Based upon substantial published literature from multiple groups, as well as unpublished studies to be presented from the applicants' laboratories, there is strong documentation that bone marrow-derived Mesenchymal Stem Cells are released into the blood stream and home to sites of primary and metastatic prostate cancer driven by the inflammatory microenvironment characteristically present within prostate cancer's stromal compartment. Thus, the hypothesis of this project is that allogeneic human bone marrow-derived MSCs (hbMSCs) can be used as a cell-based targeting vehicle to selectively deliver (i.e., home) therapeutic agents to sites of prostate cancer, thus sparing host toxicity. In this application, data will be presented validating the rationale for this "Trojan Horse" approach in which allogeneic hbMSCs are genetically-engineered to express a recombinant pro-aerolysin protein protoxin. While initially inactive, this protoxin is engineered to be selectively hydrolyzed to a picoMolar killing molecule by the enzymatic activity of a protease [i.e.. Prostate Specific Antigen (PSA)] which is only enzymatically active in high levels within the stroma at sites of prostate cancer. Since hbMSCs are being tested in clinical trials for regenerative medicine and recombinant PSA activated pro-aerolysin is in clinical testing as local therapy for prostate diseases, the proposed use of PSA-activated pro-aerolysin expressing hbMSCs could rapidly enter clinical development as systemic therapy for lethal metastatic prostate cancer based upon the successful completion of the following specific aims.