This project studies the interaction between galectin-1, a soluble galactose-binding lectin, and a variety of N-acetyllactosamine-containing oligosaccharides. A modified form of Chinese hamster ovary (CHO) cell galectin-1 has been generated and expressed in large quantities in E. coli. The initial goal was to attempt to assign all of the 1H resonances of the protein and to obtain binding constants for the ligands from a study of chemical shift and line width changes. Inter-proton distance information obtained from nuclear Overhauser effect (NOE) measurements will be used to try to identify contact areas between galectin-1 and its ligand. Another goal of this project is to generate isotope-enriched forms of the lectin for detailed NMR studies. To date, approximately 10 mg of [15N]-galectin-1 and 10 mg of [13C,15N]-doubly labeled galectin-1 have been isolated, and 1H-15N correlated spectra have been obtained to test the feasibility of using this sample to assign the 1H and 15N resonances. With the successful preparation of [13C,15N]-doubly isotope-labeled galectin-1, work began on optimizing 3D triple-resonance experiments for proton, carbon, and nitrogen assignments. Due to the high molecular weight of the homodimer, versions of the standard experiments (e.g., HNCO, HNCA, CBCACONH, HNCACB) were adopted that minimize signal loss due to relaxation. In addition, the buffer conditions were varied in order to promote solubilization and reduce aggregation. An alternate form of the galectin-1 protein has also been obtained by generating a mutant galectin-1 protein that at high concentrations forms an active monomer. Since NMR protein structure studies require protein concentrations in the millimolar range, we have generated mutations in order to form a monomer that is stable at high concentrations. The isolated monomer lectin was found to be both inactive and abnormally folded. The lectin did not agglutinate erythrocytes, it bound very poorly to immobilized asialofetuin, and it did not compete with agglutination by the dimeric C2S lectin. In addition, the protein displayed an altered circular dichroism spectrum indicating that the protein was assuming an alternate folding pattern. These data indicate that the nature of the truncation mutations that were introduced into the molecule were too severe to maintain folding of the binding domain of the monomeric lectin. Polylactosamine ligands for the lectin have been generated, including the complete series of lacto-N-neotetraose and lactosamine oligomers with up to eight sugars. NMR techniques applied to the galectin-oligosaccharide mixtures include line width analysis, magnetization transfer, and transferred NOE studies. For the ligands analyzed so far, the data are compatible with predominant binding to the terminal lactosamine unit. When no terminal galactosyl is available, the oligosaccharide slides into the binding groove such that a galactosyl residue can occupy the main binding site. It is not yet known how far into the binding groove longer oligosaccharides may extend. Differentiation between identical residues and determination of the location on the polylactosamine chain where interactions are occurring requires specific labeling of targeted residues. This work was begun by synthesizing a tetra-N-acetyllactosamine containing a terminal [1-13C]-galactosyl residue. The next synthetic step is to extend the length of the polylactosamine structure and generate a series of oligosaccharides, each with a uniquely labeled galactosyl residue at each position within the polylactosamine chain. We can then confirm whether the galectin always prefers the terminal residues when presented with a longer repeating poly-N-acetyllactosamine. A paper has been submitted for publication.
Showing the most recent 10 out of 245 publications
