The Role of Water in Protein-Ligand Binding: A Comprehensive Study by Crystallography and Isothermal Titration Calorimetry
The aim of this work is to investigate the impact of desolvation effects on protein-ligand interactions. In all complex structures with thrombin and pyridine, it is evident that preserving the original solvation state of Asp189 is a crucial and a common feature upon binding of the pyridine inhibitor...
|Online Access:||PDF Full Text|
No Tags, Be the first to tag this record!
|Summary:||The aim of this work is to investigate the impact of desolvation effects on protein-ligand interactions. In all complex structures with thrombin and pyridine, it is evident that preserving the original solvation state of Asp189 is a crucial and a common feature upon binding of the pyridine inhibitors. However, the associated entropic losses are immense. In two ligand complexes even disordered ligand portions are found in the S1 pocket, which evade full desolvation of Asp189 compared with the apo form of thrombin. The price for the desolvation of a charged amino acid is simply too large to ensure in this case a complete displacement of all waters. The determined complex structures reveal that the charged methylpyridinium derivatives do not optimally address the negatively charged Asp189 at the bottom of the S1 pocket. A short distance to the deprotonated Asp189 cannot be achieved either due to steric reasons or the bulky methyl group provides a good protection to interact in a proper way with the negatively charged Asp189. The optimal interaction geometry to Asp189 cannot be realized in this series. Therefore, the energy released from the suboptimal interaction between methylpyridinium and Asp189 is not high enough to compensate for the large desolvation price required for the charged ligands. Additionally, water effects have been studied in hydrophobic interactions in thrombin and thermolysin. The thermodynamic characterization shows a hydrophobic effect in thrombin which is clearly entropically driven. In the study, the S3/4 pocket has been gradually desolvated using increasing hydrophobic modifications in P3. In both series, the binding affinity improved by about 40-fold. The binding affinity has been optimized hydrophobically from nanomolar to low picomolar affinity. The benzamidine derivatives are even characterized by a binding mode showing two ligands to be bound simultaneously. Surprisingly, the additionally bound ligand traces remarkable well the recognition area that accommodates fibrinopeptide A (cleavage product of fibrinogen). In contrast, the examined S2' pocket of thermolysin is less well shaped, but ideally solvated because of its exposure to the protein surface. The ligands differ only by a terminal carboxylate and/or methyl group. A surprising nonadditivity of functional group contributions for the carboxylate and/or methyl groups is detected. Adding first the methyl and then the carboxylate group results in a small Gibbs free energy increase and minor enthalpy/entropy partitioning for the first modification, whereas the second involves strong affinity increase combined with huge enthalpy/entropy changes. Adding however first the carboxylate and then the methyl group yields reverse effects: now the acidic group attachment causes minor effects whereas the added methyl group provokes huge changes. The added COO- groups perturb the local water network in both carboxylated complexes and the attached methyl groups provide favorable interaction sites for water molecules. In all complexes, apart one example, a contiguously connected water network between protein and ligand functional groups is observed. In the complex with the carboxylated ligand, still lacking the terminal methyl group, the water network is unfavorably ruptured. This results in the surprising thermodynamic signature showing only minor affinity increase upon COO- group attachment. Since the further added methyl group provides a favorable interaction site for water, the network can be re-established and strong affinity increase with huge enthalpy/entropy signature is then detected. Addressing the S2' pocket of thermolysin with hydrophobic molecule portions generates also an entropically dominated signal similarly to the thrombin series. The present series of closely related thermolysin complexes shows that both thermodynamic properties are involved and many detailed structural phenomena determine the final signature. If a contiguously connected water network ruptures, an enthalpic loss and entropic gain is experienced. Particularly, in case of accommodation of ligand portions in pockets opening to the bulk solvent and exposing parts of the placed ligand to the water phase, new binding sites for water molecules can be generated, e.g. as in our study at the capping position above the carboxylate group or the site on top of the benzyl ring. Also such phenomena contribute on the molecular level to the finally determined hydrophobic effect. In summary, there are no arguments why the hydrophobic effect should be predominantly entropic or enthalpic. Small structural changes on the molecular level determine whether hydrophobic binding to hydrophobic pockets results in a more enthalpy or entropy-driven signature.|