Modes of FGF release in vivo and in vitro.

The fibroblast growth factors (FGFs) are a family of polypeptide growth regulators. The prototypes of this family are acidic and basic FGF. Unusual among their characteristics are a high affinity for the glycosaminoglycan heparin and the lack of a signal sequence for secretion. Other members of the FGF family include a number of oncogene products that also display heparin affinity but do possess signal sequences. Results from early tissue culture studies were consistent with the prediction that acidic and basic FGF would not be secreted. Investigators found that virtually no FGF was secreted into conditioned media, instead it remained cell-associated and was deposited into the basement membrane. More recently, however, a number of studies have indicated that a small amount of FGF is 'released' from cells where it is postulated to act as an autocrine regulator. Acidic and basic FGF have been localized in basement membranes both in vivo and in vitro. The mode of release to this site is also unclear but may be secondary to the mechanisms cited above with soluble FGF becoming bound to heparan sulfate molecules in the extracellular matrix. A number of observations have indicated that matrix-bound FGF is biologically active in vitro. There are no data to indicate whether the same is true for FGF bound to basement membranes in vivo. In addition to its apparent sequestration in the basement membrane, FGF has also been localized to the surface of a variety of normal and tumor cell types. In particular, endothelial cells have been shown to possess two classes of FGF-binding sites: low abundance, high-affinity receptors that mediate the biological activity as well as high abundance, low affinity binding sites. The physiologic relevance of FGF binding to these low affinity sites is not clear. The possibility of locally high concentrations of heparin released by mast cells, as well as the presence of heparan sulfate-degrading enzymes, suggests that this glycosaminoglycan bound FGF might be released from these binding sites under some circumstances. Cell surface binding of FGF has also been demonstrated in vivo; in rabbits plasma levels of the growth factor were shown to be dramatically elevated following intravenous heparinization. Since the FGFs were first noted to lack a signal sequence, cell injury has been suspected to be the most likely route for FGF release in vivo. A number of studies using different models of cell injury, including endotoxins and irradiation, have revealed that damaged cells do release FGF.(ABSTRACT TRUNCATED AT 400 WORDS)