Metabolic Requirements of Escherichia coli in Intracellular Bacterial Communities during Urinary Tract Infection Pathogenesis

Abstract

Uropathogenic Escherichia coli (UPEC) is the primary etiological agent of over 85% of community-acquired urinary tract infections (UTIs). Mouse models of infection have shown that UPEC can invade bladder epithelial cells in a type 1 pilus-dependent mechanism, avoid a TLR4-mediated exocytic process, and escape into the host cell cytoplasm. The internalized UPEC can clonally replicate into biofilm-like intracellular bacterial communities (IBCs) of thousands of bacteria while avoiding many host clearance mechanisms. Importantly, IBCs have been documented in urine from women and children suffering acute UTI. To understand this protected bacterial niche, we elucidated the transcriptional profile of bacteria within IBCs using microarrays. We delineated the upregulation within the IBC of genes involved in iron acquisition, metabolism, and transport. Interestingly, lacZ was highly upregulated, suggesting that bacteria were sensing and/or utilizing a galactoside for metabolism in the IBC. A ΔlacZ strain displayed significantly smaller IBCs than the wild-type strain and was attenuated during competitive infection with a wild-type strain. Similarly, a galK mutant resulted in smaller IBCs and attenuated infection. Further, analysis of the highly upregulated gene yeaR revealed that this gene contributes to oxidative stress resistance and type 1 pilus production. These results suggest that bacteria within the IBC are under oxidative stress and, consistent with previous reports, utilize nonglucose carbon metabolites. Better understanding of the bacterial mechanisms used for IBC development and establishment of infection may give insights into development of novel anti-virulence strategies.

Importance: Urinary tract infections (UTIs) are one of the most common bacterial infections, impacting mostly women. Every year, millions of UTIs occur in the U.S. with most being caused by uropathogenic E. coli(UPEC). During a UTI, UPEC invade bladder cells and form an intracellular bacterial community (IBC) that allows for the bacteria to replicate protected from the host immune response. In this study, we investigated genes that are expressed by UPEC within the IBC and determined how they contribute to the formation of this specialized community. Our findings suggest that galactose is important for UPEC growth in the IBC. Additionally, we found that a gene involved in oxidative stress is also important in the regulation of a key factor needed for UPEC invasion of bladder cells. These results may open the door for the development of treatments to diminish UTI frequency and/or severity.

FIG 1

Deletion of yeaR decreases type 1 piliation. (A) Type 1 piliation was assessed via hemagglutination for UTI89 and the ΔyeaR, ΔyeaR comp, LIR UTI89, and LIR ΔyeaR mutants. (B) Relative expression levels of fimI for the UTI89 and ΔyeaR strains were measured by qPCR. (C) Pili were counted on 300 cells per strain. Each cell was assigned one of the general descriptors “abundant,” “moderate,” “low,” and “bald” (no pili) to represent its piliation level. This revealed a shift to a lower piliated population in the ΔyeaR mutant that in the UTI89 strain. The cells represented in all panels were grown under conditions of type 1 induction. The representative data shown are from experiments performed in triplicate. Asterisks denote P < 0.05 (Student’s t test).

FIG 2

Tellurite and oxidate stress sensitivity. The indicated strains were incubated with increasing concentrations of hydrogen peroxide (A) or potassium tellurite (B) in combination with sodium phosphate buffer for 1 h. Percent survival was calculated against mock-treated cells incubated under the same conditions. The representative data shown are from experiments performed in triplicate. Asterisks denote P < 0.05 (Student’s t test).

FIG 3

Oxidative stress influences type 1 piliation. The indicated strains were grown under conditions of type 1 pilus induction with the addition of 0, 5, 10, or 20 mM H₂O₂ to the growth media. Hemagglutination assays were performed in duplicate. Data represent averages of results of three experiments. Asterisks denote P < 0.05 (Student’s t test).

FIG 4

Year influences 6-h IBCs and colonization. C3H/HeN mice were infected for 6 h with 1 × 10⁷ bacteria of the indicated strains. Displayed data are representative of total bladder colonization (A) or IBC enumeration by LacZ staining (B) (n = 5 to 9). Asterisks denote P < 0.05 (Mann-Whitney t test).

FIG 5

Type 1 pilus-mediated invasion. The UTI89, ΔyeaR, ΔfimA-H (Δfim), LIR UTI89, or LIR ΔyeaR strain was added to the 5637 human bladder epithelial cell line and allowed to invade for 2 h. Subsequent gentamicin treatment and cell lysis allowed the enumeration of internalized bacteria. Asterisks denote P < 0.05 (Student’s t test).

FIG 6

LacZ strain colonization and IBC formation (6 h). C3H/HeN mice were infected for 6 h with 1 × 10⁷ UTI89 or ΔlacZ mutant bacteria. Bladders were either homogenized for CFU enumeration (n = 8) (A) or stretched and fixed for IBC counting using confocal microscopy (n = 5) (B). GFP-expressing strains were used for the IBC counts. The asterisk denotes P < 0.05 (Mann-Whitney t test).

FIG 7

IBC comparisons (6 and 16 h). Representative micrographs depicting GFP-expressing UTI89 and ΔlacZ and ΔgalK mutant IBCs at 6 (top panel) or 16 (bottom panel) h postinfection. UTI89 formed significantly larger IBCs at 6 h than either mutant. At 16 h postinfection, UTI89 begins to filament just prior to dispersal. No filaments were observed at this time point with either the ΔlacZ mutant or the ΔgalK mutant. Images were taken from 5 different mice for each strain.

FIG 8

The ΔlacZ and ΔgalK mutations confer a fitness advantage during chronic cystitis. C3H/HeN mice were infected with equal numbers of differentially marked strain UTI89 bacteria in competition with ΔlacZ or ΔgalK mutant bacteria. The log competitive indexes of UTI89 versus the ΔlacZ strain (A) and UTI89 versus the ΔgalK strain (C) were determined in titers of each strain from urine samples collected 1, 3, 7, 10, 14, 21, and 28 days postinfection. Each line represents a single mouse. Log CIs are also displayed for the bladder and kidneys of mice infected with the UTI89 or ΔlacZ strain (B) or the UTI89 or ΔgalK strain (D). UTI89/UTI89 (wt/wt) control competition is also shown (B and D). WT, wild type. Asterisks denote P < 0.05 (Mann-Whitney t test).