Monoclonal antibodies targeting the FimH adhesin protect against uropathogenic E. coli UTI

Abstract

As antimicrobial resistance increases, urinary tract infections (UTIs) are expected to pose an increased burden in morbidity and expense on the health care system, increasing the need for alternative antibiotic-sparing treatments. Most UTIs are caused by uropathogenic Escherichia coli (UPEC), whereas Klebsiella pneumoniae causes a large portion of non-UPEC UTIs. Both bacteria express type 1 pili tipped with the mannose-binding FimH adhesin critical for UTI pathogenesis. We generated and biochemically characterized 33 murine monoclonal antibodies (mAbs) to FimH. Three mAbs protected mice from E. coli UTI. Mechanistically, we show that this protection is Fc independent and mediated by the ability of these mAbs to sterically block FimH function by recognizing a high-affinity FimH conformation. Our data reveal that FimH mAbs hold promise as an antibiotic-sparing treatment strategy.

Fig. 1.. FimH mAbs bind to four distinct epitope classes.

Structures of (A) E. coli FimH_LD (PDB 1KLF), (B) K. pneumoniae FimH_LD (PDB 9AT9), and (C) E. coli FmlH_LD (PDB 6AOW). Residue differences from E. coli FimH_LD are highlighted in red. (D) ELISA EC₅₀ values for each mAb to the listed protein. White cells with no values indicate EC₅₀ values were above the range measured in the assay. (E) Epitope mapping of mAbs (top labels) to a panel of FimH mutants (right labels). Binding classes were determined by shared residues that abrogated mAb binding, which are highlighted in purple in the table (bottom). Epitope residues representing a majority of mAbs within an epitope class are colored in purple on the surface of E. coli FimH_LD (top) (PDB 1KLF). Black arrows on structures indicate the binding pocket.

Fig. 2.. Structural basis of FimH mAb protection.

(A to E) Cryo-EM density maps of Fabs bound to FimH_LD. (A) Kp1 2H04 (teal), (B) Ec1 F7 (cyan), (C) Kp1 2H04 and Ec1 F7 maps superimposed on each other, (D) Kp2 2C07 (sand), and (E) Ec3 B7 (dark green) bound to FimH_LD (salmon). (F to I) Binding epitopes of the Fabs on FimH_LD colored by the same color scheme as (A) to (E). Density map overlaid on model residue interactions of (J) Kp1 2H04 (cyan) with FimH P26 and (K) Ec3 B7 (dark green) with FimH Y64. mAb heavy chain residues are labeled “HC,” and light chain residues are labeled “LC.” Black arrows on structures indicate the FimH mannose-binding pocket.

Fig. 3.. mAbs bind the relaxed conformation of FimH.

FimH mAb binding to (A) UTI89 bacteria and (B) UTI89 overexpressing conformationally shifted FimH variants (n = 3; error bars represent SEM). (C) Representative binding curves of Kp1 2H04, Ec1 F7, and Kp2 2C07 Fab binding FimH_LD and FimG_NTEH. Results are from kinetic measurements of dilution series of one experiment. (D) Observed BLI binding kinetics to Ec FimG_NTEH and Ec FimH_LD. N.D. means not determined due to the off rate being below the detection limit.

Fig. 4.. Relaxed-preferred mAbs can inhibit FimH binding.

(A) Inhibition of FimH_LD binding to BSM at a 5:1 molar ratio of mAb to protein (n ≥ 3). (B) Inhibition of UTI89 guinea pig erythrocyte hemagglutination (n ≥ 2). Kp1 2H04 mAb inhibition of (C) Ec FimH_LD and (D) Kp FimH_LD binding to C3H/HeN mouse bladders. mAb was preincubated with FimH_LD at a 10:1 molar ratio. Sections were stained with DNA dye Hoechst (blue), Ec or Kp FimH_LD (red), and antibody to uroplakin III (green). n = 4 to 5 bladder sections with n = 2 technical replicates.

Fig. 5.. FimH mAbs protect from E. coli UTI.

(A) Six- to 7-week-old C3H/HeN mice were pretreated with 0.5 mg of mAb 24 hours before infection with UTI89. ip, intraperitoneal. (B) 24 hpi bladder and (C) kidney titers (for Kp1 1B03 and Kp1 2E08, n = 5 with one independent replicate; for Ec3 B7, n = 8 with one independent replicate; for Kp2 2C07, n = 10 with two independent replicates; and for control IgG, Ec1 F7, Kp1 2H04, Kp1 1A02, and Kp1 2E02, n = 13 to 33 with three independent replicates). (D) IBC counts at 6 hpi (n = 16 for control IgG; n = 14 for Kp1 2H04 with two independent replicates). Representative 5x magnification images of IBCs (green) in splayed mouse bladders for (E) control IgG and (F) Kp1 2H04 treatments. Scale bars, 200 μm. 24 hpi bladder (G) and kidney titers (H) from the prophylactic model testing control IgG, Kp1 2H04, and Kp1 2H04_LALAPG (n = 21 for control group, n = 20 for Kp1 2H04, and n = 19 for Kp1 2H04_LALAPG with three independent replicates). For (D), the bar graph represents the median with error bars of 95% confidence interval and a Mann-Whitney U test was used to evaluate statistical significance. For (B), (C), (G), and (H), bar graphs represent geometric means with error bars of SD and statistical comparisons were made using the Kruskal-Wallis test [nonparametric analysis of variance (ANOVA)] with Dunn’s comparisons to the control group correcting for multiple comparisons. n.s., not significant; *P ≤ 0.05; **P ≤ 0.01; ****P ≤ 0.00001. The schematic in (A) was created in BioRender. E.D.B.L. (2025), https://BioRender.com/a40t157.