Interactions of neural glycosaminoglycans and proteoglycans with protein ligands: assessment of selectivity, heterogeneity and the participation of core proteins in binding

The method of affinity coelectrophoresis was used to study the binding of n ine representative glycosaminoglycan (GAG)-binding proteins, all thought to play roles in nervous system development, to GAGs and proteoglycans isolat ed from developing rat brain. Binding to heparin and non-neural heparan and chondroitin sulfates was also measured, All nine proteins-laminin-1, fibro nectin, thrombospondin-1, NCAM, L1, protease nexin-1, urokinase plasminogen activator, thrombin, and fibroblast growth factor-2-bound brain heparan su lfate less strongly than heparin, but the degree of difference in affinity varied considerably. Protease nexin-1 bound brain heparan sulfate only 1.8- fold less tightly than heparin (K-d values of 35 vs. 20 nM, respectively), whereas NCAM and L1 bound heparin well (K-d similar to 140 nM) but failed t o bind detectably to brain heparan sulfate (K-d > 3 mu M). Four proteins bo und brain chondroitin sulfate, with affinities equal to or a few fold stron ger than the same proteins displayed toward cartilage chondroitin sulfate, Overall, the highest affinities were observed with intact heparan sulfate p roteoglycans: laminin-1's affinities for the proteoglycans cerebroglycan (g lypican-2), glypican-1 and syndecan-3 were 300- to 1800-fold stronger than its affinity for brain heparan sulfate. In contrast, the affinities of fibr oblast growth factor-2 for cerebroglycan and for brain heparan sulfate were similar. Interestingly, partial proteolysis of cerebroglycan resulted in a >400-fold loss of laminin affinity, These data support the views that (1) GAG-binding proteins can be differentially sensitive to variations in GAG s tructure, and (2) core proteins can have dramatic, ligand-specific influenc es on protein-proteoglycan interactions.