Synthesis of glycomimetics of the natural Siglec-8 ligand and their pharmacodynamic characterization

Glycans are an important class of natural products, since all living cells are covered by different types of carbohydrates forming the glycocalyx. Therefore, any cell-cell recognition process is based on the ability of one cell to recognize the specific glycan pattern which is presented on the other cell surface. This is particularly important for the activation of the immune system, which needs to selectively recognize and kill invading pathogens while not damaging self cells. However, sometimes this delicate equilibrium shifts towards unwanted pathological conditions where the immune system is not able anymore to distinguish between self and non-self cells, causing autoimmune diseases. Additionally, undesired overexpression of immune cells can also lead to medical conditions, like allergic inflammation. Therefore, it is fundamental to rely on specific mechanisms to control immune system homeostasis and avoid abnormal up-regulation of these cells. For this reason, immune cells display inhibitory receptors which suppress their activation via different mechanisms, such as inducing cell death.
One type of inhibitory proteins expressed on immune cells are siglecs, a class of I-type transmembrane proteins which selectively recognize sialic acid-containing structures. Siglecs possess multifarious functions, according to the structure and cell type expressing them, but the majority of them acts as inhibitory receptor thanks to the presence of ITIM and ITIM-like motifs in their intracellular domain. One member of this family is Siglec-8, which is expressed only on eosinophils, mast cells, and to some extent on basophils. It exerts its inhibitory function by promoting eosinophils apoptosis and inhibition of mast cells degranulation, representing therefore an interesting pharmacological target for the treatment of diseases in which these cell types are overexpressed, among which asthma and allergic inflammations.
This thesis describes a medicinal chemistry approach for the development of high-affinity Siglec-8 ligands to identify potential drug candidates for the treatment of eosinophil- and mast cell-associated diseases. This drug discovery process started from the known structure of the preferred natural Siglec-8 ligand identified via a glycan array approach, the tetrasaccharide 6’-sulfo-sialyl Lewisx (6’-sulfo-sLex). The complex chemical structure of this compound neither allows an easy and straightforward synthesis of derivatives, nor do its chemical properties meet the requirements for drug-like molecules.
First, the minimal binding epitope of 6’-sulfo-sLex was identified in the corresponding Neu5Ac-α2-3-Gal6S disaccharide. Although it has a simplified chemical structure, the affinity was reduced only by a factor 2 (Paper 1). In-depth analysis of the pharmacophores, with the aid of computational studies, allowed the development of a glycomimetic analogue, where the galactose moiety was replaced by a carbocyclic structure. Affinity was improved thanks to a reduced desolvation penalty by removing non-interacting hydroxyl groups. When finally the C-9 hydroxyl group of the sialic acid was transferred into an aromatic sulfonamide, affinity was further improved almost 20-fold.
Generally, carbohydrate-binding proteins are characterized by weak affinities in the low millimolar range towards their ligands due to shallow and solvent-exposed binding sites. One way to deal with this problem is to design multivalent structures to achieve huge affinity improvements. Even in the case of a protein with a single binding site like Siglec-8, multivalency can enhance affinity through statistical rebinding, which relies on the increased local concentration of binding moiety. Thus, once the bound ligand dissociates, another one is in close proximity and immediately replaces it. To study this statistical rebinding effect in Siglec-8, different oligovalent ligands were synthesised and tested (Manuscript 1). Thermodynamic analysis showed that each binding epitope contributed enthalpically to the same extent as in the monovalent interaction, while entropic penalties were observed. Dissection of the entropic term in its components for some compounds revealed that greatly favourable solvation entropies were counterbalanced by bigger conformation entropy penalties. Additionally, to take the multivalent presentation concept to its extreme, a glycopolymer was synthesized and tested in cell-based bioassays to demonstrate the binding to Siglec-8. On the other hand, the biological outcome obtained with the best monovalent sulphonamide glycomimetic was much smaller. However, this assay proved the potential of the developed glycomimetic structures, and provided additional proof of the potential pharmacological applications that compounds with improved affinity could have.
To further improve the affinity of our glycomimetic Siglec-8 ligands, different strategies were explored (Manuscript 2). Firstly, bioisosteres of the carboxylic acid and sulfate group led to a complete loss of activity towards Siglec-8, confirming the crucial role for these two functionalities. Secondly, the previously unexplored C-4 position of the sialic acid was modified leading to amide derivatives with improved affinity. Finally, second generation sulphonamides were synthesized. While changing the electronic properties of the naphthalene moiety did not influence affinity, the extended aromatic system proved to be more important for having high-affinity compounds.