Shiga toxin (Stx)-producing Escherichia coli (STEC) can cause a broad spectrum of clinical symptoms in humans, ranging from haemolytic uraemic syndrome (HUS) to mild non-bloody diarrhoea or even asymptomatic carriage [Reference Caprioli1]. In particular, non-O157 STEC are considered emerging pathogens, despite being currently underrecognized because methods for their detection and isolation are not widely implemented. STEC infection is commonly acquired through the consumption of faecally contaminated food or water, through direct or indirect contact with animal carriers, mainly ruminants, or via secondary person-to-person transmission [Reference Caprioli1]. Enteropathogenic E. coli (EPEC) are one of the most common causes of infantile diarrhoea worldwide and are further divided into two subtypes, typical and atypical EPEC, depending on the presence or absence of the bundle-forming pilus (BFP) [Reference Trabulsi, Keller and Tardelli2]. In particular, atypical EPEC (aEPEC) are more prevalent compared to STEC in industrialized countries, where aEPEC are frequently identified both in children with diarrhoea and in healthy children [Reference Trabulsi, Keller and Tardelli2, Reference Wahl3]. Although there is no evidence of direct transmission from animals to humans, animal carriers have been suggested to be reservoirs for aEPEC infecting humans [Reference Trabulsi, Keller and Tardelli2].

On 30 May 2012, the clinical microbiological laboratory of the Hospital Complex of Navarre (CHNa) submitted a Stx1-positive stool culture to the Spanish National Reference Laboratory (SNRL) for further STEC diagnostic assays. The sample had been obtained from a 14-month-old girl reporting bloody diarrhoea. At the SNRL, both an O76:[H19] STEC and an O168:H6 aEPEC were recovered. Although STEC infections are not notifiable in Spain, because O76:H19 STEC has been associated with HUS [Reference Mellmann4] an epidemiological investigation was conducted. The girl's parents were interviewed by telephone, using a structured trawling questionnaire. The questionnaire included questions related to general food handling and hygienic procedures, as well as specific risk factors, including consumption of raw food, especially unpasteurized dairy products and potentially faecally contaminated vegetables, and non-disinfected water, as well as contacts with farm animals or pets and recent history of travel. The hypothesis-generating interview only identified consumption of vegetables grown in a family garden irrigated with well water and fertilized with sheep manure as a potential source of the girl's infection. As a consequence, single stool samples from the four household members, consisting of the index girl, her mother (age 32 years), father (33 years) and older sister (3 years), were obtained on days 36, 74, 137 and 201 (counted from the day the first STEC-positive sample was collected). Stool samples from four other family relatives, not sharing the same household but consuming the suspected vegetables, were also screened for STEC and EPEC on day 74. However, neither the suspected vegetables nor the sheep herd providing manure for the family garden could be sampled and no further action was taken.

At the CHNa, the production of Stx1 and Stx2 toxins in the stool culture from the index girl was investigated by using the Duopath Verotoxins immunochromatographic rapid test (Merck, Germany). The stool culture from the index girl, as well as all the stool samples from the follow-up on the family members, were submitted to the SNLR and screened for STEC and EPEC. For this purpose, samples were cultured on MacConkey agar (Becton Dickinson, USA) after a broth enrichment step. Bacterial growth from the first streaking area of the culture plate was tested for stx1, stx2 and eae genes by PCR [Reference Sánchez5]. When culture tested positive, individual E. coli-like colonies were tested using the same PCR to obtain the STEC or EPEC isolate, which was further confirmed biochemically as E. coli by the API 20E system (bioMérieux, France). All recovered STEC isolates were tested for the additional virulence genes ehxA and subAB by PCR [Reference Sánchez5], and the identification of stx1 and stx2 subtypes was performed using a recently developed PCR-based method [Reference Scheutz6]. All recovered EPEC isolates were tested for the presence of bfpA gene [Reference Gunzburg, Tornieporth and Riley7], in order to classify them as typical or atypical EPEC. STEC and EPEC isolates were further typed by conventional O:H serotyping, genetic H serotyping by PCR amplifying and sequencing the fliC gene [Reference Machado, Grimont and Grimont8] in non-motile isolates (results denoted in square brackets) and pulsed-field gel electrophoresis (PFGE) with XbaI according to the PulseNet protocol for E. coli O157:H7 [9]. Additionally, STEC isolates were typed by multilocus sequence typing (MLST) [Reference Wirth10]. Cluster analysis was performed using the Dice coefficient and the unweighted pair-group method with arithmetic averages (UPGMA) in InfoQuestFP v. 4·5 (Bio-Rad, UK).

On day 36, no further STEC were isolated from the girl's stool sample, but EPEC isolates were obtained. STEC and EPEC isolates were obtained from the father's stool sample and a single STEC isolate was identified in the mother's stool sample. A single EPEC isolate was obtained from the older sister (Table 1). During the follow-up period, on day 74 the father still presented with STEC and the girl with EPEC. On day 137, only the girl with EPEC remained positive (Table 1). Finally on day 201, stool samples from all four family members tested negative for both STEC and EPEC. All other relatives were found to be negative for STEC and EPEC on day 74. All recovered STEC isolates tested negative for eae but positive for ehxA and subAB and belonged to serotype O76:H19/[H19] (Table 1). Subtyping of the stx genes resulted in the detection of subtypes stx2b and/or stx1c (Table 1). The EPEC isolates belonged to multiple serotypes (O8:H25, O51:H49, O168:H6, O180:[H2], ONT:H6, ONT:H29) and were classified as aEPEC, as all of them tested negative for bfpA (Table 1).

Table 1. Characteristics and molecular typing results for STEC and aEPEC isolates from symptomatic and asymptomatic family members

STEC, Shiga toxin-producing Escherichia coli; aEPEC, atypical enteropathogenic E. coli; PFGE, Pulsed-field gel electrophoresis; MLST, multilocus sequence typing; ST, sequence type; n.d., not done; ONT, O antigen non-typable.

* Days counted from the day the first STEC-positive stool sample was collected.

† Genetic H serotyping results in non-motile isolates are given in square brackets [H].

‡ Symptomatic when the stool sample was collected.

PFGE results showed two different profiles for the O76:[H19] STEC isolate from the symptomatic girl (profile 2) and for the three O76:H19 STEC isolates from her asymptomatic parents (profile 1) (Fig. 1). It has been widely demonstrated that the loss of stx genes due to spontaneous curing of stx-carrying phages in STEC clinical isolates involves changes in the PFGE patterns, with isolates differing by 2–5 bands [Reference Bielaszewska11]. As the STEC O76:H19 isolates in the present study differed only by five bands (88·4% similarity), the two different PFGE profiles found in them could be explained by the loss of the stx2b-carrying phage from profile 2 (stx2b-positive) to profile 1 (stx2b-negative). Nevertheless, STEC O76:H19 isolates also differed in their motility (the single profile 2 isolate was non-motile while all three profile 1 isolates were motile), thus contradicting the idea that all STEC O76:H19 isolates in the present study could belong to a single strain. Moreover, MLST analysis classified all O76:H19/[H19] STEC isolates as belonging to sequence type 675 (Table 1), as do the O76:H19 reference strain (HUSEC039) in the German collection of representative HUS-associated enterohaemorrhagic E. coli (HUSEC) [Reference Mellmann4]. The seven aEPEC isolates revealed six different PFGE profiles, with one being identified on two occasions, 101 days apart, in the girl's stool samples (profile 6) (Table 1, Fig. 1).

Fig. 1. Pulsed-field gel electrophoresis (PFGE) profiles of Shiga toxin-producing Escherichia coli and atypical enteropathogenic E. coli isolates obtained from the stool samples of a girl and her asymptomatic family members. The scales at the top indicate the similarity indices (in percentages) and molecular sizes (in kilobases).

This study represents the first description of both an O76:H19 STEC infection and a mixed infection with aEPEC in Spain. In total, two different STEC strains and six aEPEC strains were isolated and characterized in a household during a 5-month period. Of the STEC-infected family members only the 14-month-old girl developed bloody diarrhoea but did not require hospitalization or antibiotic treatment, and her symptoms resolved between the first and second stool sampling. None of the other STEC-infected family members developed clinically symptomatic disease. The O76:[H19] isolate from the index girl carried both stx1 and stx2 while O76:H19 isolates from the parents only carried stx1, shown to be less frequently associated with severe human disease than stx2 [Reference Caprioli1]. Both serotypes were eae-negative and ehxA-, subAB-positive. Despite intimin production representing a common feature of STEC strains associated with severe human disease, eae-negative STEC strains have also been implicated in outbreaks and serious disease [Reference Newton12]. Moreover, it has been reported that the subtilase cytotoxin, encoded by subAB, might contribute to the virulence of eae-negative STEC strains in synergy with Shiga toxins [Reference Michelacci13], which could explain the clinical relevance in our index case. Additionally, STEC O76:H19 has been recognized to be an important non-O157 STEC associated with human illness and in particular with causing HUS [Reference Mellmann4]. Apart from the index girl, her older sister was the only aEPEC-infected family member reporting diarrhoea (before the first STEC-positive stool sample was collected), but symptoms rapidly resolved and she did not required medical care. Although the epidemiological association of aEPEC with diarrhoea is still controversial, their high prevalence worldwide and involvement in diarrhoeal outbreaks [Reference Wahl3] support the idea that some aEPEC strains are diarrhoeagenic.

The questionnaire identified consumption of vegetables fertilized with sheep manure as a likely source of infection. Sheep have been reported as a common reservoir for STEC infection and O76:H19 STEC strains with the same virulence profiles have previously been isolated from sheep [Reference Michelacci13]. Although there is no evidence of direct transmission from animals to humans, aEPEC have also been isolated from sheep and exposure to faecal pollution from a sheep herd was the suspected source of infection in a recently reported outbreak of mixed STEC and aEPEC infection in Norwegian children at a day-care centre [Reference Wahl3].

The PFGE analysis revealed prolonged carriage in two family members. It was confirmed that the father asymptomatically shed STEC (profile 1) at least for 38 days (from day 36 to day 74), with the mother being infected with the same strain on day 36 (Table 1). The index girl asymptomatically shed aEPEC (profile 6) for 101 days after resolving her STEC-associated bloody diarrhoea episode (Table 1). Prolonged asymptomatic STEC carriage has been best characterized in children, but also reported in adults, even over a 1-year period [Reference Staples14, Reference Kuusi15].

Family clusters of STEC infection have been reported to be common, with up to 50% of STEC infections being family-related, e.g. in Finland [Reference Lienemann16]. In addition, both family clusters and outbreaks of mixed STEC and EPEC infection have previously been reported [Reference Wahl3, Reference Staples14]. Although there was no conclusive evidence regarding the source of infection in this family cluster, prolonged asymptomatic carriage and close household contact between the family members pose a risk of cross-infections. This circumstance is underlined by the fact that those relatives who consumed the same vegetables but did not share the same household were not infected. Therefore, handwashing when handling food or young babies is particularly necessary to prevent STEC and other diarrhoeagenic E. coli infections in households.