Supporting information for Raman et al. (2003) Proc. Natl. Acad. Sci. USA, 10.1073/pnas.0437842100

A recently published study had identified oligosaccharide sequences derived from heparan sulfate with graded affinities for FGF-1 and -2 (1). Based on the salt concentration required to elute different octasaccharide fractions from immobilized FGF-1 column, different fractions were classified as low-affinity (0.2 M salt), moderately high-affinity (0.5–0.7M salt), and high-affinity (1 M salt) binders (Fig. 6). It is important to point out that the number of sulfate groups is about the same for the low-affinity and moderately high-affinity binders. This indicates that high-affinity binding is not just a function of nonspecific electrostatic interactions of the negatively charged sulfate groups with the basic residues on FGF-1. The kink observed in the FGF-1-binding oligosaccharide, as described in the main article, comprises a specific trisaccharide motif of an iduronic acid in the ¹C₄ conformation flanked by two glucosamines (H-I-H; H, a -D-glucosamine; I, a -L-iduronic acid), thus indicating a structural requirement for an optimal molecular fit into the HSGAG-binding site of FGF-1, in addition to the ionic interactions between the sulfate groups and the basic residues. The sulfation pattern of the kink spanning trisaccharide is such that the N- sulfate on the nonreducing end H, 2-O-sulfate on I, and 6-O-sulfate on the reducing end are aligned pointing outward from the helical axis (Fig. 3). As shown in Fig. 6, the affinity of the oligosaccharide to FGF-1 correlates directly with the frequency of occurrence of the specific trisaccharide kink along with the critical sulfate groups (indicated in red). Thus, the critical sulfate groups that provide optimal ionic contact and the kink in the helical axis of the HSGAG that maximizes the surface (van der Waals) contact influence the affinity of different oligosaccharides to FGF-1. These observations provide important structural insights to rationally interpret the graded affinities for different HSGAG oligosaccharides to FGF.

We also investigated the cocrystal structures of HSGAG oligosaccharides with antithrombin (AT-III) (2), foot and mouth disease virus (FMDV) (3), annexin V (4), and the NK1 domain of hepatocyte growth factor (HGF) (5). In the AT-III–pentasaccharide (H_NAc,6SGH_NS,3S,6SI_2SH_NS,6S) cocrystal complex, the iduronic acid residue adopts a ²S₀ conformation. Comparison of the n_L (local number of turns) and h_L (local axial rise) values between the bound and unbound pentasaccharide showed that the n_L value deviates significantly and changes from 2.0 to 3.0. This transition in the n_L value reflects a local conformational change causing a helical overwind spanning a trisaccharide. Similar to the FGF system, the trisaccharide (H_NS,3S,6SI_2SH_NS,3S,6S)-spanning kink contains the critical 3-O-sulfate group (of the nonreducing end glucosamine) required for specific interactions with AT-III (Fig. 7). Thus, in the case of AT-III–pentasaccharide complex, the optimal ionic and van der Waals contact with the protein requires the overwinding of the helical axis of the pentasaccharide, resulting in a kink.

In the FMDV cocrystal structure three individual domains that are related by threefold symmetry constitute the HSGAG-binding site (3). The coordinate dataset of a pentasaccharide (I_2SH_NS,6SI_2SH_NS,6SI_2S) was available for this cocrystal structure in the Protein Data Bank. Similar to the case of AT-III-bound pentasaccharide, the n_L and h_L calculations indicate an overwind in the HSGAG oligosaccharide with the n_L value close to 3.0. This region spans three monosaccharides (H_NS,6SI_2SH_NS,6S) with the iduronate in the ²S₀ conformation, which is very similar to the trisaccharide spanning kink in the AT-III-binding pentasaccharide. The coordinate dataset of a pentasaccharide (I_2SH_NS,6SI_2SH_NS,6SI_2S) was available for the NK1 domain cocrystal structure in the PDB. The H_NS,6SI_2SH_NS,6S trisaccharide with the iduronate adopting a ¹C₄ conformation in the HSGAG-binding region of the protein is structurally very similar to that of the kink in FGF-1- and FGF-2-bound oligosaccharides. Because annexin V was cocrystallized with only tetrasaccharides, it was difficult to make a meaningful interpretation of the n_L and h_L values for quantifying the deviations from the overall uniform helical structure. Thus, a larger oligosaccharide fragment complexed with the protein will be required to better understand this HSGAG–protein system.

1. Kreuger, J., Salmivirta, M., Sturiale, L., Gimenez-Gallego, G. & Lindahl, U. (2001) J. Biol. Chem. 276, 30744–30752.

2. Jin, L., Abrahams, J. P., Skinner, R., Petitou, M., Pike, R. N. & Carrell, R. W. (1997) Proc. Natl. Acad. Sci. USA 94, 14683–14688.

3. Fry, E. E., Lea, S. M., Jackson, T., Newman, J. W., Ellard, F. M., Blakemore, W. E., Abu-Ghazaleh, R., Samuel, A., King, A. M. & Stuart, D. I. (1999) EMBO J. 18, 543–554.

4. Capila, I., Hernaiz, M. J., Mo, Y. D., Mealy, T. R., Campos, B., Dedman, J. R., Linhardt, R. J. & Seaton, B. A. (2001) Structure (Cambridge, U.K.) 9, 57–64.

5. Lietha, D., Chirgadze, D. Y., Mulloy, B., Blundell, T. L. & Gherardi, E. (2001) EMBO J. 20, 5543–5555.