IUCr journals

Carbohydrate binding by proteins

A variety of proteins bind carbohydrates in order to perform the functions they are to fulfill. Some act on carbohydrates and alter them chemically, some sense the presence of carbohydrates and trigger the action of other proteins, some bind to carbohydrate in order to locate at particular sites of action, the list continues to expand. The lectins are one such group of proteins. Some bind carbohydrates with specificities that rival antibodies. They can also aggregate glycoproteins and other carbohydrate-containing macromolecules in solution and cause their precipitation (agglutination). A wide variety of plants, including the legumes and the grasses, have lectins that are distributed at varying concentrations throughout the plant, bark, foliage, fruit, seeds. Lectins are some of the earliest proteins studied crystallographically and concanavalin A (con A) and wheat germ agglutinin (WGA) are the two most prominent examples. Ironically, it is still not clear what roles lectins fulfill for the plants they serve.

The origins of affinity and specificity of carbohydrate binding by proteins have been a primary scientific goal from the earliest structural studies. Most of the fundamental factors controlling affinity were described by F. Quiocho and coworkers in structural studies of bacterial periplasmic binding proteins. Structures of carbohydrate complexes with WGA by C. Wright also made important contributions. An important recent contribution to the field is a paper by Ravishankar, Suguna, Surolia and Vijayan, “Structures of the complexes of peanut lectin with methyl-ß-galactose and Nacetyllactosamine and a comparative study of carbohydrate binding in Gal/GalNAc-specific legume lectins” [Acta Cryst. D55 (1999), 1375-1382]. Included for comparison are previously published structures of complexes with lactose and with T antigen from the same laboratory. One important conclusion is that the source of the 20-fold higher affinity of peanut lectin for T antigen over lactose and lactosamine comes not from direct protein-carbohydrate interactions, but rather from hydrogen-bond bridges to T-antigen carbohydrate from protein-bound water molecules, bridges that are not made to carbohydrate in other complexes even though structurally homologous water molecules are present. A detailed analysis of the role of water is presented and it emerges that in many cases key protein-carbohydrate interactions are fulfilled by water surrogates.

The legume lectins have a number of other curious attributes. Some show evidence of circular permutation that is the result of post-translational modification. In order to express their agglutination property, they form dimers and tetramers, but despite their relatively close homology, more than one kind of quaternary structure has been observed. In con A, favin and pea lectin, for example, six-stranded backbone ß-sheets in monomers fuse to form twofold symmetric twelve-stranded sheets in dimers, and tetramers have 222 symmetry. Not so in peanut lectin, GS4 and EcorL. In peanut lectin, six-stranded sheets in monomers are pressed together in dimers so that strands are close to perpendicular, but tetramers show neither 222 nor fourfold symmetry. The lectins have taught us much about protein chemistry over the years, but it appears we still have much to learn from them.

Howard Einspahr
Bristol-Myers Squibb Pharm. Res. Inst., Princeton, NJ USA