NMR experiments reveal distinct antibody-bound conformations of a synthetic disaccharide representing a general structural element of bacterial lipopolysaccharide epitopes

The recognition reactions between a synthetic disaccharide alpha-Kdo-(2-->4)-alpha-Kdo-(2-->O)-allyl and two monoclonal antibodies (mAbs) were studied by NMR, yielding two distinct bound conformations of the carbohydrate ligand. One mAb, S23-24, recognizes the disaccharides alpha-Kdo-(2-->4)-alpha-Kdo-(2-->O)-allyl and alpha-Kdo-(2-->8)-alpha-Kdo-(2-->O)-allyl with similar affinities, whereas mAb S25-2 binds to the disaccharide alpha-Kdo-(2-->8)-alpha-Kdo-(2-->O)-allyl with an approximately 10-fold higher affinity than to the disaccharide alpha-Kdo-(2-->4)-alpha-Kdo-(2-->O)-allyl. Compared to S25-2, S23-24 binds to alpha-Kdo-(2-->4)-alpha-Kdo-(2-->O)-allyl with an approximately 50-fold increased affinity. We used NMR experiments that are based on the transferred NOE effect, specifically, trNOESY, trROESY, QUIET-trNOESY, and MINSY experiments, to show that the (2-->8)-specific mAb, S25-2, stabilizes a conformation of the alpha-(2-->4)-linked disaccharide that is not highly populated in solution. S23-24 recognizes two conformations of alpha-Kdo-(2-->4)-alpha-Kdo-(2-->O)-allyl, one that is highly populated in aqueous solution and another conformation that is similar to the one bound by S25-2. This is the first example where it is experimentally shown that a carbohydrate ligand may adopt different bioactive conformations upon interaction with mAbs with different fine specificities. Our NMR studies indicate that a careful examination of spin diffusion is critical for the analysis of bioactive conformations of carbohydrate ligands.

Conformational analysis of a Chlamydia-specific disaccharide alpha-Kdo-(2-->8)-alpha-Kdo-(2-->O)-allyl in aqueous solution and bound to a monoclonal antibody: observation of intermolecular transfer NOEs

The disaccharide alpha-Kdo-(2-->8)-alpha-Kdo (Kdo: 3-deoxy-D-manno-oct-2-ulosonic acid) represents a genus-specific epitope of the lipopolysaccharide of the obligate intracellular human pathogen Chlamydia. The conformation of the synthetically derived disaccharide alpha-Kdo-(2-->8)-alpha-Kdo-(2-->O)-allyl was studied in aqueous solution, and complexed to a monoclonal antibody S25-2. Various NMR experiments based on the detection of NOEs (or transfer NOEs) and ROEs (or transfer ROEs) were performed. A major problem was the extensive overlap of almost all 1H NMR signals of alpha-Kdo-(2-->8)-alpha-Kdo-(2-->O)-allyl. To overcome this difficulty, HMQC-NOESY and HMQC-trNOESY experiments were employed. Spin diffusion effects were identified using trROESY experiments, QUIET-trNOESY experiments and MINSY experiments. It was found that protein protons contribute to the observed spin diffusion effects. At 800 MHz, intermolecular trNOEs were observed between ligand protons and aromatic protons in the antibody binding site. From NMR experiments and Metropolis Monte Carlo simulations, it was concluded that alpha-Kdo-(2-->8)-alpha-Kdo-(2-->O)-allyl in aqueous solution exists as a complex conformational mixture. Upon binding to the monoclonal antibody S25-2, only a limited range of conformations is available to alpha-Kdo-(2-->8)-alpha-Kdo-(2-->O)-allyl. These possible bound conformations were derived from a distance geometry analysis using transfer NOEs as experimental constraints. It is clear that a conformation is selected which lies within a part of the conformational space that is highly populated in solution. This conformational space also includes the conformation found in the crystal structure. Our results provide a basis for modeling studies of the antibody-disaccharide complex.

Conformational analysis of biantennary glycans and molecular modeling of their complexes with lentil lectin

Some mannose-binding legume lectins show higher affinity for fucosylated glycans than for glycans without fucose. These lectins possess a secondary binding site. Owing to the possibility of additional fucose binding, oligosaccharides adopt different conformations depending on whether they contain fucose or not. To study these conformational differences, complexes of fucosylated and unfucosylated glycans with Lens culinaris lectin have been modeled. Starting points were X-ray structures of lentil lectin and complexes of the homologous Lathyrus ochrus lectin. The SYBYL molecular modeling package with the TRIPOS force field was used. Two different models were built, displaying in both a network of hydrogen bonds between the saccharide and the binding site. Furthermore, to compare the free and bound ligand, conformational analysis in the free state has been performed. A complete analysis of all possible disaccharide fragments has been performed using the MM3 force field. A CICADA analysis employing the same force field was carried out to study the complete oligosaccharide. Low-energy conformers found by CICADA were clustered in conformational families and analyzed in terms of flexibility and rotational barriers. All values of glycosidic torsion angles are in the range as calculated by MM3 for the disaccharides.

Human complement factor B: functional properties of a recombinant zymogen of the alternative activation pathway convertase

The human complement factor B is a centrally important component of the alternative pathway activation of the complement system. Here we report the isolation, characterization and eukaryotic expression of the first full length cDNA transcript for human factor B. In a factor B dependent haemolysis assay, the recombinant human factor B generated by transient COS cell transfection was shown to reconstitute haemolytic activity of factor B depleted human serum. To study the biological activities assigned to factor B, the availability of recombinant polypeptides representing definite portions of the human factor B molecule is desirable.