We have previously shown that the chemokine fractalkine promotes the adhesion of human prostate cancer cells to bone marrow endothelial cells as well as their migration toward human osteoblasts in vitro. Thus, the interaction of fractalkine with its receptor CX3CR1 could play a crucial role in vivo by directing circulating prostate cancer cells to the bone. We found that although CX3CR1 is minimally detectable in epithelial cells of normal prostate glands, it is overexpressed upon malignant transformation. Interestingly, osteoblasts, stromal and mesenchymal cells derived from human bone marrow aspirates express the cell-bound form of fractalkine, whereas the soluble form of the chemokine is detected in bone marrow supernatants. To investigate the mechanisms regulating the levels of soluble fractalkine in the bone marrow, we focused on androgens, which play a critical role in both prostate cancer progression and skeletal metastasis. Here, we show that dihydrotestosterone dramatically increases the cleavage of fractalkine from the plasma membrane of bone cells and its action is reversed by nilutamide-an antagonist of the androgen receptor-as well as the wide-spectrum inhibitor of matrix metalloproteases, GM6001. However, dihydrotestosterone was unable to induce fractalkine-cleavage from human bone marrow endothelial cells. Thus, androgens could promote the extravasation of CX3CR1-bearing cancer cells on a fractalkine concentration gradient, while leaving unaltered their ability to adhere to the bone marrow endothelium. In conclusion, our results indicate that CX3CR1, fractalkine, and the enzymes responsible for its cleavage might represent suitable targets for therapies aiming to counteract skeletal secondary tumors from prostate adenocarcinoma. [Cancer Res 2008;68(6):1715-22]

Bone resorption by osteoclasts is thought to promote the proliferation of prostate cancer cells disseminated to the skeleton (Mundy, 2002). Using a mouse model of experimental metastasis, we found that although latestage metastatic tumors were indeed surrounded by osteoclasts, these cells were spatially unrelated to the small foci of cancer cells in early-stage metastases. This is the first evidence that survival and growth of disseminated prostate cancer cells immediately after their extravasation may not depend on osteoclast involvement. Interestingly, prostate cancer cells expressing the a-receptor for platelet-derived growth factor (PDGFRa) progress during early-stages of skeletal dissemination, whereas cells expressing lower levels or lacking this receptor fail to survive after extravasation in the bone marrow. However, non-metastatic cells acquire bone-metastatic potential upon ectopic overexpression of PDGFRa. Finally, functional blockade of human PDGFRa on prostate cancer cells utilizing a novel humanized monoclonal antibodysoon to undergo phase-II clinical trials-significantly impairs the establishment of early skeletal metastases. In conclusion, our results strongly implicate PDGFRa in prostate cancer bone tropism through its promotion of survival and progression of early-metastatic foci, providing ground for therapeutic strategies aimed at preventing or containing the initial progression of skeletal metastases in patients affected by prostate adenocarcinoma.

Supplementary Figure Legends 1-2 from CX3CR1 Is Expressed by Prostate Epithelial Cells and Androgens Regulate the Levels of CX3CL1/Fractalkine in the Bone Marrow: Potential Role in Prostate Cancer Bone Tropism

Supplementary Figures 1-2 from CX3CR1 Is Expressed by Prostate Epithelial Cells and Androgens Regulate the Levels of CX3CL1/Fractalkine in the Bone Marrow: Potential Role in Prostate Cancer Bone Tropism

Supplementary Figures 1-2 from CX3CR1 Is Expressed by Prostate Epithelial Cells and Androgens Regulate the Levels of CX3CL1/Fractalkine in the Bone Marrow: Potential Role in Prostate Cancer Bone Tropism