Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (21)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Moreno, E.
Right arrow Articles by Perez, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Moreno, E.
Right arrow Articles by Perez, R.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Glycobiology Pages 695-705  

Delineation of the epitope recognized by an antibody specific for N-glycolylneuraminic acid-containing gangliosides
Introduction
Results And Discussion
Materials And Methods
References

Delineation of the epitope recognized by an antibody specific for N-glycolylneuraminic acid-containing gangliosides

Delineation of the epitope recognized by an antibody specific for N-glycolylneuraminic acid-containing gangliosides

Ernesto Moreno1,2,4, Boel Lanne2, Ana María Vázquez1, Ikuo Kawashima3, Tadashi Tai3, Luis Enrique Fernández1, Karl-Anders Karlsson2, Jonas Ångström2, Rolando Pérez1

1Center of Molecular Immunology, P.O. Box 16040, Havana 11600, Cuba, 2Institute of Medical Biochemistry, Göteborg University, Medicinaregatan 9A, S-413 90 Göteborg, Sweden and 3Department of Tumor Immunology, Tokyo Metropolitan Institute of Medical Science, Honkomagome, Bunkyo-ku, Tokyo 113, Japan

Received on October 31, 1997; revised on February 4, 1998; accepted on February 5, 1998

P3 is a mouse monoclonal antibody (mAb) that binds to several NeuGc-containing gangliosides. It also reacts with antigens expressed in human breast tumors (Vázquez et al. (1995) Hybridoma, 14, 551-556). In this work, the binding specificity of P3 has been characterized in more detail using a panel of glycolipids that included several disialylated gangliosides and several chemical derivatives of NeuGc-GM3. The carboxyl group and the nitrogen function of sialic acid were found to play important roles in the antibody binding, whereas the glycerol tail appears to be nonrelevant. Molecular modeling was used to analyze the binding data, including the finding that P3 selectively recognizes the internal NeuGc in GD3. For this purpose, conformational studies of GD3 were performed using molecular dynamics. It was concluded that sialic acid binds the P3 antibody through its upper face (the one on which the carboxyl group is exposed) and the C4-C5 side of the sugar ring, whereas none or very little contact between the galactose residue and the protein is evident. Conformational analysis of GD3 revealed that, despite the large flexibility of the NeuGc[alpha]8NeuGc linkage, the P3 binding epitope on the external sialic acid is not well exposed for any of the possible conformations this linkage can adopt, whereas the internal sialic acid presents the epitope in a proper way for several of these conformations. As a final result, a coherent picture of the epitope that fits the wide binding data was obtained.

Key words: epitope/ganglioside/molecular modeling/ monoclonal antibody/ NeuGc

Introduction

P3 is a mouse monoclonal antibody (mAb) that binds to several N-glycolyl-containing gangliosides and also recognizes antigens expressed in human breast tumors (Vázquez et al., 1995). Only a few monoclonal antibodies (mAbs) specific for NeuGc-containing gangliosides have previously been reported (Nakamura and Handa, 1986; Furukawa et al., 1988; Miyake et al., 1988; Sanai et al., 1988; Watarai et al., 1991; Ozawa et al., 1992; Kawashima et al., 1993). As compared to these antibodies, P3 is unique in recognizing the N-glycolyl sialic acid on several different types of sugar chains, as shown in the work presented here, which was directed toward characterization of the binding specificity of the P3 mAb and delineation of the epitope that it recognizes.

Detailed characterization of the binding specificity of an antibody is important for several applications; for example, antibodies are often used as tools for testing the presence of molecules in cells, tissues or chemical preparations. This kind of information becomes especially relevant if the antibody has the ability of recognizing tumor cells and, therefore, may be used for diagnostic and therapeutic purposes.

On the other hand, studying the interactions between antibodies or other proteins and the N-acetyl or N-glycolyl types of sialic acid is also important in order to understand the molecular basis of the fine distinction that biological systems make between these two species. N-Acetylneuraminic and N-glycolylneuraminic acids are the most common types of sialic acid found in animals, occurring in a wide variety of sialoglycoproteins and gangliosides (Corfield and Schauer, 1982). In general, normal human tissues yield only NeuAc, while other mammals have significant levels of NeuGc. Despite the small structural difference between them (the methyl group in NeuAc being substituted by a hydroxymethyl group in NeuGc), the two derivatives do not appear to be biologically equivalent.

Knowledge of the way in which a molecule is recognized by a protein at the atomic level is also important for practical applications, for example, the design of inhibitors or better immunogens, or for theoretical docking studies that may provide valuable information for genetic engineering of a protein binding site. X-Ray crystallography of the receptor-ligand complex provides such information, but it is still far from being a routine procedure. At present, only one crystal structure of an anti-tumor antibody in complex with its carbohydrate antigen has been reported: the BR96 antibody in complex with LeY (Jeffrey et al., 1995). A small number of crystal structures of proteins interacting with N-acetyl sialic acid (alone or as part of a larger molecule) has been reported (Wright, 1990; Sauter et al., 1992; Varghese et al., 1992; Merritt et al., 1994), whereas no complexes containing the N-glycolyl type have been solved to date.

Several reports are found in the literature in which carbohydrate epitopes have been explored at the atomic level by chemical methods. Chemically modified glycolipids have, for example, been used to investigate the epitope recognized by the cholera toxin B-subunit on GM1 (Fishman et al., 1980; Spiegel, 1985; Schengrund and Ringler, 1989; Lanne et al., 1994) and the epitope on NeuGc-GM3 that interacts with the E.coli K99 adhesin (Lanne et al., 1995). Molecular modeling applied in conjunction with binding assays may also help to extract more information about binding epitopes (Ångström et al., 1994). In this work, a combination of different techniques was used, including binding experiments for a wide panel of NeuGc-containing gangliosides and chemical derivatives of sialic acid of NeuGc-GM3, as well as molecular modeling.

Table I. Structures and P3 mAb reactivities of glycolipids used in binding experiments
Abbreviation Structurea Binding
    TLCb ELISAc
LacCer Gal[beta]4Glc[beta]Cer - -
Gangliotri (Gg3) GalNAc[beta]4Gal[beta]4Glc[beta]Cer - -
Gangliotetra (Gg4) Gal[beta]3GalNAc[beta]4Gal[beta]4Glc[beta]Cer - -
Neolactotetra (PG) Gal[beta]4GlcNAc[beta]3Gal[beta]4Glc[beta]Cer - nd
NeuAc-GM4 NeuAc[alpha]3Gal[beta]Cer - -
NeuAc-GM3 NeuAc[alpha]3Gal[beta]4Glc[beta]Cer - -
NeuAc-GM2 GalNAc[beta]4(NeuAc[alpha]3)Gal[beta]4Glc[beta]Cer - -
NeuAc-GM1a Gal[beta]3GalNAc[beta]4(NeuAc[alpha]3)Gal[beta]4Glc[beta]Cer - -
NeuAc-GM1b NeuAc[alpha]3Gal[beta]3GalNAc[beta]4Gal[beta]4Glc[beta]Cer - -
NeuAc-NeuAc-GD3 NeuAc[alpha]8NeuAc[alpha]3Gal[beta]4Glc[beta]Cer - -
NeuAc-GD2 GalNAc[beta]4(NeuAc[alpha]8NeuAc[alpha]3)Gal[beta]4Glc[beta]Cer - -
NeuAc-GD1a NeuAc[alpha]3Gal[beta]3GalNAc[beta]4(NeuAc[alpha]3)Gal[beta]4Glc[beta]Cer - -
NeuAc-GD1b Gal[beta]3GalNAc[beta]4(NeuAc[alpha]8NeuAc[alpha]3)Gal[beta]4Glc[beta]Cer - -
NeuAc-GT1b NeuAc[alpha]3Gal[beta]3GalNAc[beta]4(NeuAc[alpha]8NeuAc[alpha]3)Gal[beta]4Glc[beta]Cer - -
NeuAc-SPG NeuAc[alpha]3Gal[beta]4GlcNAc[beta]3Gal[beta]4Glc[beta]Cer - -
NeuGc-GM3 NeuGc[alpha]3Gal[beta]4Glc[beta]Cer + ++
NeuGc-GM2 GalNAc[beta]4(NeuGc[alpha]3)Gal[beta]4Glc[beta]Cer + ++
NeuGc-GM1a Gal[beta]3GalNAc[beta]4(NeuGc[alpha]3)Gal[beta]4Glc[beta]Cer - ±
NeuGc-GM1b NeuGc[alpha]3Gal[beta]3GalNAc[beta]4Gal[beta]4Glc[beta]Cer + +
NeuGc-NeuGc-GD3 NeuGc[alpha]8NeuGc[alpha]3Gal[beta]4Glc[beta]Cer + ++
NeuAc-NeuGc-GD3 NeuAc[alpha]8NeuGc[alpha]3Gal[beta]4Glc[beta]Cer + ++
NeuGc-NeuAc-GD3 NeuGc[alpha]8NeuAc[alpha]3Gal[beta]4Glc[beta]Cer - ±
NeuGc-GD1a NeuGc[alpha]3Gal[beta]3GalNAc[beta]4(NeuGc[alpha]3)Gal[beta]4Glc[beta]Cer + +
NeuGc-GD1c NeuGc[alpha]8NeuGc[alpha]3Gal[beta]3GalNAc[beta]4Gal[beta]4Glc[beta]Cer - ±
NeuGc-SPGd NeuGc[alpha]3Gal[beta]4GlcNAc[beta]3Gal[beta]4Glc[beta]Cer + nd
VI3NeuGc[alpha]-nLc6d NeuGc[alpha]3Gal[beta]4GlcNAc[beta]3Gal[beta]4GlcNAc[beta]3Gal[beta]4Glc[beta]Cer + nd
aThe condensed sugar chain representation follows recent recommendations (Sharon, 1987).
bThe reactivities for the TLC experiments are given using a qualitative scale (positive or negative).
cBinding strengths according to results from ELISA (see Figure 4). (++), (+), and (±) denote strong, moderate, and weak binding, respectively. nd, Not determined.
dNot shown in the figures.

Results And Discussion

Binding of P3 mAb to various glycolipids by HPTLC immunostaining and ELISA

The binding of P3 mAb to different glycolipids (see Table I) was determined by HPTLC (Figures 1-3) and ELISA (Figure 4). Results are summarized in Table I. The mAb bound strongly to NeuGc-GM3 and NeuGc-GM2, as previously reported (Vázquez et al., 1995). A moderate binding was observed for NeuGc-GM1b, whereas NeuGc-GM1a, which did not bind on the TLC plate, showed a weak reactivity by ELISA (Figures 1 and 4). The corresponding NeuAc-containing monosialogangliosides did not reveal any antibody binding.


Figure 1. TLC immunostaining of monosialogangliosides with P3 mAb. (A) Ganglioside fraction from bovine brain (10 nmol/lane) and standard monosialogangliosides (1 nmol/lane) were chromatographed with chloroform/methanol/0.22% CaCl2 in water (55:45:10, v/v/v) and visualized with orcinol. (1) Bovine brain gangliosides, (2) NeuGc-GM3, (3) NeuAc-GM3, (4) NeuGc-GM2, (5) NeuAc-GM2, (6) NeuGc-GM1a, (7) NeuAc-GM1a, (8) NeuGc-GM1b, (9) NeuAc-GM1b. (B) The same gangliosides (0.1 nmol/lane) as in (A) were chromatographed under the same conditions as above and immunostained with P3 mAb.
Among the NeuGc-containing disialogangliosides that were studied, P3 bound moderately to NeuGc-GD1a, whereas a weak binding was observed to NeuGc-GD1c only by ELISA (Figures 2 and 4). Four GD3 isomers: NeuGc-NeuGc-GD3, NeuAc-NeuAc-GD3, NeuAc-NeuGc-GD3, and NeuGc-NeuAc-GD3 were also tested (Figures 3 and 4). NeuGc-NeuGc-GD3 and NeuAc-NeuGc-GD3 were strongly bound by P3 mAb, whereas the isomer having NeuGc only in the external position showed a weak binding by ELISA and NeuAc-NeuAc-GD3 was not recognized at all. These results indicate that the inner NeuGc residue in GD3 is essential for the binding of P3 mAb. A reasonable explanation for this fact was obtained through molecular modeling analysis (see below).


Figure 2. TLC immunostaining of disialogangliosides with P3 mAb. (A) The ganglioside fraction from bovine brain (10 nmol/lane) and purified disialogangliosides were chromatographed with chloroform/methanol/0.22% CaCl2 in water (55:45:10, v/v/v) and visualized with orcinol stain. (1) Bovine brain gangliosides, (2) NeuGc-NeuGc-GD3, (3) NeuAc-NeuAc-GD3, (4) NeuGc-GD1a, (5) NeuAc-GD1a. (B) The same gangliosides (0.1 nmol/lane) as in (A) were chromatographed as above and immunostained with P3 mAb.

P3 also bound strongly to NeuGc-SPG and VI3NeuGc[alpha]-nLc6 (data not shown). None of the other tested molecules having only the N-acetyl type of sialic acid (GM4, GD2, GD1a, GD1b, GT1b, NeuAc-SPG) or neutral glycolipids (LacCer, Gg3, Gg4, and PG) were recognized (data not shown).

Binding of P3 mAb to chemical derivatives of NeuGc-GM3

Figure 5 shows the positions on sialic acid where chemical substitutions were made, as well as the groups that were attached at each position. The binding responses of P3 mAb to these derivatives are shown in Figure 6.


Figure 3. TLC immunostaining of four GD3 isomers with P3 mAb. (A) TLC of four GD3 isomers (1 nmol/lane) visualized with orcinol. (1) Bovine brain gangliosides, (2) NeuGc-NeuGc-GD3, (3) NeuAc-NeuAc-GD3, (4) NeuAc-NeuGc-GD3, (5) NeuGc-NeuAc-GD3. (B) The same gangliosides (0.1 nmol/lane) as in (A) were chromatographed identically and immunostained with P3 mAb.


Figure 4. Binding of P3 mAb to purified glycolipids measured by ELISA. (A) Wells were coated with different concentrations of each glycolipid and incubated with a constant amount of P3 mAb. (B) Wells were coated with a constant amount of each glycolipid (0.2 nmol/well) and incubated with different concentrations of the antibody. Each symbol in the figure represents a group of glycolipids having the same affinity for P3 mAb: solid circles, NeuGc-GM3, NeuGc-GM2, NeuGc-NeuGc-GD3 and NeuAc-NeuGc-GD3; solid squares, NeuGc-GM1b, NeuGc-GD1a; triangles, NeuGc-GM1a, NeuGc-NeuAc-GD3, and NeuGc-GD1c; open circles, gangliosides having NeuAc as their sialic acid residue(s) (GM4, GM3, GM2, GM1a, GM1b, GD3, GD2, GD1a, GD1b, and GT1b) and neutral glycolipids (LacCer, Gg3, and Gg4).


Figure 5. Structure of the oligosaccharide part of GM3 showing naturally occurring groups and chemical substitutions made at positions C1, N5, and C7, which were tested for binding to P3 mAb. Natural derivatives are shown in bold and marked with an asterisk. Torsional angles of the glycosidic linkage between sialic acid and galactose are defined as follows: [Phi] = (C1-C2-O2-C3Gal), [Psi] = (C2-O2-C3Gal-H3Gal).


Figure 6. P3 mAb binding to derivatives and isomers of NeuGc-GM3. (A) Chemical staining of glycolipids (5 µg/lane). (B) Binding of the same compounds to P3 mAb. (1) NeuGc-GM3 alcohol, (2) NeuGc-GM3 amide, (3) NeuGc-GM3 methylamide, (4) NeuGc-GM3 ethylamide, (5) NeuGc-GM3 propylamide, (6) NeuGc-GM3 benzylamide, (7) NeuGc-GM3, (8) NeuAc-GM3, (9) NeuGc-GM3 (same as (7)), (10) de-N-acetyl-GM3 (NeuNH2-GM3), (11) TFA-GM3, (12) TFI-GM3, (13) periodate-oxidized NeuGc-GM3.

Modifications of the carboxyl group. The carboxyl group of NeuGc-GM3 was modified to alcohol and several amides. None of these derivatives was recognized by the antibody, as shown in Figure 6 (lanes 1-6). It should be pointed out, however, that the negative binding results obtained in these experiments, where bulky groups had been added to the C1 carbon (such as propyl- and benzylamide), do not necessarily indicate a direct participation of the carboxyl group in the binding interactions, since such negative responses might be due to steric interferences caused by these substituents. The nonbinding of the alcohol and amide derivatives, however, points more strongly to a direct contact of the COO- group of sialic acid with the antibody. Nonetheless, the possibility that binding might be affected as a result of conformational changes involving the glycosidic linkage close to the substituted group was also investigated.

Molecular dynamics simulations were thus carried out for the terminal disaccharide of the NeuGc-GM3 alcohol and amide derivatives, as well as for the native molecule, serving as control. Figure 7 shows the [Phi] and [Psi] trajectories obtained from these runs. For the three molecules, the values of these torsional angles corresponded to the anticlinal ([Phi] [ap] -155°, [Psi] [ap] -30°) and the synclinal ([Phi] [ap] -70°, [Psi] [ap] 10°) conformations, as was similarly obtained by Siebert et al. (1992) for NeuAc- and NeuGc-GM3. The behavior of the glycerol tail of sialic acid was the same for the three molecules (data not shown), and corresponded also to that obtained by Siebert et al. (1992). Thus, no conformational changes that might account for the nonbinding of the antibody to these derivatives were found.


Figure 7. Plots of the [Phi] (+) and [Psi] (solid circles) trajectories obtained from dynamics runs for NeuGc-GM3 alcohol (A), NeuGc-GM3 amide (B), and NeuGc-GM3 (C). See Figure 5 for definitions of angles.

It was of interest to compare these results with binding patterns shown by sialic acid with other proteins, especially in those cases where crystal data of the complex are available. No complexes with NeuGc are found in the Protein Data Bank, but four different types of protein have been reported with bound N-Acetyl sialic acid, either alone or as part of a larger ligand: wheat germ agglutinin (WGA) in complex with sialyllactose (Wright, 1990; PDB entry, 2WGC), influenza virus hemagglutinin (HA) in complex with sialyllactose (Sauter et al., 1992; PDB entry, 1HGG), influenza virus neuraminidase in complex with NeuAc (Varghese et al., 1992; PDB entry, 2BAT) and the B subunits of cholera toxin (CTB) in complex with the GM1 oligosaccharide (Merritt et al., 1994; PDB entry, 1CHB).

As for binding to P3 mAb, the carboxyl group is involved in binding in the four crystal complexes. In the complex with neuraminidase the carboxylate is firmly held by the positively charged guanidino groups of three arginines and is believed to play an important role in the cleavage reaction catalyzed by the enzyme (Varghese et al., 1992). The presence of the carboxylate has also been shown to be critical for binding of sialic acid to HA (Sauter et al., 1992). In this complex, both carboxyl oxygens participate in hydrogen bonds with a serine residue. In the binding of sialyllactose to WGA (Wright, 1990), only one of the carboxyl oxygens forms a hydrogen bond with the protein, whereas the other oxygen atom points to the solvent. In the CTB-GM1 complex (Merritt et al., 1994), the COO- group makes one direct interaction with the protein and another one through a water molecule. There is room enough in this part of the CTB binding cavity to allow several substitutions to be made at the C1 position of sialic acid (Lanne et al., 1994). Thus, involvement of the carboxyl group appears to be a general feature of sialic acid binding proteins.

Modifications of the N-function. Different derivatives having modifications of the nitrogen side chain of sialic acid in GM3, as shown in Figure 5, were tested for binding on TLC. The results are presented in Figure 6 (lanes 8-12). The incapacity of P3 to recognize NeuAc-GM3 has been shown previously, and was, furthermore, a main criterion in selecting the P3 secreting hybridoma among others (Vázquez et al., 1995). NMR and molecular modeling studies performed on both NeuAc-GM3 and NeuGc-GM3 (Siebert et al., 1992) have, moreover, shown that the conformational behavior of these two molecules is the same. Therefore, the strong distinction that P3 mAb makes between the two GM3 isomers can only be attributed to a direct contact of the antibody with the N-glycolyl function of sialic acid.

Trifluoroacetylation (Karlsson et al., 1991) of NeuAc-GM3, causing the acetamido moiety of the terminal NeuAc to be replaced by a trifluoroacetamido group, results in no or very weak binding to the antibody. Thus, the lack of binding indicates that the -CF3 group, despite its hydrophilic properties, cannot be accommodated within the same binding cavity as the distinctive -CH2OH group of the N-glycolyl form of sialic acid. Furthermore, P3 showed a weak binding to de-N-acetyl-GM3 (NeuNH2-GM3) on the TLC plate (Figure 6, lane 10), indicating not only that a very high specificity for the N-glycolyl group exists, but also that the interactions with the terminal atoms of this function are important for binding. It should be also taken into account that the amino group in NeuNH2-GM3 might be protonated, which would change the electrostatic character of this function.

Reported data show that this group also plays an important role in the binding of sialic acid to other proteins. In the interaction between sialic acid and neuraminidase, the CO and NH of the N-acetyl moiety participate in hydrogen bonds with the protein, whereas the CH3 is placed in a hydrophobic pocket. In the wheat germ agglutinin-sialyllactose complex, the N-acetyl group makes the largest number of contacts, being essential for the specificity of this protein for GlcNAc and NeuAc (Wright, 1990). Both the NH and the CO participate in hydrogen bonds, whereas the methyl group is in close contact with a tyrosine phenyl ring and the C[beta] carbon of a glutamate in the least exposed part of the binding cavity. In the complex of influenza virus hemagglutinin with sialyllactose, the acetamido nitrogen donates a hydrogen bond to a main chain carbonyl, while the methyl group is in hydrophobic contact with the indole ring of a tryptophan residue. Addition of a hydroxyl group to obtain N-glycolyl neuraminic acid abolishes the binding of the X-31 virus (Higa et al., 1985). However, extension of this side chain by an extra methyl group (N-propionyl analog) does not affect the binding (Sauter et al., 1992). In the CTB-GM1 complex, there is also a hydrogen bond donated by the NH and hydrophobic contacts of the acetyl group with a tyrosine phenyl ring (Merritt et al., 1994). In this case, however, the methyl group is also exposed to the solvent, which may explain why NeuGc-GM1 binds to CTB with approximately the same strength as the N-acetyl variant (Ångström et al., 1994).

These data indicate that the specificity for the N-acetyl isomer of sialic acid is very dependent on the hydrophobicity around the methyl group. Thus, the marked sensitivity shown by P3 towards changes of the N-glycolyl function strongly suggests that this group binds into a tight pocket that provides the necessary electrostatic interactions in a very specific manner.

Cleavage of the glycerol tail. The reactivity of P3 mAb was not affected by periodate oxidation of NeuGc-GM3, as shown in Figure 6 (lane 13), suggesting that the C8 and C9 alcohol groups do not participate in the binding. This finding agrees with the results obtained for the GD3 isomers (Figure 4), especially for NeuAc-NeuGc-GD3, in which the internal N-glycolyl sialic acid is recognized by the antibody despite the linkage of the external NeuAc to the C8 carbon of NeuGc.

The role of thisgroup in binding varies for different proteins. In both the influenza virus neuraminidase and hemagglutinin, the 8- and 9-OH of the glycerol side chain participate in hydrogen bonds with protein side chains. It has also been shown that modification of this tail reduces the binding capabilities of the virus (Sauter et al., 1992). In contrast to these two complexes, the glycerol tail in the CTB-GM1 complex is highly exposed to the solvent and the few interactions it has with the toxin subunit are water-mediated (Merritt et al., 1994). In the WGA-sialyllactose complex, this side chain makes hydrophobic contacts with a tyrosine ring, but it is also exposed to the solvent. Binding studies performed by Peters et al. (1979) demonstrated that trimming back the glycerol tail to the C7 carbon improves the binding of sialic acid to WGA.

Molecular modeling analysis of binding data for natural glycolipids

Molecular modeling analysis was carried out in order to correlate structural properties of the glycolipids with binding data, with the aim of extracting additional information about the binding epitope.

The equally high affinity of P3 mAb for NeuGc-GM3 and NeuGc-GM2 suggests that the GalNAc residue of GM2 does not participate in the binding. It should be noted that the conformational behavior of the sialic acid residue differs in these two glycolipids. The terminal, nonbranched NeuAc[alpha]3Gal[beta] disaccharide present in GM3 has been found to be flexible (Poppe et al., 1989; Sabesan et al., 1991; Siebert et al., 1992), existing mainly in two conformations (anticlinal and synclinal), whereas for the branched NeuAc[alpha]3Gal[beta] disaccharide in GM2 (Levery, 1991), GM1a (Acquotti et al., 1990), and GD1a (Sabesan et al., 1991) NMR data point to the presence of only the anticlinal orientation. For this conformation, the approach of the antibody from the side of the glycosidic and ring oxygens of sialic acid is not possible because of the presence of the GalNAc residue (see Figure 8a).


Figure 8. Models of three NeuGc gangliosides. (A) The anticlinal conformation of GM3 is represented in bold dark lines (with oxygen atoms marked in gray), whereas the synclinal orientation of sialic acid is shown in thin lines. The GM2 structure is obtained by including a GalNAc residue, which is drawn in bold gray lines. (B) GM1 oligosaccharide, represented in the same way as GM3 having the anticlinal conformation. A second possible orientation for the terminal Gal is shown in thin lines. In both panels, the glycolyl oxygen is shown as a small sphere.

On the other hand, the terminal galactose of NeuGc-GM1a, which in the preferred conformation (Figure 8b) is highly exposed at the top of the molecule, denies the antibody access to the binding epitope from this direction. Yet, a weak binding to this ganglioside was observed in the ELISA experiments, which must be due to the existence of a second possible (although much less populated) orientation for the terminal galactose (Imberty et al., 1995), where steric interference is much less pronounced, as shown in Figure 8b.

Therefore, the antibody approaches the gangliosides 'from the top" and recognizes an epitope on sialic acid that involves the carboxyl group, the N-glycolyl function (especially the terminal atoms, as can be concluded from the weak binding of the de-N-acylated derivative), and most likely also the 4-OH.

Since the internal sialic acid of NeuGc-GM1a produces only a very weak binding, whereas NeuGc-GM1b shows a moderate reactivity, the likewise moderate binding obtained also for NeuGc-GD1a must be accounted for by the external sialic acid. The weaker binding observed for these two molecules, as compared with NeuGc-GM3 and NeuGc-GM2, may result from a less favorable orientation of the external NeuGc. In both NeuGc-GM1b and NeuGc-GD1a this residue is presented mostly in an orientation (glycosidic angles were taken from Sabesan et al., 1991) where the NeuGc ring is perpendicular to the main body of the molecule, with the 4-OH side pointing downwards and thus being inaccessible to the antibody. This suggests a possible engagement of the 4-OH of NeuGc in the interactions with the protein, as occurs in the complexes between NeuAc or GlcNAc with the wheat germ agglutinin (Wright, 1990), where the 4-OH, together with the N-acetyl moiety, plays a key role in determining the binding specificity. However, this group is not important for the binding in the other known sialic acid-protein complexes. It should be also taken into account that in NeuGc-GM1b and NeuGc-GD1a the external NeuGc[alpha]3Gal[beta] disaccharide is 3-linked to the third residue (GalNAc), in contrast to NeuGc-GM3 and NeuGc-GM2, where this disaccharide is 4-linked to the corresponding residue (Glc). The different types of linkage in NeuGc-GM1b and NeuGc-GD1a produce different orientations of the external sialic acid relative to their respective internal residues, which might affect the binding of the antibody to these molecules.

The weak reactivity of P3 with NeuGc-GD1c can be understood on the basis of the considerations made above for NeuGc-GM1b and NeuGc-GD1a (unfavorable orientation of the internal sialic acid in NeuGc-GD1c), and the results of the conformational analysis carried out for NeuGc-NeuGc-GD3 being presented below, which account for the non-binding of the external NeuGc being.

The obtained definition of the epitope is illustrated in Figure 9 for the GM2 ganglioside.


Figure 9. Model of the terminal trisaccharide of GM2 showing the epitope recognized by the P3 mAb. The dots represent the solvent surface around epitope atoms, as determined from binding data and modeling analysis. The glycolyl oxygen is shown as a small sphere.

Conformational analysis of GD3

The selective recognition of the internal N-glycolyl sialic acid in GD3 by P3 mAb was quite an intriguing finding. In order to find a rationale behind this experimental fact and, at the same time, test our conclusions about the binding epitope, a conformational study of the oligosaccharide part of NeuGc-NeuGc-GD3 was carried out using molecular dynamics, as described in Materials and methods. Although the minimum energy conformers used as starting geometries were obtained for the NeuAc[alpha]8NeuAc disaccharide (Brisson et al., 1992), these results, as well as those from our simulations (see below), are not likely to be dependent on the type of N-function since in all the conformers the N-acetyl or N-glycolyl groups point away from the critical linkage region. Thus, it is reasonable to assume that the conformational behavior of the four GD3 isomers should be the same.

Fourteen 500 ps trajectories were collected (see Materials and methods). For each run, the values displayed by the torsional angles of interest ([Phi]int , [Psi]int , [Phi]ext , [Psi]ext , [omega]7, and [omega]8; see Figure 10), as well as potential energy values, were extracted from the trajectory files and plotted versus time. The obtained plots were used to identify the conformational regions that were sampled by the molecule. Thereafter, average torsional angles and average potential energies were calculated for each of these regions (Table II).


Figure 10. Structure of NeuGc[alpha]8NeuGc[alpha]3Gal[beta] (terminal part of NeuGc-NeuGc-GM3) showing the six torsional angles that determine the conformation of this trisaccharide. In the following definitions, atoms belonging to the external sialic acid are in bold. [Phi]ext = (C1-C2-O2-C8), [Psi]ext = (C2-O2-C8-C7), [omega]8 = (O2-C8-C7-O7), [omega]7 = (O7-C7-C6-O6), [Phi]int = (C1-C2-O2-C3Gal), [Psi]int = (C2-O2-C3Gal-H3Gal).

Comparison with results obtained by Brisson et al. (1992) reveals both similarities and differences. Most of the energy minima obtained by Brisson et al. were sampled in our dynamics runs for both the synclinal and anticlinal orientations of the internal sialic acid of GD3. The average values obtained by us for the torsional angles show only small differences relative to those obtained by Brisson et al., but the relative potential energies do not correlate well with the values obtained by these authors. This may result from choosing average energies to characterize the potential energy wells instead of using the energies of local minima, and it may be also a consequence of applying a different force field. In general, the energy functions used in calculations are imprecise (Halgren, 1995), and although the positions of the local minima are only slightly altered by modification of the energy function, the relative depths of the minima may vary.

In Table II, conformers are listed in order of increasing energy, making it evident that the synclinal orientation of the NeuGc[alpha]3Gal linkage was preferred over the anticlinal orientation. However, as pointed out above, these results should be analyzed with some caution. It should be also taken into account that the binding experiments were carried out on a solid phase, where other interactions may favor or disfavor the occurrence of a given geometry. Therefore, our analysis included all obtained conformers.

Table II. Conformational regions sampled by GD3 in molecular dynamics simulations
Conform. PE Total time [Phi]ext [Psi]ext w7 [omega]8 [Phi]int [Psi]int
ST1 0.0 410 -100 151 62 170 -82 10
ST2 1.1 750 -58 162 65 -177 -73 4
ST3 2.4 140 77 154 62 -178 -69 8
SG2+ 2.5 110 -69 112 63 87 -73 0
SG1+ 2.8 380 -152 94 70 99 -75 1
AT1 4.0 950 -89 150 60 172 -148 -31
SG1- 4.8 270 -158 107 73 -53 -91 -42
AG3- 4.9 650 60 137 40 -74 -160 -25
AG2+ 6.3 240 -70 115 53 69 -155 -27
AG2- 7.4 280 -74 150 49 -65 -159 -18
SG3+ 9.4 110 59 118 57 72 -78 3
AG3+ 10.7 80 64 112 55 76 -154 -26
Conformers are ordered according to the relative values of the average potential energies (PE). The standard deviations for the potential energy values oscillated between seven and eight kcal/mol for all regions (data not shown). The nomenclature used to name these conformations follows that of Brisson et al. (1992) for the NeuGc[alpha]8NeuGc linkage, with the addition of an 'A" or 'S" as a first letter to indicate the anticlinal or synclinal conformation of the NeuGc[alpha]3Gal linkage. See Figure 10 for definitions of the torsional angles.

Nonbinding of the external N-glycolyl sialic acid. In order to understand why the NeuGc-NeuAc-GD3 isomer does not bind, its external (N-glycolyl) sialic acid was superimposed onto the sialic acid of NeuGc-GM2 and NeuGc-GM1a, for each of the obtained conformers, as illustrated in Figure 11a for the ST1 geometry.


Figure 11. Models of two GD3 isomers (gray bold lines) in different conformations superimposed on GM2 (dark thin lines). Glycolyl oxygens of both molecules are represented as small spheres. (A) The external NeuGc of NeuGc-NeuAc-GD3 is superimposed on the sialic acid of GM2. Carboxyl oxygens of the two sialic acid residues of GD3 are represented as small dark spheres. In this conformation (ST1), the COO- group of the internal NeuAc stands over the COO- of the external NeuGc. (B) The internal NeuGc of NeuAc-NeuGc-GD3 (in the SG1- conformation) is superimposed on the sialic acid of GM2, resulting in the NeuAc residue lying below the glycerol tail of the GM2 sialic acid, which allows antibody access to the internal NeuGc.

In conformers ST1 and AT1, the carboxyl group of the external sialic acid is masked by the COO- of the internal one, as shown in Figure 11a. A similar situation occurs for conformers SG2+ and AG2+, where the N-acetyl group of the internal unit is placed on top of the COO- of NeuGc. Conformations ST2 and SG1- are evidently also nonbinding since in both geometries the carboxyl group of the external sialic acid points towards the galactose unit, thus being inaccessible to the antibody. The same holds true for the AG1- conformer, which was unstable during the dynamics run.

Conformer AG2- presents the carboxyl group of the external sialic acid pointing down towards the ceramide. A similar situation occurs for conformers SG1+ and AG1+ (not stable in the dynamics run), in which the 4-OH side of NeuGc points downwards.

An additional conformational region, ST3, was obtained in the dynamics run that started from the SG3+ geometry. No analogous conformation was reported by Brisson et al. (1992), although a very small energy well at the 10 kcal/mol level containing an approximate 'T3" conformer can be seen in a contour map presented by these authors. The ST3 conformation is nonbinding, since the carboxyl group of NeuGc points down, towards the ceramide moiety, with the 4-OH facing the galactose ring. The AT2 conformation was found to be unstable during the simulation, and only very short excursions from AT1 to this region were observed.

In conformers SG3+ and AG3+ the methyl group of the N-acetyl function approaches the 4-OH of the external NeuGc (distance between the methyl carbon of NeuAc and the O4 of NeuGc = 4.6 Å), which might cause some steric impediments for binding. Furthermore, the potential energies obtained for these two conformations are high, which correlates well with the NMR data of Brisson et al. (1992), according to which the population of the G3+ well must be low in order to reproduce the observed NOEs.

Binding to the internal N-glycolyl sialic acid. The possible binding conformations of the two GD3 isomers having NeuGc as the internal sialic acid were also analyzed by superimposing this sugar onto the sialic acid of GM2 and GM1.

Two of the conformers, ST3 and SG1+, cannot bind because part of the external residue covers the ring on the carboxyl side of the internal NeuGc. In a few other conformations, the external sialic acid is situated at the top of the molecule, but in a positions different from the one occupied by the terminal galactose in GM1a, making it difficult to determine whether they can bind the antibody or not. However, in at least six of the obtained geometries (SG2+, AG2+, SG1-, AG1-, AG2-, and AG3-) the epitope is presented in a favorable orientation and without steric hindrance from the external sialic acid, as illustrated in Figure 11b for the SG1- conformer.

Concluding Remarks

The combination of binding experiments for a wide panel of different structures with molecular modeling allowed us to delineate the epitope that is recognized by the P3 antibody. A coherent picture, fitting the wide range of binding data available, was obtained. For the sialic acid group, the carboxyl and the N-glycolyl functions were shown to be essential for binding, most likely by interacting directly with the protein, as in the reported crystal complexes containing bound sialic acid. Also, the binding epitope seen by the P3 antibody is consistent with an approach towards the antigen 'from the top." This is a prerequisite since in the immunization procedure followed in the antibody generation the NeuGc-GM3 molecules were inserted into liposomes, limiting the number of allowed conformations due to restrictions imposed by the membrane plane.

It is difficult to draw any conclusions about a possible participation of the Gal residue in the binding to P3 mAb, and whether the anticlinal orientation of sialic acid is needed for binding or not. Only two of the available crystal complexes, the WGA- and HA-sialyllactose complexes, provide such comparative information. In both complexes, the linkage between sialic acid and galactose exhibits a synclinal geometry. In the case of WGA, modeling analysis shows that this is the only possible conformation, since for the anticlinal geometry both the galactose and glucose residue would bump into the protein. For the influenza virus hemagglutinin, however, modeling shows that the anticlinal conformation also is possible and, moreover, allows a relative orientation between the sugar moiety and the protein that seems more appropriate for a multivalent binding of the hemagglutinin to a cell surface. In this conformation, there can be only little contact between HA and the galactose, as in the WGA-sialyllactose crystal complex.

The sequence of the variable region of P3 mAb has been recently obtained and several models of this region have been constructed. Docking studies in conjunction with site directed mutagenesis experiments are currently in progress, aimed at constructing a model of the P3-NeuGc complex. The delineation of the epitope recognized by P3 on sialic acid is of great importance for this purpose since the number of variables in the docking procedure may be considerably reduced. Furthermore, the extensive and varied binding data that has been obtained may provide a valuable test of the consistency of the model.

Materials And Methods

Monoclonal antibody

P3 mAb (an IgM antibody) was produced by immunization of Balb/c mice with liposomes containing NeuGc-GM3 and tetanus toxoid, as previously described (Vázquez et al., 1995).

Preparation of standard glycolipids

The structures of the glycolipids used in this study are shown in Table I. Gangliosides are named according to the nomenclature of Svennerholm (1964); otherwise, the nomenclature used follows IUPAC-IUB recommendations (IUPAC-IUB, 1977; Sharon, 1987).

Neutral glycolipids. LacCer and Gg4 were obtained from GM3 and NeuAc-GM1, respectively, by treatment with sialidase (Tai et al., 1983).

N-Acetylated glycolipids. GM1a, GD1a, GD1b, and GT1b were prepared from bovine brain (Hakomori, 1983). GM2 and GM4 were obtained from human brain (Svennerholm, 1976). GD2 was obtained by transformation of GD1b with bovine testis [beta]-galactosidase (Boehringer-Mannheim) (Cahan et al., 1982). GM3 was isolated from sheep spleen (Svennerholm, 1976). GM1b was kindly donated by Y. Sanai (Tokyo Metropolitan Institute of Medical Science). NeuAc-SPG was obtained from human erythrocytes (Hakomori, 1983).

N-Glycolylated glycolipids and GD3 isomers. NeuGc-GM3 was isolated from sheep spleen (Svennerholm, 1976) while NeuGc-GM2 was prepared from mouse (Balb/c) liver (Hashimoto et al., 1983). NeuGc-GM1a and NeuGc-GD1a were from mouse (ICR) liver. NeuGc-GM1b and NeuGc-GD1c, from mouse spleen and thymoma, respectively, were donated by K. Nakamura (Tokyo Metropolitan Institute of Medical Science). NeuGc-SPG and VI3NeuGc[alpha]-nLc6 were obtained from bovine erythrocytes (Chien et al., 1978). The four GD3 isomers, NeuAc-NeuAc-GD3, NeuGc-NeuGc-GD3, NeuAc-NeuGc-GD3 and NeuGc-NeuAc-GD3, isolated from bear erythrocytes, were kindly supplied by Y. Hashimoto and S. Suziki (Tokyo Metropolitan Institute of Medical Science).

Ganglioside derivatives

Derivatization of NeuGc-GM3 to amides and the primary alcohol was achieved as described previously (Nakamura and Handa, 1986; Lanne et al., 1995). Methylamine, ethylamine, propylamine, benzylamine, and methyl iodide were purchased from Aldrich, Steinheim, Germany. Trifluoro derivatives of the sialic acid and the ceramide of NeuGc-GM3 were produced as described before (Karlsson et al., 1991) yielding one derivative having the ceramide modified by a trifluoroacetimide (TFI) moiety and a second derivative in which the fatty acid is replaced by a trifluoroacetamido (TFA) moiety. In both cases the acetamido function of the sialic acid is exchanged for a trifluoroacetamido group. Periodate oxidation of the glycerol tail of NeuGc-GM3 was performed by treating the ganglioside with sodium metaperiodate, followed by borohydride reduction (Veh et al., 1976). The derivatives were purified by ion-exchange chromatography (DEAE-Sepharose CL-6B, Pharmacia, Sweden) or HPLC (silica column), and the products were identified by negative ion fast atom bombardment mass spectrometry (JEOL SX102A, JEOL, Japan) with the following settings: accelerating voltage, 10 kV; resolution, 1000; m/[Delta]m, 10% valley. Xe atom bombardment at 6 keV was employed using triethanolamine as matrix.

Thin-layer chromatography (TLC)

Pre-coated HPTLC aluminum sheets and glass plates (Merck, Darmstadt, Germany) were used for the separation of glycolipids. The solvent systems used for developing chromatography were chloroform/methanol/0.22% CaCl2 in water (55:45:10, v/v/v) or chloroform/methanol/water (68:35:8, v/v/v). Glycolipids were visualized with orcinol and resorcinol reagents.

Immunostaining on HPTLC plates

Enzyme immunostaining on HPTLC plates was performed as previously reported (Kawashima et al., 1993). Briefly, after chromatography of the glycolipids, the plates were soaked with 0.1% polyisobutylmethacrylate in cyclohexane for 75 s. After drying, the plates were incubated with phosphate-buffered saline (PBS) containing 1% bovine serum albumin for 30 min. Then, the plates were incubated with P3 mAb for 1 h at room temperature. After washing with PBS, the plates were incubated with horseradish peroxidase-conjugated goat anti-mouse IgM (Cappel, Malvern, PA) for 1 h at room temperature. The plates were washed again and incubated with the substrate solution consisting of 40 µg/ml O-phenylendiamine (Sigma, St. Louis, MO) in 80 mM citrate-phosphate buffer, pH 5.0, containing 0.12% H2O2. The reaction was stopped by dipping the plates in PBS.

To visualize the binding of the P3 mAb to the different NeuGc-GM3 derivatives, the HPTLC plates were incubated with 125I-labeled anti-mouse immunoglobulins (Dakopatts, Denmark) as second antibody for 2 h at room temperature, followed by washing with PBS. Radioactivity on HPTLC plates was detected by autoradiography (Kodak XAR-5 film, Eastman Kodak, Rochester, NY) and with an excitable phosphorescent screen, which was analyzed using a Bio-Rad GS-250 Molecular Imager (Bio-Rad Laboratories, Inc., USA). For data analysis, the Phosphor Analyst program (Bio-Rad) was used.

Enzyme-linked immunosorbent assay (ELISA)

Solid-phase ELISA was performed as previously described (Tai et al., 1984). In brief, glycolipids were serially diluted in ethanol and applied in a 96-well polystyrene microtiter plate (Dynatech, Alexandria, VA) in volumes of 50 µl/well. P3 mAb was diluted with PBS containing 1% human serum albumin. The second antibody, a peroxidase-conjugated goat anti-mouse IgM, was purchased from Cappel (Malvern, PA). Three samples of each experiment were tested. The standard deviations were less than 10% for all values. Background values of absorbance at 405 nm were less than 0.1.

Molecular modeling

Models of gangliosides carrying NeuGc (GM1a, GM1b, GM2, GM3, GD1a, and GD3) were constructed within Biograf (Molecular Simulations Inc., 1994). Starting values for the dihedral angles over the glycosidic bonds were taken from the literature (Verulaja and Rao, 1983; Acquotti et al., 1990; Sabesan et al., 1991; Siebert et al., 1992). Molecular dynamics (MD) simulations were carried out using the Dreiding-II force field (Mayo et al., 1990). MD runs for NeuGc-GM3 derivatives were started from the anticlinal conformation of the glycosidic linkage between sialic acid and galactose ([Phi] [ap] -155°, [Psi] [ap] -30°) and the torsional angle around the C1-C2 bond of the sialic acid was searched every 15° in order to choose the minimum energy conformation for the carboxyl group. Complete energy minimization was carried out before starting the dynamics runs. The ceramide part of GM3 was fixed in these simulations, which were performed at 300 K during 1000 ps, using a time step of 0.001 ps. Coordinates were saved every 500 steps. Gasteiger charges and a distance-dependent dielectric constant ([epsis] = 4r) were used. No hydrogen bond term was included in energy calculations.

MD studies of NeuGc-NeuGc-GD3 were performed using the same parameters and conditions given above, except that the total run time was 500 ps. Sixteen possible conformations were considered as starting geometries. They were given by the combination of eight sets of values for the main torsional angles ([Phi]ext , [Psi]ext , [omega]7, and [omega]8; see Figure 9) involved in the NeuAc[alpha]8NeuAc linkage, as found by Brisson et al. (1992), and two sets of values for the two torsional angles defining the linkage between the internal sialic acid and galactose ([Phi]int and [Psi]int). Briefly, in the work by Brisson et al. (1992) each of the three staggered rotamers of [omega]8 (i.e., g+, g- and anti) yielded several (3, 3, and 2, respectively) minimum energy combinations for the [Phi]ext and [Psi]ext angles, which were denoted as G1+, G2+, G3+, G1-, G2-, G3-, T1, and T2, respectively. Each of them was combined in this study with the two energy minima (anticlinal and synclinal) found for the NeuAc[alpha]3Gal linkage (Siebert et al., 1992), yielding conformers AG1+, AG2+, ... ,SG1+, SG2+, ... , ST2. Two of these starting geometries (SG2- and SG3-) were not realistic due to severe steric hindrance, and were therefore discarded. The remaining 14 conformers were energy-refined prior to the dynamics simulations.

Analysis of known complexes of proteins with sialic acid-containing molecules was carried out using the program Quanta (Molecular Simulations Inc., 1996).

References

Acquotti ,D., Poppe,L., Dabrowski,J., von der Lieth,C.-W., Sonnino,S. and Tettamanti,G. (1990) Three-dimensional structure of the oligosaccharide chain of GM1 ganglioside revealed by distance-mapping procedure: a rotating and laboratory frame nuclear overhauser enhancement investigation of native glycolipid in dimethyl sulfoxide and in water-dodecylphosphocholine solutions. J. Am. Chem. Soc., 112, 7772-7778.

Ångström ,J., Teneberg,S. and Karlsson,K.-A. (1994) Delineation and comparison of ganglioside binding epitopes for the toxins of Vibrio cholerae, Escherichia coli and Clostridium tetani: Evidence for overlapping epitopes. Proc. Natl. Acad. Sci. USA, 91, 11859-11863.

Brisson ,J.-R., Baumann,H., Imberty,A., Pérez,S. and Jennings,H.J. (1992) Helical epitope of the group B meningococcal [alpha](2-8)-linked sialic acid polysaccharide. Biochemistry, 31, 4996-5004. MEDLINE Abstract

Cahan ,L.D., Irie,R.F., Sinyh,R., Cassidenti,A. and Paulson,J.C. (1982) Identification of human neuroectodermal tumor antigen (OFA-I-2) as ganglioside GD2. Proc. Natl. Acad. Sci. USA, 79, 7929-7933.

Chien ,J.-L., Li,S.-C., Laine,R.A. and Li, Y.-T. (1978) Characterization of gangliosides from bovine erythrocyte membranes. J. Biol. Chem., 253, 4031-4035. MEDLINE Abstract

Corfield ,A.P. and Schauer,R. (1982) Occurrence of sialic acids. Cell Biol. Monogr., 10, 5-50.

Fishman ,P.H., Pacuszka,T., Hom,B. and Moss,J. (1980) Modification of ganglioside GM1. Effect of lipid moiety on choleragen action. J. Biol. Chem., 255, 7657-7664. MEDLINE Abstract

Furukawa ,K., Yamaguchi,H., Oettgen,H.F., Old,L.J. and Lloyd,K.O. (1988) Analysis of the expression of N-glycolyl neuraminic acid-containing gangliosides in cells and tissues using two human monoclonal antibodies. J. Biol. Chem., 263, 18507-18512. MEDLINE Abstract

Hakomori ,S.-i. (1983) In Kanfer,J.N. and Hakomori,S.-i. (eds.), Handbook of Lipid Research. Plenum Press, New York, pp. 1-437.

Halgren ,T.A. (1995) Potential energy functions. Curr. Opin. Struct. Biol., 5, 205-210. MEDLINE Abstract

Hashimoto ,Y., Otsuka,H., Sudo,K., Suzuki,A. and Yamakawa,T. (1983) Genetic regulation of GM2 expression in liver of mouse. J. Biochem. (Tokyo), 93, 895-901. MEDLINE Abstract

Higa ,H.H., Rogers,G.N. and Paulson,J.C. (1985) Influenza virus hemagglutinins differentiate between receptor determinants bearing N-acetyl-, N-glycolyl-, and N,O-diacetylneuraminic acids. Virology, 144, 279-282. MEDLINE Abstract

Imberty ,A., Mikros,E., Koca,J., Mollicone,R., Oriol,R. and Pérez,S. (1995) Computer simulation of histo-blood group oligosaccharides: energy maps of all constituting disaccharides and potential energy surfaces of 14 ABH and Lewis carbohydrate antigens. Glycoconjugate J., 12, 331-349.

IUPAC-IUB Commission on Biochemical Nomenclature. (1977) Eur. J. Biochem., 79, 11-21.

Jeffrey ,P.D., Bajorath,J., Chang,C.Y., Yelton,D., Hellstrom,I., Hellstrom,K.E. and Sheriff,S. (1995) The x-ray structure of an anti-tumour antibody in complex with antigen. Nature Struct. Biol., 2, 466-471.

Karlsson ,K.-A., Lanne,B., Pimlott,W. and Teneberg,S. (1991) The resolution into molecular species on desorption of glycolipids from thin-layer chromatograms, using combined thin-layer chromatography and fast-atom-bombardment mass spectrometry. Carbohydr. Res., 221, 49-61. MEDLINE Abstract

Kawashima ,I., Ozawa,H., Kotani M., Susuki,M., Kawano,T., Gomobuchi,M. and Tai,T. (1993) Characterization of ganglioside expression in human melanoma cells: immunological and biochemical analysis. J. Biochem. (Tokyo), 114, 186-193. MEDLINE Abstract

Lanne ,B., Schierbeck,B. and Karlsson,K.-A. (1994) On the role of the carboxyl group of sialic acid in binding of cholera toxin to the receptor glycosphingolipid, GM1. J. Biochem. (Tokyo), 116, 1269-1274. MEDLINE Abstract

Lanne ,B., Uggla,L., Stenhagen,G. and Karlsson,K.-A. (1995) Enhanced binding of enterotoxigenic Escherichia coli K99 to amide derivatives of the receptor ganglioside NeuGc-GM3. Biochemistry, 34, 1845-1850. MEDLINE Abstract

Levery ,S.B. (1991) 1H-NMR study of GM2 ganglioside: evidence that an interresidue amide-carboxyl hydrogen bond contributes to stabilization of a preferred conformation. Glycoconjugate J., 8, 484-492.

Miyake ,M., Ito,M., Hitomi,S., Ikeda,S., Taki,T., Kurata,M., Hino,A., Miyake,N. and Kannagi,R. (1988) Generation of two murine monoclonal antibodies that discriminate N-glycolyl and N-acetyl neuraminic acid residues of GM2 ganglioside. Cancer Res., 48, 6154-6160. MEDLINE Abstract

Mayo ,S.M., Olafson,B.D. and Goddard III,W.A. (1990) Dreiding: a generic force field for molecular simulations. J. Phys. Chem., 94, 8897-8909.

Merritt ,E.A., Sarfaty,S., van den Akker,F., L'hoir,C., Martial,J.A. and Hol,W.G.J. (1994) Crystal structure of cholera toxin B-pentamer bound to receptor GM1 pentasaccharide. Prot. Sci., 3, 166-175.

Nakamura ,K. and Handa,S. (1986) Biochemical properties of N-methylamides of sialic acid in gangliosides. J. Biochem. (Tokyo), 99, 219-226. MEDLINE Abstract

Ozawa ,H., Kawashima,I. and Tai,T. (1992) Generation of murine monoclonal antibodies specific for N-glycolylneuraminic acid-containing gangliosides. Arch. Biochem. Biophys., 294, 427-433. MEDLINE Abstract

Peters ,B.P., Ebisu,S., Goldstein,I.J. and Flashner,M. (1979) Interaction of wheat germ agglutinin with sialic acid. Biochemistry, 18, 5505-5511. MEDLINE Abstract

Poppe ,L., Dabrowski,J., von der Lieth,C.-W., Numata,M. and Ogawa,T. (1989) Solution conformation of sialosylcerebroside (GM4) and its NeuAc([alpha]2->3)Gal[beta] sugar component. Eur. J. Biochem., 180, 337-342. MEDLINE Abstract

Sabesan ,S., Duuss,J.O., Fukunaga,T., Bock,K. and Schauer,R. (1991) NMR and conformational analysis of ganglioside GD1a. J. Am. Chem. Soc., 113, 3236-3246.

Sanai ,Y., Yamasaki,M. and Nagai,Y. (1988) Monoclonal antibody directed to a Hanganutziu-Deicher active ganglioside, GM2(NeuGc). Biochim. Biophys. Acta, 294, 427-433.

Sauter ,N.K., Hanson,J.E., Glick,G.D., Brown,J.H., Crowther,R.L., Park,S.-J., Skehel,J.J. and Wiley,D.C. (1992) Binding of influenza virus hemagglutinin to analogs of its cell-surface receptor, sialic acid: analysis by proton nuclear magnetic resonance and x-ray crystallography. Biochemistry, 31, 9609-9621. MEDLINE Abstract

Schengrund ,C.-L. and Ringler,N.J. (1989) Binding of Vibrio cholerae toxin and the heat-labile enterotoxin of Escherichia coli to GM1, derivatives of GM1, and nonlipid oligosaccharide polyvaent ligands. J. Biol. Chem., 264, 13233-13237. MEDLINE Abstract

Sharon ,N. (1987) Nomenclature of glycoproteins, glycopeptides and peptidoglycans. J. Biol. Chem., 262, 13-18. MEDLINE Abstract

Siebert ,H.-C., Reuter,G., Schauer,R., von der Lieth,C.-W. and Dabrowski,J. (1992) Solution conformations of GM3 gangliosides containing different sialic acid residues as revealed by NOE-based distance mapping, molecular mechanics, and molecular dynamics simulations. Biochemistry, 31, 6962-6971. MEDLINE Abstract

Spiegel ,S. (1985) Fluorescent derivatives of ganglioside GM1 function as receptors for cholera toxin. Biochemistry, 24, 5947-5952. MEDLINE Abstract

Svennerholm ,L. (1964) The gangliosides. J. Lipid Res., 5, 145-155.

Svennerholm ,L. (1976) Ganglioside isolation. Methods Carbohydr. Chem., 6, 464-474.

Tai ,T., Paulson,J.C., Cahan,L.D. and Irie,R.F. (1983) Ganglioside GM2 as a human tumor antigen (OFA-I-1). Proc. Natl. Acad. Sci. USA, 80, 5392-5396. MEDLINE Abstract

Tai ,T., Cahan,L.D., Paulson,J.C., Saxton,R.E. and Irie,R.F. (1984) Human monoclonal antibody against ganglioside GD2: use in development of enzyme-linked immunosorbent assay for the monitoring of anti-GD2 in cancer patients. J. Natl. Cancer Inst., 73, 627-633. MEDLINE Abstract

Varghese ,J.N., McKimm-Breschkin,J.L., Caldwell,J.B., Kortt,A.A. and Colman, P.M. (1992) The structure of the complex between influenza virus neuraminidase and sialic acid, the viral receptor. Proteins, 14, 327-332. MEDLINE Abstract

Vázquez ,A.M., Alfonso,M., Lanne,B., Karlsson,K.-A., Carr,A., Barroso,O., Fernández,L.E., Rengifo,E., Lanio,M.E., Alvarez,C., Zeuthen,J. and Pérez,R. (1995) Generation of a murine monoclonal antibody specific against N-glycolylneuraminic acid-containing gangliosides that also recognizes sulfated glycolypids. Hybridoma, 14, 551-556.

Veh ,R.W., Corfield,A.P., Sander,M. and Schauer,R. (1976) Neuraminic acid-specific modification and tritium labelling of gangliosides. Biochim. Biophys. Acta, 486, 145-160. MEDLINE Abstract

Veluraja ,K., and Rao,V.S.R. (1983) Theoretical studies on the conformation of monosialogangliosides and disialogangliosides. Carbohydr. Polym., 3, 175-192.

Watarai ,S., Onuma,M. and Yasuda,T. (1991) Production of monoclonal antibodies specific for ganglioside GD3. J. Biochem. (Tokyo), 110, 889-895. MEDLINE Abstract

Wright ,C.S. (1990) 2.2 Å resolution structure analysis of two refined N-acetylneuraminyl-lactose-wheat germ agglutinin isolectin complexes. J. Mol. Biol., 215, 635-651. MEDLINE Abstract

4To whom correspondence should be addressed

This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 10 Jun 1998
Copyright©Oxford University Press, 1998.
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 J. Immunol.Home page
A. M. Hernandez, D. Toledo, D. Martinez, T. Grinan, V. Brito, A. Macias, S. Alfonso, T. Rondon, E. Suarez, A. M. Vazquez, et al.
Characterization of the Antibody Response against NeuGcGM3 Ganglioside Elicited in Non-Small Cell Lung Cancer Patients Immunized with an Anti-Idiotype Antibody
J. Immunol., November 1, 2008; 181(9): 6625 - 6634.
[Abstract] [Full Text] [PDF]

Home page
 Molecular Cancer TherapeuticsHome page
L. Roque-Navarro, K. Chakrabandhu, J. de Leon, S. Rodriguez, C. Toledo, A. Carr, C. M. de Acosta, A.-O. Hueber, and R. Perez
Anti-ganglioside antibody-induced tumor cell death by loss of membrane integrity
Mol. Cancer Ther., July 1, 2008; 7(7): 2033 - 2041.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
B. Kotlan, P. Simsa, J.-L. Teillaud, W. H. Fridman, J. Toth, M. McKnight, and M. C. Glassy
Novel Ganglioside Antigen Identified by B Cells in Human Medullary Breast Carcinomas: The Proof of Principle Concerning the Tumor-Infiltrating B Lymphocytes
J. Immunol., August 15, 2005; 175(4): 2278 - 2285.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
U. Krengel, L.-L. Olsson, C. Martinez, A. Talavera, G. Rojas, E. Mier, J. Angstrom, and E. Moreno
Structure and Molecular Interactions of a Unique Antitumor Antibody Specific for N-Glycolyl GM3
J. Biol. Chem., February 13, 2004; 279(7): 5597 - 5603.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
M. Alfonso, A. Diaz, A. M. Hernandez, A. Perez, E. Rodriguez, R. Bitton, R. Perez, and A. M. Vazquez
An Anti-Idiotype Vaccine Elicits a Specific Response to N-Glycolyl Sialic Acid Residues of Glycoconjugates in Melanoma Patients
J. Immunol., March 1, 2002; 168(5): 2523 - 2529.
[Abstract] [Full Text] [PDF]

Home page
 J. Virol.Home page
C. Delorme, H. Brüssow, J. Sidoti, N. Roche, K.-A. Karlsson, J.-R. Neeser, and S. Teneberg
Glycosphingolipid Binding Specificities of Rotavirus: Identification of a Sialic Acid-Binding Epitope
J. Virol., March 1, 2001; 75(5): 2276 - 2287.
[Abstract] [Full Text]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (21)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Moreno, E.
Right arrow Articles by Perez, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Moreno, E.
Right arrow Articles by Perez, R.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?