Insulin-like growth factor (IGF) binding to human fibroblast and glioblastoma cells: the modulating effect of cell released IGF binding proteins (IGFBPs)

The cell surface of human fibroblasts contains not only type I IGF receptors but at least two forms of IGFBPs. Studies were undertaken to analyze the mechanisms by which these IGFBPs alter IGF-I-cell surface interactions. Human fetal fibroblasts (GM10) and a human glioblastoma cell line (1690) were chosen for analysis. During assays to quantify 125I]-IGF-I binding, both cell lines were shown to release IGFBPs into the binding assay buffer. Under equilibrium conditions, 125I]-IGF-I preferentially associates with IGFBPs in the assay buffer (up to 40% of the 125I]-IGF-I added) since they have a higher affinity than type I IGF receptors or IGFBPs associated with the cell surface. Likewise the addition of increasing concentrations of unlabeled IGF-I results in preferential competition for binding to assay buffer IGFBPs. This results in a repartitioning of the 125I]-IGF-I that is bound to assay buffer IGFBPs onto cell surface binding sites. The degree of repartitioning is quantitatively related to the amount of 125I]-IGF-I bound to released IGFBPs. When cultures are exposed to cycloheximide before the binding assay, both the amount of IGFBPs that are released into the assay buffer and the amount of 125I]-IGF-I that is repartitioned are decreased. In contrast when Gln3, Ala4, Tyr15, Leu16]-IGF-I (QAYL]-IGF-I, an IGF analog that has unaltered affinity for type I IGF receptors) is iodinated and tested, the competition curve with unlabeled IGF-I shows no repartitioning effect. This form of IGF can be used to quantify type I receptor number independent of the presence of IGFBPs. IGF-I and the QAYL]-IGF-I compete equally with the 125I]-QAYL]-IGF-I for binding to cell surfaces, whereas unlabeled QAYL]-IGF-I is greater than 25-fold less potent compared to IGF-I in competing with 125I]-IGF-I for cell surface binding. Specific binding of 125I]-QAYL]-IGF-I to GM10 and 1690 cell surfaces is less than 20% of 125I]-IGF-I binding. These findings suggest that IGFBPs that are present on human fibroblast surfaces represent a large portion of the IGF binding sites. We conclude that the amount of IGFBPs released into assay buffer is a major determinant of the repartitioning of 125I]-IGF-I to cell surface binding sites and that both cell surface and assay buffer IGFBPs modulate type I IGF receptor binding.