers of chemotactic tumor cells and white cells collected from the primary tumor with tumor cells and white cells collected from elsewhere. 2. Demonstrate the utility of comparing gene expression patterns of 8 categories of cells collected from various mouse models to identify patterns unique to different subpopulations of cells. 3. Develop methods to correlate the gene expression patterns of cells collected with microneedles from the primary tumor with the histology and behavioral phenotypes of cells at the collection sites. 4. Evaluate a variety of clustering algorithms, with DNA expression patterns obtained in specific aims 2 and 3, to identify genes related to cell behaviors believed necessary for metastasis.