Introduction

Foodborne pathogens are an important public health issue in many countries and are responsible for a substantial burden of disease in the developed world. In developing countries, such as Argentina, Paraguay, and Chile, this problem is of great concern [1]. One of the microorganisms involved in foodborne diseases is Shiga toxin-producing Escherichia coli (STEC), a diverse group of bacteria capable of causing severe human diseases such as hemorrhagic colitis (HC) and the hemolytic-uremic syndrome (HUS) [2]. Cattle are the main reservoir of STEC; however, pigs, dairy products, nuts, seeds, water -which has increasingly become a concern as a source of contamination for fruits and vegetables-, and person to person contact have been shown to commonly participate in transmission [3,4,5]. STEC surveillance systems are different in each Latin American country. In general, disease incidence data relies primarily on either the foodborne disease surveillance system (Argentina and Paraguay) and/or through the acute diarrheal surveillance system (Argentina, Chile, and Paraguay, among others). Argentina has the highest number of cases caused by STEC worldwide: HUS incidence is 8.4 cases per 100,000 children less than 5 years old [6]. On the other hand, the incidence is 3.4 cases per 100,000 children in Chile (Metropolitan Region) [7]. In both countries HUS represents one of the main causes of acute renal failure in children with lethality ranges from 2 to 6% [8]. In Paraguay, routine surveillance is performed with all the stool cultures that are sent to the Central Public Health Laboratory. The incidence of HUS is low. Between 2013 and 2015, eight STEC diarrhea cases and ten HUS cases without STEC isolation were reported [8, 9].

STEC refers to the pathotype capable of producing Shiga toxin type 1 (Stx1), type 2 (Stx2), or both, which are encoded by stx1 and stx2 genes, respectively [2]. Epidemiological studies suggest that STEC strains encoding stx2 are more virulent than those harboring stx1 only [10]. Although Stx1 is less cytotoxic than Stx2, it may potentially cause disease in humans, but the information about its clinical implications is limited [11]. In addition, some STEC strains carry the Locus of Enterocyte Effacement (LEE), a pathogenicity island (PAI) which encodes genes necessary to produce attaching and effacing (A/E) lesions on the intestinal epithelium. LEE is also carried by enteropathogenic E. coli (EPEC) [12] and Escherichia albertii strains which might have been misidentified as LEE-positive STEC or EPEC because they carry the eae gene [13]. However, STEC strains lacking LEE (LEE-negative), such as those belonging to serogroups O91, O113 and O174 [14], have also been isolated from cases of severe illness, including HUS [15]. A great number of adhesins, including the pO113-encoded autoagglutinating adhesin (Saa), have been linked to the pathogenesis of LEE-negative STEC [16]. Montero et al. [17] described and characterized a member of the Heat resistant agglutinin family (Hra Family), named Hemagglutinin from Shiga toxin-producing E. coli (Hes) that participates in several colonization-associated phenotypes, including hemagglutination, adhesion and autoaggregation. Hes is encoded by a gene located in a PAI of 86-kb chromosomal mosaic element named the Locus of Adhesion and Autoaggregation (LAA), which contains 80 genes organized into four modules: module I (hes and other genes), module II (iha, lesP and others genes), module III (pagC, tpsA, tpsB and other genes) and module IV (agn43 and other genes). Moreover, LAA may be present as a “complete” (4 modules) or “incomplete” (with less than 4 modules) structure [17].

LAA presence is associated with severe human disease and although the mechanisms used by LEE-negative STEC strains to colonize the human intestine are not clear yet [17]. Recently, the role of LAA in intestinal colonization was demonstrated in a murine model of STEC infection, suggesting that LAA may be also involved in the adherence of STEC to the human intestine [18, 19]. Moreover, the association of LAA with stx1a, stx2a, stx2d and cdtB toxin genes that cause severe disease has been demonstrated [18]. In a previous study in Argentina, our group has demonstrated the presence of LAA in 46% of LEE-negative STEC strains isolated from different sources [20]. Because the incidence of HUS caused by LEE-negative STEC strains has been increased in several countries [8], in this study, we aimed to analyze the distribution of LAA and hes in LEE-negative STEC strains isolated in Chile and Paraguay from different sources.

Materials and Methods

Strains and Serotyping

A total of 128 LEE-negative STEC strains were analyzed: 72 from Chile and 56 from Paraguay. STEC strains were isolated from beef cattle (n = 94), meat (n = 31), cheese (n = 2) and wild bird (n = 1). These isolates were previously analyzed for the presence of stx1, stx2, and saa genes by PCR (Table 1) [21]. The serogroup was determined by micro-agglutination test described by Guinée et al. [22] and modified by Blanco et al. [23] (Table 1).

Table 1 Virulence profile, serogroups, and LAA presence of LEE-negative STEC strains
Full size table

PCR Amplification

STEC strains were characterized by multiplex PCR to detect LAA modules I, II, and III [17]. Additionally, the presence of agn43, as a marker of module IV, and hes were characterized by monoplex PCR [20, 24]. One LEE positive STEC strain was used as a negative control (O157:H7 EDL933). PCR reactions (multiplex and monoplex) were performed and standardized in a total volume of 50 μl by using a T-17 thermal cycler (Ivema). The reaction mixture contained 500 mM KCl, 100 mM Tris–HCl pH 9, Triton X-100, 25 mM MgCl2, 200 μM 4 deoxynucleotides (dATP, dGTP, dCTP, dTTP), 1U of TaqDNA Polymerase Highway® (Inbio), and 5 μl of DNA. The DNA was obtained by boiling bacteria suspended in sterile water for 10 min as previously described [25]. The LAA and hes primers were reported by Montero et al. [17]. Amplification products were separated by electrophoresis on 2% agarose gels containing 0.8 μg/ml of ethidium bromide in running buffer and were visualized in a UV transilluminator.

Results

STEC strains were considered as LAA-complete when they harbored modules I, II, III, and IV and were considered as LAA-incomplete when less than four modules were detected.

Complete LAA structure was identified in strains from Paraguay and Chile in similar frequencies: 41.0% (23/56) and 41.6% (30/72), respectively (Figs. 1 and 2). Similarly, the gene hes was detected in 50% of the isolates from Paraguay (28/56) and Chile (36/72) (Table 1). STEC strains carrying LAA (complete or incomplete) were distributed in 23 serogroups. The most frequent were O91 (4/4, 100%), O113 (5/6, 83%), O103 (6/9, 67%) and O174 (2/5, 40%) (Table 1).

Fig. 1
figure 1

Distribution of LAA detection in STEC strains isolated at different origins from Paraguay

Full size image
Fig. 2
figure 2

Distribution of LAA detection in STEC strains isolated at different origins from Chile

Full size image

Regarding toxins, the stx2 gene was detected in 30% (7/23) and 73% (22/30) of STEC strains harboring LAA from Paraguay and Chile, respectively (Table 1).

Discussion

Although E. coli O157:H7 is the most prevalent serotype associated with HUS, there is growing concern on the global emergence of LEE-negative STEC that have been associated with outbreaks and/or severe human illness [26]. LEE-negative STEC strains have been detected in humans, animals, food, and the environment. Still, the pathogenic mechanisms used by this group of strains to colonize humans are yet to be elucidated [27]. LAA is a PAI of an emerging group of STEC strains that cause severe diseases in humans. In fact, the complete LAA island is present in LEE-negative STEC strains which have been isolated from cases of hemorrhagic colitis and HUS [17].

In this study, the complete LAA was widely distributed in STEC strains from different origins and belonging to several serogroups. LEE-negative STEC strains isolated from Chile (41.6%) and Paraguay (41.0%) harbored the complete LAA island with a frequency of detection that is similar to that previously reported Argentina (46%) [20].

Stx2-producing strains are more often associated with HUS than strains that produce Stx1 [28]. In this study, we detected stx2 in 30% and 73% of the isolated LAA positive obtained from Paraguay and Chile, respectively. Likewise, Colello et al. [20] found a significant association between the presence of a complete LAA island and stx2 in strains isolated in Argentina. In addition, LEE-negative STEC strains of serogroups O91, O113 and O174 have been isolated from cases of severe illness [29], and in our study most of the isolates belonging to these serogroups harbored LAA. PAIs like LAA has had a remarkable role in the emergence of LEE-negative STEC strains and may contribute to the evolution and virulence of pathogenic E. coli [18].

Montero et al. [17] suggested that hes is a potential genetic marker for LAA, and this raises a new possible epidemiological scenario [17] for STEC since it is widely distributed in LEE-negative STEC strains. Our results revealed that hes was present in all LAA positive strains, carrying the complete island. In concordance with Colello et al. [20], we also observed that hes was one of the most prevalent genes in LEE-negative STEC strains [18, 20].

Therefore, our work provides new data about the presence of LAA and the hes gene in STEC strains from different sources and isolated in two Latin-American countries. These results support previous findings that suggest that the current STEC classification based on the presence of LEE might be insufficient to detect human pathogenic STEC strains, and that hes detection could complement current genetic evaluation schemes to detect strains representing risk [17, 30].

Concluding, to define a STEC strain as pathogenic is complex because there are not combinations of markers that can predict the potential of a STEC strain to cause human disease. In LEE positive strains, stx2 and eae are predictive markers for severe disease. In LEE-negative STEC strains, the scenario is even more complex because there are not additional virulence factors to stx2 associated with severe disease. In this context, hes could be used as a marker allowing classify the LEE negative STEC strains as LAA positive or negative STEC strains, and therefore, inferring the potential risk of LEE-negative strains.