Investigations on lin-Benzopurines With Respect to Dissociation Behavior, Pocket Cross-Talk, Targeting Resistance Mutants, Residual Mobility, and Scaffold Optimization
The present thesis deals with the characterization and improvement of selective antibiotics targeting the enzyme tRNA–guanine transglycosylase. Recently, the lin-benzopurine scaffold was introduced as promising starting point for the structure-based design of TGT inhibitors. Two classes, namely the...
|PDF Full Text
No Tags, Be the first to tag this record!
|The present thesis deals with the characterization and improvement of selective antibiotics targeting the enzyme tRNA–guanine transglycosylase.
Recently, the lin-benzopurine scaffold was introduced as promising starting point for the structure-based design of TGT inhibitors. Two classes, namely the lin-benzoguanines and lin-benzohypoxanthines, were found to inhibit the target enzyme in the nanomolar range. However, the analyzed molecules did not show ideal drug metabolism and pharmacokinetic features since firstly they showed poor permeation through cell membranes and secondly their attached 2-substituents were poorly defined in the difference electron density in crystal structures complicating the establishment of a structure–activity relationship.
In a comprehensive study the protonation inventory of lin-benzopurines is addressed. Initial ITC measurements performed with lin-benzoguanine-type ligands suggested the uptake of one proton by the ligand. This protonation takes place at the basic guanidine moiety of the aminopyrimidinone structure of the scaffold. The lin-benzohypoxanthines do not show this behavior. pKa Calculations support the observations. While the lin-benzohypoxanthines bind to a TGT conformation closely similar to the apo enzyme interacting with only one aspartate within the G34 recognition site, addition of the exocyclic amino functional group in case of the lin-benzoguanines induces the rotation of a second aspartate towards the binding pocket. The negatively charged environment of both aspartates in short distance provokes a pKa shift in case of the lin-benzoguanines strong enough to induce the uptake of a proton. The hypothesis is proofed by studies using site-directed mutagenesis.
Considering the binding affinities across the series of lin-benzopurines, a rather flat structure–activity relationship is observed. Therefore, additional insight into the driving forces of binding was gained by factorizing the free binding energy into enthalpy–entropy contribution. As expected, bindings of both series were found to be enthalpy-driven. Thereby, the lin-benzohypoxanthines exhibit a less pronounced enthalpic term due to their missing interaction to Asp102, which can be partly compensated by a crystallographically conserved water cluster located at the bottom of the G34 recognition site. While the thermodynamic profiles of the lin-benzohypoxanthines remain nearly unchanged, data for the lin-benzoguanines are found to be quite diverse. Obviously, the structural changes triggered by Asp102 in case of the lin-benzoguanines enable a cross-talk between U33 subpocket addressed by the 2-substituent and the G34 recognition site occupied by the parent scaffold.
Based on elevated temperature factors, a high flexibility of the 2-substituent of the lin-benzoguanines was assumed. Therefore, we tested whether a binder with high residual mobility can avoid a loss in binding affinity in case of resistant mutations compared to a binder adopting one ordered binding mode. After identification of an appropriate mutation site, different mutants were expressed and crystallized. The derived binding affinities of various 2-amino-lin-benzoguanines could be related to the binding of the parent scaffold inducing disorder of the protein in proximal distance to the mutation site rather than to the different 2-substituents. Only marginal differences could be ascribed to the properties of the substitution pattern, most likely due to electrostatic attractions and repulsions, respectively.
In further experiments MD-simulations were used to predict the binding modes of extended 2-amino-lin-benzoguanines. Subsequent crystal structure analyses unravelled novel aspects important for the further design and characterization of TGT inhibitors: Firstly, we were able to spot the 2-subsituent of the extended 2-amino-lin-benzoguanines, which bind to the ribose-32 subpocket that has never been occupied before. Secondly, the results emphasize the importance of the applied crystallization conditions. Poorly defined electron density does not always indicate ligands or substituents exhibiting high mobility in the protein-bound state. The applied crystallization protocol takes a major impact on the derived difference electron density and has to be considered in the discussion of the obtained structures.
The class of 5-azacytosines was investigated as a novel scaffold to inhibit TGT. Similarly as the lin-benzopurines, also the 5-azacytosines are deduced from the natural substrate guanine and establish similar binding features, however, the key interaction to Asp102 that was proven to be of utmost importance for substrate recognition, ligand protonation, and pocket crosstalk is poorly established. Different from all known TGT inhibitors, they do not bind to the protein in a planar fashion. In consequence, the compared to lin-benzoguanines low pKa cannot be shifted into a window appropriate for ligand protonation and binding affinity drops.