Atypical Enteropathogenic Escherichia coli: from Kittens to Humans and Beyond!

Atypical enteropathogenic Escherichia coli (aEPEC) are associated with diarrhea worldwide, yet genome-wide investigations to probe their virulome are lacking. V.

ABSTRACT

Atypical enteropathogenic Escherichia coli (aEPEC) are associated with diarrhea worldwide, yet genome-wide investigations to probe their virulome are lacking. In this issue of Infection and Immunity, V. E. Watson, T. H. Hazen, D. A. Rasko, M. E. Jacob, et al. (IAI 89:e00619-20, 2020, https://doi.org/10.1128/IAI.00619-20) sequenced aEPEC isolates from diarrheic and asymptomatic kittens. Using phylogenomics, they demonstrated that these isolates were genetically indistinguishable from human isolates, suggesting that kittens may serve as a reservoir and, perhaps, a much-needed model to interrogate aEPEC virulence. The diarrheic isolates were hypermotile, suggesting that this phenotype may distinguish virulent strains from their innocuous counterparts.

TEXT

Escherichia coli is perhaps the most exhaustively interrogated bacterial species to date. Besides being a well-known commensal resident of the gastrointestinal tract (1, 2), the bacterium has also been employed as a workhorse and has been instrumental in ushering groundbreaking discoveries in the fields of molecular biology, biochemistry, and genetics, that have aided humankind (3). In contrast to innocuous biotypes, numerous pathogenic variants (pathovars) of E. coli also exist, which have gained considerable notoriety because they cause a range of diseases that contribute to significant morbidity and mortality worldwide (4). Pathogenic isolates of E. coli can be broadly classified as intestinal (diarrheagenic) or extraintestinal on the basis of the site of infection in the host. Among diarrheagenic E. coli (DEC), there are six predominant pathovars that include enteropathogenic E. coli (EPEC), Shiga toxin-producing E. coli (STEC) (which includes enterohemorrhagic E. coli), enteroaggregative E. coli (EAEC), diffusely adherent E. coli (DAEC), enteroinvasive E. coli (EIEC), enterotoxigenic E. coli (ETEC), and adherent invasive E. coli (AIEC) (4, 5). It is estimated that, annually, DEC results in the death of ∼120,000 children under the age of 5 years, with EPEC being the most prevalent DEC that is solely responsible for ∼80,000 deaths (6–8).

EPEC belongs to the attaching and effacing (A/E) group of bacterial pathogens, so called because they attach tightly to intestinal cells to produce a pathognomonic structure, termed “A/E lesion,” on the surface of infected host cells (9–11). The formation of A/E lesions is the culmination of a three-stage infectious process. First, the bacteria adhere loosely on the surface of intestinal cells via a range of structurally diverse adhesins such as bundle-forming pilus (BFP), flagella, and intimin, among others (12). This is followed by the expression and assembly of a type III secretion system (T3SS) that directly links the cytoplasm of the bacterium to that of the host, permitting the bacterium to inject an assortment of virulence factors into the host cell (4, 5, 13). These effectors interfere with different host cell processes that lead to ultrastructural reorganization of the host actin and microtubule network, disruption of the epithelial barrier junctions, and microvillar destruction, among others (4, 5). In the final stage, bacterial cells adhere tightly to the infected host cell surface (4, 10, 11, 13). This interaction is mediated by the bacterial outer membrane protein, intimin, and its receptor, Tir, which is one of the effectors that is injected into the host cell via the T3SS (14, 15). Tir-intimin interactions initiate a signal transduction cascade that recruits actin from the depolymerized microvilli and repolymerizes the monomers at the bacterial-host interface to push the host membrane outwards, giving the infectious site the characteristic appearance of a pedestal, upon which the bacterium rests (16, 17). The fluorescence actin-staining (FAS) assay, which was developed by Knutton et al. (18), is the litmus test that identifies A/E pathogens by detecting polymerized actin beneath the pedestals at the bacterial-host interface (4, 5).

The locus of enterocyte effacement (LEE) contains the T3SS that imparts upon EPEC the ability to form A/E lesions (19). Whereas all isolates of EPEC contain the LEE, multiple reports suggest that A/E lesions are not universally formed by all strains. Currently, the EPEC pathovar is subclassified into typical EPEC (tEPEC) and atypical EPEC (aEPEC), with the distinguishing trait being the presence of a 60-MDa plasmid called the EPEC adherence factor (pEAF) in the former (20). The EAF plasmid harbors two important virulence loci, namely, the perABC operon and the 14-gene bfp operon (21–24). The bfp operon encodes proteins involved in the synthesis and assembly of the type IV bundle-forming pili that mediate bacterial-bacterial and bacterial-host attachment during the first stage of EPEC infection (23, 24). The interbacterial interactions promote the discrete clustering of bacterial cells to form microcolonies on the surface of the infecting host cell, a phenomenon termed localized adherence (LA). The perABC operon harbors two important transcriptional factors, with PerA activating transcription of the bfp operon and PerC activating the master regulator of the LEE, ler, which is specified by the first gene of the LEE1 operon, thereby synchronizing microcolony formation with the morphogenesis of A/E lesions (25, 26). Although aEPEC isolates lack the EAF plasmid, many of them are capable of localized-like adherence (LAL) (27, 28), presumably due to convergent evolution. Moreover, aEPEC exhibit more varied patterns of initial adherence than their tEPEC counterparts, such as diffused adherence and aggregative adherence (27–29). In addition to EAF, other phenotypic, molecular, epidemiological, and pathological differences have also been observed between the two types of EPEC (27). aEPEC isolates appear to possess a more varied repertoire of uncharacterized virulence factors than the common set that tEPECs share (27). From an epidemiological standpoint, tEPEC primarily causes short-term diarrheal (<14 days) episodes in infants less than 2 years of age in developing countries, with adult infections being rare (30). In contrast, aEPEC strains have been isolated from individuals of all age groups, and infections have been observed in both developing and developed nations (31–33). Infected individuals exhibit a range of clinical symptoms ranging from acute to persistent diarrhea (27). Moreover, aEPEC, unlike tEPEC, is frequently isolated from subclinical or asymptomatic individuals (34).

Up until the 1990s, most EPEC outbreaks were attributed to tEPEC; although, since then, a decline in tEPEC infections has been observed (35). In contrast, many countries have reported a sharp increase in aEPEC infections (27). In recent years, the prevalence of aEPEC infections has outnumbered that of tEPEC infections (27), and much effort is being devoted toward understanding the molecular epidemiology of this subgroup. However, perhaps the biggest obstacle into molecular investigations of aEPEC is the heterogeneity of genetic markers, which often makes it difficult to extrapolate the results of one isolate to another. Moreover, aEPEC is phylogenetically closely related to Escherichia albertii, and the two are frequently mistyped, attributing traits of one to the other (36). Lastly, whereas multiple infection model systems, such as mice, Caenorhabditis elegans, and human volunteers, have been evaluated, with various degrees of success, to understand the molecular etiology of tEPEC isolates from humans, by contrast, such model systems have not been rigorously examined for their aEPEC counterparts (1). Consequently, there is considerable paucity of information on the molecular mechanisms of virulence of aEPEC.

In a previous study, Watson et al. isolated aEPEC from kittens with and without diarrheal symptoms, although the bacterial burden was significantly higher in diseased animals than in asymptomatic kittens (37). On the basis of serotyping and pulse-field gel electrophoresis, genetically diverse isolates of aEPEC were obtained from both diarrheic and asymptomatic kittens—a finding that is consistent with observations of human aEPEC strains isolated from children (27, 37). Moreover, a statistically significant correlation was observed between aEPEC and epithelial injury and inflammatory infiltrates in the small intestines of diarrheic kittens. These histopathological effects associated with aEPEC infection in kittens are also reminiscent of the effects of human isolates of aEPEC in children (38–40). The present study by the same authors (41) is an immediate offshoot of their prior observations, and was undertaken to (i) conduct a systematic metagenomic analysis on the genomic similarity between kitten aEPEC and human aEPEC isolates and (ii) identify genotypic and phenotypic signatures unique to virulent isolates of kitten aEPEC.

The authors first sequenced and compared the genomes of 12 aEPEC isolates from kittens (6 obtained from kittens with lethal infection [LI] and 6 from kittens with nonclinical colonization [NC]) with 149 previously sequenced aEPEC genomes and reference genomes from E. coli and Shigella. Phylogenomic analysis revealed that all the kitten isolates clustered within a single phylogroup, B1; although, they were not the sole occupants and were clustered with a subset of aEPEC isolates of human origin, highlighting the relatedness between aEPEC isolates from kittens and humans. These findings were reinforced by their large-scale BLAST score ratio (LS-BSR) analysis. The LS-BSR algorithm can analyze thousands of genomes to identify coding sequences and reveal their evolutionary relatedness (42). Using this approach, no genetic biomarkers were identified that were exclusively present in all kitten aEPEC isolates and absent from human aEPEC isolates and vice versa, reinforcing the genomic similarity between isolates from the two different sources. Similarly, isolates that were associated with lethal infection in kittens also did not collectively harbor unique genetic markers that distinguished them from the nonclinical colonizers. Interestingly, LS-BSR also revealed genetic diversity among the aEPEC isolates from kittens and revealed that the largest number of unique genetic clusters were isolate specific, with many of them present in a single genome. Whereas genetic diversity has been documented in aEPEC isolates of human origin, this represents the very first report to undertake a genome-wide analysis of genetic variation in kitten isolates. The findings herein mirror observations in humans and suggest that clinical manifestations accompanying aEPEC infections in kittens may, in part, stem from differential presence of isolate-specific virulence factors and/or differential expression of conserved virulence factors.

Besides genome-wide and gene-based comparisons, the authors also phenotyped the different kitten aEPEC isolates for diverse traits that have a well-documented role in different facets of the virulent lifestyle of EPEC (28, 43). The traits assayed—biofilm formation, motility, adherence, and A/E lesion formation on host cells—were selected to identify phenotypic signatures that would distinguish isolates responsible for lethal infection from nonclinical colonizers. No statistically significant differences were evident for biofilm formation, adherence, and A/E lesion formation. Kitten aEPEC isolates displayed a range of adherence patterns, including localized-adherence like (LAL) and diffusely adherent (DA) phenotypes (28). Furthermore, adherence did not correlate with the ability to recruit and polymerize actin to form A/E lesions. These two observations are in accordance with similar observations made for human aEPEC isolates (44). Remarkably, a statistically significant difference in motility was evident between the lethal infection isolates and the nonclinical colonization isolates, with the former being far more motile. These findings, too, are in line with prior observations of human EPEC isolates in which flagella enhanced localized adherence, microcolony formation, and A/E lesion morphogenesis—traits that positively correlate with diarrheal disease (45–47). However, this is the first report that demonstrates a direct correlation between motility and virulent diarrheic isolates from kittens, thereby providing a means to distinguish them from innocuous bystander isolates of aEPEC. Future studies investigating the molecular controls of flagellar-dependent motility will provide invaluable insight into the rewiring and potential divergence of the flagellar regulatory circuitry between the two groups of isolates and whether the observed correlation with virulence translates to causation.

Collectively, the present study by Watson et al. (41) highlights 3 key findings. First, and foremost, aEPEC isolates obtained from kittens are genotypically indistinguishable from human aEPEC isolates, raising the possibility that aEPECs can be zoonotically transmitted from kittens to humans and vice versa. The isolation of aEPEC strains from asymptomatic healthy kittens raises concern that kittens could serve as a reservoir for the bacterium. Second, the histopathology and bacterial load associated with aEPEC infections in kittens resemble those of aEPEC infection in children, suggesting that kittens could potentially serve as an invaluable model system to interrogate bacterial virulence factors and their cognate host targets to provide insight into the molecular epidemiology and treatment approaches to combat aEPEC. This would represent a significant accomplishment, as the current host model systems in use have numerous limitations that range from lack of reproducibility (e.g., mouse model) (48–50) and low phylogenetic relatedness to humans (e.g., C. elegans model) to divergent histopathology (e.g., C. rodentium infection of mice) (51–54). Third, individual aEPEC isolates harbored unique gene clusters, alluding to the fact that different isolates could potentially employ a distinct subset of virulence factors to cause disease. In summary, the findings of Watson et al. (41) are significant and shed light on the prevalence, distribution, and clinical manifestations of aEPEC in kittens while revealing genomic and pathological similarities with aEPEC isolates from humans. Thus, kittens may serve as the much-needed and awaited surrogate model system to investigate the molecular mechanisms of virulence of aEPEC in vivo.

The views expressed in this Commentary do not necessarily reflect the views of the journal or of ASM.