Adenocarcinoma of the prostate exhibits a marked propensity to metastasize to the skeleton where it comonly produces osteoblastic rather than osteolytic lesions. Such lesions appear to be due to increased osteoblast numbers and activity. We previously identified growth factors for cells of the osteoblast phenotype in freshly isolated human prostatic cancer as well as in benign prostatic hyperplasia. More recently we isolated a growth factor for osteoblastlike cells from conditioned medium of a human prostate cancer cell line (PC-3) which we identified as an amino-terminal fragment (ATF) of urokinase-type plasminogen activator (uPA), a fragment containing an "EGF-like" domain. We wish to test the hypothesis that overproduction or persistence of ATFs containing this growth factor domain is important in the pathogenesis of ostooblastic metastases by prostatic cancer. In our initial series of studies we propose to compare the production of ATFs and other uPA forms by prostatic cells and tissue and by other cancer cells and tissues which are much less frequently associated with osteoblastic metastases. Cell conditioned medium and tissue content will be examined by sequence specific radioimmunoassays, Western blots and bioassays (mitogenic and plasminogen activator). The mechanism for differences in cell or tissue content which are observed will be explored by comparing uPA mRNA levels, transcription rates, tissue-specific mRNA splicing and uPA metabolism. As well, plasminogen activator inhibitor production will be compared. A second series of studies will examine the spectrum of action, biopotency, and mechanism of action of native, synthetic and recombinant ATFs on skeletal tissue in vitro. Structure-function correlations will be made and cellular specificity assessed. Effects on differentiated osteoblastic function and on in vitro mineralization will be examined and characteristics of receptor binding and of post-receptor mechanisms of action will be evaluated. The studies therefore seek to determine whether discrete molecular domains of uPA are involved in osseous growth versus osseous breakdown and could provide a novel mechanism for the differential effects on skeletal metabolism of cancer cells.