The working hypothesis of our laboratory focuses on the convergence of prosurvival, angiogenesis and motility signals at common pathways in the local tumor microenvironment for therapeutic targeting and monitoring. We continue to study two pathways we have identified: the BAG-3 stress co-chaperone protein and the ovarian cancer growth and survival factor, progranulin (prgn). We are now reporting that BAG3 is a selective survival protein for those pathways in which HSP70 is involved. It is tightly regulated in the cell by both caspase cleavage and proteosomal degradation. Mutation of the putative caspase cleavage site stabilizes the protein and protects the cell from caspase-mediated apoptosis; studies into the generalizability of this finding are underway. Through the NCI/Myriad yeast two hybrid contract, we have identified putative BAG3 binding partners, one of which is HSP27. This binding is confirmed and the function of the interaction is under investigation. Microarray analysis of cells overexpressing wt and domain-mutants of BAG3 identified up regulation of proteins involved in invasion and metastasis. We are investigating those proteins using the model we have defined in which overexpression of wt BAG3 results in reduced focal contacts, substratum adhesion and serum-induced migration. Domain deletion mutants had the opposite phenotype. The findings in this model were further confirmed using siRNA to downregulate BAG3 expression; this resulted in reduced focal contacts and adhesion. CCN1, a protein involved in matricellular communication, was upregulated in the more adhesive and migratory cells. This regulation was confirmed at the protein level and also by dysregulation by RNA silencing. CCN1 upregulation was associated with increased invasion in the motile mutants. Lastly, translational studies are ongoing examining expression of BAGs 1, 3, 4, and 6 in two venues. The first is expression of these proteins in the NCI 60 cell line screen with analysis using the COMPARE program. Preliminary findings demonstrate a potential therapeutic targeted to the BAG4 overexpressing pathway. The second direction correlates immunohistochemical staining of the BAGs, HSP70, and bcl-2. Samples include a series of ovarian cancer clinical samples from our previous clinical trial for newly diagnosed patients and also using a tissue microarray of stage and grade variant ovarian cancers of endometriod and serous histologies. Cytosolic BAG4 was statistically associated with improved survival of the ovarian cancer patients; no association was found for BAG3, either nuclear or cytoplasmic. Studies of the Drosophila homolog of BAG 3/4, evil, continue. Evil has been transfected into MDA-435 hu breast cancer cells and survival phenotype has been demonstrated for two of the Evil isoforms. We reported that prgn is a growth and survival factor for ovarian cancer. We have recently demonstrated that secretory leukocyte protease inhibitory (SLPI) is a prgn partner protein and itself is necessary for ovarian cancer cells survival. SLPI is known to be in the whey acidic protein locus on chromosome 20 in the whey acidic protein region shown to be amplified in ovarian cancer. We have now shown that SLPI is a protector of prgn; this protection is independent of the protease inhibitory activity of SLPI. Down regulation of SLPI with neutralizing antibody or siRNA induces loss of prgn and cell apoptosis. Mutations of putative prgn binding sites are under development. Xenograft studies have shown that wild type SLPI and protease inhibition site mutants are more aggressive in mouse xenografts. Further analysis of the mechanisms of that invasiveness are underway. Production of prgn and its subsequent activity may be part of a signal amplification cascade in ovarian cancer and is a logical target for molecular therapeutics. Thus, the prgn/SLPI axis and BAG3 have potential as molecular therapeutic targets in ovarian cancer and other solid tumors.
Showing the most recent 10 out of 22 publications
