Targeting Trypanosoma brucei FPPS by Fragment-based drug discovery
Trypanosoma brucei (T. brucei) is the causative agent of the Human African Trypanosomiasis (HAT), which is a neglected disease with an endemic occurrence in 36 sub-Saharan African countries. The current standard of care suffers from low efficacy and severe side effects. Therefore, new drugs with bet...
|PDF Full Text
No Tags, Be the first to tag this record!
|Trypanosoma brucei (T. brucei) is the causative agent of the Human African Trypanosomiasis (HAT), which is a neglected disease with an endemic occurrence in 36 sub-Saharan African countries. The current standard of care suffers from low efficacy and severe side effects. Therefore, new drugs with better safety and efficacy profiles are urgently needed. Nitrogen-containing bisphosphonates, a current treatment for bone diseases, have been shown to block the growth of the T. brucei parasites by inhibiting farnesyl pyrophosphate synthase (FPPS); however, due to their particular pharmacokinetic properties they are not well suited for parasitic therapy. Recently, an additional allosteric site was discovered at the surface of human FPPS that, based on sequence analysis, is likely also present in T. brucei FPPS. The high unmet medical need combined with the discovery of a potential new target site prompted a fragment-based drug discovery approach to identify non-bisphosphonate binders on T. brucei FPPS, which is presented in this work.
Fragment screening was performed by NMR and X-ray crystallography. To this end, a robust T. brucei FPPS crystallization system was established enabling high-throughput determination of crystal structures up to 1.67 Å resolution. Structural superimpositions revealed that the allosteric site found on human FPPS is in fact present in T. brucei FPPS. This observation enabled subsequent protein-observed NMR and crystal soaking experiments with established human FPPS binders resulting in three protein-ligand complex structures with bound fragments in the previously unknown allosteric site. For most of the tested binders, Kd by SPR was outside of experimental range for T. brucei FPPS and only for one fragment the Kd on T. brucei FPPS was determined three orders of magnitude higher than the IC50 value on human FPPS. Crystal structural analysis revealed a different binding mode on human and T. brucei FPPS with reduced protein-ligand interactions on T. brucei FPPS, which explains the significantly reduced binding affinity.
Encouraged by the detection of first allosteric binders on T. brucei FPPS, fragment pools were screened by ligand-observed NMR and identified hits were followed-up by single compound ligand observed NMR and protein-observed NMR resulting in 25 validated fragment hits for T. brucei FPPS. Validated hits were followed-up by crystal soaking and co-crystallization experiments and seven protein-ligand complex structures were solved using PanDDA. Out of the seven fragments, four fragments were bound in the active site, one fragment was detected in the allosteric site that was identified as part of this thesis, and two fragments were bound in surface exposed binding sites. Notably, an active site bound fragment with a four atom long flexible linker adopted an orthogonal binding mode along αD when compared to the other three ligands. Sixteen fragment analogues of the elongated flexible active site fragment were tested by SAR using additional test compounds retrieved from catalogue and archive, and one crystal structure with a fragment analogue was solved and was surprisingly found in the allosteric site.
In addition to the NMR fragment screen, an X-ray screen was performed at XChem (Diamond, UK) and at EMBL/ESRF (Grenoble, FR) resulting in seven protein-ligand structures. One fragment was positioned in the active site, three fragments in the allosteric site, two fragments in a cryptic site between helices αI and αH and one fragment at the opposite side of the allosteric site close to αG and αF. Fragment binding was further validated in protein-observed NMR.
As fragments identified by such screening approaches typically exhibit low binding affinities usually in µM to mM range, structure-based fragment optimisation based on a fragment merging and growing approach was performed. In total, ten compounds were synthesised and subjected to protein observed NMR and X-ray structural analysis. Strikingly, a fragment merger based on T. brucei and T. cruzi active-site binders bound in a new binding site close to the SARM instead to the active site.
Taken together, this work presents high-resolution structures of T. brucei FPPS and identified 19 compounds binding to seven different sites thereby paving the way for future studies aiming to identify high-affinity non-bisphosphonate inhibitors for T. brucei FPPS with pharmacokinetic properties that are suitable for parasitic indications.