Design, synthesis and biological evaluation of carbohydrate-mimetics for Siglec-8: a novel target for asthma

Sialic acids constitute a large family of monosaccharides that is characterized by a nine- carbon backbone and predominantly found as terminal moieties on glycans. These glycans are linked to glycoproteins, glycolipids, or proteoglycans located on the outer membrane of cells. The glycan layer on cell surfaces is called glycocalyx. Recognition of these endogenous glycans by their glycan-binding protein counterparts plays a key role in the immune system to discriminate between self and non-self. One important family of glycan-binding proteins involved in distinguishing self from non-self by sialic acid recognition is formed by the siglecs (sialic acid-binding immunoglobulin-type lectins). They are inhibitory receptors predominantly expressed on cells of immune system, which makes them interesting drug targets.
One member of the siglec family, Siglec-8, is exclusively expressed on cell surfaces of eosinophils, mast cells and weakly on basophils. These cell types were shown to play a crucial role in the pathophysiology of asthma. In fact, there is compelling evidence that Siglec-8 plays a key role in the progression of asthma and therefore is an interesting new pharmaceutical target.
This thesis focuses on the design, synthesis and biological evaluation of Siglec-8 ligands with the potential for a novel anti-asthmatic treatment. From literature it is known that Siglec-8 preferentially binds the tetrasaccharide 6’-sulfo-sLex. However, the complex structure of 6’- sulfo-sLex is not a suitable starting point for a medicinal chemistry approach. Therefore, we first determined the minimal binding epitope of Siglec-8 (Chapter 2) and identified the sulfated disaccharide Siaα(2-3)Gal able to bind Siglec-8 with only a minor reduction in affinity compared to the tetrasaccharide 6’-sulfo-sLex. Because a multivalent version of the sulfated disaccharide was found to bind Siglec-8 in a cell-based assay, it provided a suitable
starting point for further lead structure optimization.
To improve the affinity of our lead compound (Chapter 3) we first focused on its desolvation properties. It is well-known that desolvation, e.g. the process to remove water molecules surrounding ligand and protein, allowing formation of the ligand-protein complex, plays a crucial role in the overall binding process. Polar substrates such as carbohydrates therefore often display low affinities for their target proteins, because desolvation is costly and cannot be completely compensated by favorable binding interactions or beneficial entropic contributions. Thus, we hypothesized that by removing polar groups from our lead structure not participating in beneficial electrostatic interactions with Siglec-8, the affinity can be improved, simply by reducing the desolvation penalty. With the aid of molecular modeling, we identified such polar groups and synthesized a library of compounds that was subsequently evaluated in a competitive binding assay. By this approach, we identified a ligand binding in the µM range. 1H,15N-HSQC experiments then confirmed that the binding mode was indeed similar to 6’-sulfo-sLex.
In a second approach, we explored the protein surface in close vicinity of the carbohydrate- binding pocket. For that purpose, a small focused library containing aromatic fragments linked to the amino group in the 5-position of the sialic acid moiety was synthesized. Although the affinities of the ligands were reduced compared to the lead structure, 1H,15N- HSQC experiments with the best binder from this series revealed binding to a new pocket located close to the N-terminus of the protein. Based on molecular modeling studies, a new starting point for lead optimization processes was found. Since this pocket is highly variable
across siglecs, its targeting provides a promising strategy towards the development of selective Siglec-8 ligands.
Finally, in Chapter 3 we describe a fragment-based approach for the identification of Siglec-8 ligands. A fragment library was screened in silico for Siglec-8 binding and the best-ranked hits were further evaluated experimentally, using a differential scanning fluorimetry assay. This assay monitors the thermal unfolding of a protein in presence of a fluorescent dye. As the stability of proteins is typically increased when the protein is complexed to a ligand, a positive shift in melting temperature (Tm) indicates binding. Three hits were identified and their binding further validated with a T1ρ assay, which determines binding based on a difference in spin-spin relaxation in the rotating frame (T1ρ relaxation) of a ligand in the free form, as well as in complex with a protein. However, the results were ambiguous and further biophysical evaluation is necessary to validate binding of this fragment.