IGFBP-3 effect in prostate cancer cells and osteoblasts

YANG J; KREIMANN EL; SIKES CR; STEPHENS CL; LIN SUE-HWA; PELEG S; OLIVE M; LOGOTHETIS CJ; KARSENTY G; NAVONE NM

Washington Convention Center, Washington. D.C. USA.

Congreso; 94th Annual Meeting. American Association for Cancer Research, AACR.; 2003

American Association for Cancer Research

Advanced prostate cancer (PCa) is dominated by complications arising from bone metastases.  Bone metastases of PCa are characteristically blastic, i.e., they involve excessive bone formation, and bone is frequently the only site of progression.  These observations suggest that interaction between PCa cells and cells of the osteoblastic lineage plays a central role in PCa progression.  To study this interaction, we established an in vitro model of bone metastasis and a bicompartment co-culture system of primary mouse osteoblasts (PMOs) and human PCa cells (MDA PCa 2a and MDA PCa 2b). Several soluble factors expressed by PCa cells have been implicated as osteoblast- stimulating factors; among them, insulin-like growth factor 1 (IGF-1) plays key roles in bone formation and is an important regulator of the activity of mature osteoblasts. Moreover, free IGF-1 has been implicated in the development and metastasis of PCa. Most IGF-1s in the circulation is bound to IGF binding protein 3 (IGFBP-3), which modulates the amount of bioavailable free IGF-1 and regulates its transfer from the circulation to tissue sites of action.  We previously reported that the IGFBP-3 transcript level produced by PCa cells decreases when these cells are co-cultured with PMOs and that IGFBP-3 has an antiproliferative effect on PMOs when added to the culture medium. We report herein in that IGFBP-3 protein levels decreased in the co-culture medium as assessed using western blot and enzyme-linked immunoassay (ELISA) analysis (18.19±0.34 vs. 16.37±0.18 ng/ml).  Furthermore, PMO DNA synthesis (as assessed by tritium incorporation) decreased after the addition of IGFBP-3 and reached statistical significance at the following concentrations of IGFBP-3: 0.05, 0.1, 1, and 2 µg/ml.  After PMOs were co-culture with PCa cells, this inhibition was almost abolished at the lower two IGFBP-3 concentrations but not at the higher two (1 µg/ml and 2 µg/ml).  Under the same experimental conditions, the antiproliferative effect of IGFBP-3 is not reproduced by IGFBP-2 (0.01-1 µg/ml) or blocking IGF-1 signaling with an IGF1 receptor-neutralizing antibody at 1 µg/ml. Taken together, these results suggest that the inhibitory effect of IGFBP-3 in PMOs is specific and not attributable to sequestration of endogenous IGF-1. In contrast with the inhibition found in PMOs, IGFBP-3 had a slight antiproliferative effect in PCa cells only at the highest concentration tested (2 µg/ml) in cells growing alone or in co-culture. To assess the expression of IGFBP-3 in vivo, we performed immunohistochemical analysis of IGFBP-3 when PCa cells were growing in the femur of SCID mice.  We found that levels were high in the early stages and decreased with tumor progression.  Moreover, early lesions showed a lower apoptotic index than late stage tumors did. In conclusions, a) the effect of IGFBP-3 on PMOs seems to be IGF-1 independent, b) IGFBP-3 does not have an antiproliferative effect on PCa cells, and c) the IGFBP-3 level may have a role in the initial growth of PCa in bone.