Molecular basis of uropathogenic Escherichia coli evasion of the innate immune response in the bladder

Abstract

In the urinary tract, the innate immune system detects conserved bacterial components and responds to infection by activating the proinflammatory transcription factor NF-kappaB, resulting in cytokine secretion and neutrophil recruitment. Uropathogenic Escherichia coli (UPEC), however, has been shown to evade the host innate immune response by suppressing NF-kappaB activation in urothelial cells, which results in decreased cytokine secretion and increased urothelial apoptosis. To understand the molecular basis of UPEC modulation of inflammation, we performed a genetic screen with UPEC strain NU14 to identify genes which are required for modulation of urothelial cytokine secretion. Disruption of ampG (peptidoglycan permease), waaL (lipopolysaccharide O antigen ligase), or alr (alanine racemase) resulted in increased urothelial interleukin-8 (IL-8) and IL-6 release from urothelial cell cultures. Targeted deletion of these genes also resulted in elevated urothelial cytokine production during UPEC infection. Conditioned media from bacterial cultures of NU14 DeltaampG and NU14 DeltawaaL contained a heat-stable factor(s) which stimulated greater urothelial IL-8 secretion than that in NU14-conditioned medium. In a mouse model of urinary tract infection, NU14 DeltaampG, NU14 DeltawaaL, and NU14 Deltaalr were attenuated compared to wild-type NU14 and showed reduced fitness in competition experiments. Instillation of NU14 DeltaampG or NU14 DeltawaaL increased bladder neutrophil recruitment, indicating that enhanced urothelial cytokine secretion during urinary tract infection results in an altered host response. Thus, UPEC evasion of innate immune detection of bacterial components, such as lipopolysaccharide and peptidoglycan fragments, is likely an important factor in the ability of UPEC to colonize the urinary tract.

FIG. 1.

Four transposon insertion mutants of NU14 induced significantly increased IL-8 secretion from urothelial cell cultures and had different phenotypes in a mouse model of UTI. (A) TEU-1 cell cultures were treated with NU14 and four transposon mutants, 15E6, 22H12, 25D9, and 40G5, for 4 hours, and IL-8 secretion was determined by ELISA. The data represent the means ± standard deviations for three samples, and each experiment was performed in duplicate. (B) Mice (n = 5) were instilled via catheter with either NU14, MG1655, 15E6, 22H12, or 25D9, and bladder colonization was determined at 24 h postinfection. The asterisks (*, P < 0.05; and **, P < 0.01) indicate statistically significant differences between NU14 and each of the transposon mutants.

FIG. 2.

Targeted deletion mutants confirm the phenotypes of the transposon mutants. (A) TEU-1 cell cultures were treated for 4 hours with TNF, IL-1β, NU14, MG1655, NU14 ΔampG, NU14 ΔwaaL, or NU14 Δalr, and IL-8 secretion was determined by ELISA. (B) T24 cultures were treated for 4 hours with TNF, LPS (O55:B5), NU14, NU14-1, NU14 ΔampG, or NU14 ΔwaaL, and IL-8 secretion was determined by ELISA. (C) IL-6 secretion from TEU-1 cultures was determined after treatment with TNF, NU14, NU14 ΔampG, NU14 ΔwaaL, or NU14 Δalr. The data represent the means ± standard deviations for three samples, and each experiment was performed in duplicate. The asterisks indicate statistically significant differences (P < 0.05) between samples treated with NU14 and the targeted deletion mutants.

FIG. 3.

Deletion mutant phenotypes can be complemented through expression of the deleted gene products from plasmids. (A) TEU-1 cells were treated with TNF or infected with NU14, NU14 ΔampG/pFLAG, or NU14 ΔampG/pampG. Cell culture supernatants were harvested after 4 h of infection, and IL-8 secretion was determined by ELISA. (B) IL-8 secretion levels were determined after 4 hours of treatment of TEU-1 cultures with either TNF, NU14, NU14 ΔwaaL/pFLAG (empty vector), or NU14 ΔwaaL/pwaaL. The data represent the means ± standard deviations for three samples, and each experiment was performed in duplicate. The asterisks indicate statistically significant differences (P < 0.05) between samples treated with the deletion mutant strains carrying the empty vector control plasmid and the strains carrying the complementation construct plasmids.

FIG. 4.

Conditioned media from NU14 ΔampG and NU14 ΔwaaL cultures contain heat-stable factors which stimulate urothelial IL-8 secretion. (A) Conditioned cell culture media were prepared from cultures of NU14, NU14 ΔampG, and NU14 ΔwaaL grown overnight. TEU-1 cells were then treated with the filtered conditioned media or with conditioned media which had been heat treated at 65°C for 10 min, and IL-8 secretion was determined by ELISA. (B) Conditioned medium from NU14, NU14 ΔampG/pFLAG, NU14 ΔampG/pampG, NU14 ΔwaaL/pFLAG, or NU14 ΔwaaL/pwaaL culture was used to treat TEU-1 cultures, and the resulting IL-8 secretion was determined. The data represent the means ± standard deviations for three samples, and each experiment was performed in duplicate. The asterisks indicate statistically significant differences (P < 0.05) between samples.

FIG. 5.

Deletion of waaL results in rough-type bacterial colony morphology. NU14 (A), NU14 ΔwaaL (B), NU14 ΔwaaL/pFLAG (C), and NU14 ΔwaaL/pwaaL (D) were grown on LB agar plates containing the appropriate antibiotics overnight at 37°C, and colony morphology was photographed at a magnification of ×4, using a Leica S8AP0 microscope.

FIG. 6.

NU14 ΔampG and NU14 ΔwaaL are deficient in colonization of the mouse bladder. (A) Mice were infected via instillation into the bladder with either NU14 (n = 18), NU14 ΔampG (n = 18), NU14 ΔwaaL (n = 18), or NU14 Δalr (n = 10), and bacterial colonization was measured at 24 h postinfection. The asterisks (*, P < 0.05; and **, P < 0.0001) indicate statistically significant differences. (B and C) Mice (n = 8) were inoculated via catheter with both NU14 and NU14 ΔampG (B) or NU14 and NU14 ΔwaaL (C), and colonization of each strain was determined by differential colony counts on selective agar medium. Each data point represents the competitive index ratio, which describes the numerical ratio of NU14 to the deletion mutant strain CFU present in each mouse bladder at 24 h postinfection. The initial inoculation ratio (Innoc) is represented by a dashed line.

FIG. 7.

Infection with NU14 ΔampG and NU14 ΔwaaL stimulates enhanced MPO in mouse bladder homogenates and urines during UTI. Mice were instilled with saline or infected with NU14, NU14 ΔampG, or NU14 ΔwaaL via catheter (for bladder experiments, five saline control mice and eight infected mice were used; and for urine experiments, four control mice and eight infected mice were used). Urine samples were collected, and the bladders were removed and homogenized at 6 h postinfection. Bladder (A) and urine (B) MPO levels were determined by ELISA. Asterisks indicate statistically significant differences (P < 0.05) between saline-instilled controls and samples from mice infected with bacterial strains.