Discovery of Structure-Activity and Structure-Kinetic Relationships by in silico Methods for PDE Inhibitors Optimization

In silico methods have played an increasingly determinant role in the development of therapeutically relevant molecules. Optimization of a hit to a lead molecule requires improvements in binding affinity, kinetic profile, selectivity, and pharmacological properties. Computational techniques enable us to disclose and understand important structure-property relationships. The primary aim of this thesis is the evaluation, validation, and application of computational methods to predict multiple compound properties and understand structure-property relationships (e.g., structure-activity and structure-kinetic relationships) that could be used for the ligand optimization process. The applied methods were used to investigate Cyclic nucleotide phosphodiesterases (PDEs) ligands. In particular, this thesis focuses on the optimization of inhibitors for a parasite PDE, the Trypanosoma brucei PDEB1 (TbrPDEB1), which is genetically validated as potential drug targets for the treatment of Human African Trypanosomiasis (HAT). TbrPDEB1 is not individually essential for parasite survival, but previous studies showed that its knockdown using RNAi leads to improper cell division and, therefore, cell death. The 3D structure of the catalytic domain of TbrPDEB1 is highly similar to the one of the human enzymes, i.e., PDE4 (hPDE4) which is considered an off-target in these studies. Therefore, next to binding affinity and kinetic profile, selectivity has been a major focus of interest in this thesis. This thesis demonstrates how the computational structure-based analysis, in combination with experimental data, can be used to gain key insights for designing new optimized ligands.