PATHOGENICITY AND VIRULENCE

Typing of intimin (eae) genes from enteropathogenic Escherichia coli (EPEC) isolated from children with diarrhoea in Montevideo, Uruguay: identification of two novel intimin variants (μB and ξR/β2B)

  • 1Laboratorio de Referencia de E. coli (LREC), Departamento de Microbioloxía e Parasitoloxía, Universidade de Santiago de Compostela, 27002 Lugo, Spain
  • 2Unidade de Microbioloxía Clínica, Complexo Hospitalario Xeral-Calde, 27004 Lugo, Spain
  • 3Departamento de Bacteriología y Virología, Instituto de Higiene, Facultad de Medicina, Universidad de la República, CP 11600 Montevideo, Uruguay
  • Correspondence
    Jorge Blanco
    jba{at}lugo.usc.es
  • Journal of Medical Microbiology 2006; 55(9):1165–1174 · https://doi.org/10.1099/jmm.0.46518-0

    View at publisher PubMed

    Abstract

    A total of 71 enteropathogenic Escherichia coli (EPEC) strains isolated from children with diarrhoea in Montevideo, Uruguay, were characterized in this study. PCR showed that 57 isolates carried eae and bfp genes (typical EPEC strains), and 14 possessed only the eae gene (atypical EPEC strains). These EPEC strains belonged to 21 O : H serotypes, including eight novel serotypes not previously reported among human EPEC in other studies. However, 72 % belonged to only four serotypes: O55 : H− (six strains), O111 : H2 (13 strains), O111 : H− (14 strains) and O119 : H6 (18 strains). Nine intimin types, namely, α1 (two O142 strains), β1 (29 strains, including 13 O111 : H2 and 14 O111 : H−), γ1 (three O55 : H− strains), θ (five strains, including three strains with H40 antigen), κ (two strains), ε1 (one strain), λ (one strain), μB (six strains of serotypes O55 : H51 and O55 : H−) and ξR/β2B (22 strains, including 18 O119 : H6) were detected among the 71 EPEC strains. The authors have identified two novel intimin genes (μB and ξR/β2B) in typical EPEC strains of serotypes O55 : H51/H− and O119 : H6/H−. The complete nucleotide sequences of the novel μB and ξR/β2 variant genes were determined. PFGE typing after XbaI DNA digestion was performed on 44 representative EPEC strains. Genomic DNA fingerprinting revealed 44 distinct restriction patterns and the strains were clustered in 12 groups. Only 15 strains clustered in six groups of closely related (similarity >85 %) PFGE patterns, suggesting the prevailing clonal diversity among EPEC strains isolated from children with diarrhoea in Montevideo.

    • The GenBank/EMBL/DDBJ accession numbers for the complete nucleotide sequences of the novel μB and ξR/β2B variant intimin genes are AJ705049 and AJ715407, respectively.

    INTRODUCTION

    Enteropathogenic Escherichia coli (EPEC) was first recognized as a cause of infantile diarrhoea in the 1940s, and was associated with outbreaks in hospitals and nurseries in the UK. During one such outbreak, Bray (1945) prepared antiserum to a strain of E. coli isolated from a patient with diarrhoea and used this antiserum to show that the epidemic strains belonged to the same serogroup, later recognized as E. coli O111. In 1987, the World Health Organization (1987) recognized EPEC serotypes of 12 different O serogroups (O26, O55, O86, O111, O114, O119, O125, O126, O127, O128, O142 and O158). Although large outbreaks of infant diarrhoea due to EPEC have largely disappeared from industrialized countries, EPEC remains an important cause of potentially fatal infant diarrhoea in developing countries (Trabulsi et al., 2002). In Brazil, for example, EPEC strains are recovered from up to 30 % of cases of diarrhoea in infants of low socioeconomic level (Gomes et al., 1996). In Uruguay, diarrhoeagenic E. coli (DEC) are the most frequently identified bacterial agents associated with gastroenteritis of children in low-income population groups. Although rotavirus is increasingly recognized as a frequent cause of these infections, EPEC still accounts for a vast proportion of diarrhoea cases. Shigella flexneri and Campylobacter jejuni are also frequently isolated, especially from episodes of bloody diarrhoea (Torres et al., 2001).

    For decades, the mechanisms by which EPEC caused diarrhoea were unknown, and this diarrhoeagenic pathotype could only be identified on the basis of O : H serotyping. However, since 1979, numerous advances in our understanding of the pathogenesis of EPEC diarrhoea have been made through the application of tissue culture and molecular genetic methods (Frankel et al., 1998; Trabulsi et al., 2002; Kaper et al., 2004). One of the central mechanisms of EPEC pathogenesis is the formation of attaching and effacing (A/E) lesions, which is characterized by the intimate attachment of the bacteria to the enterocyte membrane and by the effacement of the microvilli of the enterocyte (Jerse et al., 1990; Kaper et al., 1998). The ability to produce A/E lesions has also been detected in strains of Shiga-toxin-producing E. coli (STEC) and in strains of other bacterial species. STEC and EPEC that cause characteristic A/E lesions in the intestinal mucosa are also classified as attaching and effacing E. coli (AEEC). The genetic determinants for the production of A/E lesions are located on a large chromosomal pathogenicity island, the locus of enterocyte effacement (LEE; Kaper et al., 1998). The central portion of LEE encodes intimin (Eae, a 94–97 kDa outer-membrane protein) and Tir, the intimin receptor, which is translocated into the host cell membrane by a type III secretion system. Differentiation of intimin alleles represents an important tool for EPEC and STEC typing in routine diagnostics as well as in pathogenesis, epidemiological, clonal and immunological studies. The C-terminal end of intimin is responsible for receptor binding, and it has been suggested that different intimins may be responsible for different host tissue cell tropisms (Torres et al., 2005). The 5′ regions of eae genes are conserved, whereas the 3′ regions are heterogeneous. This observation has led to the construction of universal PCR primers and allele-specific PCR primers, which have made it possible to differentiate at present 21 variants of the eae gene encoding 21 different intimin types and subtypes: α1, α2, β1, ξR/β2B, δ/β2O, κ, γ1, γ2, θ, ε1, νR/ε2, ζ, ε1, ε2, ι1, μR/ι2, λ, μB, νB, ξB and o) (Gannon et al., 1993; Adu-Bobie et al., 1998; Oswald et al., 2000; Tarr & Whittam, 2002; Zhang et al., 2002; Blanco et al., 2004a, b, d, 2005, 2006; Garrido et al., 2006).

    In 1996, EPEC strains were defined as intimin-containing DEC isolates that possess the ability to form A/E lesions on intestinal cells and do not possess Shiga toxin (stx) genes (Kaper, 1996). However, EPEC strains can be further classified as typical or atypical. Typical EPEC strains possess a virulence plasmid (EAF plasmid) that includes genes encoding the bundle-forming pilus (Bfp), which is required for localized adherence on cultured epithelial cells; atypical EPEC strains do not possess the EAF plasmid with the bfpA gene (Trabulsi et al., 2002). In industrialized countries, atypical EPEC (eae+ bfpA stx) are more frequently isolated from diarrhoeal cases, whereas typical EPEC (eae+ bfpA+ stx) dominate in developing countries (Trabulsi et al., 2002).

    The aim of this study was to establish the serotypes, intimin types and genetic diversity of typical and atypical EPEC strains isolated from children with diarrhoea in Montevideo, Uruguay.

    METHODS

    E. coli isolates and control strains.

    A total of 71 EPEC isolates from children with diarrhoea (sporadic cases) studied in Montevideo between 1990 and 1999 were characterized in this study (Torres et al. 2001; F. Schelotto and others, unpublished data). Cultures of faeces yielding EPEC strains were performed as a part of medical and microbiological surveys aimed at the determination of the aetiology of acute and persistent diarrhoea, the antibiotic resistance of identified pathogens, and the adequacy of procedures for feeding and rehydrating affected children. Most infections were acquired in the community, but were studied when the children attended Montevideo Children's Hospital.

    E. coli strains used as controls were: EPEC-E2348/69 (human, O127 : H6, bfpA, eae-α1), AEEC-IH2498a (human, O125 : H6, eae-α2), EPEC-337 (human, O111 : H2, bfpA, eae-β1), EPEC-359 (human, O119 : H6, bfpA, eae-ξR/β2B), EPEC-BL152.1 (human, O86 : H34, bfpA, eae-δ/β2O), AEEC-6044/95 (human, O118 : H5, eae-κ), STEC-EDL933 (human, O157 : H7, stx1, stx2, eae-γ1), STEC-TW07926 (human, O111 : H8, stx1, stx2, eae-θ), STEC-VTB-286 (bovine, O103 : H2, stx1, eae-ε1), AEEC-IH3205a (human, O123 : H19, eae-νR/ε2), STEC-VTO-50 (ovine, O156 : H−, stx1, eae-ζ), AEEC-CF11201 (human, O125 : H−, eae-ε1), H03/53199a (human, ONT : H45, eae-ε2), AEEC-7476/96 (human, O145 : H4, eae-ι1), AEEC-217-2 (human, O101 : H−, eae-μR/ι2), AEEC-68-4 (human, O34 : H−, eae-λ), EPEC-373 (human, bfpA, O55 : H51, eae-μB), AEEC-IH1229a (human, O10 : H−, eae-νB), STEC-B49 (bovine, O80 : H−, stx1, eae-ξB), IH2997f (human, O129 : H−, eae-o) and K12-185 (negative for stx1, stx2, bfpA and eae genes). Strains were stored at room temperature in nutrient broth with 0.75 % agar.

    Serotyping.

    The determination of O and H antigens was carried out by the method of Guinée et al. (1981), employing all available O (O1–O185) and H (H1–H56) antisera. All antisera were obtained and absorbed with the corresponding cross-reacting antigens to remove non-specific agglutinins. The O antisera were produced in the Laboratorio de Referencia de E. coli (LREC), Lugo, Spain (), and the H antisera were obtained from the Statens Serum Institut, Copenhagen, Denmark. E. coli strains representing the novel O groups O182–O185 were kindly provided by Flemming Scheutz, Statens Serum Institut.

    Detection of DEC genes.

    Detection of virulence genes of EPEC and other DEC was performed by PCR using specific primers for amplification of eight virulence genes of distinct DEC groups: STEC (stx1 and stx2) (Blanco et al., 2003), EPEC/STEC (eae) (Blanco et al., 2003), EPEC (bfpA) (Gunzburg et al., 1995), enteroinvasive E. coli (EIEC) (ipaH) (Tornieporth et al., 1995), enteroaggregative E. coli (EAEC) (pCDV432) (Schmidt et al., 1995) and enterotoxigenic E. coli (ETEC) (eltA and est) (Schultsz et al., 1994; Blanco et al., 2006).

    Typing of intimin (eae) genes.

    Typing of intimin genes into eae α1, α2, β1, ξR/β2B, δ/β2O, κ, γ1, γ2, θ, ε1, νR/ε2, ζ, ε1, ε2, ι1, μR/ι2, λ, μB, νB, ξB and o was performed by PCR as previously described (Blanco et al., 2003, 2004b, 2005, 2006). Base sequences and predicted sizes of amplified products for the specific oligonucleotide primers used in this study are shown in Table 1. The oligonucleotide primers were designed by us according to the nucleotide sequences of the virulence genes. Isolates positive for the eae gene with EAE-1 and EAE-2 primers were further analysed with all different variant primers. Due to the high sequence similarity, specific primers could not be designed to distinguish between eae-γ2 and eae-θ, eae-δ and eae-κ, or eae-ε1 and eae-ε2 genes. In these cases it was necessary to establish the nucleotide sequence of a fragment from the 3′ variable region of the eae gene.

    Table 1.

    PCR primers used for amplifying and typing of intimin eae genes

    Sequencing of the intimin (eae) genes.

    The nucleotide sequence of the amplification products purified with a QIAquick DNA purification kit (Qiagen) was determined by the dideoxynucleotide triphosphate chain-termination method of Sanger, with the BigDye Terminator v3.1 Cycle Sequencing kit and an ABI 3100 Genetic Analyser (Applied Bio-Systems).

    Nucleotide sequence accession numbers.

    The eae sequences of strains analysed were deposited in the European Bioinformatics Institute (EMBL Nucleotide Sequence Database), and the accession numbers assigned are indicated in Table 2 and in Results.

    Table 2.

    Intimin types and serotypes of typical and atypical EPEC strains isolated from children with diarrhoea in Montevideo

    Phylogenetic analyses.

    The percentage identities of nucleotides of the 21 eae variants were calculated with the clustal w program (Thompson et al., 1994) included in the EMBL software (). An unweighted pair group method with arithmetic means (UPGMA) tree for amino acid sequences was constructed from a matrix of uncorrected p distances by using mega 3.1 (Kumar et al., 2004). Robustness of the tree was tested with bootstrapping (1000 replicates). The tree was rooted using the midpoint rooting option ().

    Macrorestriction fragment analysis by PFGE.

    PFGE was performed in a CHEF Mapper system (Bio-Rad) at 14 °C in 0.5× Tris/borate/EDTA by the Enternet-proposed standard-protocol for PFGE (). Cleavage of the agarose-embedded DNA was achieved with 0.2–0.8 U μl−1 Xbal (Roche) according to the manufacturer's instructions. Run times and pulse times were 2.20–54.0 s for 22 h with linear ramping. To perform the comparison of the PFGE pulsotypes, TIFF files were analysed with BioNumerics software (Applied Maths). Cluster analysis of the Dice similarity indices based on UPGMA was done to generate a dendrogram describing the relationship among EPEC pulsotypes. A difference of at least one restriction fragment in the patterns was considered the criterion for discriminating between clones.

    RESULTS

    Virulence genes

    A total of 71 EPEC isolates were characterized in this study. PCR showed that 57 isolates carried eae and bfpA genes (typical EPEC strains), and 14 possessed only the eae gene (atypical EPEC strains). All 71 EPEC strains were negative for the other six DEC virulence genes (stx1, stx2, ipaH, pCDV432, eltA and sta) investigated.

    Serotypes

    EPEC strains belonged to 11 O serogroups, 10 H flagellar antigen types and 21 O : H serotypes, including eight novel serotypes not previously reported among human EPEC in other studies (Table 2). Of the strains, 83 % were of three O serogroups (O55, O111 and O119), 45 % expressed two H flagellar antigens (H2 and H6) and 72 % belonged to only four serotypes: O55 : H− (six strains), O111 : H2 (13 strains), O111 : H− (14 strains) and O119 : H6 (18 strains).

    Typing of eae (intimin) genes

    Nine intimin types, namely, α1 (two O142 strains), β1 (29 strains, including 13 O111 : H2 and 14 O111 : H−), ξR/β2B (22 strains, including 18 O119 : H6), γ1 (three O55 : H− strains), θ (five strains, including three strains with H40 antigen), κ (two strains), ε1 (one strain), λ (one strain) and μB (six strains of serotypes O55 : H51 and O55 : H−) were detected among the 71 EPEC strains, and none of the strains was positive for intimin types α2, γ2, δ/β2O, ζ, ε1, ε2, ι1, μR/ι2, νB, νR/ε2 or ξB (Table 2).

    Identification of two novel intimin variant genes: sequence comparison and evolutionary analysis of E. coli intimin genes

    A fragment of the 3′ variable region of the eae gene from the 24 representative EPEC strains was sequenced. We identified in typical EPEC strains of serotypes O55 : H51/H− and O119 : H6/H− two novel intimin genes eae-μB and eae-ξR/β2B that show less than 95 % nucleotide sequence identity with existing intimin genes. The complete nucleotide sequences of the novel μB (AJ705049) and ξR/β2B (AJ715407) variant genes were determined (Table 2). As the complete sequence of eae genes encoding intimins δ/β2O (AJ875027), λ (AJ715409) and ξB (AJ705051) was not available in public databases, they were also sequenced.

    We determined the genetic relationship of the 21 eae variants: α1 (M58154), α2 (AF530555), β1 (AF200363), ξR/β2B (AJ715407), δ/β2O (AJ875027), κ (AJ308552), γ1 (AF071034), γ2 (AF025311), θ (AF449418), ε1 (AF116899), νR/ε2 (AF530554), ζ (AJ271407), ε1 (AJ308550), ε2 (AJ876652), ι1 (AJ308551), μR/ι2 (AF530553), λ (AJ715409), μB (AJ705049), νB (AJ705050), ξB (AJ705051) and o (AJ876648) (Table 3). Since the nucleotide sequences analysed were of different lengths, we used clustal w (Thompson et al., 1994) for optimal sequence alignment. Identities of 93, 91 and 91 % were calculated between the novel eae-μB variant and eae-γ1, eae-γ2 and eae-θ genes, respectively. The percentage identities between the novel eae-ξR/β2B gene and the eae-β1, eae-δ/β2O and κ genes were 90, 93 and 94 %, respectively.

    Table 3.

    Pairwise alignments calculated with clustal w

    Percentage identities are shown. Values in bold type represent the maximum percentage identities between different genes.

    The phylogenetic tree for amino acid sequences of the intimin variants constructed by the UPGMA method (uncorrected p distances) of mega 3.1 revealed six groups of closely related intimins: (i) α1, α2, ζ, νB and o; (ii) λ; (iii) β1, ξR/β2B, δ/β2O and κ; (iv) ε1, ξB, νR/ε2, ε1 and ε2; (v) γ1, μB, γ2 and θ; and (vi) ι1 and μR/ι2 (Fig. 1).

    Figure image not available in archive
    Fig. 1.

    Phylogenetic tree for amino acid sequences of the intimin variants constructed by the UPGMA method (uncorrected p distances) of mega 3.1. Numbers at nodes are the percentage of bootstrap replications in which a particular node was supported. The genetic distance is shown on the scale bar. Phylogenetic analysis revealed six groups of closely related intimins: (i) α1, α2, ζ, νB and o; (ii) λ; (iii) β1, ξR/β2B, δ/β2O and κ; (iv) ε1, ξB, νR/ε2, ε1 and ε2; (v) γ1, μB, γ2 and θ; and (vi) ι1 and μR/ι2.

    Genetic diversity of EPEC: PFGE patterns

    Forty-four EPEC strains were selected to be analysed by PFGE, including the most prevalent serotypes. These included 16 O111 strains (eight O111 : H2, seven O111 : H−, one O111 : H21); 13 O119 : H6 strains; eight O55 strains (four O55 : H−, three O55 : H51, one O55 : H40), and other serotypes (O127 : H40, O33 : H6, ONT : H−, O142 : HNT, O142 : H21, O153 : H−, ONT : H51) with one strain each. Genomic DNA fingerprinting of these 44 EPEC strains by PFGE revealed 44 distinct XbaI restriction patterns, considering a difference of at least one restriction fragment in the patterns as the criterion for discriminating between them. None of the strains presented an identical PFGE pattern to another. In the dendrogram produced by the UPGMA algorithm, the isolates were clustered in 12 groups (1–12 strains per group) of 60 % similar strains according to the Dice similarity index (Fig. 2). Only 15 strains clustered in six groups of closely related (similarity >85 %) PFGE patterns: (i) FV336 and FV347; (ii) FV354, FV337, FV367, FV344 and FV345; (iii) FG187 and FV362; (iv) FG240 and FV376; (v) FV359 and FV360; and (vi) FV361 and FV385.

    Figure image not available in archive
    Fig. 2.

    Dendrogram generated by Bionumeric software, showing distance calculated by the Dice similarity index of PFGE XbaI patterns among 44 EPEC strains. The degree of similarity (%) is shown on the scale at the top left of the figure. The strains were clustered in 12 groups generated by the UPGMA algorithm of 60 % similarity according to the Dice index.

    Analysing each of the 12 groups represented in the dendrogram, all groups but two (groups I and XI) corresponded to a particular serogroup. Thus, 14 strains of serogroup O111 with intimin β1 were clustered in two clearly differentiated groups (III and IV), both including O111 strains with H2 or H− antigen, suggesting that perhaps those H− by serotyping actually carried genes for H2. Group IV included PFGE patterns more closely related (five strains with similarity >85 %). Most of the 13 strains of serotype O119 : H6 with intimin type ξR/β2B were clustered in group IX, with only two strains (FG187 and FV362; similarity >90 %) clearly differentiated in another group (V). A high heterogeneity was observed in group IX, in which genetic relatedness ranged from 65 to 92 %. Strains of serogroup O55 clustered in three different groups (II, VII and VIII). Group II included a single strain of serotype O55 : H40 with intimin θ. Group VIII included two closely related strains (FG240 and FV376; similarity 93 %) of serotype O55 : H− with intimin γ1, and group VII five strains of serotypes O55 : H−/H51 with intimin μB, of which the genetic relatedness ranged from 65 to 80 %.

    DISCUSSION

    Diarrhoeal illness is a major public health problem worldwide, with over 2 million deaths occurring each year, particularly among infants younger than 5 years (). One of the most common causes of infantile diarrhoea is EPEC. Typical EPEC strains belonging to classical EPEC serotypes were associated historically with outbreaks of infantile diarrhoea in industrialized countries, particularly during the 1940s and 1950s, but at present they are very rare (Trabulsi et al., 2002; Jenkins et al., 2003; Blanco et al., 2006). Recently, however, Gerner-Smidt et al. (2003) have reported that classical EPEC is the most common bacterial cause of diarrhoea in children less than 2 years old in Denmark. In industrialized countries, atypical EPEC are more frequently isolated from diarrhoeal cases than typical EPEC. Furthermore, Nguyen et al. (2006) have recently shown that, in contrast to patients infected with other pathogens, patients infected with atypical EPEC are far more likely to experience diarrhoea past 14 days, the point long recognized as a clinical watershed that heralds an increased risk of illness and death. Interestingly, atypical EPEC strains have also been isolated frequently from healthy infants (Afset et al., 2003; Beutin et al., 2003).

    The situation in developing countries is not well defined, but several studies in Brazil, Chile, Uruguay and Bangladesh (Albert et al., 1995; Gomes et al., 1996; Levine et al., 1996; Nunes et al., 2003; Vidal et al., 2004) have shown a high frequency of typical EPEC serotypes in stools from children with diarrhoea. However, some recent studies performed in Brazil, Thailand and Vietnam have shown a very low frequency of typical EPEC and a relatively high frequency of atypical EPEC (Pelayo et al., 1999; Vieira et al., 2001; Ratchtrachenchai et al., 2004; Rodrigues et al., 2004; Vu Nguyen et al., 2005). Our results confirm that typical EPEC strains belonging to classical EPEC serotypes O55 : H−, O111 : H2, O111 : H− and O119 : H6 are important pathogens associated with diarrhoea of children in Uruguay (Torres et al., 2001). These serotypes are the most frequent EPEC serotypes implicated in infantile diarrhoea in Brazil, and have also been frequently isolated in other countries (Trabulsi et al., 2002; Nunes et al., 2003; Ratchtrachenchai et al., 2004).

    Specific intimin subtypes may be involved in mediating both tissue tropism and host specificity, and may provide information on the association of EPEC and STEC with severe disease and on the nature of the bacterium–host relationship. In addition, host immunity to the surface-exposed proteins produced by one E. coli strain may not provide protection against intestinal colonization by E. coli strains which bear distinct intimin types (Adu-Bobie et al., 1998; Torres et al., 2005). We have identified six novel intimin variant genes that we originally designated β2, ε2, μ, ν, ξ and o when the sequences were submitted to the EMBL Nucleotide Sequence Database (Blanco et al., 2003, 2004b, 2005, 2006), and before knowing the results obtained by Ramachandran et al. (2003). The intimin β2 that we have found in all 18 typical EPEC strains of classical EPEC serotype O119 : H6 (Blanco et al., 2003; this study) is identical to intimin ξ described by Ramachandran et al. (2003) in one bovine strain of serotype ONT : HNT. Thus, in this study, our β2 intimin is referred to as ξR/β2B. We have found the ξR/β2B intimin in another four human strains of serotypes O119 : H−, ONT : H51 and O33 : H6 in the present study, in six human strains of serotypes O56 : H6 (bfpA negative), O110 : H6 (bfpA negative), O113 : H6 (bfpA negative), O137 : H6 (bfpA negative) and O167 : H6 (bfpA positive) isolated in Spain (Blanco et al., 2006), and in two E. coli strains of serotypes O139 : H14 (bfpA negative) and O167 : H6 (bfpA positive) isolated from neotropical non-human primates with diarrhoea in Brazil (Blanco et al., 2004c). The other five intimins described by us (ε2, μ, ν, ξ and o) are different to the existing intimin types and are referred to as ε2, μB, νB, ξB and o, respectively. Interestingly, all six EPEC strains positive for the novel μB intimin belonged to classical EPEC serogroup O55 (serotypes O55 : H51 and O55 : H−). However, we also detected the γ1 (O55 : H−) and θ (O55 : H40) intimin variant genes in human EPEC strains belonging to serogroup O55 characterized in this study. The novel intimin ε2 was identified in one human atypical EPEC strain of serotype ONT : H45 isolated in Spain and in three bovine typical EPEC strains of serotype ONT : H45 isolated in Switzerland (Blanco et al., 2005). The novel intimins νR and o were detected in human atypical EPEC strains of serotypes O10 : H−, O84 : H− and O129 : H− isolated in Spain (Blanco et al., 2006), whereas the intimin ξB was observed in two Spanish bovine STEC strains of serotype O80 : H− (Blanco et al., 2004b).

    Our phylogenetic analysis revealed six groups of closely related intimin genes: (i) α1, α2, ζ, νB and o; (ii) λ; (iii) β1, ξR/β2B, δ/β2O and κ; (iv) ε1, ξB, νR/ε2, ε1 and ε2; (v) γ1, μB, γ2 and θ; and (vi) ι1 and μR/ι2. This analysis is in accordance with the finding of Zhang et al. (2002). However, the intimin alleles α2, ξR/β2, ε2, μB, μR/ι2, νB, νR/ε2, ξB and o were not analysed in the study of Zhang et al. (2002) because no nucleotide sequences were available. Ramachandran et al. (2003) classified the 14 intimin genes they analysed into nine distinct phylogenetic families (α1-α2, β1-ξR/β2B, γ1, κ, ε-ε-νR/ε2, ι-μR/ι2, λ, θ and ζ).

    In recent years, DNA macrorestriction analysis by PFGE has increasingly been used for the molecular subtyping of a wide range of bacterial pathogens, and is now considered the ‘gold standard’ for the molecular subtyping of many pathogenic organisms (Mora et al., 2004). However, to our knowledge, very few studies are available concerning PFGE typing of EPEC strains. In our study we have analysed 44 EPEC strains by PFGE, including the most prevalent serotypes. None of the strains presented an identical PFGE pattern to another. In the dendrogram produced by the UPGMA algorithm, the isolates were clustered in 12 groups (1–12 strains per group) of 60 % similar strains according to the Dice similarity index. Only 15 strains were clustered in six groups of closely related (similarity >85 %) PFGE patterns, indicating the prevailing clonal diversity among EPEC strains isolated from children with diarrhoea in Montevideo.

    In conclusion, we have identified two novel intimin types in typical EPEC of serotypes O55 : H51/H− (eae-μB) and O119 : H6/H− (eae-ξR/β2B) that could become novel targets for vaccine development. Our results indicate that typical EPEC strains of serotypes O55 : H51/H− (eae-μB), O111 : H2/H− (eae-β1) and O119 : H6 (eae-ξR/β2) are important pathogens associated with diarrhoea in children in Uruguay.

    Acknowledgments

    This paper is dedicated to the memory of Dr Enrique A. González, an eminent scientist, an excellent Professor of Microbiology and a very good friend. We thank Rafael Vignoli, Laura Betancor, Alfredo Sirok, María Inés Mota and Román Vilas for their collaboration in this study. We also thank Monserrat Lamela for skillful technical assistance. This work was supported by grants from the Fondo de Investigación Sanitaria (grants FIS G03-025-COLIRED-O157 and FIS P052023), from the Xunta de Galicia (grants PGIDIT02BTF26101PR, PGIDIT04RAG261014PR and PGIDIT05BTF26101PR), from CSIC (Central Research Commission, Universidad de la República, Uruguay), and from the Manuel Pérez Foundation (Facultad de Medicina, Uruguay). G. D. acknowledges the Agencia Española de Cooperación Internacional (AECI) for a research fellowship.

    References