Interaction of Fibroblast Growth Factor-2 (FGF-2) with Free Gangliosides: Biochemical Characterization and Biological Consequences in Endothelial Cell Cultures

Exogenous gangliosides affect the angiogenic activity of fibroblast growth factor-2 (FGF-2), but their mechanism of action has not been elucidated. Here, a possible direct interaction of sialo-glycolipids with FGF-2 has been investigated. Size exclusion chromatography demonstrates that native, but not heat-denatured, (125)I-FGF-2 binds to micelles formed by gangliosides GT(1b), GD(1b), or GM(1). Also, gangliosides protect native FGF-2 from trypsin digestion at micromolar concentrations, the order of relative potency being GT(1b) > GD(1b) > GM(1) = GM(2) = sulfatide > GM(3) = galactosyl-ceramide, whereas asialo-GM(1), neuraminic acid, and N-acetylneuramin-lactose were ineffective. Scatchard plot analysis of the binding data of fluorochrome-labeled GM(1) to immobilized FGF-2 indicates that FGF–2/GM(1) interaction occurs with a K(d) equal to 6 μM. This interaction is inhibited by the sialic acid-binding peptide mastoparan and by the synthetic fragments FGF-2(112–129) and, to a lesser extent, FGF-2(130–155), whereas peptides FGF-2(10–33), FGF-2(39–59), FGF-2(86–96), and the basic peptide HIV-1 Tat(41–60) were ineffective. These data identify the COOH terminus of FGF-2 as a putative ganglioside-binding region. Exogenous gangliosides inhibit the binding of (125)I-FGF-2 to high-affinity tyrosine-kinase FGF-receptors (FGFRs) of endothelial GM 7373 cells at micromolar concentrations. The order of relative potency was GT(1b) > GD(1b) > GM(1) > sulfatide a = sialo-GM(1). Accordingly, GT(1b),GD(1b), GM(1), and GM(2), but not GM(3) and asialo-GM(1), prevent the binding of (125)I-FGF-2 to a soluble, recombinant form of extracellular FGFR-1. Conversely, the soluble receptor and free heparin inhibit the interaction of fluorochrome-labeled GM(1) to immobilized FGF-2. In agreement with their FGFR antagonist activity, free gangliosides inhibit the mitogenic activity exerted by FGF-2 on endothelial cells in the same range of concentrations. Also in this case, GT(1b) was the most effective among the gangliosides tested while asialo-GM(1), neuraminic acid, N-acetylneuramin-lactose, galactosyl-ceramide, and sulfatide were ineffective. In conclusion, the data demonstrate the capacity of exogenous gangliosides to interact with FGF-2. This interaction involves the COOH terminus of the FGF-2 molecule and depends on the structure of the oligosaccharide chain and on the presence of sialic acid residue(s) in the ganglioside molecule. Exogenous gangliosides act as FGF-2 antagonists when added to endothelial cell cultures. Since gangliosides are extensively shed by tumor cells and reach elevated levels in the serum of tumor-bearing patients, our data suggest that exogenous gangliosides may affect endothelial cell function by a direct interaction with FGF-2, thus modulating tumor neovascularization.