Design, Synthesis and Thermodynamic Studies of Galectin Ligands

Abstract: The signaling within and between cells in biology is governed by molecular recognition between natural or synthetic ligands and proteins. This thesis project aimed to investigate the thermodynamic properties of specific interaction between synthetic ligands and galectin proteins. The choice of galectins as model proteins resides in the availability of applicability in a full range of complementary experimental and theoretical methods, such as X-ray crystallography and neutron diffraction, NMR, ITC and computational simulations accompanied by design and synthesis of ligands and a rapid method to establish binding specificity and affinity (competitive fluorescent polarization assay). Furthermore, galectins are key biological mediators in numerousbiological activities playing important roles in inflammation, immunity and cancer progression, which also makes galectins interesting as a pharmaceutical target. The thermodynamic properties of binding are sensitive to minute changes in the ligandscaffold, hence we identified small modifications in the ligand design and analyzed effects of these through a multidisciplinary approach. We analyzed the importance of hydrogen bond contribution by comparing a pair of diastereomers with galectin-3Cderived by the removal of non-interacting atoms from a known potent ligand. Solvation effect and conformational entropy favored the diastereomer S over R in complex with galectin-3, even though both ligands showed similar free energy of binding.However, the affinity of the S diastereomer was higher than that of R when one of the galectin-3 arginine residues that hydrogen bound to the ligand hydroxyl group was mutated. A similar effect was observed for the closely related galectin-1: affinity for the S was higher than for R diastereomer. The same hydrogen bond between the ligand and galectin-3 was weaker when the diastereomeric hydroxylated ligands were oxidized to the corresponding ketone. Furthermore, we investigated if the ligandbinding to galectins can be achieved is only parts of the galactopyranose ring of known high-affinity ligand are kept and noninteracting atoms are removed. Indeed, replacing the galactopyranose ring with an acyclic threitol moiety with a-D-galactosemimicking stereoconfiguration and combining it with an affinity-enhancing trifluorophenyltriazole moiety reached a remarkably good affinity towards galectin-1 and 3 terminal domains. This evidenced that ligand binding to the galactoside-recognizing galectin protein family does not need the D-galactopyranose per se, only ideally positioned ligand key hydrogen-bonding groups.Finally, we turned our attention from hydrogen bonds to the close related halogen bond. The change of single ligand halogen resulted in significant differences in the thermodynamic characteristics of galectin-3 binding. The halogen atom strongly affected the solvation and the chemical shifts of the residuals involved in the binding pocket. Furthermore, the enthalpy of binding correlated with the halogen σ-hole size and showed and increasing contribution to the free energy of binding going from fluorine to iodine. However, the steric restraints imposed by the larger iodine atom resulted in a decrease of entropy, thus counterbalancing the strong enthalpy contribution by iodine. Altogether, our in-depth thermodynamic studies on ligandgalectin molecular interactions have advanced our fundamental understanding of ligand and protein dynamics, effects of hydrogen bond geometries and of ligand-protein halogen bonds. We expect that this knowledge advancement will be of interestand value for drug design not only against galectin proteins, but also for drug design in general.