Insulin-like growth factor (IGF)-binding protein release by human fetal fibroblasts: dependency on cell density and IGF peptides.

Insulin-like growth factors (IGFs) are bound to specific binding proteins in extracellular fluids in vivo and when released by cells in vitro. One class of binding protein (IGF-BP), a peptide of 26 kDa purified from amniotic fluid, has been shown to modulate IGF bioactivity on isolated human fibroblasts. We have determined the factors that control release of IGF-BP from monolayers of human fetal fibroblasts using a radioimmunoassay, and have compared this with the effects of these factors on the release of IGF-I and -II. Separation of cell-conditioned cultured medium on SDS-PAGE, and subsequent immunoblotting with antibody against IGF-BP showed that fibroblasts released a single species of immunoreactive protein of an estimated molecular weight of 30 kDa. This was not the predominant binding protein released by cells since major bands of approximately 42 kDa and 39 kDa were visualized following separation by SDS-PAGE and ligand blotting with 125I-labelled IGF-I. The 30 kDa IGF-BP was released in parallel with radioimmunoassayable IGF-I and -II over 48 h. However, a significant inverse correlation was found between the release of IGF-BP, IGF-I, IGF-II and cell density. The exposure of fibroblasts to 1.3 nmol/l or greater of IGF-I or -II caused a significant release of IGF-BP. Maximum release was seen in sparse cultures with little or no release from confluent cultures. IGF-I and -II were approximately equipotent with a fourfold increase in IGF-BP release at 19.7 nmol/l. Insulin caused a release of IGF-BP and IGF-I and -II from fibroblasts at supraphysiological concentrations (16.7 nmol/l) which again was maximal on sparse cell cultures. Increases in IGF-BP, IGF-I and -II release were also found in the presence of human placental lactogen (23.3 nmol/l), but human GH, epidermal growth factor, fibroblast growth factor and platelet-derived growth factor were without effect. The results show that human fetal fibroblasts released an IGF-BP immunologically similar to that seen in amniotic fluid together with IGF-I and -II, that IGF-BP release was enhanced by exogenous addition of IGF peptides, and that the release of all three peptides was a property of sparsely plated, rapidly growing cells. These findings strengthen the hypothesis that the cellular expression of IGF-binding proteins may represent an important level of control in IGF physiology.