OUP user menu

Role of type I fimbriae in the aggregative adhesion pattern of enteroaggregative Escherichia coli

Cristiano G. Moreira, Sylvia M. Carneiro, James P. Nataro, Luiz R. Trabulsi, Waldir P. Elias
DOI: http://dx.doi.org/10.1016/S0378-1097(03)00561-5 79-85 First published online: 1 September 2003


Enteroaggregative Escherichia coli (EAEC) is distinguished by its characteristic aggregative adherence (AA) pattern to cultured epithelial cells. In this study we investigated the role of type I fimbriae (TIF) in the AA pattern to HEp-2 cells and in biofilm formation. Accentuation of this pattern was observed when the adherence assay was performed in the absence of mannose. This effect was observed in the prototype EAEC strain 042 (O44:H18), O128:H35 strains and for other EAEC serotypes. Antiserum against TIF decreased AA by 70% and 90% for strains 042 and 18 (O128:H35 prototype strain), respectively. A non-polar knockout of fimD, the TIF usher, in strains 042 and 18 resulted in inhibition of the accentuated AA pattern of approximately 80% and 70% respectively, and biofilm formation diminution of 49% for 042::fimD and 76% for 18::fimD. Our data evidence a role for TIF in the AA pattern and in EAEC biofilm formation, demonstrating that these phenotypes are multifactorial.

  • Enteroaggregative Escherichia coli
  • Type I fimbria
  • Aggregative adherence
  • Biofilm

1 Introduction

Enteroaggregative Escherichia coli (EAEC) is a diarrheagenic E. coli category isolated from cases of acute diarrhea in developing and developed countries, and epidemiologically associated with persistent diarrhea in developing countries [1]. EAEC is distinguished by its characteristic aggregative adherence (AA) pattern to cultured epithelial cells [2]. The AA pattern is characterized by the stacked brick-like arrangement of adherent bacteria both to the cells and to the coverslip, which is distinct from the localized (LA) and diffuse (DA) patterns of adherence presented by enteropathogenic E. coli (EPEC) and diffusely adherent E. coli, respectively [1].

EAEC strains are distributed among several O serogroups [1], including those of the classical EPEC serogroups, such as O128. This serogroup contains some serotypes, such as O128:H35, which are composed of EAEC strains, i.e., strains showing the AA pattern and the presence of some EAEC-associated virulence markers [3].

EAEC strain 042 (O44:H18) has been broadly employed as the prototype for EAEC pathogenesis studies. This strain expresses the aggregative adherence fimbria II (AAF/II), which was reported to mediate AA and biofilm formation [4,5]. Other fimbrial and afimbrial adhesins have been implicated in the AA pattern of EAEC strains [69]. However, several EAEC strains do not express any of these described adhesins.

The type I fimbria (TIF) is the most common adhesin found in Enterobacteriaceae and is usually observed in commensal and pathogenic E. coli isolates [10]. This fimbria is encoded by the fim operon, which comprises nine genes (fimA, fimB, fimC, fimD, fimE, fimF, fimG, fimH and fimI), responsible for the biogenesis of TIF. fimA encodes the larger fimbrial subunit, fimFGI encode the smaller subunits, fimH encodes the pilin of the fimbria, fimC encodes the chaperone, fimD encodes the usher, and fimB and fimE are regulator genes [11].

TIF has been reported to be responsible for initial biofilm establishment in K12 E. coli [12] and is an important virulence factor described in uropathogenic E. coli infections [13,14]. A role of TIF in the LA pattern to HEp-2 cells presented by the prototype EPEC strain E2348/69 was excluded [15]. However, the role of this fimbria in the establishment of the other adhesion patterns of diarrheagenic E. coli (AA and DA) has never been examined. Therefore the main objective of this study was to investigate the role of TIF in the establishment of the AA pattern of EAEC strains.

2 Materials and methods

2.1 Bacterial strains, and culture conditions

In this study we employed seven strains belonging to the serotype O128:H35 isolated from cases of acute diarrhea and previously characterized as EAEC [3]. Strain 042 (serotype O44:H18), isolated from a child with acute diarrhea in Peru, was employed as the EAEC prototype [16]. Seven strains of the EAEC serotypes O86:H2, O111:H4, O111:H12, O111:H21, O125:H16, O125:H21 and O128:H12 [3], and one strain of serotype O125:H6, which is an atypical EPEC that displays the AA pattern [17], were also employed. E. coli S17-1(λpir) [18] was employed as host in the mutagenesis experiments.

The bacterial strains were grown in Luria–Bertani (LB) broth, LB agar, MacConkey agar (MC) or tryptic soy agar (TSA) at 37°C, and maintained at −70°C in LB broth containing 15% of glycerol. The following antibiotics were employed as indicated: ampicillin (100 µg ml−1), kanamycin (50 µg ml−1) and nalidixic acid (50 µg ml−1).

2.2 Hemagglutination assays

Hemagglutination ability was tested by glass slide agglutination as previously described [19]. Bacterial suspensions were prepared from growth in LB broth, LB agar and TSA at 37°C, mixed with an equal volume of 2% (v/v) guinea pig and rat erythrocyte suspensions in phosphate-buffered saline (PBS) pH 7.4 in the presence or absence of 1%d-mannose, and observed for hemagglutination during 1 min. Hemagglutination was designated mannose-resistant hemagglutination or mannose-sensitive hemagglutination (MSHA), in the presence or absence of mannose, respectively.

2.3 Adherence assays

The adherence assay was performed employing semi-confluent HEp-2 and HeLa cell monolayers, as previously described to characterize the AA pattern [2]. When the adhesion pattern was non-characteristic, the incubation period was prolonged for an additional 3 h (6-h assay). The adherence assay was also performed employing Dulbecco's modified Eagle's medium (DMEM) (Gibco-BRL, Gaithersburg, MD, USA) without d-mannose, in order to characterize the mannose-sensitive adherence.

2.4 Adherence inhibition assays

Adherence inhibition assays [20] employing antiserum against TIF (anti-TIF) were performed to investigate the role of TIF in forming the accentuated AA pattern. The polyclonal rabbit anti-TIF was kindly donated by M.M. Levine (Center for Vaccine Development, University of Maryland, Baltimore, MD, USA). Briefly, 40 µl of bacterial cultures grown overnight at 37°C in LB broth were pre-incubated with anti-TIF diluted 1:5 and 1:10 in 960 µl of DMEM without d-mannose for 1 h at 37°C. This mixture was added to each well with HEp-2 cell monolayers and the adhesion test was performed. Similarly, rabbit pre-immune serum was assayed as negative control of adherence inhibition. To quantify the number of bacteria adhering to the epithelial cells and to the coverslips, the following protocol was employed. After the bacteria–HEp-2 cell incubation period, non-adherent bacteria were removed from the monolayers by washing with PBS. The epithelial cells were lysed with 400 µl of 1% (v/v) Triton X-100 solution in each well of the tissue culture plate. After 10 min of incubation at room temperature, 1.6 ml of PBS was added to the wells and homogenized by pipetting several times. Serial dilutions (1:10) in PBS were plated on MC, incubated overnight and the colonies were counted to calculate the CFU ml−1. All assays were performed in triplicate.

2.5 Biofilm formation

The ability to form biofilm on a polystyrene surface was investigated as previously described by Sheikh et al. [5], using high glucose DMEM as culture medium.

2.6 TIF mutagenesis

TIF mutagenesis was achieved by constructing a non-polar knockout of the fimD gene, which encodes the usher of TIF, using the suicide vector pJP5603 [18]. This strategy was used to investigate the role of TIF in the accentuated AA pattern of strains 18 (O128:H35 prototype strain) and 042 (EAEC prototype strain).

An internal portion of fimD of 599 bp, corresponding to nucleotides 2020–2619 (GenBank accession number GI 454227), was amplified by polymerase chain reaction (PCR) using the primers forward (5′-ATGGTACCCAGAGTACATTACT-3′) and reverse (5′-ATGAGCTCGGAATATTGACGTTA-3′) and the following amplification cycle: 30×(1 min at 94°C, 1 min at 59°C and 1 min at 72°C) and 8 min at 72°C. These primers were designed such that KpnI (5′)-SacI (3′) restriction sites were inserted at the ends of each fragment. PCR products were digested with KpnI and SacI and were cloned into the corresponding sites of the suicide vector. After ligation and transformation into host S17-1(λpir), the transformants were selected on LB agar plates containing kanamycin. Selected transformants were conjugated with wild-type strains 18 and 042 by filter mating on cellulose nitrate membranes. Transconjugants were selected in LB agar plates containing kanamycin and ampicillin for strain 18 mutation, or kanamycin and nalidixic acid for strain 042 mutation. Their identity was confirmed by agglutination with O128 and O44 antisera, respectively, and the correct site of integration was confirmed by Southern blot analysis using the fimD amplicon as gene probe.

2.7 Transmission electron microscopy

Expression of TIF was examined by immunogold assay. The strains were grown statically in LB broth at 37°C overnight and examined by the immunogold technique [4], employing the anti-TIF and IgG protein-gold of 10 nm (Amersham Pharmacia Biotech, UK). Grids were examined with a LEO906E transmission electron microscope (LEO, Germany).

3 Results

All O128:H35 strains of this study were capable of agglutinating guinea pig and rat erythrocytes only in the absence of mannose, constituting the MSHA profile, when cultured in the three different culture media.

As described in Table 1, the O128:H35 strains exhibited a non-defined adhesion pattern to HEp-2 cells in the 3-h assay while the EAEC prototype strain 042 displayed the characteristic AA pattern. In the 6-h assay the AA pattern was observed for all O128:H35 and 042 strains. Because of the fact that all O128:H35 strains displayed the MSHA profile, HEp-2 adherence tests were performed without mannose, in order to investigate the possible role of a mannose-sensitive adhesin in the AA pattern of these strains. Employing such a variation in the adherence assay, an accentuated AA pattern was observed for all tested strains, including the prototype 042 (Table 1). These results were reproducible in HeLa cells (data not shown).

View this table:
Table 1

Adhesion patterns to HEp-2 cells presented by the strains of this study

StrainSerotypeAdhesion pattern to HEp-2 cells
3-h assay6-h assay3-h assay (without mannose)
14O128:H35NDAAaccentuated AA
15O128:H35NDAAaccentuated AA
16O128:H35NDAAaccentuated AA
18O128:H35NDAAaccentuated AA
19O128:H35NDAAaccentuated AA
21O128:H35NDAAaccentuated AA
22O128:H35NDAAaccentuated AA
10O86:H2AAaccentuated AA
38O111:H4AAaccentuated AA
30O111:H21AAaccentuated AA
01O125:H6NDAAaccentuated AA
09O125:H16AAaccentuated AA
06O125:H21AAaccentuated AA
20O128:H12AAaccentuated AA
042O44:H18AAAAaccentuated AA
  • ND, non-defined pattern of adhesion.

The adherence patterns exhibited by strain 18 (O128:H35) in the 3- and 6-h assays, and by strain 042 in the 3-h assay are illustrated in Fig. 1A,B,F. As observed, the accentuated AA patterns of strains 18 and 042 were similar (Fig. 1C,G). Interestingly, although the number of adherent bacteria was increased, the characteristic aggregative pattern was maintained on the coverslip and the number of adherent bacteria covering the cell surfaces was increased.

Figure 1

HEp-2 cell adherence assays with (A) strain 18 (O128:H35), 3-h assay with mannose, (B) strain 18, 6-h assay with mannose, (C) strain 18, 3-h assay without mannose, (D) strain 18 pre-incubated with anti-TIF (1:10), 3-h assay without mannose, (E) strain 18::fimD, 3-h assay without mannose, (F) strain 042 (O44:H18), 3-h assay with mannose, (G) strain 042, 3-h assay without mannose, (H) strain 042 pre-incubated with anti-TIF (1:10), 3-h assay without mannose, (I) strain 042::fimD, 3-h assay without mannose. Magnification: 1000×.

The accentuated AA of strains 18 and 042 was dramatically inhibited by anti-TIF (Fig. 1D,H), as shown by CFU counts (Fig. 2). The antiserum inhibited approximately 70% of adhesion in strain 042, and in strain 18 the inhibition was nearly 90%. The rabbit pre-immune serum, employed as negative control, did not affect the amount of adherent bacteria displayed by the wild-type strains (data not shown).

Figure 2

Inhibition of HEp-2 cell adherence by anti-TIF. A: Strain 042 without pre-incubation with anti-TIF. B: Strain 042 pre-incubated with anti-TIF. C: Strain 18 without pre-incubation with anti-TIF. D: Strain 18 pre-incubated with anti-TIF.

The cloning of an internal portion of fimD into the suicide vector pJP5603 generated the recombinants pCGM-1 and pCGM-2, harboring the fimD amplicon of strains 18 and 042, respectively. Using these recombinants, fimD non-polar mutants were obtained and designated 18::fimD and 042::fimD. The mutants had their O serogroups and the lactose-fermenting capacity on MC agar confirmed. The insertion in fimD was verified by acquired kanamycin resistance and the correct localization of insertion was checked by Southern blot analysis (data not shown). The lack of expression of TIF in these two mutants was examined by immunogold assay. As presented in Fig. 3, anti-TIF decorated type I fimbriae in strains 18 and 042 (A and C) in contrast to the absence of anti-TIF labelling in the respective mutants (B and D).

Figure 3

Transmission electron microscopy of immunogold staining for type I fimbriae. A: Strain 18 (magnification: 60 000×, scale bar: 201 nm). B: Strain 18::fimD (magnification: 77 500×, scale bar: 156 nm). C: Strain 042 (magnification: 60 000×, scale bar: 201 nm). D: Strain 042::fimD (magnification: 46 460×, scale bar: 260 nm).

The accentuated AA pattern to HEp-2 cells exhibited by both prototype wild-type strains in the absence of mannose was notably reduced, but not completely abolished, in their respective mutants (Fig. 1E,I). This decrease of HEp-2 cell-adherent bacteria was quantified by CFU counts as presented in Fig. 4, demonstrating a decrease of approximately 70% of inhibition for 18::fimD and 80% for 042::fimD, compared to the respective wild-type strains. Biofilm formation was also reduced in the TIF mutants (Fig. 5), by approximately 76% for 18::fimD and 49% for 042::fimD, in comparison to the 18 and 042 wild-type strains.

Figure 4

HEp-2 cell-adherent bacterial counts. A: Strains 18 and 18::fimD. B: Strains 042 and 042::fimD.

Figure 5

Biofilm formation in high glucose DMEM medium. A: Strains 18 and 18::fimD. B: Strains 042 and 042::fimD.

The adherence assay in the absence of mannose was also performed with seven other strains belonging to EAEC serotypes and one atypical EPEC that presents the AA pattern. The results obtained confirmed the possible role of TIF in other serotypes, since all of these strains presented an increased number of adherent bacteria, characterizing an accentuation in the AA pattern, similar to that observed for strains 18 and 042 (Table 1).

4 Discussion

In this study we investigated the involvement of TIF in the characteristic adherence pattern of EAEC. Type I fimbriae are characterized as rigid rod-shaped bacterial surface structures, sensitive to mannose and analogous oligosaccharides, and are found in approximately 70% of E. coli isolates [21]. The role of TIF has been demonstrated in K12 E. coli adherence to abiotic surfaces [12]. However, a role of TIF in adherence of diarrheagenic E. coli to epithelial cells in culture has been excluded for EPEC [15].

Initially, we investigated the presence of adhesins in the O128:H35 serotype, since these strains do not express AAF/I and AAF/II adhesins of EAEC and exhibit the AA pattern [3]. The O128:H35 strains presented a MSHA hemagglutination profile, indicating the presence of mannose-sensitive adhesins. Taking these results into consideration, a modification in the adherence assay described by Cravioto et al. [22] was performed, which included removal of mannose during the bacteria–HEp-2 cell incubation period, to investigate the possible role of mannose-sensitive adhesins in the AA pattern of these strains. All O128:H35 strains showed an accentuated AA pattern in the absence of mannose, as well as the 042 strain. Interestingly, instead of a non-specific adhesion pattern, in the absence of mannose bacteria were found completely covering the epithelial cells and in a honeycomb arrangement on the coverslip, demonstrating that the AA pattern was maintained, although exacerbated. The anti-TIF serum was capable of inhibiting the accentuated AA pattern supporting the role of TIF in this adherence pattern.

A knockout of TIF confirmed its involvement in the accentuated AA pattern of the prototype strains 042 and 18, since it was notably reduced in comparison to their respective wild-type strains, but not abolished. That was thought to be due to the presence of AAF/II in strain 042 [4] and probably to an additional adhesin in strain 18, as yet not characterized. This was an important finding, since TIF has never been associated with the adhesion to epithelial cells displayed by EAEC or even other enteroadherent E. coli. A role of TIF in the establishment of the AA pattern was excluded by Qadri et al. [19], simply because of the observation that the adherence assay was performed in the presence of mannose, which inhibits TIF.

Biofilm formation capacity has been described for EAEC strains grown in high glucose cell culture media [5]. The role of TIF in biofilm formation of EAEC was also evaluated in this study. Biofilm formation was highly reduced, but not abolished in the TIF mutants, similarly to the observation in the adherence assay to epithelial cells. Lack of involvement of TIF in biofilm formation was reported by Sheikh et al. [5], pre-incubating strain 042 with antiserum against TIF. However, such pre-incubation might not be sufficient to inhibit the whole bacterial population during the 24-h incubation of the biofilm formation test. In contrast, in our study the inhibition of TIF was achieved by obtaining isogenic mutants, unable to express the fimbriae during the whole incubation period of adhesion and biofilm formation tests.

Eight other strains belonging to different serotypes were also tested in the absence of mannose, to examine the significance of our findings to other strains that display the AA pattern. The results obtained confirm the importance of TIF in other serotypes, since all of these strains presented an accentuated AA pattern in the absence of mannose, showing the amplitude and importance of our findings to the EAEC category. Considering the data obtained in our study, we propose that the inclusion of a complementary adherence assay performed in the absence of mannose should be performed in the detection of diarrheagenic E. coli strains displaying non-defined adherence patterns to HEp-2 cells.

Other studies are necessary to explain the real role of TIF-mediated adherence to the intestinal mucosa, because mucosal adherence is complex and probably multifactorial in many enteric pathogens. Evaluation of the adherence capacity by the prototype strains and the isogenic TIF mutants in an in vitro organ culture model [23] could demonstrate such involvement of TIF. On the other hand, it is difficult to establish the real role of TIF in EAEC pathogenesis, since there is no animal model for EAEC diarrhea.

The intercellular adherence among the bacteria shown in the AA pattern of EAEC strains could be explained by mannose residues present in the O antigen of bacterial lipopolysaccharide [24], which could act as a receptor to TIF. We hypothesize that TIF is responsible because for the first attachment to the intestinal mucosa followed by the expression of other accessory fimbrial or afimbrial adhesins, responsible for the establishment of the definitive AA pattern. This characterizes the multifactorial conditions for the AA phenotype and biofilm formation, and consequently for the development of EAEC pathogenesis.


We wish to thank Dr. Tânia A.T. Gomes for performing the hybridization assays and for useful discussions, and Dr. Myron M. Levine for providing the anti-TIF serum. This work was supported by the Fundação de Amparo à Pesquisa do Estado de São Paulo (Grant 01/08570-3 to L.R.T. and fellowships to C.G.M. and W.P.E.) and the Conselho Nacional de Desenvolvimento Cientı́fico e Tecnológico (Grant 521160/98-7 to L.R.T.).


  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].
View Abstract