OUP user menu

Molecular and ultrastructural characterisation of EspA from different enteropathogenic Escherichia coli serotypes

Bianca C Neves, Stuart Knutton, Luiz R Trabulsi, Vanessa Sperandio, James B Kaper, Gordon Dougan, Gad Frankel
DOI: http://dx.doi.org/10.1111/j.1574-6968.1998.tb13301.x 73-80 First published online: 1 December 1998

Abstract

Enteropathogenic Escherichia coli (EPEC) encode a type III secretion system located on a pathogenicity island known as the locus for enterocyte effacement. Four proteins are known to be exported by this type III secretion system – EspA, EspB and EspD required for subversion of host cell signal transduction pathways and a translocated intimin receptor protein (Tir) required for intimin-mediated intimate attachment and attaching and effacing lesion formation. The espA gene is located within the locus for enterocyte effacement and the EspA polypeptide from the prototype EPEC strain E2348/69 (O127:H6) has recently been shown to be a component of a filamentous structure involved in bacteria-host cell interaction and locus for enterocyte effacement-encoded protein translocation involved in attaching and effacing lesion formation. In this study we have extended our investigation of EspA to strains belonging to other classical EPEC serotypes. DNA sequencing demonstrated that the espA gene from the different EPEC strains share at least 65% DNA identity. In addition, we detected morphologically and antigenically similar EspA filaments in all but one of the bacterial strains examined including recombinant, non-pathogenic E. coli expressing espA from a cloned locus for enterocyte effacement region (HB101(pCVD462)).

Key words
  • Enteropathogenic Escherichia coli
  • EspA
  • EspA filament

1 Introduction

Enteropathogenic Escherichia coli (EPEC) is a common cause of diarrhoea, particularly among young infants in developing countries [1]. Small bowel biopsies of children infected with EPEC reveal discrete colonies of bacteria attached to the mucosa [2]. Binding of EPEC to the brush border triggers a cascade of trans-membrane and intracellular signals leading to cytoskeletal reorganisation and formation of a specific attaching and effacing (A/E) lesion [3]. This A/E lesion is characterised by destruction of microvilli and intimate adherence of bacteria to cup-like pedestals formed by accumulation of polymerised actin beneath the EPEC bacillus [4, 5].

In vitro experiments using cultured epithelial cells have implicated several genes in A/E lesion formation. All of these genes map to a pathogenicity island termed the locus of enterocyte effacement (LEE) region [6, 7]. The LEE region encodes intimin, an outer membrane protein adhesin required for intimate attachment of EPEC to host cells [8, 9] and for the organisation of polymerised actin into a cup-like pedestal beneath each attached bacterium [4]. The LEE region also encodes a type III secretion system [10], and three EPEC-secreted proteins EspA, EspB and EspD [1113], which are required for induction of host cell protein phosphorylation, including tyrosine phosphorylation of a fourth secreted protein Tir (for translocated intimin receptor) (formerly Hp90) which serves as a host cell intimin receptor [14, 15]. Previous reports [16, 17] suggested that expression of some of the Esps depends on the per (plasmid encoded regulatory) locus [18] which also regulates the expression of intimin and the BFP (bundle forming pilus) [1820]. Both the per locus and the bfp gene cluster reside on a 90-kb EAF (EPEC adherence factor) plasmid (pMAR2), commonly found in EPEC isolates [5].

In a recent study we showed that EspA is a major component of a large extracellular filamentous appendages (EspA filament) on the surface of EPEC which appears transiently during A/E lesion formation, forms a direct link between the bacterium and the host cell surface and is required for the translocation of EspB into infected host cells [19, 20]. Here we report the molecular and ultrastructural characterisation of EspA, and EspA filaments from different EPEC strains. We found 65% overall identity between the different EspA polypeptides and detected EspA filaments on all but one of the clinical EPEC strains. In addition, we have demonstrated that E. coli K-12 containing the cloned LEE region is also capable of producing the EspA filaments.

2 Materials and methods

2.1 Bacterial strains

The bacterial strains used in this study are listed in Table 1.

View this table:
Table 1

Bacterial strains and plasmids used in this study

Serotype/strainDescriptionReference/originEspAa
O127:H6 (E2348/69)EPEC (prototype)[8]+
O127:H6 (JPN15)EPEC (EAF plasmid-cured E2348/69)[8]+
O127:H6 (UMD 872)EPEC (espA-deficient E2348/69)[12]
O128ab:H2/(20)EPECBrazil+
O119:H6/(36)EPECBrazil+
O119:H6/(16)EPECBrazil+
O111ab:H2/(19)EPECChile
O55:H7/(58)EPECBrazil+
O55:H6/(30)EPECBrazil+
O119:H2/(79)EPECBrazil
HB101(pCVD462)E. coli K-12 containing the cloned LEE region from E2348/69[7]+
HB101(pCVD462/pMAR7)Like HB101(pCVD462) but with pMAR7[7]+
  • aResults of immunofluorescence staining of EspA filaments.

2.2 Cloning of espA genes

Long-range polymerase chain reaction (PCR) was used to amplify a fragment of the LEE region spanning from orfU to espB. The espA gene is located on this fragment downstream of the eae gene and upstream of espD[12]. EPEC chromosomal DNA, purified from 1-ml overnight bacterial cultures using QIAamp tissue kit (Qiagen), was used as a template. PCR was performed with Gene Amp XL PCR kit (Perkin Elmer) using orfU-F (5′-TTATCTGACACTAATGACGAATATATGATG) and espB-R (5′-GTCATATCACGTAGACGGCTAGAGA) primers and 28 cycles of 94°C for 1 min and 60°C for 10 min. The PCR products were cloned into pGEM-T vector (Promega). Recombinant plasmids were screened by PCR using internal espA primers, as described [19].

2.3 DNA sequencing of espA

Recombinant plasmids were purified as described by Adu Bobie et al. [21]. DNA sequencing was performed using a DyeDeoxy Terminator Cycle Sequencing Kit (Applied Biosystems), according to the manufacturer's instructions, and Perkin Elmer ABI/Prism 377 automated DNA sequencer. On the basis of the emerging DNA sequence, additional (walking) primers were synthesised in the forward and reverse orientations (for sequencing of both DNA strands). Sequence analysis and contig assembly was carried out using Genejockey II in an Apple Macintosh computer. The nucleotide sequences encoding EspA from the different strains have been submitted to the EMBL database, under accession numbers: AJ225015 [O119:H2/(79)], AJ225016 [O119:H6/(16)], AJ225017 [O119:H6/(36)], AJ225018 [O111:H2/(19)], AJ225019 [O55:H6/(30)], AJ225020 [O55:H7/(58)] and AJ225021 [O128:H2/(20)].

2.4 Preparation of EPEC-secreted proteins (Esp)

The analysis of culture supernatants for EPEC-secreted proteins was performed as previously described [10, 19]. A volume of concentrated samples corresponding to 4×106 colony-forming units was loaded and separated by 12% SDS-PAGE. Separated proteins were transferred to nitrocellulose (Bio-Rad Laboratories, Richmond, CA) membrane as described previously [19]. Membranes were blocked with 2% bovine serum albumin in phosphate-buffered saline (PBS) containing 0.05% Tween 20 for 2 h and probed with anti-EspA antiserum at a dilution of 1:1000 in PBS-0.05% Tween 20. Subsequently, goat anti-rabbit secondary antibody, horseradish peroxidase conjugate, was used at a concentration of 1:2000 in PBS-0.05% Tween for 1 h. 4-Chloro-1-naphthol (Sigma) was used for the development of the immunoblots.

2.5 Electron microscopy and immunofluorescence staining of bacterial cells

For immunolabelling of bacteria, 10-ml samples of washed bacterial suspensions were applied to carbon-coated grids for 5 min, excess liquid removed, and grids immediately placed face down on drops of EspA antiserum (1:80 dilution) for 30 min. After washing, grids were placed on drops of 10-nm gold-labelled goat anti-rabbit sera (1:20 dilution) (British BioCell International) for 30 min. After further washing in PBS and distilled water grids were either air dried or negatively stained with 1% ammonium molybdate and dried. For immunogold labelling of cell-associated bacteria, HEp-2 cell monolayers were briefly fixed for 10 min in 0.1% glutaraldehyde, washed, and incubated with EspA antiserum for 2 h at room temperature. Cells were washed and incubated with gold-labelled goat anti-rabbit serum for 12 h at 4°C. After further thorough washing, cells were fixed in 3% buffered glutaraldehyde and processed for thin-section electron microscopy using standard procedures. Samples were examined in a Jeol 1200EX electron microscope operated at 80 kV.

For scanning electron microscopy HEp-2 cell monolayers were briefly fixed for 10 min in 0.1% glutaraldehyde, washed, incubated with EspA antiserum for 2 h, washed again and fixed in 3% buffered glutaraldehyde. Monolayers were post-fixed in 1% osmium tetroxide, dehydrated through graded acetone solutions and critical point dried. Mounted specimens were sputter-coated with gold (Polaron Ltd) and examined in a Jeol 1200EX scanning transmission EM operated at 40 kV.

Immunofluorescence staining of bacterial cells was performed on infected HEp-2 cells monolayers, following 3 h incubation with an overnight bacterial culture as previously described [19].

3 Results

3.1 Analysis of the EspA DNA sequence from different EPEC strains

Long-range PCR amplification was employed, using purified DNA from the different EPEC strains as templates, to clone a fragment of the LEE region containing the eae, espA, espD and espB genes. The DNA sequence of the espA genes was determined by automatic DNA sequencing and use of walking primers. Alignment of the EspA amino acid sequences, obtained from the different strains and including the published EspA sequence from the prototype EPEC strain E2348/69 (O127:H6) [12], revealed 65% overall identity (Fig. 1). However, careful examination of the amino acid sequences revealed that EspA of strain O119:H6/(16) not only had a two-amino acid deletion (at positions 13, 14) but also contributed most to the sequence diversity (Fig. 1). Indeed, omitting the O119:H6/(16) strain from the sequence alignment revealed 81% overall identity between the other EspA polypeptides. In order to rule out the possibility that the EspA sequence of stain O119:H6/(16) is unique to this isolate, we cloned and sequenced the espA gene from a different O119:H6 isolate, strain 36. Alignment of this sequence with EspA of O119:H6/(16) revealed 100% identity, suggesting the O119:H6 strains encode a more divergent EspA polypeptide.

Figure 1

Alignment of the EspA amino acid sequences from the different EPEC serotypes (dots on top of the sequence represent conserved residues). 65% overall identify was recorded. EspA from the O119:H6 strains (16 and 36) contains a two-amino acid deletion and amino acid substitutions. Eliminating this EspA from the alignment increases the overall identity between the other EspA sequences to 81%.

3.2 Characterisation of EPEC-secreted EspA

Secreted EPEC proteins were made from each of the EPEC strains. Probing the secreted EPEC proteins with the EspA antiserum revealed a variation in the level of reactivity. Strong reaction with the EspA antiserum was obtained with strains E2348/69, O128:H2(20), O55:H6/(30) and O55:H7/(58). The two O119:H6 strains, although appearing to secrete a high level of the protein (data not shown), showed only a moderate reaction with the EspA antiserum (Fig. 2) whilst strain O111:H2/(19) secreted a low level of the protein (data not shown) and reacted poorly with the antiserum; no reactivity was observed with EPEC strain UMD872 (Fig. 2), harbouring a deletion mutation in the espA gene [12]. Similar levels of EspA were observed for E2348/69 and JPN15, a pMAR2-cured derivative of E2348/69 (data not shown).

Figure 2

Western blot detection of EspA in the mid-exponential-phase bacterial culture supernatants. Strong reaction with the EspA antiserum was seen from strains E2348/69, HB101(pCVD462), O128:H2/(20), O55:H6/(30) and O55:H7/(58) (lanes 1, 3, 5, 9 and 10, respectively), intermediate levels of reactivity from strains HB101(pCVD462/pMAR7), O119:H6/(16), O119:H6/(36), and O119:H2/(79) (lanes 4, 6, 7 and 11 respectively) and low levels of reactivity from strain O111:H2/(19) (lane 8); no EspA was detected in the bacterial supernatants of UMD872 (lane 2).

3.3 Ultrastructural examination of the EspA filaments

By immunoelectron microscopy, the EspA antiserum labelled numerous ∼50 nm diameter and up to 2 mm long filamentous structures on the surface of wild-type and plasmid-cured EPEC strain E2348/69 (Fig. 3A). In cell adhesion assays, and prior to A/E lesion formation, these structures were seen to form a bridge between bacteria and the eukaryotic cell surface (Fig. 3B); similar structures linking bacteria and the eukaryotic cell surface were seen by scanning electron microscopy (Fig. 3C).

Figure 3

By immunogold labelling an EspA antiserum stained ∼50 nm diameter and up to 2 μm long filamentous surface structures on wild-type and plasmid-cured E2348/69 (A) which, in cell adhesion assays, were seen to form a bridge between bacteria and the HEp-2 cell surface (B). Similar filamentous structures connecting bacteria and the HEp-2 cell surface were also seen by scanning electron microscopy (C). A, ×25 000; B, 35 000; C, 30 000.

In this study we used immunofluorescence with the EspA antiserum to detect EspA filaments on the surface of E. coli strains belonging to different EPEC serotypes during infection of HEp-2 cells. Clearly defined EspA filaments, similar to those produced by wild-type and plasmid-cured E2348/69 (Fig. 4A), were observed on the surface of EPEC strains O128:H2/(20), O55:H6/(30), O55:H7/(58) and O119:H2/(79) (Table 1). Somewhat shorter and less defined EspA filaments were observed on the surface of the two O119:H6 strains (16 and 36) (Fig. 4D; Table 1); EspA filaments could not be detected on O111:H2/(19) (Table 1) or UMD872, the negative control strain (Fig. 4B; Table 1).

Figure 4

Fluorescence micrographs showing EPEC strains stained with the EspA antiserum. The antiserum stained clearly defined filamentous surface structures (arrows) on strains E2348/69 (A) and HB101(pCVD462) (C) and short filaments on strains O119:H6/(36) (D). No staining of EspA filaments was detected on the negative control strain UMD872 (B).

3.4 Expression of EspA by HB101(pCVD462) and HB101(pCVD462/pMAR7)

EPEC-secreted protein preparations were also made from E. coli K-12 (HB101) strains containing either the cloned LEE region, HB101(pCVD462), or the cloned LEE region and EPEC virulence plasmid pMAR7, HB101(pCVD462/pMAR7) [7]. Probing the Esp preparations with the EspA antiserum revealed, surprisingly, a higher level of secreted EspA from HB101(pCVD462) than from HB101(pCVD462/pMAR7) (Fig. 2). Consistent with this result, we also observed longer EspA structures on HB101(pCVD462) (Fig. 4C) and shorter EspA filaments on the surface of HB101(pCVD462/pMAR7) (data not shown).

4 Discussion

EPEC exert their pathogenic effects by binding intimately to host intestinal epithelial cells and producing A/E lesions. Three EPEC-secreted proteins, EspA, EspB and EspD, have been shown to be required to stimulate host cell signal transduction events required for A/E lesion formation [1113] but the mechanisms by which these proteins stimulate signal transduction in host cells remain unknown. Recently, we have shown that one EPEC-secreted protein, EspA, is a major component of a transiently expressed surface organelle which forms a direct link between the bacterium and the host cell, and which is required for the translocation of another secreted protein, EspB, into host cells [19]; EspA has also been shown to be required for translocation to the host cell of an intimin receptor, Tir [14]. In this study, we compared the amino acid sequence and expression of EspA of several clinical EPEC isolates. Our results show that, overall, the EspA polypeptides share at least 65% identity, and that one of the EPEC serotypes tested (O119:H6) encodes a more divergent EspA polypeptide. The differences in this EspA, compared with the prototype EspA (that of E23248/69), included amino acid substitutions and deletion of two amino acids close to the amino-terminus. However, we found that the sequence of EspA polypeptides from the two strains belonging to EPEC clone 1 (O127:H6 and O55:H6) [21] were identical, while EspA from strains belonging to EPEC clone 2 (O111:H2, O119:H2 and O128:H2) were at least 95% identical. These results complement previous studies showing that in EPEC clone 1 the LEE region was inserted into the selC locus [22], and that these strains express a distinct intimin type (intimin α) [21], while in strains belonging to EPEC clone 2 the LEE was inserted at a different site [22] and these strains specifically express a different intimin type, intimin β[21].

Polyacrylamide gel electrophoresis and Western blots using EspA antiserum, made using recombinant EspA from E2348/69 (O127:H6) as immunogen, cross-reacted with the EspA polypeptide of all the serotypes tested despite the inter-bacterial variation in the levels of the secreted Esps. The Western blot analysis also correlated with observation of EspA filaments revealed by immunofluorescence straining of infected HEp-2 cells. Strains secreting high levels of EspA and showing high cross-reactivity with the antiserum showed clearly defined EspA filaments, whereas strains secreting low levels of EspA (O111:H2/(19)), or showing low cross-reactivity (O119:H6/(16 and 36)), showed either short EspA filaments or no filaments at all. Most interesting were the two E. coli K-12 strains possessing either the LEE, or the LEE and pMAR7 plasmids. A lower level of secreted EspA was seen with HB101(pCVD462/pMAR7), and EspA filaments were barely detectable. In contrast, higher levels of secreted EspA were seen from HB101(pCVD462) and this strain produced EspA filaments which appeared longer even than those produced by E2348/69. This difference could be due to the fact that the LEE region in HB101(pCVD462) is encoded on a multi-copy-number plasmid. Together with the fact that we were unable to detect any difference in the level of secreted EspA and EspA filaments between E2348/69 and JPN15 [19], this difference in EspA expression between HB101(pCVD462) and HB101(pCVD462/pMAR7) indicates that expression of EspA is not dependent on the EAF plasmid-encoded per genes, although the presence of the pMAR7 plasmid may influence expression/secretion of EspA from a multi-copy plasmid in K-12.

The presence of an extracellular polymeric EspA-associated organelle, which seems to bridge and mediate cross-talk between pathogenic bacteria and target host cells, makes it an attractive target for any future EPEC vaccine. Indeed, in recent studies we found high levels of IgA antibodies in human colostrum from mothers in Sao Paulo, Brazil [23], a region where EPEC diarrhoea is endemic among low social level infants, and it is one of the main causes of morbidity and mortality in this group.

Acknowledgements

This work was supported by Wellcome Trust grants to G.D. and S.K. and a FINEP/MCT/PRONEX grant (41.96.0881.00) to L.R.T. B.C.N.'s visit to Imperial College was supported in part by the British Council, Brazil.

References

View Abstract