Multiple subfamilies of carbohydrate recognition domains in animal lectins.

Calcium ion-dependent carbohydrate recognition domains (CRDs) are found in a range of proteins including receptors for serum glycoproteins and proteoglycans of the extracellular matrix. These C-type CRDs have homologous amino acid sequences characterized by the presence of certain invariant residues. Analysis of the genes for five of the proteins reveals that in each case the CRD-coding sequence is separated from the rest of the gene by an intron. The genes fall into two groups: those in which the coding sequence for the CRD is interrupted by two introns, and those in which the coding sequence is contained in a single exon. The sequences of domains in each category are consistent with the suggestion that the different gene structures reflect early evolutionary divergence of two subfamilies of C-type CRDs in animal lectins. However, carbohydrate-binding specificity does not directly parallel the evolutionary categorization. Comparison of the primary structures of CRDs in each subfamily which have related binding specificities may help to identify residues involved in ligating carbohydrates. This type of analysis is being extended by the use of bacterial expression systems to investigate in greater detail the binding characteristics of the CRDs.