Glycan-binding proteins, also referred to as lectins, participate in regulating complex biological processes such as bacterial and viral infection, inflammation, and cancer progression through interaction with glycans displayed in high density on the cell surface. The type and density of presentation of glycan structures, as well as changes in these properties, often provide a readout of the cell state: for example, changes in glycosylation are associated with tumor progression and metastasis. For these reasons, obtaining a complete map of cell glycosylation and correlating glycan composition to physiological and pathologic cell functions represents a major goal in glycoscience, and has the potential to impact several fields. Unfortunately, identification of complex glycan structures, especially when presented at high density on a two-dimensional cell surface, represents a major challenge. Because of the inherent difficulties in studying the glycome, this area of investigation is restricted to relatively small number of highly trained specialists. The development of robust, accessible, inexpensive methods to identify, quantify, and characterize complex glycans will encourage other scientists to investigate the glycome. Here, we address these issues in two related objectives: first, we will integrate computational methods and directed evolution methods to generate a panel of structurally related lectins using the well-studied lectin cyanovirin as a template. The resulting lectin protein library will be assessed for their ability to bind a panel of different glycans. Affinities and structures will be determined to the extent possible. These data will be used in an iterative process to refine our glycan docking, and directed evolution approaches. Our second objective is to increase the multivalency of these protein-carboyhydrates through rational engineering of the glycan binding modules as we have done in the past. This creates a multivalent display optimized for tight interaction with clustered surface-displayed glycans. As a proof of concept, we will initially focus on mannose-rich glycans such as those found on the surface of enveloped viruses. We anticipate obtaining a set of specific reagents for target glycans with wide specifities and profound affinities. Our approach will result in a panel of stable, robust, structurally homogeneous designer lectins specific for glycans of choice, beyond the initial targets described here. Because of these properties, the reagents will be accessible for non-experts users and will have a broad range of applications. Additionally, we will have developed a robust platform that is highly adaptable to other protein scaffolds and glycan targets, which will be pursued in future applications. Finally, through iterative feedback of experimental results, we will advance our current BP-dock methods for gylcans and develop a webserver for glycan docking predictions.
The application of natural lectins and antibodies in glycoscience is currently hindered due to their low binding affinity and poor specificity. We propose to develop a class of rationally designed high affinity, high specificity reagents targeted to glycans by integrating rational design of lectins using computational methods with guided directed evolution. These new reagents will be integrated into available platforms such as lectin microarrays, and address an urgent need in for robust, reliable, and inexpensive tools to identify glycans.
