There are several components to this project. Development of new imaging techniques for detection and characterization of preinvasive and early invasive mammary cancers in mice: We are developing magnetic resonance (MR) imaging, ultrasound (US) imaging, and intravital microscopy (IVM) techniques for the detection of early stage mammary cancers in mice. The purpose of developing these new imaging techniques is two-fold: (i) to perform in vivo characterization of early mammary tumorigenesis in mouse models of breast cancer (see next project below) for which imaging techniques are necessary since early stage cancers are too small to be visible by eye or palpated, and (ii) to use mouse models as a test-bed for developing new imaging strategies that can be applied in women. In addition to developing new imaging acquisition techniques, we are also developing a Mulitmodality Mouse Imaging Workstation to improve image analysis and data storage. We are currently performing a study evaluating noncontrast MR and US imaging in wild-type mice, xenograft models and transgenic models;in doing so, we have successfully developed a new strategy for performing MR/US/histology correlations using an agar grid. Upon successful completion of this pilot study, we will begin to develop US contrast-enhanced techniques for quantifying lesion vasculature in a collaboration with Paul Dayton. We have also established a collaboration with Roberto Weigert to begin IVM of the mouse mamamry glands. Investigate cell-type susceptibility in new mouse models of breast cancer: The normal human breast can be broadly characterized as being comprised of epithelial cells that express cytokeratins K18/K8 (luminal cells) and K5/K14 (basal and some epithelial cells). Some intermediate epithelial cells are known to co-express K5/K18 as well as other keratins such as K19. To determine the role of cell-type susceptibility in mammary tumorigenesis, we have developed mouse models which drive expression of a conditional T121 allele thus functionally inactivating the pRb pathway, known to be frequently altered in human breast cancer--to specific cell types under control of K18, K19 and K5 regulation (please see Project 1 for a description of these models, and crosses with prostate-specific Cre lines to drive prostate tumorigenesis). We are currently breeding these mice to mammary specific Cre lines to drive cell-type-specific, mammary-specific tumorigenesis, and have also crossed the mice to p53 conditional knockout mice to determine if the tumor phenotype can be accelerated with p53 loss. Mice with tumors will be dissected and tissues processed for pathologic and molecular analysis. We will also be applying the imaging techniques developed above to characterize the timescales of early stage tumorigenesis and progression in these different mouse models. While waiting for these mice to develop, we are performing a study of the distribution of K5, K18 and K19 cells in the normal mouse mammary gland. Quantify lesion physical properties and assess role in progression: In addition to studying the molecular and genetic characteristics of early stage tumors, we are also interested in determining the role of lesion physical properties during progression of preinvasive lesions. In general, most imaging techniques provide a direct reflection of the physcial properites of tissues: for MRI, the tissue magnetic relaxation parameters T1, T2 or proton density p;for US, their acoustic properties. Thus, the noncontrast MR and US imaging we are currently developing already provide a physical characterization. We are also interested in developing advanced techniques to quantify physical lesion properties--such as stiffness (from US elastography) and basement membrane permeability (from IVM). Such physical characteristics of lesions can be noninvasively assessed with in vivo imaging, and thus changes in these properties can be assessed over time to determine if any are predictive of progression. We will commence this project upon successful completion of the MR/US pilot study outlined above.
