Locus of enterocyte effacement from Citrobacter rodentium: sequence analysis and evidence for horizontal transfer among attaching and effacing pathogens

Abstract

The family of attaching and effacing (A/E) bacterial pathogens, which includes diarrheagenic enteropathogenic Escherichia coli (EPEC) and enterohemorrhagic E. coli (EHEC), remains a significant threat to human and animal health. These bacteria intimately attach to host intestinal cells, causing the effacement of brush border microvilli. The genes responsible for this phenotype are encoded in a pathogenicity island called the locus of enterocyte effacement (LEE). Citrobacter rodentium is the only known murine A/E pathogen and serves as a small animal model for EPEC and EHEC infections. Here we report the full DNA sequence of C. rodentium LEE and provide a comparative analysis with the published LEEs from EPEC, EHEC, and the rabbit diarrheagenic E. coli strain RDEC-1. Although C. rodentium LEE shows high similarities throughout the entire sequence and shares all 41 open reading frames with the LEE from EPEC, EHEC, and RDEC-1, it is unique in its location of the rorf1 and rorf2/espG genes and the presence of several insertion sequences (IS) and IS remnants. The LEE of EPEC and EHEC is inserted into the selC tRNA gene. In contrast, the Citrobacter LEE is flanked on one side by an operon encoding an ABC transport system, and an IS element and sequences homologous to Shigella plasmid R100 and EHEC pO157 flank the other. The presence of plasmid sequences next to C. rodentium LEE suggests that the prototype LEE resided on a horizontally transferable plasmid. Additional sequence analysis reveals that the 3-kb plasmid in C. rodentium is nearly identical to p9705 in EHEC O157:H7, suggesting that horizontal plasmid transfer among A/E pathogens has occurred. Our results indicate that the LEE has been acquired by C. rodentium and A/E E. coli strains independently during evolution.

FIG. 1

Gene map for C. rodentium LEE. Gene designation is according to that of EPEC E2348/69 LEE, which is shown for comparison (17). The orientation of each gene is shown by the direction of the arrow. Note the different locations of the rorf1 (r1) and rorf2 (espG/r2) genes in EPEC and C. rodentium LEE, as well as the association of several IS's or IS remnants with the C. rodentium LEE. The major operons encoded by the LEE (LEE1, -2, -3, and -4, Tir, and R1/R2) and their transcriptional directions are shown and adapted from references and . Please see Fig. 3 for details of the LEE flanking regions. The map is not drawn completely to scale.

FIG. 2

Alignment of EspF proteins from C. rodentium, EHEC O157:H7 strain EDL933, EPEC O127:H6 strain E2348/69, and rabbit diarrheagenic E. coli strain RDEC-1 (O15:H⁻). Identical amino acids are indicated by dots, while dashes show absent amino acids. Note the different numbers of the 47 amino acid repeats and the different sizes of the proteins.

FIG. 3

Insertion sites of the LEE in the chromosomes of C. rodentium, EPEC, EHEC, and RDEC-1. The major operons and their transcriptional directions in the LEE are shown. Note the unique gene organization of rorf2 (espG) and rorf1 genes in C. rodentium LEE. The diagram is not drawn to scale.

FIG. 4

C. rodentium carries at least three plasmids, and the LEE is not located on these plasmids. (A) Ethidium bromide-stained 0.7% agarose gel, with 1-kb DNA ladder from GIBCO BRL as molecular marker (in kilobase pairs); (B) ethidium bromide-stained 0.7% agarose gel, with λHindIII as standard; (C) Southern blot hybridization of gel B using the C. rodentium escC gene as a probe. Locations of the three plasmids, pCRP1/pDBS1, pCRP2, and pCRP3, in C. rodentium and the EAF plasmid are indicated. DBS100, wild-type C. rodentium strain; DBS231, C. rodentium strain cured of the plasmid pDBS1/pCRP1; E2348/69, EPEC O127:H6 strain E2348/69. Chr stands for sheared chromosomal DNA band. Sizes are given in kilobases.

FIG. 5

A schematic model to explain the different gene organization seen for the rorf1 and rorf2/espG genes in the C. rodentium LEE versus that of EPEC, EHEC, and RDEC-1. The model is based on the hypothesis that the prototype LEE was located on an R100- and pO157-like plasmid. When it was disseminated and introduced into different A/E pathogens during evolution, the insertion of the plasmid into the bacterial chromosome can occur at at least two sites, I and II. These different recombination events, coupled with the retention or loss of the plasmid sequences (dashed line for C. rodentium LEE), resulted in the different gene orders for rorf1 and rorf2 (espG) genes and the different LEE lineages observed for EPEC, EHEC, RDEC-1, and C. rodentium, as well as other A/E pathogens. The major operons encoded by the LEE (LEE1, -2, -3, and -4, Tir, and R1/R2) and their transcriptional directions are shown and adapted from references and .