NMR methodologies to determine the structure of fast exchanging carbohydrate protein complex
MetadataShow full item record
Nuclear magnetic resonance (NMR) spectroscopy has proven to be one of the most important techniques for determining the structure and dynamics of protein-carbohydrate complexes. However, the study of ligand-protein complexes, particularly when the ligands are carbohydrates, is not without its difficulties. Historically, structure determination by NMR relies heavily on the availability of inter- or intramolecular distance constraints from Nuclear Overhauser Effects (NOEs). However, the dominance of hydrogen bonding networks in carbohydrate recognition, and an inability to observe the rapidly exchanging hydrogen-bonding protons makes observation of intermolecular NOEs rare. As a result, this thesis looks to distance-independent residual dipolar couplings (RDC) and long range pseudo contact shifts (PCSs) to constrain the global structure of carbohydrate-protein complexes. It makes application of methods based on these observables to characterize carbohydrate interactions with galectin-3, a protein of considerable interest because of its cell-surface recognition roles and correlation of these roles with a number of human diseases. The affinities of most carbohydrates to galectin-3 are very low. The dissociation constant for lactose, the model carbohydrate used in most studies presented, from galectin-3 is 0.2 mM. In the case of weakly-binding fast-exchanging systems, the application of RDCs is inhibited by the dominant contribution from free-state ligands. In order to accurately extract bound-state RDCs from the observed average, significant enhancement of bound-state RDCs is needed. Novel methods for the enhancement of bound-state RDCs of lactose are presented in chapters 2 and 3 of this thesis. These include increasing the association of galectin-3 with a surrounding liquid crystal medium through a hydrophobic propyl chain and electrostatic interactions between His-tagged galectin-3 and a nickel doped alignment medium. Unfortunately, these methods do not allow the study of hydrophobic ligands. A more universal method (described in chapter 4) relies on the application of paramagnetism-based constraints, including self-oriented RDCs and pseudo-contact shifts. The results clearly show that accurately-measured RDCs and PCSs can position lactose in the galectin-3 carbohydrate binding pocket well enough to provide useful data for the rational design of competitive inhibitors of natural ligands.