Bi-directional interactions of prostate cancer cells and bone marrow endothelial cells: Involvement of transforming growth factor beta and its type II receptor

Prostate cancer preferentially metastasizes to bone, yet little is known about the cellular and molecular factors involved. Endothelial cells are the initial contact for circulating prostate cells entering the bone microenvironment; and they can also be stimulated to undergo angiogenesis to promote metastatic growth. In this study, significant bi-directional interactions were observed for bone marrow endothelial cells and highly metastatic prostate cancer cells. In vitro co-culture and conditioned media experiments revealed that HBME-1 cells stimulated proliferation and migration of the highly metastatic PC-3 cell line. Concomitantly, PC-3 cells inhibited HBME-1 growth and induced them to differentiate into cord-like structures, mimicking angiogenesis.; PC-3 cells mediated growth inhibition and differentiation of bone marrow endothelial cells through production and activation of TGFbeta and alteration of TGFbetaRII-mediated signal transduction. Complimentary DNA microarray analysis suggested that TGFbeta was involved in PC-3 cell-induced responses in HBME-1 cells. Subsequent functional analyses revealed that low concentrations of TGFbeta cytokines inhibited HBME-1 cell growth to a similar level as PC-3 conditioned media and partially induced differentiation. Additionally, TGFbeta blockade reversed HBME-1 growth inhibition conferred by PC-3 cells, yet it apparently did not inhibit angiogenesis; and some TGFbeta inhibitors actually induced further angiogenesis. In stark contrast, TGFbetaRII inhibition significantly decreased PC-3 cell-induced angiogenesis. These discrepant results may be largely due to TGFbeta superfamily receptor/ligand promiscuity and cross-talk with other signaling pathways.; Additional studies are needed to identify other angiogenic factors that work in concert with the TGFbeta signaling pathway to mediate angiogenesis induced by highly metastatic prostate cancer cells. Blockade of TGFbeta signaling, using Fc:sTGFbetaRII in the SCID-hu prostate cancer (PC-3) bone metastasis model was ineffective at reducing tumor growth and angiogenesis. This approach may have been too broad; and thus, blocking TGFbeta specifically in bone marrow endothelial cells should be employed. In addition, cytokine array analyses revealed other factors, including IL-8 and GRO, which may help regulate angiogenesis induced by prostate cancer cells. Ultimately, a combinatorial approach of targeted, cell-specific blockade of TGFbeta signaling and inhibition of other angiogenesis signaling pathways may prove to be a very potent means of treating human prostate cancer bone metastasis.