Structure of a selectin-like mutant of mannose-binding protein complexed with sialylated and sulfated Lewis(x) oligosaccharides

Molecular Characteristics, Receptor Specificity, and Pathogenicity of Avian Influenza Viruses Isolated from Wild Ducks in Russia

Avian influenza viruses (AIV) of wild ducks are known to be able to sporadically infect domestic birds and spread along poultry. Regular surveillance of AIV in the wild is needed to prepare for potential outbreaks. During long-year monitoring, 46 strains of AIV were isolated from gulls and mallards in Moscow ponds and completely sequenced. Amino acid positions that affect the pathogenicity of influenza viruses in different hosts were tested. The binding affinity of the virus for receptors analogs typical for different hosts and the pathogenicity of viruses for mice and chickens were investigated. Moscow isolates did not contain well-known markers of pathogenicity and/or adaptation to mammals, so as a polybasic cleavage site in HA, substitutions of 226Q and 228G amino acids in the receptor-binding region of HA, and substitutions of 627E and 701D amino acids in the PB2. The PDZ-domain ligand in the NS protein of all studied viruses contains the ESEV or ESEI sequence. Although several viruses had the N66S substitution in the PB1-F2 protein, all Moscow isolates were apathogenic for both mice and chickens. This demonstrates that the phenotypic manifestation of pathogenicity factors is not absolute but depends on the genome context.

Solution structure, glycan specificity and of phenol oxidase inhibitory activity of Anopheles C-type lectins CTL4 and CTLMA2

Malaria, the world's most devastating parasitic disease, is transmitted between humans by mosquitoes of the Anopheles genus. An. gambiae is the principal malaria vector in Sub-Saharan Africa. The C-type lectins CTL4 and CTLMA2 cooperatively influence Plasmodium infection in the malaria vector Anopheles. Here we report the purification and biochemical characterization of CTL4 and CTLMA2 from An. gambiae and An. albimanus. CTL4 and CTLMA2 are known to form a disulfide-bridged heterodimer via an N-terminal tri-cysteine CXCXC motif. We demonstrate in vitro that CTL4 and CTLMA2 intermolecular disulfide formation is promiscuous within this motif. Furthermore, CTL4 and CTLMA2 form higher oligomeric states at physiological pH. Both lectins bind specific sugars, including glycosaminoglycan motifs with β1-3/β1-4 linkages between glucose, galactose and their respective hexosamines. Small-angle x-ray scattering data supports a compact heterodimer between the CTL domains. Recombinant CTL4/CTLMA2 is found to function in vivo, reversing the enhancement of phenol oxidase activity in dsCTL4-treated mosquitoes. We propose these molecular features underline a common function for CTL4/CTLMA2 in mosquitoes, with species and strain-specific variation in degrees of activity in response to Plasmodium infection.

Structural Aspects of Carbohydrate Recognition Mechanisms of C-Type Lectins

Carbohydrate recognition is an essential function occurring in all living organisms. Lectins are carbohydrate-binding proteins and are classified into several families. In mammals, Ca²⁺-dependent C-type lectins, such as β-galactoside-binding galectin and sialic acid-binding siglec, play crucial roles in the immune response and homeostasis. C-type lectins are abundant and diverse in animals. Their immunological activities include lymphocyte homing, pathogen recognition, and clearance of apoptotic bodies. C-type lectin domains are composed of 110-130 amino acid residues with highly conserved structural folds. Remarkably, individual lectins can accept a wide variety of sugar ligands and can distinguish subtle structural differences in closely related ligands. In addition, several C-type lectin-like proteins specifically bind to carbohydrate ligands in Ca²⁺-independent ways. The accumulated 3D structural evidence clarifies the unexpected structural versatility of C-type lectins underlying the variety of ligand binding modes. In this issue, we focus on the structural aspects of carbohydrate recognition mechanisms of C-type lectins and C-type lectin-like proteins.

Stabilization of branched oligosaccharides: Lewis(x) benefits from a nonconventional C-H···O hydrogen bond

Although animal lectins usually show a high degree of specificity for glycan structures, their single-site binding affinities are typically weak, a drawback which is often compensated in biological systems by an oligovalent presentation of carbohydrate epitopes. For the design of monovalent glycomimetics, structural information regarding solution and bound conformation of the carbohydrate lead represents a valuable starting point. In this paper, we focus on the conformation of the trisaccharide Le(x) (Gal[Fucα(1-3)]β(1-4)GlcNAc). Mainly because of the unfavorable tumbling regime, the elucidation of the solution conformation of Le(x) by NMR has only been partially successful so far. Le(x) was therefore attached to a (13)C,(15)N-labeled protein. (13)C,(15)N-filtered NOESY NMR techniques at ultrahigh field allowed increasing the maximal NOE enhancement, resulting in a high number of distance restraints per glycosidic bond and, consequently, a well-defined structure. In addition to the known contributors to the conformational restriction of the Le(x) structure (exoanomeric effect, steric compression induced by the NHAc group adjacent to the linking position of L-fucose, and the hydrophobic interaction of L-fucose with the β-face of D-galactose), a nonconventional C-H···O hydrogen bond between H-C(5) of L-fucose and O(5) of D-galactose was identified. According to quantum mechanical calculations, this C-H···O hydrogen bond is the most prominent factor in stabilization, contributing 40% of the total stabilization energy. We therefore propose that the nonconventional hydrogen bond contributing to a reduction of the conformational flexibility of the Le(x) core represents a novel element of the glycocode. Its relevance to the stabilization of related branched oligosaccharides is currently being studied.

Mannosylated liposomes as antigen delivery vehicles for targeting to dendritic cells

The immune stimulating ability of mannosylated liposomes containing FITC-ovalbumin as a model antigen and displaying either a branched tri-mannose or a mono-mannose ligand on the liposome surface was investigated in human monocyte-derived dendritic cells (MoDCs) and murine bone-marrow-derived dendritic cells (BMDCs). Uptake of liposomes, dendritic cell activation and proliferation of CD8+ T cells from OT-I transgenic mice were determined by flow cytometry. Uptake of liposomes displaying the tri-mannose ligand was enhanced in human MoDCs compared with both non-mannosylated liposomes and liposomes displaying mono-mannose ligands. However, this increased uptake did not result in an increase in expression of CD80 or CD86 on the surface of the MoDCs. In contrast, neither tri-mannose- nor mono-mannose-containing liposomes were taken up by murine BMDCs to a greater extent than non-mannose-containing liposomes. The expression of CD86 and CD40 on the surface of BMDCs was not increased after exposure to mannosylated lipo-somes and BMDCs incubated with mannosylated liposomes were not able to stimulate proliferation of CD8+ T cells to any greater extent than BMDCs incubated with non-mannosylated liposomes. These findings suggest that while mannose-containing ligands can enhance the uptake of antigen-containing liposomes by some dendritic cells, important differences in the affinity of carbohydrate-binding receptors for mannose-containing ligands do exist between species. In addition, the increase in uptake of antigen by dendritic cells using mannosylated liposomes does not necessarily result in enhanced dendritic cell activation.

6-sulfo sialyl Lewis X is the common receptor determinant recognized by H5, H6, H7 and H9 influenza viruses of terrestrial poultry

Background

Influenza A viruses of domestic birds originate from the natural reservoir in aquatic birds as a result of interspecies transmission and adaptation to new host species. We previously noticed that influenza viruses isolated from distinct orders of aquatic and terrestrial birds may differ in their fine receptor-binding specificity by recognizing the structure of the inner parts of Neu5Acα2-3Gal-terminated sialyloligosaccharide receptors. To further characterize these differences, we studied receptor-binding properties of a large panel of influenza A viruses from wild aquatic birds, poultry, pigs and horses.

Results

Using a competitive solid-phase binding assay, we determined viral binding to polymeric conjugates of sialyloligosaccharides differing by the type of Neu5Acα-Gal linkage and by the structure of the more distant parts of the oligosaccharide chain. Influenza viruses isolated from terrestrial poultry differed from duck viruses by an enhanced binding to sulfated and/or fucosylated Neu5Acα2-3Gal-containing sialyloligosaccharides. Most of the poultry viruses tested shared a high binding affinity for the 6-sulfo sialyl Lewis X (Su-SLe^x). Efficient binding of poultry viruses to Su-SLe^x was often accompanied by their ability to bind to Neu5Acα2-6Gal-terminated (human-type) receptors. Such a dual receptor-binding specificity was demonstrated for the North American and Eurasian H7 viruses, H9N2 Eurasian poultry viruses, and H1, H3 and H9 avian-like virus isolates from pigs.

Conclusion

Influenza viruses of terrestrial poultry differ from ancestral duck viruses by enhanced binding to sulfated and/or fucosylated Neu5Acα2-3Gal-terminated receptors and, occasionally, by the ability to bind to Neu5Acα2-6Gal-terminated (human-type) receptors. These findings suggest that the adaptation to receptors in poultry can enhance the potential of an avian virus for avian-to-human transmission and pandemic spread.

The C-type lectin L-SIGN differentially recognizes glycan antigens on egg glycosphingolipids and soluble egg glycoproteins from Schistosoma mansoni

Recognition of pathogen-derived carbohydrate constituents by antigen presenting cells is an important step in the induction of protective immunity. Here we investigated the interaction of L-SIGN (liver/lymph node specific ICAM-3-grabbing nonintegrin), a C-type lectin that functions as antigen receptor on human liver sinusoidal endothelial cells, with egg-derived glycan antigens of the parasitic trematode Schistosoma mansoni. Our data demonstrate that L-SIGN binds both schistosomal soluble egg antigens (SEA) and egg glycosphingolipids, and can mediate internalization of SEA by L-SIGN expressing cells. Binding and internalization of SEA was strongly reduced after treatment of SEA with endoglycosidase H, whereas defucosylation affected neither binding nor internalization. These data indicate that L-SIGN predominantly interacts with oligomannosidic N-glycans of SEA. In contrast, binding to egg glycosphingolipids was completely abolished after defucosylation. Our data show that L-SIGN binds to a glycosphingolipid fraction containing fucosylated species with compositions of Hex(1)HexNAc(5-7)dHex(3-6)Cer, as evidenced by mass spectrometry. The L-SIGN "gain of function" mutant Ser363Val, which binds fucosylated Lewis antigens, did not bind to this fucosylated egg glycosphingolipid fraction, suggesting that L-SIGN displays different modes in binding fucoses of egg glycosphingolipids and Lewis antigens, respectively. Molecular modeling studies indicate that the preferred binding mode of L-SIGN to the respective fucosylated egg glycosphingolipid oligosaccharides involves a Fucalpha1-3GalNAcbeta1-4(Fucalpha1-3)GlcNAc tetrasaccharide at the nonreducing end. In conclusion, our data indicate that L-SIGN recognizes both oligomannosidic N-glycans and multiply fucosylated carbohydrate motifs within Schistosoma egg antigens, which demonstrates that L-SIGN has a broad but specific glycan recognition profile.

Cooperative Hydrogen Bonds of Macromolecules. 3. A Model Study of the Proximity Effect

Experimental and theoretical evidence for the proximity effect as a basic mechanism of the hydrogen bond cooperativity was obtained in a model system. Hydrogen bond (HB) interaction between poly(4-vinylpyridine) (PVP) and selected acids as HB donors was studied using PFG NMR self-diffusion measurements, 1H NMR longitudinal relaxation and quantum-mechanical DFT calculations. Bivalent HB donors, such as glutaric (GA) and adipic acid (AA), were compared to univalent donors 4-chlorobutyric acid, 4-acetylbutyric acid and 5-chlorovaleric acid. PFG NMR established substantially larger HB equilibrium constants for AA and, in particular, GA than for univalent donors, thus indicating cooperativity of COOH groups in bivalent donors. According to the values of these constants, the fraction of the transiently bound GA and AA molecules, which are bound by two hydrogen bonds, is 0.70 and 0.63, respectively. This result, which means substantial cooperativity in particular in GA, was then independently verified by a relaxation study comparing longitudinal relaxation rates of univalent and bivalent donors. Analysis of relaxation led to the same probabilities that HB of one COOH group of a bivalent donor will be accompanied by HB of another COOH group of the same molecule, namely 0.70 for GA and 0.63 for AA. Such cooperativity must be due to the proximity effect, i.e., the lowering of the entropy demand of the next binding by the motional restriction imposed by the already existing bonds. This conclusion is in excellent agreement with DFT calculations on the interaction of GA with a model vinylpyridine dimer, which predict preference of double binding of the same GA molecule over that of two GA molecules and show that this preference is due to a substantially lower entropy demand.

Cooperative Hydrogen Bonds of Macromolecules. 2. Two-Dimensional Cooperativity in the Binding of Poly(4-vinylpyridine) to Poly(4-vinylphenol)

The hydrogen bond interaction of poly(4-vinylphenol) (PVF), ligated by a 20 mol/mol excess of pyridine-d(5) (PD) in tetrahydrofuran-d(8), with poly(4-vinylpyridine) (PVP) was studied using liquid and solid-state NMR and quantum mechanical calculations. Because of its cooperative interaction, PVP substitutes PD in its hydrogen bond with PVF, thus forming a PVF-PVP complex, which gradually precipitates from solution. On the basis of the 1H/13C NMR spin-diffusion experiments and density functional theory quantum calculations, the complex is shown to have the fairly regular structure of a polymer sheet with intermittent H-bond links between PVF and PVP chains. The cooperativity of PVP interaction with PVF was studied by measuring the dependence of the binding degree alpha of PVP on its polymerization degree (P(n), being 10, 17, 30, 36, 48, 65, and 84) at various PVP/PVF molar ratios. The value of alpha was established indirectly by measuring the fraction of liberated PD using its 2H quadrupolar relaxation and pulsed field-gradient spin-echo measurement of self-diffusion. The cooperativity is shown to be of a higher order and two-dimensional, that is, dependent on both the polymerization degree of PVP and its ratio to PVF. A mathematical model of such two-dimensional cooperativity based chiefly on a proximity effect is suggested.

Aryl O- and S-galactosides and lactosides as specific inhibitors of human galectins-1 and -3: Role of electrostatic potential at O-3

Phase transfer catalyzed reaction was used for the high yielding synthesis of aryl 1-thio-beta-d-galacto- and lacto-pyranosides carrying a panel of substituents on the phenyl groups. Best galectin-1 inhibitors were simple p-nitrophenyl thiogalactoside 5a for the monosaccharide and o-nitrophenyl thiolactoside 6f or napthylsulfonyl lactoside 8c, both being 20 times better relative to natural ligands. Relative inhibitory properties as low as 2500 and 40 microM were observed, respectively. The electronic effects of the lactoside aglycons directly influenced the electrostatic potential at O-3, which was associated with the inhibitory potencies against galectin-1.

The C-type lectin-like domain superfamily

The superfamily of proteins containing C-type lectin-like domains (CTLDs) is a large group of extracellular Metazoan proteins with diverse functions. The CTLD structure has a characteristic double-loop ('loop-in-a-loop') stabilized by two highly conserved disulfide bridges located at the bases of the loops, as well as a set of conserved hydrophobic and polar interactions. The second loop, called the long loop region, is structurally and evolutionarily flexible, and is involved in Ca2+-dependent carbohydrate binding and interaction with other ligands. This loop is completely absent in a subset of CTLDs, which we refer to as compact CTLDs; these include the Link/PTR domain and bacterial CTLDs. CTLD-containing proteins (CTLDcps) were originally classified into seven groups based on their overall domain structure. Analyses of the superfamily representation in several completely sequenced genomes have added 10 new groups to the classification, and shown that it is applicable only to vertebrate CTLDcps; despite the abundance of CTLDcps in the invertebrate genomes studied, the domain architectures of these proteins do not match those of the vertebrate groups. Ca2+-dependent carbohydrate binding is the most common CTLD function in vertebrates, and apparently the ancestral one, as suggested by the many humoral defense CTLDcps characterized in insects and other invertebrates. However, many CTLDs have evolved to specifically recognize protein, lipid and inorganic ligands, including the vertebrate clade-specific snake venoms, and fish antifreeze and bird egg-shell proteins. Recent studies highlight the functional versatility of this protein superfamily and the CTLD scaffold, and suggest further interesting discoveries have yet to be made.

Receptor-binding properties of swine influenza viruses isolated and propagated in MDCK cells

To study the receptor specificities of H1 and H3 influenza viruses isolated recently from pigs, we employed the analogues of natural receptors, namely sialyloligosaccharides conjugated with polyacrylamide in biotinylated and label free forms. All Madin-Darby canine kidney (MDCK) cell-propagated viruses with human H3 or classical swine H1 hemagglutinins bound only to Neu5Acalpha2-6Galbeta1-bearing polymers, and not to Neu5Acalpha2-3Galbeta1-bearing polymers. This receptor-binding pattern is typical for human influenza viruses and it differs from the previously described receptor-binding specificity of egg-adapted swine influenza viruses. Swine virus isolates with avian-like H1 and H3 hemagglutinins displayed distinct receptor specificity by binding to both Neu5Acalpha2-6Gal- and Neu5Acalpha2-3Gal-containing receptors. These viruses, as well as egg-adapted swine and turkey viruses with a classical swine HA, differed from the related duck viruses by increased affinity to sulfated sialyloligosaccaride, Su-SiaLe(x). Except for avian-like H3 viruses, none of the studied swine viruses bound to Neu5Gc-containing sialoglycopolymers, suggesting that binding to these sialic acid species abundantly expressed in pigs may not be essential for virus replication in this host.

Evolution of the receptor binding phenotype of influenza A (H5) viruses

Receptor specificity of influenza A/H5 viruses including human 2003-04 isolates was studied. All but two isolates preserved high affinity to Sia2-3Gal (avian-like) receptors. However, two isolates (February, 2003, Hong Kong) demonstrated decreased affinity to Sia2-3Gal and moderate affinity to a Sia2-6Gal (human-like) receptors. These two viruses had a unique Ser227-Asn change in the hemagglutinin molecule. Thus, a single amino acid substitution can significantly alter receptor specificity of avian H5N1 viruses, providing them with an ability to bind to receptors optimal for human influenza viruses. Asian 2003-04 H5 isolates from chickens and humans demonstrated highest affinity to the sulfated trisaccharide Neu5Acalpha2-3Galbeta1-4(6-HSO3)GlcNAcbeta (Su-3'SLN) receptor but, in contrast to 1997 isolates, had increased affinity to fucosylated Su-3'SLN. American poultry H5 viruses also had increased affinity to Su-3'SLN. These data demonstrate that the genetic evolution of avian influenza A(H5N1) viruses is accompanied during adaptation to poultry by the evolution of their receptor specificity.

Receptor specificity of influenza viruses from birds and mammals: new data on involvement of the inner fragments of the carbohydrate chain

We studied receptor-binding properties of influenza virus isolates from birds and mammals using polymeric conjugates of sialooligosaccharides terminated with common Neu5Ac alpha2-3Gal beta fragment but differing by the structure of the inner part of carbohydrate chain. Viruses isolated from distinct avian species differed by their recognition of the inner part of oligosaccharide receptor. Duck viruses displayed high affinity for receptors having beta1-3 rather than beta1-4 linkage between Neu5Ac alpha2-3Gal-disaccharide and penultimate N-acetylhexosamine residue. Fucose and sulfate substituents at this residue had negative and low effect, respectively, on saccharide binding to duck viruses. By contrast, gull viruses preferentially bound to receptors bearing fucose at N-acetylglucosamine residue, whereas chicken and mammalian viruses demonstrated increased affinity for oligosaccharides that harbored sulfo group at position 6 of (beta1-4)-linked GlcNAc. These data suggest that although all avian influenza viruses preferentially bind to Neu5Ac alpha2-3Gal-terminated receptors, the fine receptor specificity of the viruses varies depending on the avian species. Further studies are required to determine whether observed host-dependent differences in the receptor specificity of avian viruses can affect their ability to infect humans.

Molecular basis of the differences in binding properties of the highly related C-type lectins DC-SIGN and L-SIGN to Lewis X trisaccharide and Schistosoma mansoni egg antigens

The dendritic cell-specific C-type lectin DC-SIGN functions as a pathogen receptor that recognizes Schistosoma mansoni egg antigens through its major glycan epitope Galbeta1,4(Fucalpha1,3)GlcNAc (Lex). Here we report that L-SIGN, a highly related homologue of DC-SIGN found on liver sinusoidal endothelial cells, binds to S. mansoni egg antigens but not to the Lex epitope. L-SIGN does bind the Lewis antigens Lea, Leb, and Ley, similar as DC-SIGN. A specific mutation in the carbohydrate recognition domain of DC-SIGN (V351G) abrogates binding to all Lewis antigens. In L-SIGN Ser363 is present at the corresponding position of Val351 in DC-SIGN. Replacement of this Ser into Val resulted in a "gain of function" L-SIGN mutant that binds to Lex, and shows increased binding to the other Lewis antigens. These data indicate that Val351 is important for the fucose specificity of DC-SIGN. Molecular modeling and docking of the different Lewis antigens in the carbohydrate recognition domains of L-SIGN, DC-SIGN, and their mutant forms, demonstrate that Val351 in DC-SIGN creates a hydrophobic pocket that strongly interacts with the Fucalpha1,3/4-GlcNAc moiety of the Lewis antigens. The equivalent amino acid residue Ser363 in L-SIGN creates a hydrophilic pocket that prevents interaction with Fucalpha1,3-GlcNAc in Lex but supports interactions with the Fucalpha1,4-GlcNAc moiety in Lea and Leb antigens. These data demonstrate for the first time that DC-SIGN and L-SIGN differ in their carbohydrate binding profiles and will contribute to our understanding of the functional roles of these C-type lectin receptors, both in recognition of pathogen and self-glycan antigens.

CORCEMA refinement of the bound ligand conformation within the protein binding pocket in reversibly forming weak complexes using STD-NMR intensities

We describe an intensity-restrained optimization procedure for refining approximate structures of ligands within the protein binding pockets using STD-NMR intensity data on reversibly forming weak complexes. In this approach, the global minimum for the bound-ligand conformation is obtained by a hybrid structure refinement method involving CORCEMA calculation of intensities and simulated annealing optimization of torsion angles of the bound ligand using STD-NMR intensities as experimental constraints and the NOE R-factor as the pseudo-energy function to be minimized. This method is illustrated using simulated STD data sets for typical carbohydrate and peptide ligands. Our procedure also allows for the optimization of side chain torsion angles of protein residues within the binding pocket. This procedure is useful in refining and improving initial models based on crystallography or computer docking or other algorithms to generate models for the bound ligand (e.g., a lead compound) within the protein binding pocket compatible with solution STD-NMR data. This method may facilitate structure-based drug design efforts.

Structural characterization of a mannose-binding protein-trimannoside complex using residual dipolar couplings

The ligand-binding properties of a 53 kDa homomultimeric trimer from mannose-binding protein (MBP) have been investigated using residual dipolar couplings (RDCs) that are easily measured from NMR spectra of the ligand and isotopically labeled protein. Using a limited set of 1H-15N backbone amide NMR assignments for MBP and orientational information derived from the RDC measurements in aligned media, an order tensor for MBP has been determined that is consistent with symmetry-based predictions of an axially symmetric system. 13C-1H couplings for a bound trisaccharide ligand, methyl 3,6-di-O-(alpha-D-mannopyranosyl)-alpha-D-mannopyranoside (trimannoside) have been determined at natural abundance and used as orientational constraints. The bound ligand geometry and orientational constraints allowed docking of the trimannoside ligand in the binding site of MBP to produce a structural model for MBP-oligosaccharide interactions.

Design and synthesis of sialyl Lewis(x) mimics as E- and P-selectin inhibitors

The selectins are a family of cell-adhesion proteins that mediate the rolling of leukocytes on activated endothelial cells through the recognition of the carbohydrate epitope sialyl Lewis(x) (sLe(x)). Control of the leukocyte-endothelial cell adhesion process may prove useful in cases where excess recruitment of leukocytes can contribute to acute diseases such as stroke and reperfusion injury and chronic diseases such as psoriasis and rheumatoid arthritis. The development of molecules that block the interactions between sLe(x) and the selectins has become an active area of research. In this review, we will highlight the various approaches taken toward the development of sLe(x) mimetics as antagonists of E- and P-selectin, including the use of structural information about the selectins and their interactions with sLe(x) that have been revealed through the use of NMR, protein crystallography and molecular modeling.

Effect of protonation of the N-acetyl neuraminic acid residue of sialyl Lewis(X): a molecular orbital study with insights into its binding properties toward the carbohydrate recognition domain of E-selectin

Semiempirical molecular orbital (MO) calculations with an implicit treatment of the water environment were employed in order to assess whether the sialyl Lewis(X) (sLe(X)) tetrasaccharide binds to E-selectin in the anionic or neutral (i.e., protonated) state. The analysis of the frontier molecular orbitals, electrostatic potential surfaces, and conformational behavior of the sugar indicates that its neutral form possesses the necessary characteristics for binding. In particular, the LUMO level of the neutral sLe(X) molecule is localized both on the carboxylic group of the N-acetyl neuraminic acid (NeuNAc) residue and on the nearby glycosidic linkage. These two moieties interact with the Arg97 residue of E-selectin, as revealed by a recent crystal structure analysis of the E-selectin/sLe(X) complex. The energetics of this specific interaction was investigated with the aid of ab initio Hartree-Fock MO calculations, which resulted in a BSSE-corrected binding energy of 16.63 kcal/mol. Our observations could open up new perspectives in the design of sLe(X) mimics.

An endogenous Drosophila receptor for glycans bearing alpha 1,3-linked core fucose residues

The genome of Drosophila melanogaster encodes several proteins that are predicted to contain Ca(2+)-dependent, C-type carbohydrate-recognition domains. The CG2958 gene encodes a protein containing 359 amino acid residues. Analysis of the CG2958 sequence suggests that it consists of an N-terminal domain found in other Drosophila proteins, a middle segment that is unique, and a C-terminal C-type carbohydrate-recognition domain. Expression studies show that the full-length protein is a tetramer formed by noncovalent association of disulfide-linked dimers that are linked through cysteine residues in the N-terminal domain. The expressed protein binds to immobilized yeast invertase through the C-terminal carbohydrate-recognition domain. Competition binding studies using monosaccharides demonstrate that CG2958 interacts specifically with fucose and mannose. Fucose binds approximately 5-fold better than mannose. Blotting studies reveal that the best glycoprotein ligands are those that contain N-linked glycans bearing alpha1,3-linked fucose residues. Binding is enhanced by the additional presence of alpha1,6-linked fucose. It has previously been proposed that labeling of the Drosophila neural system by anti-horseradish peroxidase antibodies is a result of the presence of difucosylated N-linked glycans. CG2958 is a potential endogenous receptor for such neural-specific carbohydrate epitopes.

Molecular dynamics simulations of glycoclusters and glycodendrimers

Protein-carbohydrate recognition plays a crucial role in a wide range of biological processes, required both for normal physiological functions and the onset of disease. Nature uses multivalency in carbohydrate-protein interactions as a strategy to overcome the low affinity found for singular binding of an individual saccharide epitope to a single carbohydrate recognition domain of a lectin. To mimic the complex multi-branched oligosaccharides found in glycoconjugates, which form the structural basis of multivalent carbohydrate-protein interactions, so-called glycoclusters and glycodendrimers have been designed to serve as high-affinity ligands of the respective receptor proteins. To allow a rational design of glycodendrimer-type molecules with regard to the receptor structures involved in carbohydrate recognition, a deeper knowledge of the dynamics of such molecules is desirable. Most glycodendrimers have to be considered highly flexible molecules with their conformational preferences most difficult to elucidate by experimental methods. Longtime molecular dynamics (MD) simulations with inclusion of explicit solvent molecules are suited to explore the conformational space accessible to glycodendrimers. Here, a detailed geometric and conformational analysis of 15 glycodendrimers and glycoclusters has been accomplished, which differ with regard to their core moieties, spacer characteristics and the type of terminal carbohydrate units. It is shown that the accessible conformational space depends strongly on the structural features of the core and spacer moieties and even on the type of terminating sugars. The obtained knowledge about possible spatial distributions of the sugar epitopes exposed on the investigated hyperbranched neoglycoconjugates is detailed for all examples and forms important information for the interpretation and prediction of affinity data, which can be deduced from biological testing of these multivalent neoglycoconjugates.

Complete relaxation and conformational exchange matrix (CORCEMA) analysis of intermolecular saturation transfer effects in reversibly forming ligand-receptor complexes

A couple of recent applications of intermolecular NOE (INOE) experiments as applied to biomolecular systems involve the (i) saturation transfer difference NMR (STD-NMR) method and (ii) the intermolecular cross-saturation NMR (ICS-NMR) experiment. STD-NMR is a promising tool for rapid screening of a large library of compounds to identify bioactive ligands binding to a target protein. Additionally, it is also useful in mapping the binding epitopes presented by a bioactive ligand to its target protein. In this latter application, the STD-NMR technique is essentially similar to the ICS-NMR experiment, which is used to map protein-protein or protein-nucleic acid contact surfaces in complexes. In this work, we present a complete relaxation and conformational exchange matrix (CORCEMA) theory (H. N. B. Moseley et al., J. Magn. Reson. B 108, 243-261 (1995)) applicable for these two closely related experiments. As in our previous work, we show that when exchange is fast on the relaxation rate scale, a simplified CORCEMA theory can be formulated using a generalized average relaxation rate matrix. Its range of validity is established by comparing its predictions with those of the exact CORCEMA theory which is valid for all exchange rates. Using some ideal model systems we have analyzed the factors that influence the ligand proton intensity changes when the resonances from some protons on the receptor protein are saturated. The results show that the intensity changes in the ligand signals in an intermolecular NOE experiment are very much dependent upon: (1) the saturation time, (2) the location of the saturated receptor protons with respect to the ligand protons, (3) the conformation of the ligand-receptor interface, (4) the rotational correlation times for the molecular species, (5) the kinetics of the reversibly forming complex, and (6) the ligand/receptor ratio. As an example of a typical application of the STD-NMR experiment we have also simulated the STD effects for a hypothetical trisaccharide bound to a protein. The CORCEMA theory for INOE and the associated algorithm are useful in a quantitative interpretation of the intensity changes in the ligand in both the STD-NMR and ICS-NMR, provided the identity of the receptor protons experiencing direct RF saturation is known. The formalism presented here is likely to be useful in the design of bioactive ligands to a specific target protein and in the quantitative mapping of binding epitopes and interfaces between molecules in complexes.

Minimal requirements for the binding of selectin ligands to a C-type carbohydrate-recognition domain

The C-type carbohydrate-recognition domains of E-selectin and rat serum mannose-binding protein have similar structures. Selectin/mannose-binding protein chimeras created by transfer of key sequences from E-selectin into mannose-binding protein have previously been shown to bind the selectin ligand sialyl-Lewis(X) through a Ca(2+)-dependent subsite, common to many C-type lectins, and an accessory site containing positively charged amino acid residues. Further characterization of these chimeras as well as analysis of novel constructs containing additional regions of E-selectin demonstrate that selectin-like interaction with sialyl-Lewis(X) can be faithfully reproduced even though structural evidence indicates that the mechanisms of binding to E-selectin and the chimeras are different. Selectin-like binding to the nonfucosylated sulfatide and sulfoglucuronyl glycolipids can also be reproduced with selectin/mannose-binding protein chimeras that contain the two subsites involved in sialyl-Lewis(X) binding. These results indicate that binding of structurally distinct anionic glycans to C-type carbohydrate-recognition domains can be mediated by the Ca(2+)-dependent subsite in combination with a positively charged region that forms an ionic strength-sensitive subsite.

Enhanced 3-O-sulfation of galactose in Asn-linked glycans and Maackia amurensis lectin binding in a new Chinese hamster ovary cell line

We report the characterization of two Chinese hamster ovary cell lines that produce large amounts of sulfated N-linked oligosaccharides. Clones 26 and 489 were derived by stable transfection of the glycosaminoglycan-deficient cell mutant pgsA-745 with a cDNA library prepared from wild-type cells. Peptide:N-glycanase F released nearly all of the sulfate label, indicating that sulfation had occurred selectively on the Asn-linked glycans. Hydrazinolysis followed by nitrous acid treatment at pH 4 and borohydride reduction yielded reduced sulfated disaccharides that comigrated with standard Gal3SO4beta1-4anhydromannitol. The disaccharides were resistant to periodate oxidation but became sensitive after the sulfate group was removed by methanolysis, indicating that the sulfate was located at C3 of the galactose residues. Maackia amurensis lectin bound to the sulfated glycopeptides on the cell surface and in free form, even after sialidase treatment. This finding indicates that the lectin requires only a charged group at C3 of the galactose unit and not an intact sialic acid. Growth of cells with chlorate restored sialidase sensitivity to lectin binding, indicating that sulfation and sialylation occurred largely at the same sites. The enhanced sulfation was due to elevated sulfotransferase activity that catalyzed transfer of sulfate from phosphoadenosine-5'-phosphosulfate to Galbeta1-4(3)GlcNAcbeta-O-naphthalenemethanol.

Trinuclear Zn(II) and Cu(II) homo and heterotrimetallic complexes involving D-glucopyranosyl and biscarboxylate bridging ligands. A substrate binding model of xylose isomerases

Reactions of MCl(2).nH(2)O with N,N'-bis(D-glucopyranosyl)-1,4,7-triazacyclononane ((D-Glc)(2)-tacn), which was formed from D-glucose and 1,4,7-triazacyclononane (tacn) in situ, afforded a series of mononuclear divalent metal complexes with two beta-D-glucopyranosyl moieties, [M((D-Glc)(2)-tacn)Cl]Cl (M = Zn (11), Cu (12), Ni (13), Co (14)). Complexes 11-14 were characterized by analytical and spectroscopic measurements and X-ray crystallography and were found to have a distorted octahedral M(II) center ligated by the pentacoordinate N-glycoside ligand, (beta-D-glucopyranosyl)(2)-tacn, and a chloride anion. Each D-glucose moiety is tethered to the metal center through the beta-N-glycosidic bond with tacn and additionally coordinated via the C-2 hydroxyl group, resulting in a lambda-gauche five-membered chelate ring. When L-rhamnose (6-deoxy-L-mannose) was used instead of D-glucose, the nickel(II) complex with two beta-L-rhamnopyranosyl moieties, [Ni((D-Man)(2)-tacn)(MeOH)]Cl(2) (15), was obtained and characterized by an X-ray analysis. Reactions of 11 (M = Zn) with [Zn(XDK)(H(2)O)] (21) or [Cu(XDK)(py)(2)] (22) (H(2)XDK = m-xylylenediamine bis(Kemp's triacid imide)) yielded homo and heterotrimetallic complexes formulated as [Zn(2)M'((D-Glc)(2)-tacn)(2)(XDK)]Cl(2) (M' = Zn (31), Cu (32)). The similar reactions of 12 (M = Cu) with complex 21 or 22 afforded [Cu(2)M'((D-Glc)(2)-tacn)(2)(XDK)]Cl(2) (M' = Cu (33), Zn (34)). An X-ray crystallographic study revealed that complexes 31 and 34 have either Zn(II)(3) or Cu(II)Zn(II)Cu(II) trimetallic centers bridged by two carboxylate groups of XDK and two D-glucopyranosyl residues. The M...M' separations are 3.418(3)-3.462(3) A (31) and 3.414(1)-3.460(1) A (34), and the M...M'...M angles are 155.18(8) degrees (31) and 161.56(6) degrees (34). The terminal metal ions are octahedrally coordinated by the (D-Glc)(2)-tacn ligand through three nitrogen atoms of tacn, two oxygen atoms of the C-2 hydroxyl groups of the carbohydrates, and a carboxylate oxygen atom of XDK ligand. The central metal ions sit in a distorted octahedral environment ligated by four oxygen atoms of the carbohydrate residues in the (D-Glc)(2)-tacn ligands and two carboxylate oxygen atoms of XDK. The deprotonated beta-D-glucopyranosyl unit at the C-2 hydroxyl group bridges the terminal and central ions with the C-2 mu-alkoxo group, with the C-1 N-glycosidic amino and the C-3 hydroxyl groups coordinating to each metal center. Complexes 31-34 are the first examples of metal complexes in which D-glucose units act as bridging ligands. These structures could be very useful substrate binding models of xylose or glucose isomerases, which promote D-glucose D-fructose isomerization by using divalent dimetallic centers bridged by a glutamate residue.

Substrate-dependent chemoselective aldose-aldose and aldose-ketose isomerizations of carbohydrates promoted by a combination of calcium ion and monoamines

Epimerization of aldoses at C-2 has been extensively investigated by using various metal ions in conjunction with diamines, monoamines, and aminoalcohols. Aldoses are epimerized at C-2 by a combination of alkaline-earth or rare-earth metal ions (Ca(2+), Sr(2+), Pr(3+), or Ce(3+)) and such monoamines as triethylamine. In particular, the Ca(2+)-triethylamine system proved effective in promoting aldose-ketose isomerization as well as C-2 epimerization of aldoses. 13C NMR studies using D-(1-(13)C)glucose and D-(1-(13)C)galactose with the CaCl(2) system in CD(3)OD revealed that the C-2 epimerization proceeds via stereospecific rearrangement of the carbon skeleton, or 1,2-carbon shift, and ketose formation proceeds partially through an intramolecular hydrogen migration or 1,2-hydride shift and, in part, via an enediol intermediate. These simultaneous aldose-aldose and aldose-ketose isomerizations showed interesting substrate-dependent chemoselectivity. Whereas the mannose-type aldoses having 2,3-erythro configuration (D-mannose, D-lyxose, and D-ribose) showed considerable resistance to both the C-2 epimerization and the aldose-ketose isomerization, the glucose-type sugars having 2,3-threo and 3,4-threo configurations, D-glucose and D-xylose, are mainly epimerized at C-2 and those having the 2,3-threo and 3,4-erythro configurations, D-galactose and D-arabinose, were mostly isomerized into 2-ketoses. These features are of potential interest in relevance to biomimic sugar transformations by metal ions.

A microcantilever device to assess the effect of force on the lifetime of selectin-carbohydrate bonds

A microcantilever technique was used to apply force to receptor-ligand molecules involved in leukocyte rolling on blood vessel walls. E-selectin was adsorbed onto 3-microm-diameter, 4-mm-long glass fibers, and the selectin ligand, sialyl Lewis(x), was coupled to latex microspheres. After binding, the microsphere and bound fiber were retracted using a computerized loading protocol that combines hydrodynamic and Hookean forces on the fiber to produce a range of force loading rates (force/time), r(f). From the distribution of forces at failure, the average force was determined and plotted as a function of ln r(f). The slope and intercept of the plot yield the unstressed reverse reaction rate, k(r)(o), and a parameter that describes the force dependence of reverse reaction rates, r(o). The ligand was titrated so adhesion occurred in approximately 30% of tests, implying that >80% of adhesive events involve single bonds. Monte Carlo simulations show that this level of multiple bonding has little effect on parameter estimation. The estimates are r(o) = 0.048 and 0.016 nm and k(r)(o) = 0.72 and 2.2 s(-1) for loading rates in the ranges 200-1000 and 1000-5000 pN s(-1), respectively. Levenberg-Marquardt fitting across all values of r(f) gives r(o) = 0.034 nm and k(r)(o) = 0.82 s(-1). The values of these parameters are in the range required for rolling, as suggested by adhesive dynamics simulations.

The molecular mechanism of sulfated carbohydrate recognition by the cysteine-rich domain of mannose receptor

The mannose receptor (MR) binds foreign and host ligands through interactions with their carbohydrates. Two portions of MR have distinct carbohydrate recognition properties. One is conferred by the amino-terminal cysteine-rich domain (Cys-MR), which plays a critical role in binding sulfated glycoproteins including pituitary hormones. The other is achieved by tandemly arranged C-type lectin domains that facilitate carbohydrate-dependent uptake of infectious microorganisms. This dual carbohydrate binding specificity enables MR to bind ligands by interacting with both sulfated and non-sulfated polysaccharide chains. We previously determined crystal structures of Cys-MR complexed with 4-SO(4)-N-acetylglucosamine and with an unidentified ligand resembling Hepes (N-[2-hydroxyethyl]piperazine-N'-[2-ethanesulfonic acid]). In continued efforts to elucidate the mechanism of sulfated carbohydrate recognition by Cys-MR, we characterized the binding affinities between Cys-MR and potential carbohydrate ligands using a fluorescence-based assay. We find that Cys-MR binds sulfated carbohydrates with relatively high affinities (K(D)=0.1 mM to 1.0 mM) compared to the affinities of other lectins. Cys-MR also binds Hepes with a K(D) value of 3.9 mM, consistent with the suggestion that the ligand in the original Cys-MR crystal structure is Hepes. We also determined crystal structures of Cys-MR complexed with 3-SO(4)-Lewis(x), 3-SO(4)-Lewis(a), and 6-SO(4)-N-acetylglucosamine at 1.9 A, 2.2 A, and 2.5 A resolution, respectively, and the 2.0 A structure of Cys-MR that had been treated to remove Hepes. The conformation of the Cys-MR binding site is virtually identical in all Cys-MR crystal structures, suggesting that Cys-MR does not undergo conformational changes upon ligand binding. The structures are used to rationalize the binding affinities derived from the biochemical studies and to elucidate the molecular mechanism of sulfated carbohydrate recognition by Cys-MR.

Binding of glycosulfopeptides to P-selectin requires stereospecific contributions of individual tyrosine sulfate and sugar residues

P-selectin glycoprotein ligand-1 (PSGL-1) is a mucin on leukocytes that binds to selectins. P-selectin binds to an N-terminal region of PSGL-1 that requires sulfation of at least one of three clustered tyrosines (TyrSO(3)) and an adjacent core-2-based O-glycan expressing sialyl Lewis x (C2-O-sLe(x)). We synthesized glycosulfopeptides (GSPs) modeled after this region of PSGL-1 to explore the roles of individual TyrSO(3) residues, the placement of C2-O-sLe(x) relative to TyrSO(3), the relative contributions of fucose and sialic acid on C2-O-sLe(x), and the function of the peptide sequence for binding to P-selectin. Binding of GSPs to P-selectin was measured by affinity chromatography and equilibrium gel filtration. 2-GSP-6, which has C2-O-sLe(x) at Thr-57 and TyrSO(3) at residues 46, 48, and 51, bound to P-selectin with high affinity (K(d) approximately 650 nm), whereas an isomeric trisulfated GSP containing C2-O-sLe(x) at Thr-44 bound much less well. Non-sulfated glycopeptide (2-GP-6) containing C2-O-sLe(x) at Thr-57 bound to P-selectin with approximately 40-fold lower affinity (K(d) approximately 25 microm). Proteolysis of 2-GP-6 abolished detectable binding of the residual C2-O-sLe(x)-Thr to P-selectin, demonstrating that the peptide backbone contributes to binding. Monosulfated and disulfated GSPs bound significantly better than non-sulfated 2-GP-6, but sulfation of Tyr-48 enhanced affinity (K(d) approximately 6 microm) more than sulfation of Tyr-46 or Tyr-51. 2-GSP-6 lacking sialic acid bound to P-selectin at approximately 10% that of the level of the parent 2-GSP-6, whereas 2-GSP-6 lacking fucose did not detectably bind; thus, fucose contributes more than sialic acid to binding. Reducing NaCl from 150 to 50 mm markedly enhanced binding of 2-GSP-6 to P-selectin (K(d) approximately 75 nm), demonstrating the charge dependence of the interaction. These results reveal a stereospecific interaction of P-selectin with PSGL-1 that includes distinct contributions of each of the three TyrSO(3) residues, adjacent peptide determinants, and fucose/sialic acid on an optimally positioned core-2 O-glycan.

Mechanism of pH-dependent N-acetylgalactosamine binding by a functional mimic of the hepatocyte asialoglycoprotein receptor

Efficient release of ligands from the Ca(2+)-dependent carbohydrate-recognition domain (CRD) of the hepatic asialoglycoprotein receptor at endosomal pH requires a small set of conserved amino acids that includes a critical histidine residue. When these residues are incorporated at corresponding positions in an homologous galactose-binding derivative of serum mannose-binding protein, the pH dependence of ligand binding becomes more like that of the receptor. The modified CRD displays 40-fold preferential binding to N-acetylgalactosamine compared with galactose, making it a good functional mimic of the asialoglycoprotein receptor. In the crystal structure of the modified CRD bound to N-acetylgalactosamine, the histidine (His(202)) contacts the 2-acetamido methyl group and also participates in a network of interactions involving Asp(212), Arg(216), and Tyr(218) that positions a water molecule in a hydrogen bond with the sugar amide group. These interactions appear to produce the preference for N-acetylgalactosamine over galactose and are also likely to influence the pK(a) of His(202). Protonation of His(202) would disrupt its interaction with an asparagine that serves as a ligand for Ca(2+) and sugar. The structure of the modified CRD without sugar displays several different conformations that may represent structures of intermediates in the release of Ca(2+) and sugar ligands caused by protonation of His(202).

Minimal sulfated carbohydrates for recognition by L-selectin and the MECA-79 antibody

Sulfated forms of sialyl-Le(X) containing Gal-6-SO(4) or GlcNAc-6-SO(4) have been implicated as potential recognition determinants on high endothelial venule ligands for L-selectin. The optimal configuration of sulfate esters on the N-acetyllactosamine (Galbeta1-->4GlcNAc) core of sulfosialyl-Le(X), however, remains unsettled. Using a panel of sulfated lactose (Galbeta1-->4Glc) neoglycolipids as substrates in direct binding assays, we found that 6',6-disulfolactose was the preferred structure for L-selectin, although significant binding to 6'- and 6-sulfolactose was observed as well. Binding was EDTA-sensitive and blocked by L-selectin-specific monoclonal antibodies. Surprisingly, 6', 6-disulfolactose was poorly recognized by MECA-79, a carbohydrate- and sulfate-dependent monoclonal antibody that binds competitively to L-selectin ligands. Instead, MECA-79 bound preferentially to 6-sulfolactose. The difference in preferred substrates between L-selectin and MECA-79 may explain the variable activity of MECA-79 as an inhibitor of lymphocyte adhesion to high endothelial venules in lymphoid organs. Our results suggest that both Gal-6-SO(4) and GlcNAc-6-SO(4) may contribute to L-selectin recognition, either as components of sulfosialyl-Le(X) capping groups or in internal structures. By contrast, only GlcNAc-6-SO(4) appears to contribute to MECA-79 binding.

Collectin structure: a review

Collectins are animal calcium dependent lectins that target the carbohydrate structures on invading pathogens, resulting in the agglutination and enhanced clearance of the microorganism. These proteins form trimers that may assemble into larger oligomers. Each polypeptide chain consists of four regions: a relatively short N-terminal region, a collagen like region, an alpha-helical coiled-coil, and the lectin domain. Only primary structure data are available for the N-terminal region, while the most important features of the collagen-like region can be derived from its homology with collagen. The structures of the alpha-helical coiled-coil and the lectin domain are known from crystallographic studies of mannan binding protein (MBP) and lung surfactant protein D (SP-D). Carbohydrate binding has been structurally characterized in several complexes between MBP and carbohydrate; all indicate that the major interaction between carbohydrate and collectin is the binding of two adjacent carbohydrate hydroxyl group to a collectin calcium ion. In addition, these hydroxyl groups hydrogen bond to some of the calcium amino acid ligands. While each collectin trimer contains three such carbohydrate binding sites, deviation from the overall threefold symmetry has been demonstrated for SP-D, which may influence its binding properties. The protein surface between the three binding sites is positively charged in both MBP and SP-D.

Structure of a C-type carbohydrate recognition domain from the macrophage mannose receptor

The mannose receptor of macrophages and liver endothelium mediates clearance of pathogenic organisms and potentially harmful glycoconjugates. The extracellular portion of the receptor includes eight C-type carbohydrate recognition domains (CRDs), of which one, CRD-4, shows detectable binding to monosaccharide ligands. We have determined the crystal structure of CRD-4. Although the basic C-type lectin fold is preserved, a loop extends away from the core of the domain to form a domain-swapped dimer in the crystal. Of the two Ca(2+) sites, only the principal site known to mediate carbohydrate binding in other C-type lectins is occupied. This site is altered in a way that makes sugar binding impossible in the mode observed in other C-type lectins. The structure is likely to represent an endosomal form of the domain formed when Ca(2+) is lost from the auxiliary calcium site. The structure suggests a mechanism for endosomal ligand release in which the auxiliary calcium site serves as a pH sensor. Acid pH-induced removal of this Ca(2+) results in conformational rearrangements of the receptor, rendering it unable to bind carbohydrate ligands.

An unusual carbohydrate binding site revealed by the structures of two Maackia amurensis lectins complexed with sialic acid-containing oligosaccharides

Seeds from the legume tree Maackia amurensis contain two lectins that can agglutinate different blood cell types. Their specificity toward sialylated oligosaccharides is unique among legume lectins; the leukoagglutinin preferentially binds to sialyllactosamine (alphaNeuAc(2-3)betaGal(1-4)betaGlcNAc), whereas the hemagglutinin displays higher affinity for a disialylated tetrasaccharide (alphaNeuAc(2-3)betaGal(1-3)[alphaNeuAc(2-6)]alphaG alNAc). The three-dimensional structure of the complex between M. amurensis leukoagglutinin and sialyllactose has been determined at 2.75-A resolution using x-ray crystallography. The carbohydrate binding site consists of a deep cleft that accommodates the three carbohydrate residues of the sialyllactose. The central galactose sits in the primary binding site in an orientation that has not been observed previously in other legume lectins. The carboxyl group of sialic acid establishes a salt bridge with a lysine side chain. The glucose residue is very efficiently docked between two tyrosine aromatic rings. The complex between M. amurensis hemagglutinin and a disialylated tetrasaccharide could be modeled from the leukoagglutinin/sialyllactose crystal structure. The substitution of one tyrosine by an alanine residue is responsible for the difference in fine specificity between the two isolectins. Comparison with other legume lectins indicates that oligosaccharide specificity within this family is achieved by the recycling of structural loops in different combinations.