Monique A. Pond, Rebecca A. Zangmeister



For quality assurance and control purposes there is a need to quantify changes observed in glycan (carbohydrate moiety) structure during the production of monoclonal antibody drug therapies.  The development of analytical methods to evaluate carbohydrate interactions has been difficult in part because of the enormous structural complexity and diversity of carbohydrates.  In addition, carbohydrate-binding proteins generally form weak monovalent carbohydrate interactions with low selectivity.  Therefore, carbohydrate-binding proteins typically possess multiple binding sites which allow them to bind two or more carbohydrate ligands simultaneously.  The multivalent complexes that form are highly selective with an overall strong interaction between the protein and the bound carbohydrates.  To successfully investigate the glycan interactions that form these complexes, tools must be capable of monitoring binding events that vary with the spacing, or density, of carbohydrates in a particular region.  One strategy is to use lectins, proteins with highly specific carbohydrate binding functionalities, for carbohydrate analysis.  Because lectins are ubiquitous in nature and capable of differentiating between similar glycan structures, there is great potential for using them to monitor glycosylation.  However, at present only a subset of lectin-carbohydrate binding interactions have been studied. 

The aim of this project is to develop a tool for rapidly characterizing the unique binding properties of unknown lectins.  Our approach is to use catanionic surfactant vesicles to control the carbohydrate density of glycan arrays.  We prepared vesicles that spontaneously form in water and remain stable at room temperature for months.  By varying the amount of glycoconjugate added during preparation, glycans were incorporated onto the surface of the vesicles in a controlled range of densities.  The vesicles were applied to commercially-available, nitrocellulose-coated glass slides to generate glycan arrays.  As proof of concept, the binding of two lectins, Concanavalin A and peanut agglutinin, to the arrays was quantified using a biotin-avidin fluorescence sandwich assay.  We observed a difference in lectin binding based on the carbohydrate density of the vesicles applied to a particular glycan array surface.  Our method for preparing glycan arrays can be expanded to provide a high-throughput platform for characterizing unknown lectins.  This will lead to the discovery of a diverse lectin library that can be used to detect slight changes in carbohydrate structure during the manufacturing of monoclonal antibody drug products.