Synthesis and evaluation of glycomimetics : tool compounds binding to the FimH adhesin for analytical applications and new antagonists of the PapG-II adhesin

The spreading of bacterial resistance is promoting global research efforts toward the development of new therapeutic alternatives. Antivirulence therapy seems to be a valid, new avenue for discovering innovative medicines. In this context, anti-adhesive drugs, which block the first step of bacterial colonization of the host’s tissues are particularly attractive, as they do not kill the pathogens, and thus do not contribute to the selection of resistant strains.
Urinary tract infections (UTIs) are among the most frequent reasons for antibiotic intake, thus playing a pivotal role in spreading bacterial resistance. Moreover, their recurrent nature reduces consistently patient’s quality of life. As the most common pathogen involved in UTIs is E. coli (about 80% in otherwise healthy patients), an anti-adhesive therapy against it would be highly valuable. E. coli uses filamentous structures called pili to adhere to the host’s tissues. In UTIs concerning the lower urinary tract (cystitis), type 1 pili are mainly involved. At the tip of type 1 pili, the lectin FimH is expressed, which recognizes mannosylated glycoproteins, abundant in the urinary bladder. A large body of literature is dedicated to antagonizing FimH. Despite nanomolar antagonists have been long discovered, suitable clinical candidates are lacking. One important determinant for a successful drug is the target occupancy time. Using surface plasmon resonance, we demonstrated that our lead structures have excellent kinetic profiles, when tested against the FimH lectin domain (paper 1). However, one crucial limiting factor was the poor pharmacokinetic profile of these antagonists. We therefore successfully tailored the physicochemical properties of a set of promising lead structures (paper 2).
In order to support therapy, a detection system for FimH-expressing E. coli is of great importance. Biosensors offer several advantages, including reliability, low cost, and ease of use. Using a FimH antagonist as recognition element and FimH as analyte, we developed a FimH sensitive biosensor, providing the first proof of concept of label-free detection of a pathologically relevant protein, by field-effect, silicon nanoribbons-based sensors (SiNR-BioFET, paper 4).
However, most research efforts have until very recently focused on the isolated lectin domain of FimH, which exists in a high-affinity state. To finally clarify if the high-affinity state is the appropriate therapeutic target, a study based on crystallography, molecular dynamics, and kinetics was undertaken on the full-length FimH protein, which exists prevalently in a low-affinity state (paper 3). The results support the use of the full-length protein as the most appropriate model for anti-adhesive therapy, thus opening a completely new research path for medicinal chemistry studies.
In UTIs involving the human upper urinary tract (pyelonephritis), E. coli type P pili have been shown to play an important role. The adhesive properties of these pili arise from the PapG-II adhesin, which recognizes the tetrasaccharide epitope of tetraosyl galactosyl globosides (GbO4). Although the incidence of upper UTIs as compared to cystitis is rather low, the risk of serious organ damage is high. Moreover, the increasing frequency of resistant strains requires new therapeutic alternatives. Medicinal chemistry has so far focused on the modification of the minimal binding epitope, i.e. Galα(14)Gal. However, the best published lead compound exhibits affinity only in the mid-micromolar range. Based on a critical analysis of the present literature on antagonists of PapG-II and of the closely related PapG-I, a new, not yet explored sub-binding site was identified and explored. Disappointingly, no improvement in affinity could be achieved (chapter 3.2.2), confirming the challenging nature of the target.
Fragment-based approaches have been shown to have a great potential for hard-to-drug targets. In our group, second-site ligand search using fragments had been successfully applied on other lectin targets. The same strategy was applied to PapG-II, albeit with scarce success (manuscript 1).
The observation that the hexasaccharide epitope of the sialosyl galactosyl globoside shows 5-fold increased affinity for PapG-II as compared to the epitope of GbO4, from which it differs by an added disaccharid units at the non-reducing end, led us to study the details of the interaction. Crystallographic and thermodynamic investigations suggested that the improvement in affinity arises from an entropic contribution, due to the non-binding, terminal saccharidic units (paper 5).
The data collected during the development of this thesis added important information on PapG-II and will assist further medicinal chemistry research toward the development of high-affinity antagonists.