Traditionally, binding proteins are known to regulate the activity of ligands by prolonging their half-life, and insulin-like growth factor (IGF)-binding proteins (IGFBPs) are no exception to this. The IGFBP family contains six high-affinity members with variable functions and mechanisms of actions. In addition to functioning as simple carrier proteins, IGFBPs in serum function to regulate the endocrine actions of IGFs by regulating the amount of IGF available to bind to signaling IGF-I receptors, whereas locally produced IGFBPs act as autocrine/paracrine regulators of IGF action. Furthermore, recent in vitro and in vivo findings that IGFBPs function independently of the IGFs as growth modulators are particularly exciting. Regarding the role of IGFBPs as ligand-independent growth modulators, our recent data that IGFBP-5 stimulates markers of bone formation in osteoblasts lacking functional IGFs provide evidence that IGFBP-5 itself is a growth factor that can act independently of IGFs to regulate bone formation. In terms of the mechanism by which certain IGFBPs mediate their effects in a ligand-independent manner, the binding of IGFBP to its putative receptor on the cell membrane may stimulate the signaling pathway independent of an IGF receptor, to mediate the effects of IGFBPs in certain target cell types. IGFBPs may also exert IGF-independent effects by transcriptional activation of genes by IGFBPs transported into the nucleus via their nuclear localization signal. In conclusion, IGFBPs are unusually pleotrophic molecules with functions ranging from the traditional role of prolonging the half-life of the IGFs to functioning as growth factors independent of the IGFs. In this regard, it was surprising to find that the human genome contains only about 35 000 genes. One mechanism to account for such complexity with a relatively small number of genes is strikingly illustrated by the multifunctional IGFBP class of proteins.

Our previous findings suggest that binding of IGF binding protein-4 (IGFBP-4) to IGFs is essential for the inhibitory effect of IGFBP-4 on the activity of IGFs, both in vitro and in vivo. Therefore, understanding the structural determinants of IGF binding in IGFBP-4 is important to the general understanding of the biology of the IGF system. This study sought to further localize the IGF binding domain and to evaluate the role of Cys residues in IGF binding. Our data revealed that full-length IGFBP-4 peptides lacking the residues Leu(72)-Ser(91) or Leu(72)-His(74) or Gly(75)-Ser(91) failed to bind to IGF-I or IGF-II, whereas deletion of the residue Leu(72) or residues Met(80)-Ser(91) led to a 2- to 3-fold reduction in IGF-I and IGF-II binding activity. The IGF-I and IGF-II binding activities were dramatically reduced by the single mutation, Cys9/Arg (>25-fold), and to a lesser degree, by the single mutation, Cys12/Arg (the first N-terminal Cys residue was designated Cys1). The mutation Cys17/Ser or Cys18/Tyr or Cys20/Ser each resulted in a similar but moderate ( approximately 5-fold) reduction in IGF-II binding activity. The IGF-I binding activity was also dramatically reduced by the mutation Cys18/Tyr, and to a lesser extent, by the mutation Cys17/Ser or Cys20/Ser. These data suggest: 1) the IGF-I and IGF-II binding domain in IGFBP-4 involves a hydrophobic motif (Leu(72)-Met(80)) located in the distal part of the conserved N-terminal region, and 2) the N-terminal Cys residues (Cys9 and Cys12) are more critical than the C-terminal Cys residues (Cys17 and Cys20) in affecting the IGF-I and IGF-II binding. Based on these data, we speculate that the structural determinants of IGF-I and IGF-II binding in IGFBP-4 are very similar, if not identical.