Nature of the interaction of growth factors with suramin.

Suramin inhibits the binding of a variety of growth factors to their cell surface receptors. The direct interaction of suramin with acidic fibroblast growth factor has been detected by the enhancement of the drug's fluorescence in the presence of the protein with the maximum effect occurring at a molar ratio of suramin to aFGF of 2:1. This interaction stabilizes aFGF to thermal denaturation and partially protects a free thiol in its polyanion binding site from oxidation. The binding of suramin to aFGF also induces aggregation of the growth factor to at least a hexameric state as detected by static and dynamic light scattering as well as by gel filtration studies. Both CD and amide I' FTIR spectra of aFGF in the presence and absence of suramin suggest that the drug may also be causing a small conformational change in the growth factor. Suramin produces an even greater aggregation of bFGF and PDGF but not of EGF or IGF-1. Evidence for a suramin-induced conformational change in IGF-1 but not EGF is found by CD, however. It is concluded that suramin binds to many growth factors and that this induces microaggregation and, in some cases, conformational changes. In the case of aFGF, suramin interacts at or near its heparin binding site. The relationship between these phenomena and the anti-growth factor activity of suramin remains to be clearly elucidated.

Both homodimeric isoforms of PDGF (AA and BB) have mitogenic and chemotactic activity and stimulate phosphoinositol turnover.

Platelet-derived growth factor (PDGF) occurs as three dimeric isoforms, AA, BB, and AB, which were previously shown to bind to two receptors with different isoform-specificity, the A/B-type (binds all three isoforms) and the B-type (binds only PDGF-BB). Results from competition binding experiments with Swiss 3T3 cells suggest the existence of a third receptor type, which recognizes PDGF-AB and PDGF-BB. Furthermore, Swiss 3T3 cells and human dermal fibroblasts express different relative and absolute levels of these receptor types. In particular, Swiss 3T3 cells express 90,000 PDGF-AA binding sites (A/B-receptors) per cell, whereas human fibroblasts express only 20,000 A/B-receptors per cell. All three PDGF isoforms were tested in either cell type for their effect on DNA synthesis. PDGF-BB and PDGF-AA were also tested in Swiss 3T3 cells for their effect on inositol phospholipid metabolism and chemotaxis. Each isoform promoted all three processes dose-dependently, but there were differences in the maximum cellular responses elicited. These responses reflect the capacity of the cells to bind the individual isoforms. These results demonstrate that the previous distinctions in responsiveness to the different PDGF isoforms are primarily a consequence of the differences in the levels of surface expression of the various isoform-specific receptor types, rather than of the differences in the intrinsic activity of these isoforms. Furthermore, these results suggest that all types of PDGF receptors are capable of responding to their respective ligands by mediating phosphoinositide breakdown, chemotactic responses, and DNA synthesis. Whether they exhibit other functional differences remains to be seen.

Different effects of homo- and heterodimers of platelet-derived growth factor A and B chains on human and mouse fibroblasts.

Binding sites for platelet-derived growth factor (PDGF) differ in their selectivity for the AA, AB and BB forms of PDGF. Human fibroblasts bind BB well and AA poorly, whereas Swiss 3T3 cells bind more similar quantities of each ligand. We found that AA PDGF was weakly mitogenic for human fibroblasts, but strongly mitogenic for 3T3 cells. Tyrosine phosphorylation of human fibroblast receptors was stimulated most by BB and least by AA, whereas the phosphorylation of 3T3 cell receptors was stimulated more uniformly by the three dimers. The receptor polypeptides that were phosphorylated were very similar. We suggest that phosphorylation of the receptor is proportional to the number of binding sites available for each ligand. Tyrosine phosphorylation of a number of other cell proteins was also proportional to receptor phosphorylation. In contrast, protein kinase C (PKC)-dependent serine and tyrosine phosphorylations were stimulated maximally by low level occupancy of PDGF binding sites, and phosphorylation of p36 required high occupancy. These data raise the possibility that differences in biological potency of AA, AB and BB forms of PDGF may be due simply to differences in the numbers of binding sites, rather than to different biochemical functions of their receptors.

Human peripheral blood monocytes secrete a unique form of PDGF.

PDGF isolated from platelets and forms of PDGF produced by cells transformed with the v-sis gene (PDGF B chain) or U-20S osteosarcoma cells which express the PDGF A chain gene are known to be processed as disulfide-linked dimers of approximately 30 kd. Western blot analysis with anti human PDGF antibody of the PDGF-like factor synthesized and secreted by human blood monocytes (MDGF) on SDS gels indicates that it lacks interchain disulfide bridges and behaves as a 16-kd monomer under nonreducing conditions. Additionally, MDGF exhibits different sensitivities to either formic acid or CNBr cleavage compared to PDGF A or B chain molecules indicating it may have a different primary structure. These data suggest that MDGF represents a unique form of PDGF which lacks interchain disulfide bridges and may represent a new member of the PDGF family of growth factors.