Background: Glycoprotein non-metastatic B (GPNMB), also known as Osteoactivin, is a cell surface bound, transmembrane protein that is expressed in 41-74% of breast cancers. We have shown that it promotes migration, invasion and metastasis of breast cancer cells, that it is associated with shorter recurrence-free and overall survival times, and that it is most commonly expressed among breast tumors of the basal/triple-negative subtypes. Patients with basal/triple-negative breast cancer have aggressive disease and are not amenable to currently available targeted therapies. For these reasons GPNMB is an attractive target for therapeutic intervention in breast cancer; indeed, a GPNMB-targeted antibody-drug conjugate (CDX-011) has shown impressive clinical response among patients with GPNMB-positive breast cancer and is being investigated in Phase IIb clinical trials. Despite its promise as a therapeutic target in breast cancer, little is known about the functional role of GPNMB the primary tumor microenvironment.

Materials and Methods: We have employed 66cl4 murine mammary carcinoma cells and human BT549 breast cancer cells to investigate the effects of GPNMB on tumor growth and angiogenesis, respectively in vivo. Control and GPNMB-expressing tumors were analyzed using immunohistochemistry for degree of proliferation apoptosis and angiogenesis. To better characterize functionally important domains of GPNMB we employed immunoblot and ELISA analyses to assess levels of full length, intact GPNMB and shed GPNMB ectodomain in breast cancer cells. Transient siRNA-mediated knockdown studies were used to identify sheddases responsible for GPNMB shedding. Finally, we employed in vitro migration assays to assess function of the shed extracellular domain (ECD) of GPNMB.

Results: In this study we show that GPNMB promotes the outgrowth of mammary tumors in vivo and enhances their microvascular density. In an effort to investigate the functional domains of GPNMB that are responsible for this pro-angiogenic effect, we found that that the extracellular domain (ECD) of GPNMB is constitutively shed in breast cancer cells and show that this ECD is capable of inducing endothelial migration. Finally, we have implicated a disintegrin and metalloprotease-10 (ADAM10) as a novel sheddase responsible for liberation GPNMB ectodomain.

Conclusion: In this study we have described ADAM10-mediated shedding of GPNMB ECD - a protein fragment that induces endothelial cell migration - as a novel mechanism by which GPNMB can promote angiogenesis in breast cancer. Further efforts to characterize the effect of GPNMB on endothelial cell function will be discussed.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P1-02-02.