Tissue pathology is a manifestation of the genomic aberrations that define cancer (i.e. stage, grade, type). In prostate cancer, the best prognostic marker is Gleason score, the composite tumor grading system that summarizes primary tumor morphology. Subsets of patients with localized disease inevitably develop metastases, a point where it is often too late for curative treatment. Thus, we propose to model and therapeutically target molecular subtypes that drive aggressive primary prostate cancer. We start from large, ?omic datasets and use tissue pathology and gene expression as the readout of disease development. To evaluate the direct effect of molecular drivers of aggressive primary disease, we use a novel prostate stem cell, tissue recombination mouse model that recapitulates prostate development and tissue pathology. To model specific molecular subtypes, we will genetically engineer mouse prostate stem cells to introduce gene knockouts or Tet-regulated genes. These engineered cells will then be combined with fetal urogenital mesenchyme and engrafted under the mouse kidney capsule to produce prostate structures. We will evaluate tissue pathology and isolate regions of interest with aggressive tissue morphology (e.g., cribriform patterning) for RNA sequencing. Using this approach to model three mutually exclusive molecular subtypes, we can directly compare and contrasts gene expression and pathways changes to identify common and subtype- specific effects. Using these data, along with ?omic data from patients, we will use two network-based computational models to identify novel therapeutic treatments and capture the systems-level gene regulation and functional relationships. The therapeutic predictions will be screened in vitro using engineered mouse and human cells with promising treatments being promoted for testing in the tissue recombination mouse model. Overall, through this project, we will study drivers of aggressive primary prostate cancer, characterize mechanisms of disease development, and identify pre-clinical pharmacological treatment strategies. Computational models and the prostate stem cells will be valuable resources for the CSBC and larger research communities.
Primary prostate cancer tissue morphology is highly predictive of disease aggressiveness. Leveraging novel computational and in vivo experimental approaches, we will model, characterize and pharmacologically target multi-genic molecular subtypes that promote aggressive primary prostate cancer pathology.
