Glycosaminoglycan-protein interactions: definition of consensus sites in glycosaminoglycan binding proteins.

Although interactions of proteins with glycosaminoglycans (GAGs), such as heparin and heparan sulphate, are of great biological importance, structural requirements for protein-GAG binding have not been well-characterised. Ionic interactions are important in promoting protein-GAG binding. Polyelectrolyte theory suggests that much of the free energy of binding comes from entropically favourable release of cations from GAG chains. Despite their identical charges, arginine residues bind more tightly to GAGs than lysine residues. The spacing of these residues may determine protein-GAG affinity and specificity. Consensus sequences such as XBBBXXBX, XBBXBX and a critical 20 A spacing of basic residues are found in some protein sites that bind GAG. A new consensus sequence TXXBXXTBXXXTBB is described, where turns bring basic interacting amino acid residues into proximity. Clearly, protein-GAG interactions play a prominent role in cell-cell interaction and cell growth. Pathogens including virus particles might target GAG-binding sites in envelope proteins leading to infection.

Interaction of fibroblast growth factor-1 and related peptides with heparan sulfate and its oligosaccharides.

Fibroblast growth factors (FGFs) are a family of angiogenic and mitogenic proteins that promote cell division. The binding of FGFs to the heparan sulfate of cell-surface-bound proteoglycans appears to be critical for their activity. The interaction of fibroblast growth factor-1 (FGF-1 or aFGF) using heparin lyase-derived oligosaccharides from heparan sulfate was investigated. FGF-1 was also shown to protect sequences in heparan sulfate from heparin lyase digestion and protected oligosaccharide products of octasaccharide and decasaccharide size were recovered by FGF-1 affinity chromatography, suggesting that the high-affinity binding of heparan sulfate to FGF-1 resides within an octasaccharide sequence. The FGF-1 binding affinity of heparan sulfate is reduced compared to heparin presumably due to the absence of 6-sulfate groups in heparan sulfate. Inspection of the FGF-1 heparan sulfate binding domain shows that the majority of interacting amino acids are contained within a 20-amino-acid sequence that folds back upon itself (because of three turns) forming a triangular shaped cup of positive charge. The importance of FGF-1 binding site topology was investigated using three synthetic peptide mimics of the FGF-1 glycosaminoglycan (GAG) binding site. Heparan sulfate affinity chromatography and isothermal titration calorimetry, used to measure binding thermodynamics, demonstrated that a synthetic peptide analogous to the GAG binding site in FGF-1 bound tightly to heparan sulfate. A peptide containing a D-proline in place of L-proline bound with considerably reduced affinity, presumably due to the altered structure of the second turn in the binding site. A cyclic peptide, expected to be topologically most similar to the triangular GAG binding site in FGF-1, bound with the highest affinity to heparan sulfate. These data suggest the triangular topology of the GAG binding site in FGF is critical for its interaction with heparan sulfate. Analysis of known GAG binding sites in 25 proteins using the Chou-Fasman algorithm show that these sites commonly contain turns.

Dengue virus infectivity depends on envelope protein binding to target cell heparan sulfate.

Dengue virus is a human pathogen that has reemerged as an increasingly important public health threat. We found that the cellular receptor utilized by dengue envelope protein to bind to target cells is a highly sulfated type of heparan sulfate. Heparin, highly sulfated heparan sulfate, and the polysulfonate pharmaceutical Suramin effectively prevented dengue virus infection of target cells, indicating that the envelope protein-target cell receptor interaction is a critical determinant of infectivity. The dengue envelope protein sequence includes two putative glycosaminoglycan-binding motifs at the carboxy terminus; the first could be structurally modeled and formed an unusual extended binding surface of basic amino acids. Similar motifs were also identified in the envelope proteins of other flaviviridae. Developing pharmaceuticals that inhibit target cell binding may be an effective strategy for treating flavivirus infections.

Pattern and spacing of basic amino acids in heparin binding sites.

Glycosaminoglycan (GAG)-protein interactions regulate a myriad of physiologic and pathologic processes, yet an understanding of how these molecules interact is lacking. The role of the pattern and spacing of basic amino acids (arginine (R) and lysine (K)) in heparin binding sites was investigated using peptide analogs as well as by examining known heparin binding sites. Peptides having the general structure R(n)W (n = 3-9, where tyrosine (W) was added for peptide detection) were synthesized and their interaction with heparin was determined by isothermal titration calorimetry. Binding affinity increased with increasing number of R residues. A 9-mer of R (R9W) bound as tightly to heparin as acidic fibroblast growth factor under physiologic conditions. Despite their high affinity for heparin, long stretches of basic amino acids are uncommon in heparin binding proteins. Known heparin binding sites most commonly contain single isolated basic amino acids separated by one nonbasic amino acid. Peptides having the structure, H3CCONH-GRRG(m)RRG(5-m)-CONH2 (denoted as the RRG(m)RR peptide series) and H3CCONH-GRRRG(m)RG(5-m)-CONH2 (denoted as the RRRG(m)R peptide series), where m = 0-5, were synthesized to test the hypothesis that the spacing of basic amino acids in heparin binding sites is optimally arranged to interact with different GAGs. The peptides, in both the -RRG(m)RR- and -RRRG(m)R- peptide series, when m = 0, bound most tightly with heparin, as measured by affinity chromatography. In contrast, the -RRG(m)RR-peptide series interacted most tightly with heparan sulfate when m = 0 or 1, whereas the -RRRG(m)R- peptide series bound tightest when m = 3. These results are consistent with our understanding of heparin and heparan sulfate structure. A highly sulfated GAG, such as heparin, interacts most tightly with peptides (or peptide sequences within proteins) containing a complementary binding site of high positive charge density. Heparan sulfate, having fewer and more highly spaced negatively charged groups, interacts most tightly with a complementary site on a peptide (or peptide sequences with proteins) that has more widely spaced cationic residues.

Structural differences and the presence of unsubstituted amino groups in heparan sulphates from different tissues and species.

This study presents a comparison of heparan sulphate chains isolated from various porcine and bovine tissues. 1H-NMR spectroscopy (500 MHz) was applied for structural and compositional studies on intact heparan sulphate chains. After enzymic digestion of heparan sulphate using heparin lyase I (EC 4.2.2.7) II and III (EC 4.2.2.8), the compositions of unsaturated disaccharides obtained were determined by analytical capillary electrophoresis. Correlations between the N-sulphated glucosamine residues and O-sulphation and between iduronic acid content and total sulphation were discovered using the data obtained by NMR and disaccharide analysis. Heparan sulphate chains could be classified into two groups based on the sulphation degree and the iduronic acid content. Heparan sulphate chains with a high degree of sulphation possessed also a significant number of iduronic acid residues and were isolated exclusively from porcine brain, liver and kidney medulla. The presence and amount of N-unsubstituted glucosamine residues (GlcNp) was established in all of the heparan sulphates examined. The structural context in which this residue occurs was demonstrated to be: high sulphation domain --> 4)-beta-D-GlcAp-(1 --> 4)-alpha-D-GlcNp-(1 --> 4)-beta-D-GlcAp-(1 --> low sulphation domain (where GlcNp is 2-amino-2-deoxyglucopyranose, and GlcAp is glucopyranosyluronic acid), based on the isolation and characterization of a novel, heparin lyase III-derived, GlcNp containing tetrasaccharide and hexasaccharide. The results presented suggest that structural differences may play a role in important biological events controlled by heparan sulphate in different tissues.

Identification of a heparin binding peptide on the extracellular domain of the KDR VEGF receptor.

Vascular endothelial growth factor (VEGF), a potent and specific activator of endothelial cells, is expressed as multiple homodimeric forms resulting from alternative RNA splicing. VEGF121 does not bind heparin while the other three isoforms do, and it has been documented that the binding of VEGF165 to its receptor is dependent upon cell surface heparin sulfate proteoglycans. Little is known about the biochemical mechanism that allows for heparin regulation of growth factor binding. For example, it is not clear whether heparin interactions with growth factor or with cell surface receptors or both are essential for VEGF binding to its receptor. In this manuscript we provide results which are consistent with the hypothesis that an interaction between heparin and a site on the KDR receptor subtype is essential for VEGF165 binding. First, we demonstrate that expression of KDR into a CHO cell line deficient in heparan sulfate biosynthesis does not allow VEGF165 binding unless heparin is exogenously added during the binding assay. Secondly, we show that a ten amino acid synthetic peptide, corresponding to a sequence from the extracellular domain of the KDR, both inhibits VEGF165 binding to the receptor and also binds heparin with high avidity. Third, affinity purification of heparin molecules on a KDR-derived peptide affinity column, together with capillary electrophoresis and polyacrylamide electrophoresis analysis, was used to show that the KDR-derived peptide interacts with a specific subset of polysaccharide chains contained in the unfractionated heparin. Taken together, these results are consistent with the hypothesis that interactions between cell surface heparan sulfate proteoglycans and the VEGF receptor contribute to allowing maximal VEGF binding.

Heparin structure and interactions with basic fibroblast growth factor.

Crystal structures of heparin-derived tetra- and hexasaccharides complexed with basic fibroblast growth factor (bFGF) were determined at resolutions of 1.9 and 2.2 angstroms, respectively. The heparin structure may be approximated as a helical polymer with a disaccharide rotation of 174 degrees and a translation of 8.6 angstroms along the helix axis. Both molecules bound similarly to a region of the bFGF surface containing residues asparagine-28, arginine-121, lysine-126, and glutamine-135, the hexasaccharide also interacted with an additional binding site formed by lysine-27, asparagine-102, and lysine-136. No significant conformational change in bFGF occurred upon heparin oligosaccharide binding, which suggests that heparin primarily serves to juxtapose components of the FGF signal transduction pathway.

Importance of specific amino acids in protein binding sites for heparin and heparan sulfate.

Heparin and heparan sulfate bind a variety of proteins and peptides to regulate many biological activities. Past studies have examined a limited number of established heparin binding sites and have focused on basic amino acids when modeling binding site structural motifs. This study examines the prevalence of individual amino acids in peptides binding to heparin or heparan sulfate. A 7-mer random peptide library was synthesized using the 20 common amino acids. This 7-mer library was affinity separated using both heparin and heparan sulfate-Sepharose. Bound peptide populations were eluted with a salt step gradient (pH 7) and analysed for amino acid composition. Peptides released from heparin-Sepharose by 0.3 M NaCl were enriched in arginine, lysine, glycine and serine; and depleted in methionine and phenylalanine. In contrast, peptides released from heparan sulfate-Sepharose were enriched in arginine, glycine, serine, and proline (at 0.15 M NaCl). These peptides were depleted in histidine, isoleucine, methionine (not detectable) and phenylalanine. In the heparin binding sites of proteins, which have been published, the enriched amino acids were arginine, lysine and tyrosine. Depleted amino acids include aspartic acid, glutamic acid, glutamine, alanine, glycine, phenylalanine, serine, threonine and valine. This study demonstrates that heparin and heparan sulfate bind different populations of peptide sequences. The differences in amino acid composition indicate that the positive charge density and spacing requirements differ for peptides binding these two glycosaminoglycans.

Differences in the interaction of heparin with arginine and lysine and the importance of these basic amino acids in the binding of heparin to acidic fibroblast growth factor.

Although the interaction of proteins with glycosaminoglycans (GAGs) such as heparin are of great importance, the general structural requirements for protein- or peptide-GAG interaction have not been well characterized. Electrostatic interactions between sulfate and carboxylate groups on the GAG and basic residues in the protein or peptide dominate the interaction, but the thermodynamics of these electrostatic interactions have not been studied. Arginine residues occur frequently in the known heparin binding sites of proteins. Arginine is also more common than lysine in randomly synthesized 7-mer peptides that bind to immobilized heparin and heparan sulfate. We have used heparin affinity chromatography, equilibrium dialysis, and isothermal titration calorimetry techniques to further investigate these interactions. A 7-mer of arginine eluted from a heparin-affinity column at 0.82 M NaCl, whereas the analogous 7-mer of lysine eluted at 0.64 M. Similarly, the putative heparin binding site peptide (amino acid residues 110-130) from acidic fibroblast growth factor, which contained four lysine and two arginine residues, eluted at 0.50 M, whereas the analogous peptide with six lysine residues eluted at 0.41 M and one with six arginine residues eluted at 0.54 M. At 25 degrees C in 10 mM sodium phosphate, pH 7.4, carboxy and amino termini blocked arginine (blocked arginine) bound to heparin twice as tightly as blocked lysine as measured by equilibrium dialysis Similarly, at 30 degrees C in 10 mM sodium phosphate, pH 7.4, and in water, blocked arginine bound 2.5 times more tightly to anions in heparin than blocked lysine. Using titration calorimetry, the enthalpy of blocked arginine and lysine binding to heparin was 1.14 +/- 0.24 and 0.45 +/- 0.35 kJ/mol, respectively, under identical conditions. Our observations show that blocked arginine- and arginine-containing peptides bound more tightly to GAGs than the analogous lysine species and suggest that the difference was due to the intrinsic properties of the arginine and lysine side chains. The greater affinity of the guanidino cation for sulfate in GAGs is probably due to stronger hydrogen bonding and a more exothermic electrostatic interaction. This can be rationalized by soft acid, soft base concepts.

Lectin affinity electrophoresis.

Lectin affinity electrophoresis is a powerful technique to investigate the interaction between a lectin and its ligand. Affinity electrophoresis results from the reduced mobility of a charged species owing to its interaction with an immobile species. In this protocol, a two-dimensional lectin affinity electrophoresis experiment is described that affords separation of oligosaccharides. The first-dimension is composed of a weak, polyacrylamide, capillary tube gel containing a lectin. The example described involves a mixture of fluorescently labeled disaccharides. The mobility of only the lectin-binding disaccharide is reduced affording a separation in the first-dimension. The tube gel is then extruded and placed onto the second-dimension gradient polyacrylamide gel and subjected to electrophoresis. Mobility in the second-dimension is dependent on molecular size and visualization si by fluorescence under transillumination. This method is also applicable, with appropriate modifications, for the separation and analysis of glycopeptides and glycoproteins.

Two-dimensional affinity resolution electrophoresis demonstrates that three distinct heparin populations interact with antithrombin III.

Heparin is a polydisperse, highly sulfated polysaccharide consisting of repeating 1-->4 linked uronic acid and glucosamine sugar residues that binds to coagulation proteins, complement proteins, and growth factors to regulate a variety of biological activities. Heparin is best known as an anticoagulant, an activity that results largely from a specific pentasaccharide sequence in heparin that interacts with a unique site in antithrombin III. Little is known about additional structures within heparin that might interact with antithrombin III or the heparin structures that interact with the myriad of other heparin-binding proteins and peptides. Unfractionated glycosaminoglycan heparin that had been prepared from porcine intestinal mucosa was examined for its capacity to bind antithrombin III using a new technique developed to quantitate that interaction. Two-dimensional affinity resolution electrophoresis is a powerful method that allows assessment of unique species of heparin molecules that bind to protein, allowing determination of heparin molecular weight for each protein-binding heparin species as well as the dissociation constant of each interaction. This study provides the first definitive evidence that glycosaminoglycan heparin contains at least three populations of molecules with affinity for antithrombin III. Furthermore, the affinity of each heparin species for antithrombin III appears to vary inversely with the size of the heparin chain, with some smaller oligosaccharides having greater affinity for antithrombin III than larger oligosaccharides.

Nature of the interaction of heparin with acidic fibroblast growth factor.

The binding of human acidic fibroblast growth factor (aFGF) to heparin has been analyzed by a variety of different approaches to better elucidate the nature of this protein/sulfated polysaccharide interaction. Static and dynamic light scattering as well as analytical ultracentrifugation analyses indicates that 14-15 molecules of a FGF can bind to a 16-kDa heparin chain, with approximately 10 of these bound relatively uniformly to high-affinity sites. The dissociation constants of these latter sites are estimated to be approximately 50-140 nM on the basis of surface plasmon resonance experiments in which the association and dissociation rates of aFGF interaction with immobilized heparin were measured. The size of the binding site of a FGF on heparin was also determined by heparin lyase digestion of a FGF/heparin complexes followed by isolation and characterization of protected oligosaccharides. The smallest aFGF-protected oligosaccharide comigrated with delta UA2S(1-->4)-alpha-D-GlcNp2S6S(1-->4)-alpha-L-IdoAp-2S( 1-->4)-alpha-D-GlcNp2S6S (where delta UA represents 4-deoxy-alpha-L-threo-hex-4-enopyranosyluronic acid and S is sulfate). Thus, aFGF appears to bind at high density (one molecule every 4-5 polysaccharide units) and with high affinity to heparin. This potentially provides a concentrated, stabilized storage form of the growth factor that can be released for receptor-mediated cellular activation in response to the proper stimuli. It is also possible that close proximity of aFGF molecules on the highly sulfated regions of heparan chains may be involved in the induction of receptor aggregation as suggested by Ornitz et al. [Ornitz, D. M., Yayon, A., Flanagan, J. G., Svahn, C. M., Levi, E., & Leder, P. (1992) Mol. Cell. Biol. 12, 240-247].