Manipulating flagellar gene regulation in uropathogenic Escherichia coli to explore its effect(s) on the urothelial proinflammatory response

Abstract

Urinary tract infections (UTIs) are a global problem and the major pathogen associated with such infections is Uropathogenic Escherichia coli (UPEC). UTIs are classified according to the site of infection: cystitis describes a bladder infection and pyelonephritis a kidney infection. If not treated symptomatic UTIs can lead to bacteraemia, sepsis and death. An additional condition is asymptomatic bacteriuria (ASB), which mainly affects older and/or immunocompromised populations. ASB is associated with UPEC colonising the lower urinary tract (UT) without causing symptoms and often underlies symptomatic infections. A panel of clinical uro-associated E. coli isolates recovered from cystitis, pyelonephritis, ASB and bacteraemia patients were characterised for motility and their ability to induce an innate response in urothelial cells, determined by the induction of NF-κB and/or IL-8 synthesis. Data suggested a mix of motile and non-motile strains, but the motility patterns and abilities of the motile strains to induce an innate response were very variable. The aim of this thesis was to understand the factors underpinning this variation. Lower UT tissues defend themselves from bacterial assaults via innate mechanisms involving urothelial receptors with the most important being Toll-like Receptor 5 (TLR5). TLR5 detects the bacterial flagellar subunit, flagellin and activation results in the synthesis of host defence agents and bacterial killing. Hence, flagellar appear key in inducing the urothelial host defences. Synthesising flagella is an energy-consuming process initiated only when motility is advantageous. This led to the hypothesis that growth fitness is important in the recognition of uro-associated E. coli by urothelial cells. Chapter 3 describes experiments, measuring bacterial doubling times, and comparing the growth fitness of motile and non-motile clinical isolates. The motile (M) strain mean doubling time was 17.4±3.0 min versus 19.7±2,5 min for the non-motile (NM) strains (P=0.011). However, the range of doubling times was similar (M: 13.9-26.9 min; NM: 14.7- 25.3min). These latter data argued against fitness being a primary driving factor in the variation of host recognition. To examine if variation linked to flagellar abundance strains were engineered, using pSE_flhDC to hyperexpress flagellar. The mutant strains showed increased motility and innate responses (2-fold) indicating that variability in the host responses linked to fluctuations in bacterial flagellar numbers and subsequent motility. Chapter 4 confirmed these data by constructing and using ∆flhDC and ∆clpP mutants, which resulted in either a loss of flagellar or unregulated flagellar synthesis. Six strains were analysed NCTC10418, CFT073, 3408, MG1655, 5469 and 5489. To quantify flagellar numbers the FlgE of three strains NCTC10418, CFT073 and 3408 was replaced with FlgEA240C and the mutated cells stained with Alexa Fluor 488 maleimide. These cells were examined using fluorescence microscopy. Visualisation of the FlgEA240C foci of the ∆clpP mutants supported increases in flagellar numbers as well as an increase in the bacterial population carrying flagellar. These data further supported the concept that variability in the bladder host response is linked to population heterogeneity and flagellar abundance of the invading uropathogens. Data presented in Chapters 3 and 4 related to bacteria cultured in enriched media, which physiologically does not reflect the UT environment. Experiments were therefore performed using three strains, 3408, 5469 and 5489 and their ∆flhDC and ∆clpP mutants grown in artificial urine media (AUM) at pH 5.5 (acid urine) and pH 6.5 (alkaline urine). Data showed growth in AUM was associated with a reduced motility and bladder innate response, and that pH did not impact these data.

In summary results have shown that uro-associated E. coli isolates, generally, regulate flagellar synthesis to allow colonisation or infection, but appear to exploit population heterogeneity to prevent recognition by urothelial cell TLR5 receptors and bacterial killing.