Molecular modelling studies on wheat germ agglutinin-saccharide interactions

Abstract

Thus, the present study provides many insights into the conformational dynamics of the saccharides in vacuum and in water. It shows that at room temperature, both in vacuum and in water, the sugar ring exhibits limited flexibility with fluctuations centered around the 4C1 conformation for GlcNAc and GalNAc, and the 2C5 conformation for NeuNAc. However, at high temperature, the sugar ring can readily undergo transitions from one chair conformation to another chair conformation via different boat conformations. The present study demonstrates the high flexibility of the side group hydroxyls and indicates a preference for them to form hydrogen bonds with water molecules rather than forming intramolecular hydrogen bonds. It also reveals the possibility of the glycerol side chain of NeuNAc existing in an extended as well as a bent conformer, with the extended conformation being the predominant one. In conclusion, the present study indicates that these molecules are flexible and the inter-unit dihedral angles (<, >) generally fluctuate very rapidly in a limited region. Hence, these molecules may not exist in any discrete conformation in solution. This study also shows that the solvent water affects significantly the orientation of the hydroxymethyl group and the possible inter- and intra-residue hydrogen bonds. It also dampens the fluctuations in and and restricts them to a very narrow region.

These results also suggest that the use of a distance-dependent dielectric constant (4r) may not completely reflect the effect of a solvent like water, but reflects the gross features of the molecule satisfactorily.

Thus, this study provides the following insights into the conformational dynamics of monosialogangliosides: i) The pyranose ring in the saccharide chain of GM4, GM3, GM2, and GM1 persists as in the monosaccharide in a chair conformation with a limited distortion in ring geometry. ii) Inclusion of water in the simulation restricts the frequency of transition from one conformer to another. iii) In both GM4 and GM3, the (2 3)Gal fragment favors conformations around (-80°, -10°) and (-150°, -20°), while only the latter is favored in the case of GM2 and GM1. iv) For GM4 and GM3 in vacuum, a hydrogen bond is possible between the ring oxygen of NeuNAc and the C-2 hydroxyl of Gal, particularly when the (, ) is around (-80°, -10°). Such a hydrogen bond is not possible in GM1 and GM2. A hydrogen bond between the C-3 hydroxyl of Glc and the ring oxygen of Gal is possible in all the monosialogangliosides. v) In GM1 and GM2, in vacuum, a strong hydrogen bond is possible between the carboxylic oxygen of NeuNAc and the amide hydrogen of GalNAc. This hydrogen bond becomes weaker in the presence of water.

These studies have, therefore, provided the following insights into the binding of some monosaccharides to WGA: i) The calculated binding energies of GlcNAc, GalNAc, and ManNAc explain the experimental studies, showing that GlcNAc binds better to WGA than GalNAc and ManNAc. ii) GalNAc may bind to WGA in nearly the same orientation as GlcNAc at both the primary and secondary sites, whereas ManNAc binds in totally different orientations. iii) NeuNAc binds equally well at the primary binding site and weakly at the secondary binding site compared to GlcNAc. iv) In the preferred conformer, NeuNAc and GlcNAc have similar hydrogen bonding schemes only at the primary site. v) The secondary site is more specific than the primary site. The modes of binding of NeuNAc to WGA and HA are quite different, and the glycerol side chain does not play a significant role while binding to WGA. vi) The energetically favored orientations of GlcNAc or NeuNAc are not significantly affected when they are part of a disaccharide fragment, as in (GlcNAc)2 or (NeuNAc-Gal). vii) The Trp150 residue at the secondary site is involved in sugar binding, in agreement with fluorescence and NMR studies, but in disagreement with X-ray studies.

In conclusion, the present study clearly identifies Trp150 at the secondary site as being involved in sugar binding. This explains the fluorescence and NMR results. Such an interaction is not indicated in the model of C.S. Wright from X-ray crystallographic studies. However, in the (GlcNAc)3-WGA complex, having the saccharide at either the primary or the secondary site, the hydrogen bond schemes predicted for the lowest energy conformer agree with the models proposed by C.S. Wright for the non-reducing sugar with the protein. In addition to the hydrogen bonds proposed by C.S. Wright, the present study also suggests the possibility of additional hydrogen bond interactions between the middle and reducing sugar residues at the primary site, and between the middle sugar residue and the protein at the secondary site. These interactions perhaps explain the higher affinity of (GlcNAc)3 and (GlcNAc)2 over GlcNAc.

The hydrogen bonds proposed by C.S. Wright between (GlcNAc)2 and WGA do not agree with the lowest energy conformer when the sugar is bound either at the primary or secondary sites. At the primary site, the agreement is better with conformer1, and at the secondary site with conformer3. Conformer2 at the primary site has about 2.4 kcal/mol higher energy than conformer1. Unlike conformer1, for conformer2 at the primary site, both sugar residues of (GlcNAc)2 participate in hydrophobic interactions with the His66 and Tyr64 rings. Additionally, the close proximity of the CH3 group to the top face of the Tyr73 ring leads to the possibility of additional hydrophobic interactions. Similarly, at the secondary site, the minimum energy conformer of (GlcNAc)2 is not the one present in the solid state. In the lowest energy conformer, (GlcNAc)2 does not interact with Trp150, and hydrogen bond interactions are quite different from those reported by X-ray diffraction studies.

This clearly suggests that conformer1 is not consistent with earlier studies either in solid state or in solution. In conformers 2 and 3, Trp150 is placed close to the CH3 group of the non-reducing sugar, leading to possible hydrophobic interactions. Though conformer3 has about 4 kcal/mol higher energy than conformer2, Trp150 stacks with the reducing sugar only in conformer3, and hence from the point of hydrophobic interactions, conformer3 is more favored than conformers1 and 2. It is well known that such interactions are largely responsible for driving the molecule to the binding site. Once the molecule reaches the binding site, hydrogen bonds and electrostatic interactions stabilize the complex. Generally, stacking interactions are not reflected in the present calculations. Perhaps they are mainly responsible for the occurrence of conformer2 at the primary site and conformer3 at the secondary site in the solid state.

These results, therefore, suggest that if the calculated potential energy differences between two conformers are not very large, hydrophobic interactions may play a major role in the binding of saccharides to WGA. In all the complexes at both primary and secondary sites, the glycerol side chain of NeuNAc favors an extended conformation, which is accessed frequently in the MD results described in Chapter 5. In all the complexes, the NeuNAc residue of the monosialogangliosides favorably interacts with WGA both at the primary and secondary sites, suggesting the importance of the binding of NeuNAc at the interior of the binding site. However, these saccharides preferentially bind at the primary binding sites of WGA, as hydrophobic interactions are possible only at this site.

In all the low-energy conformers of WGA-GM1 and WGA-GM2 complexes, the NeuNAc-Gal fragment favors a conformation close to (-150°, -20°), whereas in GM4 and GM3 complexes, the NeuNAc-Gal fragment favors (, ) values around (-70°, -10°) and (-150°, -20°). In the primary site, the conformers of the WGA-GM4 and WGA-GM3 complexes agree with the energetically favorable mode of binding in the WGA-NeuNAc complex as described in Chapter 6 and also with the X-ray study of the WGA-GM3 complex (C.S. Wright, 1990).