Abstract
Thermodynamic analysis of carbohydrate binding by Artocarpus integrifolia (jackfruit) agglutinin (jacalin) shows that, among monosaccharides, Me.alpha.GalNAc (methyl-.alpha.-N-acetylgalactosamine) is the strongest binding ligand. Despite its strong affinity for Me.alpha.GalNAc and Me.alpha.Gal, the lectin binds very poorly when Gal and GalNAc are in .alpha.-linkage with other sugars such as in A- and B-blood-group trisaccharides, Gal-.alpha.-3Gal and Gal.alpha.1-4Gal. These binding properties are explained by considering the thermodynamic parameters in conjunction with the minimum energy conformations of these sugars. It binds to Gal-.beta.-3GalNAc.alpha.Me with 2800-fold stronger affinity over Gal.beta.1-3GalNAc.beta.Me. It does not bind to asialo-GM1 (monosialoganglioside) oligosaccharide. Moreover, it binds to Gal.beta.1-3GalNAc.alpha.Ser, the authentic T (Thomsen-Friedenreich)-antigen, with about 2.5-fold greater affinity as compared with Gal.beta.1-3GalNAc. Asialoglycophorin A was found to be about 169,333 times stronger an inhibitor than Gal.beta.1-3GalNAc.alpha.Ser. The present study thus reveals the exquisite specificity of A. integrifolia lectin for the T-antigen. Appreciable binding of disaccharides Glc.beta.1-3GalNAc and GlcNAc.beta.1-3Gal and the very poor binding of .beta.-linked disaccharides, which instead of Gal and GalNAc contain other sugars at the reducing end, underscore the important contribution made by Gal and GalNAc at the reducing end for recognition by the lectin. The ligand-structure-dependent alterations of the c.d. spectrum in the tertiary structural region of the protein allows the placement of various sugar units in the combining region of the lectin. These studies suggest that the primary subsite (subsite A) can accommodate only Gal or GalNAc or .alpha.-linked Gal or GalNAc, whereas the secondary subsite (subsite B) can associate either with GalNAc.beta.Me or Gal.beta.Me. Considering these factors a likely arrangement for various disaccharides in the binding site of the lectin is proposed. Its exquisite specificity for the authentic T-antigen, Gal.beta.1-3GalNAc.alpha.Ser, together with its virtual non-binding to A- and B-blood-group antigens, Gal.beta.1-3GalNAc.beta.Me and asialo-GM1 should make A. integrifolia lectin a valuable probe for monitoring the expression of T-antigen on cell surfaces.