Developments in Mannose-Based Treatments for Uropathogenic Escherichia coli-Induced Urinary Tract Infections

Abstract

During their lifetime almost half of women will experience a symptomatic urinary tract infection (UTI) with a further half experiencing a relapse within six months. Currently UTIs are treated with antibiotics, but increasing antibiotic resistance rates highlight the need for new treatments. Uropathogenic Escherichia coli (UPEC) is responsible for the majority of symptomatic UTI cases and thus has become a key pathological target. Adhesion of type one pilus subunit FimH at the surface of UPEC strains to mannose-saturated oligosaccharides located on the urothelium is critical to pathogenesis. Since the identification of FimH as a therapeutic target in the late 1980s, a substantial body of research has been generated focusing on the development of FimH-targeting mannose-based anti-adhesion therapies. In this review we will discuss the design of different classes of these mannose-based compounds and their utility and potential as UPEC therapeutics.

Keywords: FimH; anti-adhesion; carbohydrates; mannose; urinary tract infections (UTIs).

Figure 1

a) Structure of the urothelium with basal cells attached to a basement membrane, an intermediate layer and a layer of umbrella cells. b) Structure of the uroplakin (UP) plaque consisting of six heterodimeric units with each unit composed of two dimers UPIa/II and UPIb/IIIa. c) Structure of the heterodimer units in which a green ellipse represents a high‐mannose‐containing N‐glycan and the grey ellipses on UPIb/IIIa represent complex N‐glycans. d) Structure of UPIa and e) structure of UPIb. Both structures consist of four transmembrane domains and have an approximate molecular weight of 30 kDa. ^[9] The green circles represent mannose residues, the blue squares represent glucosamine residues and the yellow circles represent galactose residues.

Figure 2

The pathogenesis cycle for UPEC consists of six stages: Stage 1) colonization of the periurethral areas and the urethra, Stage 2) movement of UPEC up the urethra, Stage 3) UPEC adherence, Stage 4) biofilm formation, Stage 5) epithelial cell invasion and formation of an intracellular bacterial population, and Stage 6) colonization of the urinary tract and kidneys by UPEC followed by entry into the blood stream.

Figure 3

Schematic diagram showing the three‐stage membrane zippering mechanism thought to be used by UPEC during the invasion of the urothelium; Stage 1) binding of UPEC to the urothelium, Stage 2) localized rearrangement of the urothelium actin cytoskeleton, Stage 3) envelopment and internalization of the bound UPEC.

Figure 4

Structural organisation of the type 1 pilus which includes the following subunits: FimA (blue), FimC (grey), FimD (green rectangle), FimF (purple), FimG (green oval), and FimH (red). The position of the N‐terminal domain (N, orange) and two C‐terminal domains (C1, pink; C2, cyan) relative to the transmembrane pore in FimD are also indicated.

Figure 5

Crystal structure of FimH (orange) with a heptylmannoside (green) ligand bound in the N‐terminal lectin domain (FimH_LD; PDB ID: 4LOV). ^[2] Residues with side chains shown as sticks (blue) are involved in noncovalent interactions with heptylmannoside, sodium ions (purple spheres) or water molecules (red spheres). The dashed lines (purple) denote short‐range, noncovalent interactions.

Figure 6

Catch bond mechanism for FimH binding to mannose in urethral lumen under no shear flow conditions, where unbound and bound forms exist at equilibrium, and moderate shear flow conditions, where the bound state is favoured. Under moderate shear flow conditions, the hydrodynamic drag on the micron‐sized UPEC bacterium (not to scale) results in a physical force on the tether which activates the catch bond mechanism in FimH. The higher affinity binding to mannose under these conditions prevents the UPEC bacterium from being flushed out of the urinary tract during urination.

Figure 7

A depiction of the main interactions that occur between mannose‐based pentasaccharide 1 and the extended FimH binding site. ^[46] Red indicates interactions mediated by van der Waals, aromatic stacking and hydrophobic interactions, blue indicates interactions mediated by hydrogen bonding.

Figure 8

a) Structure of FimH the lectin domain with the tyrosine gate open (PDB ID: 4AV0). ^[51] b) Structure of the FimH lectin domain when the tyrosine gate is closed (PDB ID: 4ATT). ^[52] Note the movement of the Tyr48 residue (blue) from facing the Asp47 residue (yellow) in the open conformer to facing the Thr51 residue (pink) in the closed conformer. The Arg98 residue (green) plays a role in stabilizing the surface loop conformation on which the Thr51 residue is located.

Figure 9

Comparison of the structures of potent oligosaccharide‐based FimH inhibitors (oligosaccharides 2, 3, and trisaccharide 4) to weak oligosaccharide‐based FimH inhibitors (mannotriose 5, mannopentaose 6 and Manα‐1,3Man 7).

Figure 10

Structure of the potent FimH inhibitor oligomannoside 1.

Figure 11

Structure of the alkyl mannoside scaffold where n=0–7.

Scheme 1

Reaction scheme for the synthesis of alkyl mannosides by glycosylation with aliphatic alcohols using an AgOTf activation system.

Figure 12

Structure of p‐nitrophenyl‐α‐mannoside (8), p‐nitro‐O‐chlorophenyl‐α‐mannoside (pNoClPαMan; 9) and 4‐methylumbelliferyl‐α‐mannoside (MeUmbαMan; 10).

Figure 13

Structure of biphenyl mannoside 11.

Figure 14

Structure of biphenyl mannoside 12 and biphenyl mannoside 13.

Figure 15

Structure of biphenyl mannoside 11, biphenyl mannoside 14, biphenyl mannoside 15 and biphenyl mannoside 16 where R=CF₃, Cl, Me, OMe, F and R’=CF₃, Me, Cl.

Figure 16

Structure of biphenyl mannoside 17.

Figure 17

Structure of some substituted biphenyl mannoside analogues with promising FimH activity and pharmacokinetic properties, ^[1] such as dissociation constants (K _D) and log p _e values assessed using a parallel artificial membrane permeability (PAMP) assay.

Scheme 2

Synthesis of ortho‐substituted biphenyl mannosides. Pathways A and B: a) BF₃ ⋅ Et₂O, CH₂Cl₂, reflux, 45 h, (25–75 %); b) 3‐substituted phenylboronic acid derivatives, cat. Pd(PPh₃)₄, Cs₂CO₃, dioxane/water (5 : 1), 80 °C, 1 h; c) cat. MeONa, MeOH, RT, 12 h, (b+c 3–64 %): Pathway C: d) MeNH₂/EtOH, RT, e) bis(pinacolato)diboron, cat. Pd(dppf)Cl₂, KOAc, DMSO, 80 °C.

Figure 18

Structure of squarate mannosides 18 and 19.

Scheme 3

Reaction scheme for proposed covalent bond formation between squarate mannoside 18 and the N terminus of the FimH lectin (residue Phe1).

Figure 19

Structure of diamide squarate mannoside 20.

Figure 20

Structure, IC₅₀ values and K _D values of 2‐O‐n‐heptyl‐1,6‐anhydro‐d‐glycero‐d‐galactitol 21 and n‐heptyl α‐d‐mannopyranoside 22. ^[69]

Figure 21

Structure of thiazolylaminomannosides scaffold 1 and neothiazolylaminomannoside scaffold 2 as well as the structure and IC₅₀ values of thiazolylaminomannosides 23 and neothiazolylaminomannoside 24. ^[70a]

Figure 22

Structure and IC₅₀ value of indolinylphenyl 25. ^[71]

Figure 23

Structures of the α‐d‐mannopyranoside‐based inhibitors, n‐heptyl α‐d‐mannopyranoside 22, biphenyl α‐d‐mannopyranoside derivatives 26 and 27, indolylphenyl mannoside derivative 28 and squarate mannosides derivative 29.

Figure 24

Diagram showing two potential mechanisms whereby multivalent ligands can increase apparent binding affinity. a) Clustering effect where a multivalent ligand binds to one receptor initially and then captures additional receptors as they diffuse into close proximity resulting in clustering of the ligand ‐bound receptors. b) Induced aggregation whereby multivalent ligands bind to multiple lectins on different bacterial units cross linking them together.

Figure 25

Structure of multivalent HM glycoconjugate series designed using a carbohydrate core where n=0, 1,2 and 6 (when using β‐cyclodextrin core).