No abstract

Both the dendritic cell receptor DC-SIGN and the closely related endothelial cell receptor DC-SIGNR bind human immunodeficiency virus and enhance infection. However, biochemical and structural comparison of these receptors now reveals that they have very different physiological functions. By screening an extensive glycan array, we demonstrated that DC-SIGN and DC-SIGNR have distinct ligand-binding properties. Our structural and mutagenesis data explain how both receptors bind high-mannose oligosaccharides on enveloped viruses and why only DC-SIGN binds blood group antigens, including those present on microorganisms. DC-SIGN mediates endocytosis, trafficking as a recycling receptor and releasing ligand at endosomal pH, whereas DC-SIGNR does not release ligand at low pH or mediate endocytosis. Thus, whereas DC-SIGN has dual ligand-binding properties and functions both in adhesion and in endocytosis of pathogens, DC-SIGNR binds a restricted set of ligands and has only the properties of an adhesion receptor.

C-type (Ca(2+)-dependent) animal lectins such as mannose-binding proteins mediate many cell-surface carbohydrate-recognition events. The crystal structure at 1.7 A resolution of the carbohydrate-recognition domain of rat mannose-binding protein complexed with an oligomannose asparaginyl-oligosaccharide reveals that Ca2+ forms coordination bonds with the carbohydrate ligand. Carbohydrate specificity is determined by a network of coordination and hydrogen bonds that stabilizes the ternary complex of protein, Ca2+ and sugar. Two branches of the oligosaccharide crosslink neighbouring carbohydrate-recognition domains in the crystal, enabling multivalent binding to a single oligosaccharide chain to be visualized directly.

Lectins are responsible for cell surface sugar recognition in bacteria, animals, and plants. Examples include bacterial toxins; animal receptors that mediate cell-cell interactions, uptake of glycoconjugates, and pathogen neutralization; and plant toxins and mitogens. The structural basis for selective sugar recognition by members of all of these groups has been investigated by x-ray crystallography. Mechanisms for sugar recognition have evolved independently in diverse protein structural frameworks, but share some key features. Relatively low affinity binding sites for monosaccharides are formed at shallow indentations on protein surfaces. Selectivity is achieved through a combination of hydrogen bonding to the sugar hydroxyl groups with van der Waals packing, often including packing of a hydrophobic sugar face against aromatic amino acid side chains. Higher selectivity of binding is achieved by extending binding sites through additional direct and water-mediated contacts between oligosaccharides and the protein surface. Dramatically increased affinity for oligosaccharides results from clustering of simple binding sites in oligomers of the lectin polypeptides. The geometry of such oligomers helps to establish the ability of the lectins to distinguish surface arrays of polysaccharides in some instances and to crosslink glycoconjugates in others.

Protein-carbohydrate interactions serve multiple functions in the immune system. Many animal lectins (sugar-binding proteins) mediate both pathogen recognition and cell-cell interactions using structurally related Ca(2+)-dependent carbohydrate-recognition domains (C-type CRDs). Pathogen recognition by soluble collections such as serum mannose-binding protein and pulmonary surfactant proteins, and also the macrophage cell-surface mannose receptor, is effected by binding of terminal monosaccharide residues characteristic of bacterial and fungal cell surfaces. The broad selectivity of the monosaccharide-binding site and the geometrical arrangement of multiple CRDs in the intact lectins explains the ability of the proteins to mediate discrimination between self and non-self. In contrast, the much narrower binding specificity of selectin cell adhesion molecules results from an extended binding site within a single CRD. Other proteins, particularly receptors on the surface of natural killer cells, contain C-type lectin-like domains (CTLDs) that are evolutionarily divergent from the C-type lectins and which would be predicted to function through different mechanisms.