| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
INRA, UR454 Unité de Microbiologie, F-63122 Saint-Genès Champanelle, France
Correspondence
Christine Martin
cmartin{at}clermont.inra.fr
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
A supplementary table listing genetic features of the identified GlpheV-CRICC168 ORFs is available with the online version of this paper.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Typically, GEIs are blocks of DNA with signatures of mobile genetic elements, i.e. they possess genes encoding factors that are involved in genetic mobility, such as integrases, transposases, phage genes and origins of replication, are frequently flanked on one side by a tRNA gene and by direct repeat sequences, and are often associated with a genetic instability (Hacker & Kaper, 2000; Hacker & Carniel, 2001). GEIs have a modular organization and encode various functions related to virulence, symbiosis, metabolism, resistance to antibiotics and degradation of xenobiotic compounds (Hacker & Carniel, 2001). Pathogenicity islands (PAIs) represent a subclass of GEIs that contain similar features to GEIs but encode various virulence factors and are normally absent from non-pathogenic strains of the same or closely related species (Hacker & Kaper, 2000; Gal-Mor & Finlay, 2006; Schmidt & Hensel, 2004).
Enteropathogenic Escherichia coli (EPEC), enterohaemorrhagic E. coli (EHEC) and the mouse enteropathogen Citrobacter rodentium belong to the family of attaching and effacing (A/E) bacterial pathogens. A/E lesions are mediated by components of a type III secretion apparatus encoded by a PAI called the locus of enterocyte effacement (LEE) (Deng et al., 2004; Frankel et al., 1998; Garmendia et al., 2005). EHEC and EPEC are poorly pathogenic in mice but infect humans and domestic animals. In contrast, C. rodentium is a natural mouse pathogen that is related to E. coli, hence providing an in vivo model for A/E pathogens (Mundy et al., 2005). While C. rodentium infection of mice results in colonic hyperplasia, EHEC O157 : H7 strains and other Shiga toxin-producing E. coli (STEC) cause a spectrum of human illnesses such as watery diarrhoea, haemorrhagic colitis (HC), haemolytic–uraemic syndrome (HUS) and thrombotic thrombocytopenic purpura. HUS in EHEC infection is the leading cause of acute renal failure in children, and is mainly caused by the production of Shiga toxins (Beutin, 2006; Clarke et al., 2003; Paton & Paton, 1998).
Although the LEE seems to confer enhanced virulence, LEE-negative STEC strains are also associated with severe human disease (Girardeau et al., 2005; Johnson et al., 1996; WHO, 1999). These observations suggest that other unknown factors, possibly GEIs or PAIs, enhance the virulence potential of STEC strains (Boerlin et al., 1999; Kaper et al., 1999; Karmali et al., 2003). Compared with the E. coli K-12 genome, O157 : H7 strains EDL933 and Sakai contain several additional GEIs (O islands) including the 87 kb O island 48 (OI-48) and the 23 kb OI-122 (Perna et al., 2001). To investigate the structural diversity of GEI OI-122 between the O157 : H7 E. coli EDL933 and the LEE-negative O113 : H21 E. coli CL3, a region of 27 kb was sequenced by Shen et al. (2004). These authors describe, in E. coli CL3 and in other STEC strains, a novel hybrid genomic region composed of three physically distinct portions: a 13 kb segment (left part), which carries a Yersinia pestis-like haemolysin/adhesin gene cluster predicted to encode members of the ShlA/HecA/FhaA exoprotein family (secreted by the two-partner secretion pathway); a central 4.4 kb segment bracketed by two 190 bp direct repeat sequences, which carries five ORFs (including transposase genes); and a composite 10 kb segment (right part), which carries three EDL933 OI-122 genes (including the virulence gene Z4321; pagC-like), a 2.2 kb fragment of the enteroaggregative E. coli (EAEC) O42 and a 5.3 kb segment of the EDL933 OI-48. Its left and right termini (Z1635 and Z1644, respectively) contain ORFs that show homology to EDL933 OI-48 genes. Given the presence of putative virulence genes (a haemolysin/adhesin gene cluster and pagC) and mobile genetic elements, this GEI was considered to be a PAI. Because this was the first PAI found in the CL3 genome, it was designated PAI ICL3 (Shen et al., 2004).
A large variety of STEC serotypes have been implicated in disease. However, certain STEC serotypes recovered from animals and food have never been associated with serious human disease. For a better understanding of the apparent differences in virulence between groups of STEC serotypes, STEC strains have been classified into five seropathotypes (A–E) by Karmali et al. (2003), according to their incidence and association with HUS and outbreaks. Recent studies have demonstrated that determination of the seropathotype distribution of virulence factors allows identification of DNA targets for selective detection of strains that present a risk to public health (Gilmour et al., 2006; Karmali et al., 2003). As we have shown, there is a link between seropathotype, prevalence of various virulence factors, phylogeny and Shiga toxin gene expression (de Sablet et al., 2008; Girardeau et al., 2005).
The initial objective of the present study was to determine whether the PAI ICL3 locus is a suitable DNA target for the selective detection of LEE-negative STEC strains belonging to seropathotype C that represent a significant risk of disease in humans. With this aim, the presence of PAI ICl3 elements among LEE-negative STEC strains and other E. coli pathotypes as well as commensal E. coli isolates was investigated in a collection of 469 E. coli strains. We show the widespread dissemination of the PAI ICL3 element among LEE-negative STEC strains, and its absence from other E. coli pathotypes. In addition, we discover that the PAI ICL3 is also present in C. rodentium. We further report that the C. rodentium PAI ICL3 is borne on a 53 kb GEI termed in this study GIpheV-CRICC168.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
; Pradel et al., 2000; Vernozy-Rozand et al., 2002). The second set, which comprised 54 STEC strains of diverse geographical origins isolated from human disease, was provided by the STEC Center, National Food Safety and Toxicology Center (NFSTC), Michigan State University, USA. Of these 287 STEC isolates, 44 are LEE-positive and 243 are LEE-negative. The third set comprised 77 extra-intestinal pathogenic E. coli (ExPEC) isolated from humans with bacteraemia or urinary tract infections and from calves with intestinal or extraintestinal colibacillosis (Girardeau et al., 2003). The fourth set comprised 35 ExPEC strains isolated in Scotland from lambs with nephropathy (Bertin et al., 1998). The last set consisted of 50 enterotoxigenic E. coli (ETEC) isolates collected from calves with intestinal colibacillosis (Contrepois et al., 1993). In addition, seven E. coli intestinal commensal strains were isolated from healthy human and animal subjects. Thirteen reference strains of different pathotypes, including the neonatal meningitis E. coli (NMEC) strain RS218, adherent–invasive E. coli (AIEC) strain LF82, EAEC strain O42, ETEC strain H10407, human EPEC strains E2348/69, E56/54 and B171-8, rabbit EPEC strains RDEC-1 and E22, ExPEC strains 536 and CFT073, and EHEC O157 : H7 strains EDL933 and Sakai, were kindly provided by Josée Harel (GREMIP, Faculté de Médecine Vétérinaire, Université de Montréal, Canada), Arlette Darfeuille Michaud (Pathogénie Bactérienne Intestinale, Université d'Auvergne, Clermont-Ferrand, France), Chantal Le Bouguénec (unité de Pathogénie Bactérienne des Muqueuses, Institut Pasteur, Paris), and the NFSTC. C. rodentium strain DBS100 was a kind gift from K. T. Wilson (Vanderbilt University, USA) with authorization of D. Schauer (MIT, USA).
Investigation of E. coli strains for the presence of PAI ICL3 elements by Southern hybridization.
The 469 E. coli strains and C. rodentium strain DBS100 were surveyed for the presence of five marker sequences specific to the prototypic PAI ICL3. Marker sequences included a first region spanning the 5' end junction with the EDL933 GEI OI-48 (ms-1) and the 5' end-haemolysin/adhesin gene (ms-2), a second region spanning elements of the EDL933 GEI OI-122 (ms-3 and ms-4), and a third region spanning the 3' end junction with the EDL933 GEI OI-48 (ms-5) (Table 1, Fig. 1). Strains were screened by dot-blot hybridization assay using DIG-labelled probes generated with a PCR DIG Probe Synthesis kit (Roche Diagnostics). Southern hybridization was carried out with DNA probes generated from PCR products using E. coli CL3 as the template. Primers shown in Table 1 were designated on the basis of the sequence of the prototypic PAI ICL3 (GenBank accession no. AY275838). Hybridizations were carried out overnight at 42 °C. ECL labelling and signal detection systems (Amersham/Pharmacia Biotech) were used for detection. E. coli CL3 and MG1655 were used as positive and negative controls, respectively.
|
|
). This database contains all of the complete Buchnera, E. coli, Photorhabdus, Serratia, Shigella, Salmonella and Yersinia genome sequences, together with preliminary data from a number of sequencing projects currently in progress at the Wellcome Trust Sanger Institute. The nucleotide sequences of C. rodentium strain ICC168 were found in the coliBASE database. The C. rodentium genome sequencing project is being carried out by the Wellcome Trust Sanger Institute in collaboration with Gad Frankel and Gordon Dougan of the Centre for Molecular Microbiology and Infection, Imperial College, UK. Sequence data were produced by the Micro-organisms Sequencing Group at the Sanger Institute and can be obtained from http://www.sanger.ac.uk/Projects/C_rodentium/.
Each coding sequence annotated in the published PAI ICL3 of the archetypal E. coli strain CL3 (Shen et al., 2004) was analysed by BLAST searches on the coliBASE server, supplemented as needed by visualization of the genomic context of homologues using an Artemis applet (Rutherford et al., 2000) within coliBASE, by multiple alignments on the EBI CLUSTAL W server (http://www.ebi.ac.uk/clustalw), and by searches of the National Center for Biotechnology Information conserved domain database (http://www.ncbi.nlm.nih.gov/Structure/cdd/cdd.shtml). BLAST, genomic BLAST (with microbial genome) and FASTA searches with the non-redundant protein and nucleotide databases were accessed at http://www.ncbi.nlm.nih.gov and http://www.ebi.ac.uk/fasta3/. The IS Finder database was accessed at http://www-is.biotoul.fr.
Statistical analyses.
Statistical analyses were performed with SAS for Unix Windows (version 8.01, SAS Institute). Comparison of the prevalence of a particular characteristic in different populations was evaluated with the chi-squared test and odds ratios (ORs), and 95 % confidence intervals (CI) were determined.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). The five PAI ICL3 marker sequences were found in C. rodentium and among 102 of the 243 (42 %) LEE-negative STEC strains belonging to 29 serotypes. PCR linkage analysis (not shown) carried out on 18 STEC strains of different serotypes demonstrated that the PAI ICL3 element carried by these strains has a structure similar to that of the prototypic PAI ICL3. Interestingly, the PAI ICL3 appears to be unique to C. rodentium and LEE-negative STEC strains, since it was not detected in intestinal commensal isolates nor in any of the strains belonging to the other E. coli pathotypes.
Seropathotype distribution of the PAI ICL3
Consistent with an earlier study (Shen et al., 2004), PAI ICL3-related elements were not found among strains in seropathotypes A and B that included the LEE-positive strains known to be associated with outbreaks (serotypes O157 : H7, O111 : H2, O111 : H–, O103 : H2 and O26 : H11). However, the prevalence of strains carrying PAI ICL3 was significantly higher among isolates in seropathotype C linked to disease [58 of 87 strains (67 %)], than among isolates in both seropathotypes D and E not linked to disease [44 of 156 strains (28 %)]. This finding revealed PAI ICL3 as a significant predictor of ‘virulent’ status among the LEE-negative strains [P<0.0001; OR=5.5 (95 % CI=2.7–8.4)].
Our previous findings, with the same STEC strains, showed that the distribution pattern of virulence factors differed considerably between clonal groups (Girardeau et al., 2005). Accordingly, the present study revealed a close association between PAI ICL3 and certain clonal groups. Indeed, PAI ICL3 was detected in most (90 %) strains belonging to serotypes ON : H21, O91 : H21, O113 : H21 and O174 : H21, which are the archetypal strains of the virulent clonal group STEC-1 associated with HUS. Therefore, when only STEC-1 strains in seropathotype C were compared with isolates in seropathotypes D and E, the presence of the PAI ICL3 appeared as an even stronger predictor of virulent status [P<0.0001; OR=25.7 (95 % CI=8.6–76.6)].
Identification of a similar PAI ICL3 element in the C. rodentium genome
Sequence data available in the coliBASE database indicate that the PAI ICL3 gene cluster is virtually identical at the nucleotide level (97 % identical within 17.5 kb) to a genomic fragment identified in the genome of C. rodentium ICC168 (positions 5 236 633 to 5 256 813 in the genome sequence) (Fig. 1, bold type in Supplementary Table S1). The C. rodentium PAI ICL3 contains 15 ORFs; ORFs ROD49881 and ROD50041 define its left and right boundaries, respectively (corresponding in the E. coli CL3 PAI ICL3 to ORFs Z1635 and S14; accession nos AAQ19121.1 and AAQ19137.1, respectively). In contrast to the conserved ‘core’ of the PAI, the central segment bracketed by the two 190 bp direct repeat sequences differed completely in size and sequence between C. rodentium and E. coli (Fig. 1).
Sequence comparison of PAI ICL3 elements in E. coli and C. rodentium
Whereas the prototypic PAI ICL3 of E. coli CL3 is 21 925 bp in length, the PAI ICL3 of C. rodentium spans 20 181 bp (Fig. 1). The difference can be accounted for by a larger central segment in E. coli CL3 (ORFs S5–S9) than in the corresponding region of C. rodentium (ORFs CR33–CR28). Comparisons of the nucleotide sequences showed that this variable central segment contains a unique complement of genes. They encode short peptides that show homology with the C terminus of haemagglutinin-like proteins and with several proteins of unknown function (Supplementary Table S1). The fact that this segment is bracketed by direct repeat sequences (Shen et al., 2004) and has a low G+C content (42.8 and 40.2 % in E. coli and C. rodentium, respectively) suggests that it has been horizontally transferred.
A second structural difference in the C. rodentium PAI ICL3 is the size of the gene that defines its right-hand boundary (CR21/22 and S14 in C. rodentium and E. coli, respectively). These two ORFs show high similarity to the gene that encodes the putative ImpA-like ATPase from the EAEC strain O42 (1368 bp in length) (locus EcO42-4550). However, while the C. rodentium sequence lying within CR21 and CR22 ORFs includes 1365 bp from E. coli O42 ImpA, the E. coli S14 is much shorter (801 bp) (Fig. 1).
C. rodentium PAI ICL3 is carried by a 53 kb region with characteristics of a GEI
Using the sequence data available in the coliBASE server for C. rodentium strain ICC168, the genomic localization of the PAI ICL3 was investigated and the flanking regions were explored for signatures of mobile genetic elements. We found that the C. rodentium PAI ICL3 is carried by a 53 kb GEI inserted into the pheV-tRNA locus. Moritz & Welch (2006) have proposed that islands be given unique names that comprise their chromosomal location relative to the sequence of E. coli strain MG1655 and the host strain number. In addition, when there is no evidence that a PAI is necessary for complete virulence of a pathogen, they propose that the phenotypically neutral abbreviation GI, for ‘genomic island’, be used. Therefore, the novel genomic island described here was named GIpheV-CRICC168 [for pheV-associated genomic island of the C. rodentium (CR) strain ICC168]. Its genetic organization is described in Supplementary Table S1 and Fig. 2.
|
Insertion sequence (IS) and prophage elements
Ten ORFs of GIpheV-CRICC168 were homologous to whole or partial IS elements or transposons of the IS3, IS4 or IS66 type (Supplementary Table S1, Fig. 2). Data from the IS Finder database revealed that a 2704 bp stretch of sequence lying within the CR10–CR12 gene cluster shares high similarity (96.5 %) with the IS679 element. The inverted repeat sequence GTAAGCGNNTCANNNAACCGTNTT was found at both ends of this sequence, suggesting that the IS element is intact and potentially functional in GIpheV-CRICC168. Additionally, it was flanked by an 8 bp direct repeat (CCCTGATG), which could be the target sequence of the insertion. The identification of an IS679-like element in virulence plasmid pB171 of EPEC B171 (Tobe et al., 1999), PAI AGI-3 of avian pathogenic E. coli (APEC) (Chouikha et al., 2006), LEE of C. rodentium (Deng et al., 2001) and PAI of S. flexneri (Jin et al., 2002) raises the possibility that IS679 is involved in the transfer of virulence determinants between different bacterial strains and species. Other IS elements appeared to be defective in their transposition capacity, since these sequences were disrupted either by frameshift or by integration of another IS element. Additionally, five non-functional ORFs were homologous to prophage CP4-57 and CP-933L elements. The distribution of many of these mobile elements on GIpheV-CRICC168 suggested that extensive rearrangements have occurred during the evolution of this GEI.
Genetic features of the GEI GIpheV-CRICC168
The G+C content variations observed throughout the complete sequence of GIpheV-CRICC168 reflect the mosaic structure of the GEI. These detected variations indicated that GIpheV-CRICC168 comprises four distinct modules bound by insertion elements or bacteriophage sequences (Supplementary Table S1, Fig. 2).
Module I spans a 12399 bp segment (including CR07 to CR18) that is nearly identical (98.2 %) to a region (c4509–c4517) of the selC-associated GEI of the uropathogenic E. coli strain CFT073 (termed in this study GIselCCFT073) (Welch et al., 2002). The complete size of this segment in GIpheV-CRICC168 is 2635 bp larger than that of the corresponding region in GIselCCFT073. This difference is due to the presence of the IS679 element (CR10–CR12). The CR07 gene encodes a putative protein that shares similarity with Bacillus licheniformis tRNA nucleotidyl transferase. The putative CR14 gene product is identical to the E. coli CFT073 TatD-like DNase. Proteins that might be encoded by CR15 and CR17 genes are highly similar to the PP-loop superfamily ATPase proteins. In addition, this segment comprises seven genes of mobile genetic elements, such as a bacteriophage gene (CR18) and IS elements (CR04–CR06 and CR10–CR12), as well as ORFs of unknown function.
Module II spans the PAI ICL3 elements previously described in E. coli CL3. This 20.2 kb segment that contains 15 ORFs (from CR21 to CR43) represents the ‘core’ of the island. The G+C content (52.5 %) of module II differs markedly from that of module III located at its left flank, which averages only 45 %.
Modules III and IV were found to be present with the same organization in the selC-associated GEI AGI-3, involved in carbohydrate assimilation and virulence of the APEC strain BEN2908 (Chouikha et al., 2006). Module III, which contains 12 ORFs (CR45–CR65), spans a 9.4 kb segment which is nearly identical (94 % identity) to module 4 (ORFs aec-59 to aec-64) of the GEI AGI-3 and to a segment (ORF29–ORF34) of the 111 kb PAI IRW1374 of the EHEC strain RW1374 (Jores et al., 2005). The CR64 gene encodes a putative ERA-like GTP-binding protein involved in 16S rRNA maturation, regulation of the cell cycle, and protein synthesis in E. coli (Meier et al., 2000). The CR45 gene flanks the left junction of the PAI ICL3; its predicted product is highly related (97 %) to a transposase of the IS3 family. Although it was truncated at its 5' end by insertion of two IS elements (CR46 and CR47), this IS-related sequence may be a remnant of the PAI ICL3 association with module 4 of the GEI AGI-3.
Module IV has a significantly higher G+C content, which averages 56.6 %. This module spans an 8.5 kb segment (from CR67 to CR82) highly homologous to module 5 of the GEI AGI-3. This gene cluster is also present in two other pheV-associated PAIs [ORF35–ORF47 of the 111 kb PAI IRW1374 of EHEC strain RW1374 (Jores et al., 2005) and the she PAI of S. flexneri (Al-Hasani et al., 2001)], and in two different GEIs of E. coli strain CFT073, inserted at the serX and pheV loci (Chouikha et al., 2006). The order and orientation of the corresponding ORFs in GIpheV-CRICC168 are identical to those of their homologues in E. coli and S. flexneri. The CR67 gene product has high identity (97 %) to the E. coli autotransporter adhesin adhesin-involved-in-diffuse-adherence (AIDA-I), which belongs to the autotransporter protein of the type V secretion system family (Maurer et al., 1999). In the GEI AGI-3 of the APEC strain BEN2908, the aec-67 gene that encodes AIDA-I is truncated at its 5' end by the insertion of an IS911 transposase gene (Chouikha et al., 2006). In contrast, GIpheV-CRICC168 possesses an intact and potentially functional AIDA-I gene (CR67) similar to the native AIDA-I precursor of E. coli MG1655.
Other than the AIDA-I-like gene, Module IV contains only phage- and plasmid-related sequences. CR68 is highly related to a gene adjacent to the origin of transfer in plasmids F and R100, while CR69 has clear identity to antirestriction proteins of conjugative plasmids. At the right end of the GIpheV-CRICC168, four ORFs are present (CR77–CR80) with similarities to ORFs L007–L0012 of the putative prophage CP-933L in the EHEC LEE. As reported for the she PAI of S. flexneri (Al-Hasani et al., 2001), these phage- and plasmid-related sequences may be remnants of the PAI association with self-transmissible elements.
PAI ICL3 is carried in C. rodentium and E. coli by different GEIs
The O island 48 (OI-48EDL933), also termed the tellurite resistance- and adherence-conferring island (TAI), is inserted 2 bp from the serW-tRNA gene in E. coli EDL933 (Perna et al., 2001). As reported elsewhere (Shen et al., 2004), the prototypic E. coli PAI ICL3 contains EDL933 OI-48 genes Z1635, Z1636 and Z1637 at the left terminus, and Z1642, Z1643 and Z1644 at the right terminus (Fig. 1). From this finding, the E. coli PAI ICL3 is believed to reside within an E. coli CL3-homologous OI-48 GEI (termed in this study OI-48CL3). Although the C. rodentium PAI ICL3 closely resembles the E. coli PAI ICL3, it differs in its location. We show here that the C. rodentium PAI ICL3 is carried by the GEI GIpheV-CRICC168 inserted into the pheV-tRNA gene (position 5272 Mb in the C. rodentium genome sequence). IS elements homologous to the putative transposase IS3A and sequence similar to the putative prophage CP-933L flank the left and right ends of the C. rodentium PAI ICL3, respectively (Fig. 1). We hypothesize that these elements have played a role in the integration of the PAI ICL3 element into the GEI of C. rodentium. Taken together, our observations indicate that the C. rodentium PAI ICL3 was inserted differently into the bacterial genome and thus represents an evolutionary lineage different from that of the E. coli PAI ICL3.
The genomes of other E. coli pathotypes contain deleted versions of the PAI ICL3
BLAST searches on the coliBASE server revealed that other genome-sequenced E. coli strains possess deleted PAI ICL3 sequences in which all that remain are the genes located at the extremities of the island. Six of the 16 E. coli genome sequences showed evidence of PAI ICL3 being inserted in at least three different chromosomal sites, which is compatible with the idea that PAI ICL3 ancestors entered E. coli and C. rodentium genomes at multiple times through independent events. Three patterns of deletion, which removed almost all of the PAI ICL3 sequences, were noted (Fig. 3).
|
). We have identified the rearrangements observed in EDL933 as a deletion that fuses the three segments found in S1, S3 and S14, to generate the chimaeric gene Z1640 (Fig. 3a). The presence of the putative functional insertion sequence IS629 (ORFs Z1638 and Z1639) contiguous to Z1640 may indicate that the chimaeric Z1640 gene arose following an IS-mediated deletion. This deletion event has removed a large part (19 kb) of the PAI ICL3 gene cluster carried by the OI-48EDL933 ancestor. An identical rearrangement was found in the EHEC O157 : H7 Sakai strain (ECs1382 gene) and in the EPEC O111 : H9 E110019 strain (EcolE1_01002508 gene), demonstrating that a similar pattern of deletion has occurred in these strains.
In the ExPEC strain CFT073, the PAI ICL3 ancestor was inserted within a selC-associated GEI (GIselCCFT073) which contains a S. flexneri shiA homologue and the ethanolamine utilization gene cluster (Welch et al., 2002). As described above (Fig. 2), the genetic organization and DNA content of module I of GIpheV-CRICC168 are similar to those of the corresponding region (c4509–c4517) of GIselCCFT073. The homology ends immediately upstream of the c4517 locus. Preceding c4517, the c4518 locus is identical to CR47 located at the 5' end of the PAI ICL3 element of GIpheV-CRICC168 (Fig. 3b). As the region located between c4517 and c4518 appears considerably shorter than its equivalent in the full PAI ICL3 gene cluster, we concluded that a deletion including a large part (17.5 of 22 kb) of the PAI ICL3 gene cluster had occurred. A 56 bp sequence corresponding to the IS911 5' end was found adjacent to c4514 (Fig. 3). This IS segment, which contains the canonical C-terminal inverted repeat, might be the remnant of an IS-assisted deletion in a GIselCCFT073 ancestor. A homologous pattern of deletion was also identified within the APEC O1 strain.
In the EAEC strain O42, the PAI ICL3 ancestor was inserted at position 3 405 569 bp within a pheV-associated GEI (GIpheVO42) which carries the gene cluster encoding the antibiotic peptide microcin H47 (ORFs, EcO42-3195/mchF-EcO42-3199/mchC). Following the microcin gene cluster, a 2 kb segment (region EcO42-3191–EcO42-3188) is identical to the region CR37–CR43 located at the 5' end of the C. rodentium PAI ICL3 gene cluster (Fig. 3c). ORF EcO42-3188 is followed by a 3.5 kb segment (including the locus EcO42-3187) identical to region CR21–CR24 located at the 3' end of the C. rodentium PAI ICL3 gene cluster. The region located between EcO42-3188 and EcO42-3187 appears considerably shorter than its equivalent in the full PAI ICL3 gene cluster. We concluded that a deletion removing a large part (17 of 22 kb) of the PAI ICL3 gene cluster had occurred between the two loci CR37 and CR24. The intergenic sequence between EcO42-3187 and EcO42-3188 contains a 305 bp sequence nearly identical to the 3' end of the Shigella dysenteriae insertion sequence IS911. The presence of IS911 sequences may indicate that the EcO42-3188–EcO42-3187 region arose following an IS911-mediated deletion event that has removed a large part of the PAI ICL3 element in the microcin region of EAEC strain O42.
Different variants of the PAI ICL3 gene cluster are present in C. rodentium and E. coli
On the basis of the host GEI, the integration site and the pattern of deletion, five variant types of PAI ICL3 were identified. Their characteristics are summarized in Table 2. The first variant (CL3-PAI ICL3) contains the full complement of PAI ICL3 genes previously identified in E. coli CL3 (Shen et al., 2004) and resides within an E. coli CL3-homologous OI-48 GEI (OI-48CL3). The second variant (CR-PAI ICL3) contains the C. rodentium PAI ICL3 gene cluster carried by the GEI GIpheV-CRICC168 inserted at the pheV-tRNA locus, described above. In a third variant (EDL933-del/PAI ICL3), the O island OI-48-EDL933 (inserted at the serW-tRNA locus) carries the deleted derivative of the PAI ICL3 gene cluster found in different evolutionary lineages of A/E pathogens (EHEC strains EDL933 and Sakai, and atypical EPEC strain E110019). In the fourth variant (CFT073-del/PAI ICL3), the GEI GIselCCFT073 (inserted at the selC-tRNA locus) carries the deleted derivative of the PAI ICL3 gene cluster found in ExPEC strains CFT073 and APEC 01. In the last variant (O42-del/PAI ICL3), the GEI GIpheVO42 (inserted at the pheV-tRNA locus) carries the deleted derivative of the PAI ICL3 gene cluster found in the EAEC strain O42.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Consistent with these findings, this study showed a significantly higher prevalence of the PAI ICL3 in the virulent seropathotype C than in the non-virulent seropathotypes D and E. In addition, the presence of the PAI ICL3 element in most strains belonging to the virulent clonal group STEC-1 indicates that PAI ICL3 is a common component of the genome of certain E. coli lineages. Therefore, when LEE-negative STEC strains in the virulent seropathotype were compared with isolates in non-virulent seropathotypes, PAI ICL3 appeared to be a significant predictor of the virulent status.
In C. rodentium, a 53 kb GEI carries a PAI ICL3-equivalent gene cluster
A PAI ICL3-equivalent gene cluster was also detected in the genome of C. rodentium. The fact that in both E. coli and C. rodentium the PAI ICL3 element is extremely conserved at the nucleotide level and in genetic organization strongly supports the idea that the PAI ICL3 from E. coli CL3 and C. rodentium ICC168 shares a common origin. Deviation in G+C content of a gene compared with the whole genome is often a valuable marker for identifying genes recently acquired by horizontal transfer. The G+C content of the PAI ICL3 haemolysin/adhesin gene cluster is 56.6 %, compared with 
50.1 % for the E. coli genome, suggesting that this DNA segment was recently acquired by E. coli.
We further report that C. rodentium PAI ICL3 is borne on a 53 kb GEI termed GIpheV-CRICC168. The presence of a putative functional phage P4-like integrase, the insertion into a tRNA locus (pheV), and the presence of 23 bp direct repeat sequences carrying an attB-like site, suggest that the GEI GIpheV-CRICC168 was acquired from a bacteriophage via horizontal transfer. This novel C. rodentium GEI exhibits marked similarities to certain chromosomal regions of ExPEC CFT073 (module I), STEC CL3 (module II) and APEC BEN2908 or EHEC RW1374 (modules III and IV) strains, particularly with regard to certain GEIs of these strains. All of these regions were surrounded by intact or fragmented IS elements, suggesting that GIpheV-CRICC168 acquired these regions through the horizontal transfer of mobile elements. In module I of GIpheV-CRICC168, located upstream of the PAI ICL3 element, is the intact 2.7 kb IS679. IS679, also found next to the C. rodentium LEE, is thought to have played a role in acquisition of the LEE by C. rodentium (Deng et al., 2001).
We could identify 82 ORFs within GIpheV-CRICC168. In addition to the putative virulence factors, including the Yersinia haemagglutinin/adhesin gene cluster and pagC carried by the prototypic E. coli PAI ICL3 (Shen et al., 2004), the C. rodentium GEI contains a gene encoding the diffusely adhering E. coli adhesin AIDA-1. Given the presence of these putative virulence genes, this GEI could be considered a PAI. However, because there is no evidence that the GEI GIpheV-CRICC168 is necessary for complete virulence of C. rodentium, according to the Moritz nomenclature (Moritz & Welch, 2006), the phenotypically neutral abbreviation GI, for ‘genomic island’, was used.
C. rodentium and E. coli possess different evolutionary lineages of the PAI ICL3 elements
The PAI ICL3 variant types of both E. coli CL3 and C. rodentium possess the full complement of PAI ICL3 genes and are highly conserved. However, they have been shown to be inserted into different host GEIs. While the C. rodentium variant type (CR-PAI ICL3) is borne on the GEI GIpheV-CRICC168, the E. coli CL3 variant type (CL3-PAI ICL3) is believed to reside within a homologous OI-48 GEI (OI-48CL3). It is believed that the LEE has been acquired by A/E pathogens at multiple times, possibly via the horizontal transfer of a putative plasmid (Deng et al., 2001). As reported for the LEE, it appears that the PAI ICL3 element was acquired by E. coli and C. rodentium after horizontal transfer through independent events. IS elements, similar in sequence to putative prophage CP-933L and other putative DNA-binding proteins that flank the ends of the PAI ICL3 from C. rodentium and E. coli CL3, may have played a role in the integration of PAI ICL3 into the host GEI.
The PAI ICL3 element is present in other E. coli pathotypes, but it has been subjected to extensive deletions
Other than STEC, we discovered that other E. coli pathotypes (including A/E and non-A/E pathogens) possess deleted PAI ICL3 sequences in which all that remain are the genes located at the extremities of the genomic segment. In the deleted derivatives of PAI ICL3, the junction boundaries are often marked by truncated genes fused to generate a chimaeric gene (i.e. EDL933 Z1640), revealing that a deletion has occurred. The PAI ICL3 variant types EDL933-del/PAI ICL3, CFT073-del/PAI ICL3, and O42-del/PAI ICL3 (from EHEC/EPEC, ExPEC and EAEC O42, respectively), have undergone extensive deletions that removed almost all of the PAI ICL3. ISs of several classes (IS3, IS911 and IS629) are found at the sites of deletion, suggesting that homologous recombination between such elements following deletion accounts for the multigene deletions. Such deletion events explain the absence of PAI ICL3 elements among LEE-positive isolates. Hence, the difference between LEE-positive and LEE-negative STEC strains is a deletion in LEE-positive rather than an insertion into LEE-negative STEC strains.
It remains unclear whether IS elements found at the junction sites are responsible for (or remnants of) the observed deletions. However, IS-mediated deletions and genetic instability have been frequently observed in well-known PAIs and GEIs, e.g. Y. pestis HPI (Schubert et al., 1998), ETT2 (Ren et al., 2004) and Salmonella Spi (Amavisit et al., 2003). The ETT2 cluster (encoding a second cryptic type III secretion system) simultaneously provides a model of gene flux and of genetic loss, and shows a whole spectrum of reductive evolution, from an apparently intact 27.5 kb cluster in E. coli O42 to only two residual gene fragments in S. flexneri (Ren et al., 2004). As reported here, the PAI ICL3 cluster provides a new example of gain and loss of genetic elements from one lineage to another.
PAI ICL3 ancestors entered the E. coli genomes at multiple times, through independent events
Many tRNA genes are frequently used as integration sites for GEIs (Gal-Mor & Finlay, 2006; Hacker & Carniel, 2001). Among six sequenced E. coli genomes, as well as in the genome of C. rodentium, the PAI ICL3 gene cluster was found to be carried by five different GEIs associated with three different tRNA loci (pheV, selC and serW). The observed relationships between the E. coli pathotype and the integration site in a tRNA locus (EHEC/EPEC/serW, ExPEC/selC, EAEC/pheV) are compatible with the idea that different host GEIs carrying the PAI ICL3 genes entered the E. coli genomes at multiple times, through independent events, and then integrated in different tRNA loci. Various mechanisms could account for the generation of the deleted PAI ICL3 variant types. One scenario is that a host GEI carrying the full complement of PAI ICL3 genes entered the genome first, and then has been subject to IS-mediated deletion. In another scenario, PAI ICL3 assembled on another element (i.e. a resident plasmid or phage) has been subject first to deletion and then integrated into a host genomic island in an insertional hotspot.
PAI ICL3 dissemination throughout E. coli strains of different pathotypes (EHEC, EPEC, ExPEC and EAEC) and its wide distribution among distantly related LEE-negative STEC strains imply an efficient mechanism of transfer. Temperate phages or transmissible plasmids are candidates for PAI ICL3 vehicles. The putative functional phage-like integrase, the att sites, the direct repeat sequences, and the targeting of tRNA loci as integration sites shared by the different PAI ICL3-host GEIs, suggest that these GEIs may have been acquired from a bacteriophage via horizontal transfer.
A B1 genetic background appears necessary for the maintenance of a full complement of PAI ICL3 genes
We noted an important distinction between C. rodentium and LEE-negative STEC which contain a complete PAI ICL3 gene cluster and other E. coli pathotypes which contain deleted PAI ICL3 gene clusters. A striking phylogenetic distribution of the PAI ICL3 genotypes could explain this distinction. While six major phylogenetic groups of E. coli (A, B1, C, E, D and B2) form the core of the E. coli species (Escobar-Paramo et al., 2004), we showed in a previous study that LEE-negative STEC strains in seropathotype C belong exclusively to the ECOR group B1 (Girardeau et al., 2005). Therefore, the observed link between complete PAI ICL3 genotype and seropathotype C demonstrates a striking phylogenetic link between intact PAI ICL3 and ECOR group B1. Consistent with previous studies that suggest that the arrival and stability of a PAI in a genome require a particular genetic background (Escobar-Paramo et al., 2004; Reid et al., 2000), a B1 genetic background seems to be necessary for the maintenance of a full complement of PAI ICL3 genes. The finding that E. coli strains belonging to phylogenetic groups E (EHEC EDL933 and Sakai), B2 (ExPEC CFT073 and APEC O1) and D (EAEC O42) carry exclusively the deleted versions of PAI ICL3 supports this concept.
Conclusion
Taken together, our results indicate that (i) the PAI ICL3 gene cluster is a common component of the genome of LEE-negative STEC strains linked to disease, and could provide a new marker for these strains; (ii) a PAI ICL3-equivalent gene cluster is present in the genome of C. rodentium; (iii) the C. rodentium PAI ICL3 is borne on a genomic region with characteristics typical of a horizontally transferable GEI; (iv) other E. coli pathotypes (including A/E and non-A/E pathogens) possess deleted subtypes of PAI ICL3; (v) the PAI ICL3 gene cluster entered E. coli genomes at multiple times, through independent events; and (vi) a B1 genetic background is necessary for the maintenance of a full complement of PAI ICL3 genes in E. coli.
Edited by: R. J. Lamont
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Amavisit, P., Lightfoot, D., Browning, G. F. & Markham, P. F. (2003). Variation between pathogenic serovars within Salmonella pathogenicity islands. J Bacteriol 185, 3624–3635.
Bertin, Y., Martin, C., Girardeau, J. P., Pohl, P. & Contrepois, M. (1998). Association of genes encoding P fimbriae, CS31A antigen and EAST 1 toxin among CNF1-producing Escherichia coli strains from cattle with septicemia and diarrhea. FEMS Microbiol Lett 162, 235–239.[CrossRef][Medline]
Beutin, L. (2006). Emerging enterohaemorrhagic Escherichia coli, causes and effects of the rise of a human pathogen. J Vet Med B Infect Dis Vet Public Health 53, 299–305.[Medline]
Boerlin, P., McEwen, S. A., Boerlin-Petzold, F., Wilson, J. B., Johnson, R. P. & Gyles, C. L. (1999). Associations between virulence factors of Shiga toxin-producing Escherichia coli and disease in humans. J Clin Microbiol 37, 497–503.
Bonnet, R., Souweine, B., Gauthier, G., Rich, C., Livrelli, V., Sirot, J., Joly, B. & Forestier, C. (1998). Non-O157 : H7 Stx2-producing Escherichia coli strains associated with sporadic cases of hemolytic–uremic syndrome in adults. J Clin Microbiol 36, 1777–1780.
Chaudhuri, R. R., Khan, A. M. & Pallen, M. J. (2004). coliBASE : an online database for Escherichia coli, Shigella and Salmonella comparative genomics. Nucleic Acids Res 32, D296–D299.
Chouikha, I., Germon, P., Bree, A., Gilot, P., Moulin-Schouleur, M. & Schouler, C. (2006). A selC-associated genomic island of the extraintestinal avian pathogenic Escherichia coli strain BEN2908 is involved in carbohydrate uptake and virulence. J Bacteriol 188, 977–987.
Clarke, S. C., Haigh, R. D., Freestone, P. P. & Williams, P. H. (2003). Virulence of enteropathogenic Escherichia coli, a global pathogen. Clin Microbiol Rev 16, 365–378.
Contrepois, M., Bertin, Y., Girardeau, J. P., Picard, B. & Goullet, P. (1993). Clonal relationships among bovine pathogenic Escherichia coli producing surface antigen CS31A. FEMS Microbiol Lett 106, 217–222.[CrossRef][Medline]
Deng, W., Li, Y., Vallance, B. A. & Finlay, B. B. (2001). Locus of enterocyte effacement from Citrobacter rodentium: sequence analysis and evidence for horizontal transfer among attaching and effacing pathogens. Infect Immun 69, 6323–6335.
Deng, W., Puente, J. L., Gruenheid, S., Li, Y., Vallance, B. A., Vazquez, A., Barba, J., Ibarra, J. A., O'Donnell, P. & other authors (2004). Dissecting virulence: systematic and functional analyses of a pathogenicity island. Proc Natl Acad Sci U S A 101, 3597–3602.
de Sablet, T., Bertin, Y., Vareille, M., Girardeau, J. P., Garrivier, A., Gobert, A. P. & Martin, C. (2008). Differential expression of stx2 variants in Shiga toxin-producing Escherichia coli belonging to seropathotypes A and C. Microbiology 154, 176–186.
Escobar-Paramo, P., Clermont, O., Blanc-Potard, A. B., Bui, H., Le Bouguenec, C. & Denamur, E. (2004). A specific genetic background is required for acquisition and expression of virulence factors in Escherichia coli. Mol Biol Evol 21, 1085–1094.
Frankel, G., Phillips, A. D., Rosenshine, I., Dougan, G., Kaper, J. B. & Knutton, S. (1998). Enteropathogenic and enterohaemorrhagic Escherichia coli: more subversive elements. Mol Microbiol 30, 911–921.[CrossRef][Medline]
Gal-Mor, O. & Finlay, B. B. (2006). Pathogenicity islands: a molecular toolbox for bacterial virulence. Cell Microbiol 8, 1707–1719.[CrossRef][Medline]
Garmendia, J., Frankel, G. & Crepin, V. F. (2005). Enteropathogenic and enterohemorrhagic Escherichia coli infections: translocation, translocation, translocation. Infect Immun 73, 2573–2585.
Gilmour, M. W., Tracz, D. M., Andrysiak, A. K., Clark, C. G., Tyson, S., Severini, A. & Ng, L. K. (2006). Use of the espZ gene encoded in the locus of enterocyte effacement for molecular typing of Shiga toxin-producing Escherichia coli. J Clin Microbiol 44, 449–458.
Girardeau, J. P., Lalioui, L., Said, A. M., De Champs, C. & Le Bouguenec, C. (2003). Extended virulence genotype of pathogenic Escherichia coli isolates carrying the afa-8 operon: evidence of similarities between isolates from humans and animals with extraintestinal infections. J Clin Microbiol 41, 218–226.
Girardeau, J. P., Dalmasso, A., Bertin, Y., Ducrot, C., Bord, S., Livrelli, V., Vernozy-Rozand, C. & Martin, C. (2005). Association of virulence genotype with phylogenetic background in comparison to different seropathotypes of Shiga toxin-producing Escherichia coli isolates. J Clin Microbiol 43, 6098–6107.
Hacker, J. & Carniel, E. (2001). Ecological fitness, genomic islands and bacterial pathogenicity. A Darwinian view of the evolution of microbes. EMBO Rep 2, 376–381.[Medline]
Hacker, J. & Kaper, J. B. (2000). Pathogenicity islands and the evolution of microbes. Annu Rev Microbiol 54, 641–679.[CrossRef][Medline]
Jin, Q., Yuan, Z., Xu, J., Wang, Y., Shen, Y., Lu, W., Wang, J., Liu, H., Yang, J. & other authors (2002). Genome sequence of Shigella flexneri 2a: insights into pathogenicity through comparison with genomes of Escherichia coli K12 and O157. Nucleic Acids Res 30, 4432–4441.
Johnson, R., Clarkes, R. C., Wilson, J. B., Read, S. C., Rhan, K., Renwick, S. A., Sandhu, K. A., Alves, D., Karmali, M. A. & other authors (1996). Growing concerns and recent outbreaks involving non-O157 : H7 serotypes of verotoxigenic Escherichia coli. J Food Prot 59, 1112–1122.
Jores, J., Wagner, S., Rumer, L., Eichberg, J., Laturnus, C., Kirsch, P., Schierack, P., Tschape, H. & Wieler, L. H. (2005). Description of a 111-kb pathogenicity island (PAI) encoding various virulence features in the enterohemorrhagic E. coli (EHEC) strain RW1374 (O103 : H2) and detection of a similar PAI in other EHEC strains of serotype O103 : H2. Int J Med Microbiol 294, 417–425.[CrossRef][Medline]
Kaper, J. B., Mellies, J. L. & Nataro, J. (1999). Pathogenicity islands and other mobile genetic elements of diarrheagenic Escherichia coli. In Pathogenicity Islands and Other Mobile Virulence Elements, pp. 33–58. Edited by J. B. Kaper & J. Hacker. Washington, DC: American Society for Microbiology.
Karmali, M. A., Mascarenhas, M., Shen, S., Ziebell, K., Johnson, S., Reid-Smith, R., Isaac-Renton, J., Clark, C., Rahn, K. & other authors (2003). Association of genomic O island 122 of Escherichia coli EDL 933 with verocytotoxin-producing Escherichia coli seropathotypes that are linked to epidemic and/or serious disease. J Clin Microbiol 41, 4930–4940.
Maurer, J., Jose, J. & Meyer, T. F. (1999). Characterization of the essential transport function of the AIDA-I autotransporter and evidence supporting structural predictions. J Bacteriol 181, 7014–7020.
Meier, T. I., Peery, R. B., McAllister, K. A. & Zhao, G. (2000). Era GTPase of Escherichia coli: binding to 16S rRNA and modulation of GTPase activity by RNA and carbohydrates. Microbiology 146, 1071–1103.
Moritz, R. L. & Welch, R. A. (2006). The Escherichia coli argW–dsdCXA genetic island is highly variable, and E. coli K1 strains commonly possess two copies of dsdCXA. J Clin Microbiol 44, 4038–4048.
Mundy, R., MacDonald, T. T., Dougan, G., Frankel, G. & Wiles, S. (2005). Citrobacter rodentium of mice and man. Cell Microbiol 7, 1697–1706.[CrossRef][Medline]
Paton, J. C. & Paton, A. W. (1998). Pathogenesis and diagnosis of Shiga toxin-producing Escherichia coli infections. Clin Microbiol Rev 11, 450–479.
Perna, N. T., Plunkett, G. R., Burland, V., Mau, B., Glasner, J. D., Rose, D. J., Mayhew, G. F., Evans, P. S., Gregor, J. & other authors (2001). Genome sequence of enterohaemorrhagic Escherichia coli O157 : H7. Nature 409, 529–533.[CrossRef][Medline]
Pradel, N., Livrelli, V., De Champs, C., Palcoux, J. B., Reynaud, A., Scheutz, F., Sirot, J., Joly, B. & Forestier, C. (2000). Prevalence and characterization of Shiga toxin-producing Escherichia coli isolated from cattle, food, and children during a one-year prospective study in France. J Clin Microbiol 38, 1023–1031.
Reid, S. D., Herbelin, C. J., Bumbaugh, A. C., Selander, R. K. & Whittam, T. S. (2000). Parallel evolution of virulence in pathogenic Escherichia coli. Nature 406, 64–67.[CrossRef][Medline]
Ren, C. P., Chaudhuri, R. R., Fivian, A., Bailey, C. M., Antonio, M., Barnes, W. M. & Pallen, M. J. (2004). The ETT2 gene cluster, encoding a second type III secretion system from Escherichia coli, is present in the majority of strains but has undergone widespread mutational attrition. J Bacteriol 186, 3547–3560.
Rutherford, K., Parkhill, J., Crook, J., Horsnell, T., Rice, P., Rajandream, M. A. & Barrell, B. (2000). Artemis: sequence visualization and annotation. Bioinformatics 16, 944–945.
Schmidt, H. & Hensel, M. (2004). Pathogenicity islands in bacterial pathogenesis. Clin Microbiol Rev 17, 14–56.
Schubert, S., Rakin, A., Karch, H., Carniel, E. & Heesemann, J. (1998). Prevalence of the "high-pathogenicity island" of Yersinia species among Escherichia coli strains that are pathogenic to humans. Infect Immun 66, 480–485.
Shen, S., Mascarenhas, M., Rahn, K., Kaper, J. B. & Karmali, M. A. (2004). Evidence for a hybrid genomic island in verocytotoxin-producing Escherichia coli CL3 (serotype O113 : H21) containing segments of EDL933 (serotype O157 : H7) O islands 122 and 48. Infect Immun 72, 1496–1503.
Tobe, T., Hayashi, T., Han, C. G., Schoolnik, G. K., Ohtsubo, E. & Sasakawa, C. (1999). Complete DNA sequence and structural analysis of the enteropathogenic Escherichia coli adherence factor plasmid. Infect Immun 67, 5455–5462.
Vernozy-Rozand, C., Montet, M. P., Lequerrec, F., Serillon, E., Tilly, B., Bavai, C., Ray-Gueniot, S., Bouvet, J., Mazuy-Cruchaudet, C. & other authors (2002). Prevalence of verotoxin-producing Escherichia coli (VTEC) in slurry, farmyard manure and sewage sludge in France. J Appl Microbiol 93, 473–478.[CrossRef][Medline]
Welch, R. A., Burland, V., Plunkett, G., III, Redford, P., Roesch, P., Rasko, D., Buckles, E. L., Liou, S. R., Boutin, A. & other authors (2002). Extensive mosaic structure revealed by the complete genome sequence of uropathogenic Escherichia coli. Proc Natl Acad Sci U S A 99, 17020–17024.
WHO (1999). Zoonotic non-O157 Shiga toxin-producing Escherichia coli (STEC). In Report of a WHO Scientific Working Group Meeting. Berlin, Germany: World Health Organization.
Received 17 December 2008;
revised 15 January 2009;
accepted 15 January 2009.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | MICROBIOLOGY | J GEN VIROL |
| J MED MICROBIOL | ALL SGM JOURNALS | |
