Uropathogenic Escherichia coli as a model of host–parasite interaction

Resistance to mucosal infection varies greatly in the population, but the molecular basis of disease susceptibility is often unknown. Studies of host-pathogen infections are helpful to identify virulence factors, which characterise disease isolates, and successful defence strategies of hosts that resist infection. In the urinary tract infection (UTI) model, we have identified crucial steps in the pathogen-activated innate host response, and studied the genetic control of these activation steps. Furthermore, genetic variation in the innate host-response defence is investigated as a basis of disease susceptibility. The Toll-like receptor 4 (TLR4) controls initial mucosal response to uropathogenic Escherichia coli (UPEC). Bacterial TLR4 activation in epithelial cells leads to chemokine secretion and neutrophil recruitment and TLR4 mutant mice develop an asymptomatic carrier state. The chemokine receptor CXCR1 determines the efficiency of neutrophil migration and activation, and thus of bacterial clearance. CXCR1 mutant mice become bacteremic and develop renal scars and studies in UTI prone children have detected low CXCR1 expression, suggesting that CXCR1 is also essential for human disease susceptibility.

Introduction

Urinary tract infections (UTIs) provide an excellent model to study how the host recognises and deals with mucosal pathogens [1, 2, 3]. The infecting strain encounters a microbially naïve mucosal environment where it may cause acute, potentially life-threatening infections. Chronic sequels are prevalent, and there is a link between acute infection and chronicity [4, 5]. The mechanisms underlying commensalism can also be studied in the urinary tract, because asymptomatic bacteriuria (ABU) occurs in at least 1% of the population, who can carry >10⁵ cfu/ml of Escherichia coli in their urine for months or years with no or few symptoms [6, 7]. Studies in the UTI model have identified molecular mechanisms that initiate tissue attack by mucosal pathogens and that trigger the innate host response [8^•]. On the basis of these mechanisms the genetics of disease susceptibility are beginning to be understood [9, 10].

The severity of UTI reflects the virulence of the infecting strain. In the 1940s, hemolysin was identified as a characteristic of E. coli that cause extra-intestinal infections [11]. Uropathogenic E. coli (UPEC) strains were later shown to belong to a restricted set of serotypes or ‘clones’ [12], and strains causing acute pyelonephritis and ABU strains were found to differ in surface-antigen repertoire [7]. In the 1970s, this information was extended to involve attachment to the urinary tract mucosa [13]. High tissue-attachment was shown to characterise the most virulent strains but not the asymptomatic carrier strains. Attachment was thus proposed as the first step in the pathogenesis of UTI, and the epithelial cell was recognised as the first sensor of tissue attack [14, 15, 16]. Since then, the molecular basis of virulence has been extensively studied, and several essential virulence factors have been identified [17]. The virulence genes are encoded on pathogenicity islands, and their expression, regulation and evolution have been elegantly characterised [18, 19, 20]. Recent studies have suggested that ABU strains might be attenuated pathogens, carrying deletions in the virulence genes involved in attachment and tissue attack (Figure 1) [21].

The virulence factors enable UPEC to trigger epithelial cell responses leading to inflammation, cell detachment and apoptosis, or invasion and bacteremia [14, 15, 22, 23, 24, 25, 26, 27, 28•]. The tissue response is lethal for the organism, however, and it is unlikely that invasion offers a significant advantage that would drive the evolution of the virulent phenotype. It is more likely that the adaptation of uropathogenic clones occurs mainly in the large intestine, and that virulence is ‘co-incidental’ [29]. For example, the mucosal receptors for P-fimbriated uropathogens are expressed in both the intestinal and urinary tracts, but ligand-binding has different consequences at the two sites. In the urinary tract, attaching P-fimbriated strains evokes the host response, and P-fimbrial signalling through TLR4 leads either, to elimination of the bacteria or to disease [30]. In the large intestine, however, P-fimbriae promote bacterial persistence, but there is no evidence of inflammation. ABU resembles the intestinal carrier state as the bacteria persist without provoking a host response. Recent studies have suggested that ABU strains carry virulence genes, but do not express them. ABU might thus represent a successful adaptation, because the bacteria can persist without competition in a niche with a rich nutrient source, often for several years. In this case, the host might produce signals that attenuate bacterial virulence.

Susceptibility to UTI varies greatly among the population, as does the severity of disease among susceptible individuals [31]. Studies in pyelonephritis-prone children have identified ‘high responders’ which exhibit abnormalities that exaggerate the damaging rather than protective aspects of innate immunity [3, 32]. Experimental infections in different mutant mice have identified a single gene defect that causes the high responder phenotype [5]. Neutrophils are crucial effectors of host defence in the urinary tract, and neutrophil dysfunctions — owing to defective IL-8 receptor expression — cause acute pyelonephritis and renal scarring [10]. In ‘low responders’, however, bacteriuria establishes without evoking a response, showing that suppression of inflammatory signals may be protective even though the infection remains. Furthermore, experimental infections have identified genetic control mechanisms that decide if the host will remain asymptomatic or develop disease. Mice carrying a mutation in the signalling domain of Toll-like receptor 4 (TLR4) develop an asymptomatic carrier state resembling human ABU infection [33, 34, 35].

This review focuses on these two steps in disease pathogenesis and their consequences for human disease.