Previous Article | Next Article ![]()
Infection and Immunity, May 2002, p. 2681-2689, Vol. 70, No. 5
0019-9567/02/$04.00+0 DOI: 10.1128/IAI.70.5.2681-2689.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.
Laboratório Especial de Microbiologia, Instituto Butantan, São Paulo, Brazil,1 Institute of Infections and Immunity, Nottingham University, Nottingham,2 Institute of Child Health, Birmingham University, Birmingham, United Kingdom3
Received 2 July 2001/ Returned for modification 31 October 2001/ Accepted 8 January 2002
|
|
|---|
|
|
|---|
E. coli O55 is one of the most frequently isolated serogroups of EPEC implicated in infantile diarrhea (11). Two major electrophoretic types (ETs) designated ET1 and ET5 have been identified in this serogroup by multilocus enzyme electrophoresis analysis (26). ET1 corresponds to typical O55:H6 EPEC strains, while ET5 comprises O55:H- and O55:H7 strains with different combinations of virulence genes. In addition to A/E lesion formation, some O55:H-/O55:H7 EPEC strains displayed a characteristic diffuse adherence (DA) pattern, whereas others displayed both LA and DA patterns (Fig. 1). Hybridization with a DNA probe (daaC probe) developed to detect DA E. coli (DAEC) (5) indicated that such strains possess a specific DA adhesin. The aim of this study was to characterize the DA adhesin expressed by O55 DA-EPEC strains.
![]() View larger version (61K): [in a new window] |
FIG. 1. Micrographs illustrating the different adherence pattern phenotypes produced by O55 ET5 EPEC strains. (a) Strain 384 showing LA; (b) strain 5E showing DA; (c) strain 135/12 showing combined LA and DA. Magnification, x100.
|
|
View this table: [in a new window] |
TABLE 1. Genotypic and phenotypic characteristics of the O55 EPEC strains studieda
|
Adhesin purification and N-terminal sequencing.
For adhesin purification bacteria were grown in Luria-Bertani agar at 37°C to optimize adhesin production, and bacterial surface proteins were extracted as described by Stirm et al. (30). The isolated surface proteins in the cleared supernatant were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), using a 15% polyacrylamide gel and proteins visualized using Coomassie brilliant blue staining. For N-terminal amino acid sequencing, proteins in the supernatant were precipitated overnight at 4°C with 80% ammonium sulfate and collected by centrifugation. Following dialysis, proteins were separated by SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membranes by electroblotting. The isolated adhesin protein was visualized by Western blot analysis using a rabbit antiserum raised against whole 135/12 bacteria that had been repeatedly absorbed with 135/12 grown at 25°C (absorbed 135/12 antiserum). The position of the protein which reacted strongly with the antiserum was visualized on duplicate PVDF membranes by amide black staining and excised for N-terminal amino acid sequencing. The presence of a common
16-kDa protein in 135/12 and all ET5 O55 DA-EPEC strains examined (Table 1) suggested the presence of a distinct DA adhesin. N-terminal sequencing of this protein showed identity with the N-terminal portion of AfaE-1, an afimbrial adhesin of uropathogenic E. coli (UPEC) (18).
Cloning and sequence analysis of EPEC DA adhesin genes.
Genomic DNA of the 135/12 EPEC strain was partially digested with restriction endonuclease Sau3A I (Gibco) to obtain a range of DNA fragments between 11 and 20 kb which were cloned into
BlueSTAR using the BamHI Arms Kit (Novagen). The library was screened using the absorbed 135/12 antiserum (28), and plasmids from 24 positive plaques were rescued and used for transformation of E. coli strain DH5
. All 24 clones displayed DA. Plasmids extracted from each adherent clone were cut with a range of restriction enzymes, and the banding patterns were compared by agarose gel electrophoresis. This yielded 12 independent clones of the same chromosomal region, which conferred a typical DA pattern. Clone DH5
(pRK-1), which contained an 11.6-kb insert, expressed a protein of
16 kDa, and promoted DA and hemagglutination of human red blood cells, was selected for further analysis.
The DNA sequence of pRK-1 was determined by using a Dyedeoxy Terminator Cycle Sequencing kit (Applied Biosystems). On the basis of the emerging DNA sequence, additional (walking) primers were synthesized in the forward and reverse orientations (for sequencing of both DNA strands). Sequence analysis and contig assembly was carried out using DNAstar. The lefthand region of pRK-1 consisted of five open reading frames (ORFs) which, on comparison with published sequences, revealed a very high degree of identity to the Afa afimbrial adhesin operon (18) (Fig. 2). The five genes were designated EPEC afaA to -E, in line with the adhesin designation Afa (afimbrial adhesin) and comparison with the afa gene nomenclature.
![]() View larger version (25K): [in a new window] |
FIG. 2. Linear restriction map of the 11.6-kb fragment from the pRK-1 clone, showing the constructions obtained in this study (pRK-7 and pRK-51) and the genetic organization of EPEC afa operon. The vertical bars indicate the enzyme restriction sites. The arrows indicate different ORFs and the horizontal bars indicate genetic sequences. The percentages indicate the degree of identity between the EPEC Afa and AfaE-1 amino acid sequences. F and R were the primers used to amplify the EPEC afaE gene to generate pRK-51. (Primer sequences: forward, CCGATATCCGCAGACCGTGGAATAAGGCATCAC; reverse, CCGCTCGAGGCCTTGACATCCACGTTTGCACCGTC; underlining indi-cates restriction sites).
|
Downstream of EPEC, afaE ORFs without homology to published sequences were localized. The right end of pRK-1 contained sequences with homology to E. coli K-12 (Fig. 2). Upstream were sequences showing 97% identity to the phenylalanyl tRNA locus pheV that is inserted at 67 min on the K-12 E. coli chromosome, indicating that the EPEC afa genes are inserted into the chromosome at this point. This region is known to be a hot spot for insertion of retrotransposons, phages, and pathogenicity islands (12, 31).
Upstream of pheV we identified three ORFs in an opposite frame to the afa operon (Fig. 2) and which showed homology with integrase genes. ORF 1 (104 aa) had 72% identity with this locus of the Shigella flexneri SHI-2 pathogenicity island (GenBank accession number AF141323) (20). ORF 2 (128 aa) had 68% identity with the integrase gene of the S. flexneri SHI-2 pathogenicity island (accession number AF141323) (20) and 66% identity with the CP4-like integrase present in the LEE pathogenicity island of E. coli O157:H7 (accession number AF071034) (25). ORF 3 (126 aa) had 77% identity to this gene of the S. flexneri SHI-2 pathogenicity island and 78% identity with the CP4-like integrase. These ORFs showed more than 50% identity with other integrases from a number of known bacteriophages, including the putative prophage SF6-like integrase (accession number P37326) and the integrase-phage phi-R73 (accession number A42465M), and with the prophage CP4-57 integrase (accession number U03737). These sequence data suggest that the EPEC afa operon is likely to have been acquired by horizontal transmission in the presence of a putative integrase and may be part of a pathogenicity island; sequences upstream of the afa operon are currently being investigated to determine if this is the case.
Primers derived from an internal afaE-1 sequence (33) amplified a 487-bp fragment from strain 135/12 and from all of the ET5 DA-EPEC strains, suggesting that the DA component of the DA phenotype of these strains could be mediated by an Afa-like adhesin (Table 1). The amplicon obtained from strain 135/12 was cloned into the pCR II vector to generate pRK7 (Fig. 2) and sequenced using a Perkin-Elmer ABI Prism 373 automated DNA sequencer. Analysis of the sequence of the EPEC afaE gene confirmed 98% identity at the amino acid level with AfaE-1.
Mutagenesis and complementation of the EPEC afaE gene.
pRK-1 cloned into E. coli K-12 strain DH5
exhibited DA to HEp-2 cells (Fig. 3c). To demonstrate that EPEC afaE was required for expression of the EPEC Afa adhesin and, consequently, for the diffuse pattern of HEp-2 cell adherence, an internal fragment of this ORF was cloned into the multiple cloning site of the suicide vector pJP5603 (designated pRK-20) (24). This construction was then mobilized into parent strain 135/12, and integration into homologous genes was confirmed by Southern blotting (data not shown). In standard HEp-2 adherence assays (7) the mutant, designated 135/12(afaE::pRK-20), showed only LA; DA was abolished (Fig. 3e), as was the ability to agglutinate human erythrocytes (data not shown). Western blot analysis showed that the band corresponding to the
16-kDa surface protein expressed by the EPEC afaE gene (Fig. 4, lanes 1 and 4) was absent in strain 135/12(afaE::pRK-20) (Fig. 4, lane 2). Complementation in trans of the mutated afaE gene was achieved by transforming 135/12(afaE::pRK-20) with pRK-51 to produce strain 135/12(afaE::pRK20)/pRK-51. This plasmid was constructed from a PCR product amplified from the flanking regions of the EPEC afaE gene with reverse and forward primers (Fig. 2). The 1,039-bp fragment obtained was digested with EcoRV and XhoI (Gibco) and ligated into pBC phagemid SK+ (Stratagene Cloning Systems, La Jolla, Calif.). DA (Fig. 3g), agglutination of human erythrocytes, and expression of the
16-kDa surface protein (Fig. 4, lane 3) were all restored when 135/12(afaE::pRK-20) was complemented with the EPEC afaE gene [strain 135/12(afaE::pRK20)/pRK-51].
![]() View larger version (105K): [in a new window] |
FIG. 3. Micrographs illustrating the adherence patterns on HEp-2 cells (left) and the inhibition of the DA pattern with anti-EPEC Afa serum (right). (a) LA and DA patterns in strain 135/12; (b) inhibition of the DA pattern with persistence of the LA pattern. (c) The DA pattern of DH5 (pRK-1); this pattern is inhibited in panel d. (e) The DA pattern was abolished in the mutant 135/12 (afaE::pRK-20), which displays only the LA pattern; this pattern was maintained in panel f. (g) The DA pattern was restored in the complemented strain 135/12(afaE::pRK-20)/pRK51, displaying both the LA and DA patterns. (h) The DA component of the LA-DA pattern was inhibited after addition of anti-EPEC Afa serum. Magnification, x100.
|
![]() View larger version (35K): [in a new window] |
FIG. 4. SDS-PAGE and immunoblot analysis with anti-EPEC Afa serum of the bacterial surface protein extracts of strains 135/12 (lane 1), 135/12(afaE::pRK-20) (lane 2), 135/12 (afaE::pRK-20)/pRK51 (lane3); and DH5 (pRK-1) (lane 4). (a) The 16-kDa protein was expressed by strains 135/12, 135/12(afaE::pRK-20)/pRK51, and DH5 (pRK-1) but was absent in the mutant 135/12(afaE::pRK-20) strain. (b) As indicated by the arrow, the anti-EPEC Afa shows recognition of the 16-kDa protein from strains 135/12, 135/12(afaE::pRK-20)/pRK51, and DH5 (pRK-1), whereas this protein was not recognized on the mutant 135/12(afaE::pRK-20).
|
(pRK-2) and repeatedly absorbed with DH5
was also used to examine inhibition of adherence of wild-type 135/12, DH5
(pRK-1), 135/12(dafaE::pRK-20), and 135/12 (dafaE::pRK-20)/pRK51 (Fig. 3). The DA component of the LA-DA pattern displayed by both wild-type 135/12 and EPEC afaE-complemented strain 135/12 (dafaE::pRK-20)/pRK51 (Fig. 3a and g) was inhibited by the antiserum, leaving only an LA pattern (Fig. 3b and h); inhibition of DA in the cloned EPEC afa strain DH5
(pRK-1) resulted in no adherence (Fig. 3d), while the EPEC afaE deletion mutant 135/12(dafaE::pRK-20) displayed only LA that was unaffected by the EPEC Afa antiserum (Fig. 3f). Reduced LA-DA adherence and expression of the 16-kDa protein of 135/12 at 30°C and no adherence and the absence of the 16-kDa protein at 25°C (Fig. 5a and b), together with the inhibition of bacterial adherence by proteins extracted from the surface of strain 135/12 grown at 37°C but not at 25°C, indicated temperature-dependent expression of both the LA (Bfp) and DA (EPEC Afa) adhesins.
![]() View larger version (72K): [in a new window] |
FIG. 5. (A) Micrographs illustrating reduced LA and DA produced by strain 135/12 at 25 and 30°C, in comparison with LA and DA produced at 37°C. (B) SDS-PAGE of the surface proteins extracted from strain 135/12 and immunoblot with anti-EPEC Afa serum, showing the presence of the 16-kDa adhesin as indicated by the arrow, only after growth at 30 and 37°C.
|
(pRK-1) in comparison with AfaE-1 expressed by strain KS52, the prototype AfaE-1-expressing UPEC strain (18). By negative staining (15), strains 135/12 (Fig. 6a) and DH5
(pRK-1) (Fig. 6c) were seen to express very fine fibrillar material at the bacterial surface, although it was very difficult to resolve individual fibrils; similar fibrillar material was produced by strain KS52 in addition to numerous rigid rod-like fimbriae (Fig. 6e). This fine fibrillar surface material specifically stained with an antiserum raised against strain 135/12 surface proteins, and some immunogold-labeled preparations of strain 135/12 (Fig. 6b) and DH5
(pRK-1) (Fig. 6d) clearly revealed a fibrillar pattern of gold labeling; a similar staining pattern was also seen with strain KS52 (Fig. 6f).
![]() View larger version (127K): [in a new window] |
FIG. 6. Negative-stained (a, c, and e) and immunogold-labeled (b, d, and f) electron micrographs of EPEC strain 135/12 (a and b), DH5 (pRK-1) (c and d), and UPEC strain KS52 (e and f). All three strains expressed fine fibrillar material at the bacterial surface (a, c, and e; arrows), while strain KS52 simultaneously expressed rod-like fimbriae (e). The fibrillar structure was more clearly revealed following immunogold staining (b, d, and f; arrows). Bars, 0.2 µm.
|
![]() View larger version (201K): [in a new window] |
FIG. 7. Transmission electron micrographs showing HEp-2 cells infected with EPEC strain 135/12 (a and b) and UPEC strain KS52 (c and d). Strain 135/12 produced A/E lesions on HEp-2 cells (a and b), whereas strain KS52 did not exhibit intimate attachment (c and d). In each case, immunogold labeling stained surface material which frequently showed a fibrillar structure and which connected bacteria to the HEp-2 cell surface (b and d; arrows) except for where intimate EPEC attachment had occurred (a and b). Bars, 0.2 µm.
|
Afa gene clusters encoding afimbrial adhesins have previously been identified in DAEC associated with both intestinal and urinary tract infections (1), but this is the first characterization of afa genes associated with typical A/E EPEC. Afa determinants belong to a family of gene clusters that includes the dra and daa genes, encoding the Dr and F1845 adhesins, respectively, collectively called the Afa/Dr DAEC family (23). Afa gene clusters share a highly conserved region, including afaA, afaB, afaC, afaD, and afaF genes, genes which encode regulatory, chaperone, and usher functions (8, 9, 17). The structural adhesin-encoding gene, afaE, however, is highly heterogeneous, leading to the production of antigenically distinct adhesins (AfaE-I to AfaE-IV in humans) (16, 8). Bacteria expressing AfaE-1 and AfaE-III specifically attach to epithelial cells by recognizing the CD55 decay-accelerating factor (DAF) (22); this is also likely to be the case for EPEC Afa but it has yet to be confirmed.
The EPEC afa gene cluster was localized to genes homologous to the daaC probe. Sequence analysis of the daaC probe used in this study showed 97.7% identity within its 390 bp with the EPEC afa usher gene (afaC), suggesting that the daaC probe, because of its high degree of identity among ushers of the Dr family of adhesins, should be considered a probe for the Afa/Dr family of adhesins and not just for the presence of the F1845 fimbrial adhesin as originally described (5).
Afa-E1 is reported to be an afimbrial adhesin. The ultrastructural studies performed here indicate that EPEC Afa has a fine fibrillar structure. It was conceivable that a 2-aa difference in the primary sequences could result in different quaternary adhesin structures. Indeed, AfaE-III and the Dr hemagglutinin differ in only 3 aa, and yet AfaE-III is reported to be afimbrial and the Dr hemagglutinin is reported to be fimbrial (33). However, we thought it unlikely that EPEC Afa and AfaE-1 would have radically different structures, and this was confirmed when we examined AfaE-1 expressed by KS52, the prototype AfaE-1 isolate; KS52 possessed a fine fibrillar surface structure indistinguishable from that of strain 135/12. Fine fibrillar surface structures such as these are very difficult to visualize by electron microscopy, probably because they easily collapse or are sensitive to drying during preparation for electron microscopy; based on negative staining alone, we might have concluded that EPEC Afa was afimbrial. However, stabilization with antibody made it much easier to demonstrate the fibrillar structures. In view of this study, a reexamination of other AfaE adhesins might also reveal fibrillar structures and thereby stimulate a review of current Afa nomenclature.
As far as we are aware this is the first description of an Afa/Dr adhesin associated with EPEC, and it may be characteristic of the O55 serogroup, since the screening of our comprehensive EPEC collection did not identify this adhesin in any other EPEC serogroup. Furthermore, EPEC Afa is not restricted to the DA-EPEC isolated in Brazil. We also found Afa in four daaC+ O55:H- EPEC strains isolated in the United Kingdom (kindly provided by Henry Smith, Central Public Health Laboratory, London, England). These O55 strains showed the same ET5 profile when analyzed by randomly amplified polymorphic DNA (unpublished observations).
Of interest is the possible role of Afa in EPEC pathogenicity. The presence of both LA and DA patterns in the in vitro cell adherence assays indicates that both Bfp and Afa are being expressed, with Bfp-expressing bacteria giving rise to localized bacterial microcolonies and the LA phenotype and Afa-expressing bacteria giving rise to the DA phenotype. The simultaneous expression of both LA and DA phenotypes might confer some advantage during EPEC colonization of the gut, particularly in the initial stage of adhesion and infection; the presence of EPEC Afa might be a particularly useful marker as an additional adhesin in those atypical EPEC strains that lack EAF plasmids and Bfp. Phase variation of Afa expression in vivo (32) might further assist strains to evade host immune responses and, thus, enhance intestinal colonization. Members of the Afa/Dr family of adhesins have been shown to bind to a CD55 DAF receptor, an interaction that triggers brush border microvillus injury and actin cytoskeletal alterations (3). Furthermore, AfaD proteins have been shown to promote invasion of bacteria into epithelial cells (13). Under the conditions employed we did not detect any invasion of HEp-2 cells by strain 135/12, nor was it possible to assess cell damage or cytoskeletal alterations due to the presence of the Afa adhesin in an EPEC background. A/E lesion formation appeared to be the dominant phenotype, and the ultrastructural observations showed that EPEC Afa only functions as an initial adhesin but is eventually eliminated from the region of bacteria-host cell contact to allow the typical intimate intimin-Tir interaction and A/E lesion formation characteristic of EPEC.
Nucleotide sequence accession number. The DNA sequence of the EPEC Afa operon has been deposited in the GenBank database (accession no. AF325672).
This work was supported by Fundação de Amparo à Pesquisa do Estado de São Paulo 95/7033-1 and 96/4148-5 and by The Wellcome Trust (S.K.).
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»