Biological characterization of the bacterial Adhesin FimH — implications for drug design

At first glance the pathogenesis of a urinary tract infection (UTI) may seem straightforward, but it involves layers of complexity that extend beyond the simple notion of bacterial binding to and subsequent invasion of urothelial host cells. UTIs are governed by a plethora of various processes where, on one side, the host contrives defense mechanisms to fight against bacterial invasion and on the other side bacteria acquire virulence factors to ensure their survival. In the majority of UTI cases, the bacteria engaged in this struggle belong to Escherichia coli. As a commensal constituent of the intestinal flora, E. coli finds itself outside of its natural habitat in the urinary tract. However, several strains of E. coli have acquired pathogenic traits to allow them not just to survive in the urinary tract, but also to grow and persist in it.
The very first step of a UTI is the adhesion of uropathogenic E. coli (UPEC) to the urothelium. After adhesion, UPEC invade urothelial cells and form intracellular biofilm-like communities, in which they are protected from host defenses and from which they can re-emerge. When UPEC flux out of the intracellular communities, they can either spread out on the luminal surface or they can penetrate into underlying cell layers where they can form quiescent intracellular reservoirs. These intracellular reservoirs are a major driving force of UPEC persistence in the urinary tract. The intracellular component of UTI pathogenesis is predicated on initial adhesion, which is mediated by the bacterial lectin FimH. Located at the tip of type 1 pili, the bacterial adhesin FimH binds to terminal mannose residues on high-mannose N-glycans of urothelial transmembrane glycoproteins. The adhesive properties of FimH and its specificity for α-D-mannose have spurred studies into the development of anti-adhesive glycomimetics, also known as FimH antagonists. The mode of action of FimH antagonists is simple, but elegant in its simplicity. By outcompeting natural mannose glycans on host urothelial cells, antagonists can saturate bacterial FimH and block bacteria from latching onto the urothelial surface. Without this contact, UPEC cannot trigger the pathogenic cascade that ultimately establishes a UTI. Accordingly, as they lack firm adherence to the urothelium, bacteria are simply washed away through micturition. The effect of FimH antagonists is neither bacteriolytic nor bacteriostatic, making it highly likely that this anti-adhesive approach to UTI treatment will not impose selection pressure on UPEC. This, in turn, makes FimH antagonists an attractive alternative to conventional antibiotics, which have been plagued by the emergence of antibiotic resistance.
The development of FimH antagonists has advanced markedly since their inception over four decades ago. High affinity FimH antagonists are now identified routinely and as a consequence, the drug development process has shifted from being purely affinity driven to being a balancing act between pharmacodynamics (PD) and pharmacokinetics (PK). This shift in focus is exemplified in Publication 1 and Publication 5: In Publication 1, bioisosteric iterations on key residues of the canonical biphenyl aglycone scaffold lead to the identification of a FimH antagonist with an optimized PK/PD profile. In a preventive UTI mouse model, this FimH antagonist was able to reduce the bacterial load in the bladder by a factor of 500, even outclassing one of the commonly prescribed antibiotics for UTI, ciprofloxacin. In Publication 5, we further experimented on the biphenyl scaffold to balance two ostensibly diametrical key determinants of pharmacokinetics: solubility and permeability. FimH antagonists were investigated with regards to those pharmacokinetic properties after changes in the substitution pattern of their biphenyl moiety or after the introduction of heterocyclic structures—adding to the existing body of research for this prominent class of FimH antagonists.
In Manuscript 2, the focus is back on binding affinity. Individual hydroxyl groups of α-D-mannose were investigated for their energy contributions to the overall binding affinity for FimH. Among other things, this study demonstrated why the only potent FimH antagonists are modified at (C1)-OH and how modifying any other hydroxyl group will inevitably lead to a diminished binding affinity. Indeed, drug design of FimH antagonists has been constrained almost exclusively to attaching an aglycone moiety to (C1)-OH of α-D-mannose. Strategies to improve the binding affinity have mostly centered on improving interactions between the aglycone of the FimH antagonist and a hydrophobic element adjacent to the mannose-binding pocket of the FimH lectin. Known as the tyrosine gate, this hydrophobic element forms π-π interactions with the aromatic aglycone of a FimH antagonist. In Manuscript 1 we took a closer look at the tyrosine gate, which is a crucial determinant of antagonist design, and turned our focus away from synthetic antagonists towards natural mannose glycans. We hypothesize that the tyrosine gate is not significantly participatory in the binding of natural mannose glycans and that its existence and proximity to the mannose-binding pocket might be serendipity more than physiological functionality.
Publication 2–4 deal with the conformational heterogeneity of the FimH lectin. Publication 2 marked the first time that all conformational states of FimH were characterized structurally. We explored the implications of this conformational heterogeneity on the binding kinetics of the FimH lectin at both the molecular and cellular level. In Publication 3, we investigated the conformational variability of different FimH variants by generating fimbriated E. coli strains with an isogenic background. These bacterial strains fimbriated with FimH of differential degrees of conformational variability were tested in a flow chamber assay in an attempt to observe bacteria in an environment that emulated the physiological conditions encountered by UPEC in the urinary tract. All FimH variants showed very distinct binding profiles under flow conditions. We then analyzed the inhibitory potency of FimH antagonists for these varying FimH variants, essentially measuring binding affinity as a function of the conformational variability of FimH.
The evaluation of FimH antagonists in target-based assays has mostly relied on a FimH construct that is locked in a particular conformation. Publication 4 represents our effort to take the conformational heterogeneity of FimH into account when evaluating FimH antagonists in target-based assays and reveals that the corresponding affinities were overestimated almost consistently by 100-fold.