Development and synthesis of nanomolar e-selectin antagonists using a fragment-based approach

By mediating the extravasation of leukocytes from the blood stream, selectins are involved in a key step of the inflammatory cascade. The excessive recruitment of leukocytes into inflamed tissue is responsible for the onset and the progression of various inflammatory diseases (e.g. rheumatoid arthritis, reperfusion injuries or asthma). Furthermore, selectins contribute to the development and metastasis of cancer. Given the health impact of these diseases, selectins represent a valuable drug target.
Many research efforts for the development of small-molecule E-selectin antagonists have been based on sialyl Lewisx (sLeX), the minimal carbohydrate epitope recognized by the selectins. From this lead with affinities in the millimolar range, antagonists with more drug-like properties and affinities in the low micromolar range have been developed. However, the selectins have shaped up as difficult drug targets with only relatively few successful applications in vivo and no marketed anti-selectin drug. The comparably low affinities of E-selectin antagonists are an important reason for the lack of clinical success in this field.
In this work, the problem of modest affinity was approached from a new direction using a fragment-based approach, as the possibility of forming high-affinity ligands from low-affinity fragments is one particular advantage of fragment-based drug discovery. Using a known sLeX mimic (first-site ligand) as starting point, an NMR-based screening was performed to identify small fragments (second-site ligands) binding to a proximate second binding site. This led to the identification of 5-nitroindole as a second-site ligand. In situ click experiments based on the Huisgen 1,3-dipolar cycloaddition were performed to screen for a suitable linker to connect the two fragments. As this approach failed, a library of triazole-nitroindole antagonists was synthesized, and a ranking was performed using a specifically designed Biacore experiment.
The detailed investigation of the five most potent ligands identified in the screening revealed potent E-selectin antagonists with affinities ranging from 30 to 89 nM and improved binding kinetics, i.e. prolongued ligand-receptor half-life times in the range of minutes. Derivatives of the most potent antagonist were synthesized providing first insights into structure-activity relationships and a basis for the future development of these antagonists, also with respect to their physicochemical and pharmacokinetic properties.