OUP user menu

Virulence-related DNA sequences and adherence patterns in strains of enteropathogenic Escherichia coli

Saeid Bouzari , Mohammad N. Jafari , Forough Shokouhi , Mostafa Parsi , Anis Jafari
DOI: http://dx.doi.org/10.1111/j.1574-6968.2000.tb09045.x 89-93 First published online: 1 April 2000


The presence of the genes for Escherichia coli adherence factor (EAF), attaching and effacing lesion (eae) and bundle-forming pili (bfp) in 72 strains identified as enteropathogenic E. coli (EPEC) by slide agglutination was evaluated using hybridization and PCR. The adherence property of these strains was assayed using 3h HeLa cells adherence assay. The results obtained indicated that virulence-associated genes were present in 65% of the strains but only ten (13.9%) isolates were positive for all the three markers (typical EPEC), 37 (51.4%) isolates carried either one or two of these determinants (atypical EPEC) and the remaining 25 (34.7%) were negative for all these genes. In vitro adherence assay showed that 44 (61.1%) strains adhered to HeLa cells with a defined pattern, 13 (18.1%) isolates adhered loosely with no definite pattern and the remaining 15 (20.8%) were non-adherent. Analysis of the results showed a statistically significant association between the presence of the virulence-related genes with adherence of the strains with a defined pattern (P≤0.0001). These results indicated that since over 60% of the strains identified by serogrouping carried at least one of the putative virulence markers, it therefore seems that this simple test is still of value in our setting although the need for a confirmatory test is also indicated.

  • Escherichia coli
  • Enteropathogenic Escherichia coli
  • Virulence marker
  • Adhesion

1 Introduction

Enteropathogenic Escherichia coli (EPEC) are an important cause of acute diarrhea especially among young children in developing countries [1]. These strains have historically been recognized not by the possession of a specific virulence factor, but by the presence of certain O:H antigens [2]. Although strains with these antigens are still routinely identified as EPEC by agglutination tests, recently it has been shown that these O serogroups may include not only true EPEC, but also strains belonging to other E. coli pathotypes [2].

The exact mechanism by which EPECs cause diarrhea is unknown, but adherence to mucosal surfaces and the formation of attaching and effacing (A/E) lesions in intestinal epithelial cells are considered important characteristics of infection by these microorganisms [3]. Initially Cravioto et al. [4] showed that about 80% of classical EPEC strains adhere to HEp-2 cells in the presence of D-mannose and based on these studies three distinct patterns of adherence have so far been identified [5]. Localized adherence (LA) in which microcolonies are formed at one or more places on the epithelial cells, diffuse adherence (DA) where bacteria cover the cell uniformly and enteroaggregative (AA) in which bacteria have a characteristic stacked-brick appearance on the surface of the cells. Localized adherence which has been associated with EPEC strains is mediated by a fimbrial structure called bundle-forming pili (BFP) which is encoded by a large plasmid known as the EPEC adherence factor (EAF) [6]. A 1-kb probe has been developed from gene sequences carried on a plasmid in strain E2348/69 (serotype 0127:H6) which has been used extensively in epidemiological studies in various parts of the world [7]. In addition a chromosomal gene cluster of 35 kb designated the locus of enterocyte effacement (LEE) necessary for the production of the A/E lesion was subsequently identified and sequenced and a 1-kb fragment from within the eae open reading frame has been used as a DNA probe [8]. It has been proposed that the presence of these genes could be used as a possible marker for EPEC identification (typical EPEC) in addition to or as a replacement for the O serogrouping [9,10]. However, recent studies have shown that not all serotypes harbor EAF plasmid and that the EAF-negative EPEC strains are also adherent and cause A/E lesions [1113]. Furthermore it has also been suggested that not all the virulence-associated factors are evenly distributed in the wild-type EPEC strains circulating in different geographical areas [1417]. Therefore, this study was undertaken in order to examine a sample of Iranian isolates belonging to classical serogroups for the presence of EAF plasmid, eae and bfp genes and their relation to the adherence ability of these strains.

2 Materials and methods

2.1 Strains and serogrouping

We studied 72 E. coli strains belonging to ten different serogroups that were isolated from children with diarrhea. These strains were cultured on nutrient agar (Difco, USA) and smooth colonies were used for serogroup identification by slide agglutination using commercially available antisera (Pasteur-Merieux, France).

2.2 Adherence to HeLa cells

Adherence was tested by the method described by Cravioto et al. [4] using 3-h bacterial cell contact. Adherence patterns were identified according to Scaletsky et al. [18]. Standard strains E2348/69, serotype O127:H6 isolated from an infant diarrhea outbreak in England showing LA, E. coli strain O42 isolated from a child with acute diarrhea and the hemolytic strain 17-2 serotype O3:H2 isolated from an infant with diarrhea in Chile both exhibiting AA and strain C1845 serotype O75:NM isolated from a child with chronic diarrhea showing DA obtained from the Center for Vaccine Development, University of Maryland, School of Medicine (kindly provided by Dr. J.B. Kaper) were used as controls.

2.3 DNA extraction and hybridization

DNA was extracted using the alkaline lysis method [19], which was then used for both PCR and dot-blot. Dot-blot hybridization was performed under stringent conditions with DIG-labeled probes for EAF (EAF plasmid) and eae (E. coli attaching and effacing) virulence-associated factors.

Plasmid pJPN16 containing a 1-kb fragment derived from pMAR2 of strain E2348/69 [7] was digested with BamHI–SalI and pCVD434 (EPEC adherence factor plasmid) was digested with SalI–KpnI for the 1-kb fragment used as EAF and eae probes, respectively. These plasmids were a gift from Dr. C. LeBouguenec, Pasteur Institute, Paris. These fragments were eluted, DIG-labeled and then used for hybridization according to manufacturer's instruction (Boehringer Mannheim, Germany).

2.3 PCR

Primers capable of amplifying an internal 326-bp region of the sequence encoding BFP protein were synthesized (GeneSet, France) according to Gunzberg et al. [20] with the following sequences: 5′-AAT GGT GCT TGC GCT TGC TGC-3′ and 5′-GCC GCT TTA TCC AAC CTG GTA-3′.

DNA was amplified for 30 cycles consisting of denaturation at 94°C for 30 s, annealing at 56°C for 1 min and polymerization at 72°C for 2 min. The amplified product was resolved by agarose gel electrophoresis and visualized by UV after staining with ethidium bromide. Reference strains E2348/69 and HB101 were used as positive and negative controls, respectively. All experiments were performed in duplicate.

2.4 Statistical analysis

A X2 test was used for analysis of data.

3 Results

The E. coli strains included in this study belonged to ten different serogroups and only ten (13.9%) of the total 72 isolates tested with probes and PCR carried EAF, eae and bfp virulence-associated markers simultaneously (typical EPEC), 37 (51.4%) strains had either of these determinants alone or in combination (atypical EPEC) and the remaining 25 (34.7%) were non-EPEC in the sense that they were negative for all three factors (Table 1).

View this table:

Distribution of virulence-related DNA sequences among different serogroups

In this study irrespective of the adherence of the EAF probe hybridized with 32 (44.4%) strains, the eae gene was detected in 36 (50%) isolates and only 14 (19.4%) strains gave positive results for the bfp gene using PCR (Table 1).

The results of the adherence assay showed that 57 (79.2%) strains adhered to HeLa cells and 15 (20.8%) isolates were non-adherent. Of the adherent strains 17 (29.8%) exhibited LA, 13 (22.8%) AA, 9 (15.8%) DA, 5 (8.8%) LA/DA and 13 (22.8%) adhered to the cells in udefined pattern (Table 2). Among the ten typical EPEC strains 9 (90%) exhibited LA and one strain manifested LA/DA type of adherence, whereas of the atypical strains, LA was manifested by 4 (10.8%) isolates, AA and DA by 8 (21.6%) strains each, LA/DA by 4 (10.8%), an undefined pattern was observed in 6 (16.2%) and the remaining 7 (18.9%) were non-adherent. Of the non-EPEC 4 (16.%) strains exhibited LA, 1 (4%) DA, 5 (20%) AA, 7 (28%) UDP and 8 (32%) were non-adherent (Table 2).

View this table:

Distribution of various adherence patterns among different classes of E. coli

It has to be mentioned that overall 47 strains reacted with all or some of the DNA-related sequences, of which 34 (72.3%) strains adhered to HeLa cells with a defined pattern (LA, DA, AA, LA/DA), 6 (12.8%) showed an undefined pattern and the remaining 7 (14.9%) were non-adherent (P≤0.0001).

Of the 15 (20.8%) non-adherent strains, 7 (46.7%) belonged to atypical EPEC, in which the eae gene was detected in five and the remaining two reacted with the probe for EAF (Table 2).

Typical EPECs were mostly encountered among the strains belonging to serogroup 0119 (4/10) and the least among strains of serogroup 055 (1/17), respectively, and non-EPEC mostly among serogroups 0126 (6/12), 0111 (4/12) and 055 (5/17) (Table 1).

4 Discussion

The results of this study showed that 65% of the E. coli strains identified as EPEC by slide agglutination with commercially available antisera carried at least one of the EPEC virulence-associated DNA sequences. However, only 12.5% (9/72) of these isolates could be classified as typical EPEC on the basis of simultaneous carriage of EAF, eae and bfp sequences and localized adherence to HeLa cells (Table 1).

Comparison of the circulation rate of DNA sequences associated with A/E lesions and EAF plasmid in Italy [15] and those previously reported from UK [21] has led to the conclusion that eae might be a more suitable marker for epidemiological studies in Europe, while the EAF-related probe has been proposed as more appropriate for developing countries. However, in the present study in which isolates were classified, as EPEC on the basis of serogrouping and serotyping was not performed, no significant difference between the circulation rate of EAF and eae sequences was observed (EAF=44.4%, eae=50%). Therefore, both might be used, although eae had a slightly higher occurrence.

EAF plasmids have been implicated in the initial adhesion of EPEC strains to intestinal mucosal surfaces [22], but data concerning the association between LA and EAF plasmids are varied. Results reported from Italy and UK [15,21] have shown low circulation of EAF plasmids among strains showing LA, whereas in a study conducted in Sao Paulo 100% association between LA and EAF had been observed [17]. In the present study this association was observed in 76% of the isolates, but similar to the data reported by Rosa et al. [23] some LA strains were negative for all three genetic markers described for EPECs (Table 2). The mechanism by which these strains adhere to HeLa cells was not elucidated, however, since strains containing a large plasmid other than EAF mediating similar functions had previously been reported [24], therefore a more detailed study of the plasmid content and also adhesin(s) produced by these strains is required.

It is interesting to note that EAF plasmid was also detected in isolates showing DA and AA patterns. Codetection of this plasmid and AA sequences among ETEC strains was recently reported [25], but further studies are necessary to clarify the significance of these novel combinations.

A small number of strains in this study exhibited both LA and DA simultaneously which contained either all or some of the genetic markers described for EPECs, but genetical determinants of DA in this group were not investigated. However in a study carried out by Scaletsky et al. [26] it has been shown that DA was not encoded by genes homologous to DNA probes used for identification of this group. Furthermore it was postulated that this pattern may be linked to a new E. coli virulence category, therefore further studies are being undertaken to elucidate the genetic determinants of DA in this group.

The occurrence of the eae gene alone among non-adherent atypical EPEC strains was not unexpected, since in the two-stage model of EPEC pathogenesis proposed by Knutton et al. [27] adherence and A/E are considered independent as far as their expression is concerned. Furthermore, lack of adherence in the strains possessing EAF sequences could be due to dysfunction, deletion or rearrangement in various elements responsible for each stage of adhesion located on this large plasmid.

Heterogeneity with respect to virulence markers among EPEC serogroups has previously been reported [2,1214,24], in this study typical EPEC were more frequently encountered among the strains belonging to serogroup 0119 similar to the findings of Goncalves et al. [28] and atypical EPEC among strains of serogroup 055.

Although identification of EPEC strains has remained controversial with different results reported from various geographical areas, the results obtained in this study indicated that 65% of the isolates identified by commercial antisera are probably a putative EPEC and although O serogrouping is not a perfect tool, in our setting it is still a useful method for presumptive identification of these strains.


We gratefully acknowledge the help of Mr. A.H. Salmanian and N. Shahrokhi in the preparation of the manuscript and we are also indebted to Mr. R. Jirsaraei and A. Soleimani for their technical help.


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
View Abstract