Glycoproteins

Carbohydrate recognition is one of the most sophisticated recognition processes in biological
systems, mediating many important aspects of cell-cell recognition, such as inflammation, cell
differentiation, and metastasis. Consequently, lectin-glycan interactions have been intensively
studied in order to mimic their actions for potential bioanalytical and biomedical applications.
Galectins, a class of ß-galactoside-specific animal lectins, have been strongly implicated in
inflammation and cancer. Galectin-3 is involved in carbohydrate-mediated metastatic cell
heterotypic and homotypic adhesion via interaction with Thomsen-Friedenreich (TF) antigen on
cancer-associated MUC1. However, the precise mechanism by which galectin-3 recognizes TF
antigen is poorly understood. Our thermodynamic studies have shown that the presentation of the
carbohydrate ligand by MUC1-based peptide scaffolds can have a major impact on recognition,
and may facilitate the design of more potent and specific galectin-3 inhibitors that can be used as
novel chemical tools in dissecting the precise role of galectin-3 in cancer and inflammatory
diseases. Another lectin, odorranalectin (OL), has been recently identified from Odorrana grahami
skin secretions as the smallest cyclic peptide lectin, has a particular selectivity for L-fucose and
very low toxicity and immunogenicity, rendering OL an excellent candidate for drug delivery to
targeted sites, such as: (1) tumor-associated fucosylated antigens implicated in the pathogenesis
of several cancers, for overcoming the nonspecificity of most anticancer agents; (2) the olfactory epithelium of nasal mucosa for enhanced delivery of peptide-based drugs to the brain.

Follistatin (FS) proteins are highly conserved inhibitors of Activins, members of the Transforming Growth Factor beta (TGF-beta) family, which play prominent roles in patterning and cell proliferation, and can contribute to tumor formation. Comparison of FS from Drosophila (dFS) and humans (hFS) in flies shows that hFS is less active. The goal of this thesis is to test three possible mechanisms: dFS might be more stable and turn over at a lower rate, exhibit a stronger affinity for ligands, or diffuse less because of stronger interaction with the extracellular matrix. We generated chimeric proteins of dFS and hFS by exchanging individual protein domains. Our results suggest that the increased activity is likely due to ligand binding. Based on the recent structure of the hFS-Activin complex, we speculate that stronger interactions with heparin sulfate in the extracellular matrix may also contribute to the increased activity of dFS.