EnteroaggregativeEscherichia coli(EAEC): A Cause of Acute and Persistent Diarrhea of Worldwide Importance

Enteroaggregative Escherichia coli (EAEC) was first described in 1987 [1] in a child with acute diarrhea in Lima, Peru, and has since been linked with persistent diarrhea in children living in areas where EAEC is endemic [2], individuals with human immunodeficiency virus infection [3], and as a cause of diarrhea in travelers from industrialized countries visiting less-developed areas of the world. In several regions of the world, EAEC surpasses enterotoxigenic E. coli (ETEC) as the most common bacterial pathogen identified in diarrheal stool samples and in the United States; this emerging pathogen is becoming increasingly recognized as a leading cause of sporadic diarrhea in otherwise healthy adults and children [4,5]. Contaminated food appears to be the main source of EAEC infection and has been implicated in several foodborne outbreaks of diarrhea [6,7]. In Mexico, EAEC is commonly found in food, including desserts and salsa [8,9], and may explain the high rates of EAEC-related travelers' diarrhea seen in US visitors to that country. In contrast with ETEC, to which travelers are most susceptible during the first weeks of exposure, travelers remain susceptible to EAEC infection throughout their stay in Mexico [10], likely reflecting the particular ability of EAEC to evade the immune system and cause persistent infection. Importantly, EAEC is more likely to be resistant to antibiotics than are other diarrheagenic pathotypes. An emerging concept is the notion that certain enteric pathogens, including EAEC, cause not only acute and persistent diarrheal disease but can also persist in the human intestine subclinically, inducing chronic inflammation in the absence of dysentery. The chronic disruption in gut function by persistent infection has been linked to malnutrition and decreased physical and intellectual development in children [11]. In turn, malnutrition, including deficiencies in micronutrients, further facilitates infection and perpetuates the cycle of infection, malnutrition, and developmental deficits, increasing the burden of the disease by adding developmental deficits to the complications of acute diarrheal disease. Although animal models for the study of EAEC diarrhea are lacking, the development of the neonatal and weaned mouse murine model of chronic infection, as described in this issue by Roche et al [12], provides new insights on the interplay between persistent EAEC infection, developmental stunting, and malnutrition, and is an important advance in the field. EAEC pathogenesis is determined by the organism's ability to adhere to intestinal cells, produce enterotoxins and cytotoxins, and induce inflammation. EAEC are characterized by the ability to aggregate intimately with each other, adhere to human HEp-2 cells, and also attach to abiotic surfaces when grown in tissue culture plates. In this growth environment, EAEC colonies resemble “stacked bricks” when viewed by light microscopy. At the molecular level, the genes that encode for the aggregative phenotype are contained in a large plasmid (pAA). The aggR regulon present in pAA controls a number of plasmid genes coding for virulence factors and at least 2 pathogenicity islands in the EAEC chromosome. Dispersin, an antigenic antiaggregative protein encoded by the aap gene in the pAA plasmid and regulated by aggR, modulates fimbrial adhesion and facilitates penetration of the microorganism through intestinal mucus by binding to lipopolysaccharide and altering the electrostatic properties of the EAEC outer membrane surface. Although aap is not present in all EAEC and has been recently described in a proportion of E. coli strains not harboring genes characteristic of other diarrheagenic E. coli [13], this study provides evidence that undernutrition increases the bacterial load of EAEC infection as evidenced by quantitative polymerase chain reaction (PCR) of app. Additional studies that dissect how dispersin facilitates colonization and contributes to enteropathogen burden and how these properties correlate with outcomes of acute and chronic infection are needed. Shortly after the initial description of EAEC, the empirically derived molecular probe CVD432, also known as the AA probe, was developed by Baudry et al [14]. This probe provided the basis for a PCR-based assay a few years later [15]. The AA probe was later shown to hybridize with a region of the pAA plasmid coding for an ATP binding cassette transporter apparatus that is needed to translocate dispersin across the bacterial cell membrane [5]. Epidemiologic studies showed that the majority of isolates hybridizing with the AA probe also contained aggR and that EAEC isolates that hybridized with the AA probe were found to be associated with persisting diarrhea in case-controlled studies. It was also noted, however, that not all HEp-2 adherent EAEC strains isolated from humans with diarrhea hybridized with primers specific for aggR. This finding led to the suggestion that EAEC be classified into “typical” EAEC when aggR is present and “atypical” EAEC when aggR is absent [16]. The medical literature thus contains studies of EAEC that rely solely on the aggregative HEp-2 assay, studies that depend solely on the use of molecular probes, and yet others that use a combination of these 2 methods to identify EAEC. Studies of human intestinal biopsy samples and challenge studies involving volunteers exposed to various EAEC demonstrated that EAEC adherence and pathogenicity are variable, with some strains preferentially infecting the small bowel, whereas others involve both the small bowel and colon. Additionally, fecal inflammatory markers appear to be more commonly seen in stools of subjects with EAEC-related travelers' diarrhea acquired in Africa than in cases acquired in the Americas [17], likely reflecting infection of different sites in the gut. This heterogeneity among typical and atypical strains from geographically diverse regions further complicates the sorting of EAEC into discrete groups [18] and will influence organism reclassification. Although atypical EAEC has been identified in calves, piglets, and horses [19], animals have not been implicated as a source or reservoir for human infection. In the study by Roche et al [12], 2 well-characterized strains of typical EAEC possessing distinct adhesin fimbriae, demonstrate in the mouse model a difference in pathogenicity, suggesting that EAEC host species-specificity may be adhesinspecific. In humans, EAEC elicits an inflammatory response that is initiated by the interaction of flagellin and Toll-like receptor 5 that in turn results in the recruitment of inflammatory cells to the gut by interleukin 8 and other cytokines. In mice, the EAEC pic serine protease is not only mucinolytic and assists with establishing infection but may actually facilitate the use of mucin as a source carbon by EAEC [20] when other nutrients are unavailable or exhausted. A related serine protease, the plasmid-encoded toxin pet causes rounding and extrusion of enterocytes in human tissue explants. How EAEC survives the hostile environment of the gut is unknown. Evasion of host immunity may be related to the ability of EAEC to produce tightly aggregated colonies in the absence of dispersin and aggregin and its ability to form a layer of thick biofilm that can interfere with the penetration of colicins, phages, antibacterial peptides, such as lactoferrin, specific and nonspecific immunoglobulin A, and possibly antibiotics. In contrast to the streptomycin-treated murine model of EAEC infection [20], in which pathologic changes do not occur and which thus serves as a model for the study of EAEC colonization in the colon, an important observation of the Roche study is the finding that in colonic tissues, chronic infection with EAEC leads to inflammatory changes in the mucosa and depletion of mucin in goblet cells. Nevertheless, the development of additional models for the study of EAEC in the small bowel is still needed. With the recent publication of the complete genome of the prototypic strain EAEC 042 and its comparison with commensal and laboratory E. coli strains, we have learned that 551 genes are EAEC-specific [21] and that many of these genes are known or are putative novel virulence factors not previously identified in EAEC, including the presence of elements characteristic of type 2 secretion systems typically associated with the export of the heat-labile toxin of ETEC and colonization factors that are similar to the acfD of Vibrio. In addition, analysis of the EAEC genome has shown the presence of type 3 secretory system components that include intimin-like genes of enteropathogenic E. coli and relatives of the ipaH protein of Shigella, as well as evidence for type 5 and type 6 secretory system elements. How well these and other genes are represented in other strains that show preferential selectivity for the small intestine or the colon and originate from diverse regions in the world remains to be determined. The availability of an animal model to study the impact that EAEC has on growth and development and how undernutrition facilitates the persistence of EAEC infection is an important step in the development of interventions for the control and prevention of EAEC diarrhea and its long-term complications.