The majority of deaths due to cancer are caused not by the primary tumor but by distant metastases. The process by which an indolent primary tumor progresses to become metastatic involves a dramatic reprogramming of the tumor cell's gene expression and chromatin modification patterns. Effectively analyzing these changes across the entire genome, and identifying robust pro-metastatic gene expression and chromatin modification signatures, will be essential to understanding tumor progression and treating human cancer. Our ability to profile epigenetic and gene expression patterns on a genome-wide scale has advanced enormously in recent years. However, the applicability to mouse models of cancer is complicated by the difficulties in isolating pure populations of tumor cells that are free from contaminating stroma. We will develop technology for specifically profiling tumor cells from solid tumors in mice, free from stroma and other contaminating cell types. This will allow the accurate comparison of pure populations of nuclei from primary tumors and metastases. Specifically, we will: 1) Develop methods for biotin tagging a nuclear envelope protein in human cells lines and for isolating the biotin tagged nuclei from a xenograft tumor in nude mice. As proof of principle, we will test whether we can perform a genome-wide comparison of primary tumors and metastases from these models. 2) Generate and test a genetically engineered mouse model in which expression of a nuclear envelope biotinylation cassette can be induced by the Cre recombinase. When combined with mouse models of human cancer in which the tumor is driven by Cre-mediated activation of an oncogene or deletion of a tumor suppressor, this will allow for the isolation and analysis of pure populations of tumor cell nuclei. This work will generate tools and methods that will be widely applicable to the study of tumor progression and metastasis in mouse models of human cancer.
The dissemination of primary tumor cells to form metastases at distant sites is a major problem for treating human cancer. Understanding how metastasis is controlled will be key to effectively treating human cancer, but is complicated by the difficulty i isolating pure tumor cells from a solid cancer. We will develop methods and widely applicable tools for specifically purifying tumor cell nuclei from mouse models of human cancer.