• criterion;
  • diarrheagenicity;
  • diffusely adherent Escherichia coli (DAEC);
  • flagella;
  • interleukin 8 (IL-8)


  1. Top of page

DAEC is considered potentially diarrheagenic. For diffuse adhesion, the role of the Afa, which was originally identified as a uropathogenic factor, is now understood. However, the role of DAEC in diarrheal disease remains controversial because DAEC is often isolated not only from patients but also from healthy individuals. Previously, we suggested that Afa/Dr DAEC, which can induce high levels of IL-8 secretion in cultures of human carcinoma epithelial cells (HEp-2, Caco-2), is enterovirulent. In the present study, we examined whether IL-8 secretion induced by certain Afa/Dr DAEC strains was primarily due to flagella via TLR5. All IL-8 high-inducing strains were highly motile in swarming tests. Partially purified flagella induced IL-8 in a dose-dependent manner. However, IL-8 induction was inhibited by small-interfering RNA against TLR5 or by treating flagella with disialoganglioside-GD1a, a TLR5 blocker. TLR5 is reportedly located on the basolateral side of intestinal epithelia; flagella should not have reached TLR5 from the apical side beyond tight junctions. Reduction in the number of intracellular organisms by wortmannin, a PI3K inhibitor, did not reduce IL-8 secretion. Afa/Dr DAEC seemed to loosen the tight junctions because it quickly reduced transepithelial electrical resistance after infection. Decreased resistance led to increased IL-8 production. In conclusion, diffuse adhesion itself is insufficient to induce high levels of IL-8, and simultaneous stimulation by flagella via TLR5 is likely required for additional induction. Clinically, high motility may be a candidate criterion for predicting the ability of Afa/Dr DAEC strains to induce higher levels of IL-8 secretion.

List of Abbreviations: 

afimbrial adhesive sheath


colony-forming units


diffusely adherent Escherichia coli


diarrheagenic Escherichia coli


enteroaggregative Escherichia coli


Eagle's minimum essential medium


fetal bovine serum




Luria broth




phosphatidyl-inositol-3 kinase


small-interfering RNA


transepithelial electrical resistance


tight junction



Infectious diarrheal disease remains an important public health problem and results in 1.5–2.5 million deaths/year worldwide, mostly in children under 5 years of age in developing countries (1, 2). Escherichia coli is the predominant facultative anaerobe of the normal colon flora. However, particular clones can cause sepsis, urinary tract infections, and diarrheal disease. DEC is a major pathogen associated with enteric disease (3). DEC is classified according to pathogenicity as enteropathogenic E. coli, enterotoxigenic E. coli, enteroinvasive E. coli, enterohemorrhagic E. coli, and EAggEC (3).

Colonization of the human intestine is an essential part of the infection process. DAEC shows diffuse adhesion to HEp-2 cells (4) and represents a putative sixth DEC class. Its role in diarrheal disease, however, has been controversial. DAEC was similarly isolated from children with and without diarrhea and showed no diarrheagenicity in adult volunteers (5–7). Several groups have suggested that age-dependent susceptibility may explain epidemiological discrepancies because a positive association with diarrhea was found when study populations were age-stratified (4, 5, 8–11). Clinical microbiologists have hesitated to classify DAEC isolates as causative agents. DAEC probably comprises a heterogeneous group of organisms with variable enteropathogenicity (12). Measuring diffuse adhesion alone is insufficient to evaluate the diarrheagenicity of strains and, therefore, other distinguishing characteristics are needed.

DAEC possesses no other typical virulence markers associated with diarrhea. The only known virulence factors of DAEC strains are adhesins. Approximately 90% of uropathogenic E. coli colonize on urinary tract epithelial cells by recognizing the antigenic determinants of the human P blood-group system with P adhesin (13). However, 10% of mannose-resistant hemagglutination-positive pyelonephritic strains recognize receptors other than the P antigen. Afa are the colonization antigens in these cases (14). The Dr adhesins recognize the Dr antigen of the Cromer blood-group system. This Dr antigen consists of the exposed domain of the decay-accelerating factor (15). Although F1845 was initially found to be a fimbria that was unique to DAEC (16), Dr adhesins (Dr, DrII, Afa-I, and Afa-III), including F1845, are encoded in the same gene clusters and are considered to be subtypes of Afa (17). In a recent review, DAEC was classified into two groups: Afa/Dr and non-Afa/Dr DAEC (18). In fact, we found that nearly half of our DAEC strains express the Afa/Dr gene (19, 20).

Germani et al. reported that DAEC isolation rates differ significantly between patients and controls only when the presence of Afa sequences in the strains is considered (21). Furthermore, Bétis et al. (22, 23) found that two DAEC strains induce IL-8 in the human colonic epithelial cell line T84. Diarrheagenic EAggEC causes a large amount of the proinflammatory chemokine IL-8 to be released from human intestinal epithelial cells. Fecal IL-8 elevation may correlate with the severity of clinical symptoms (24–26). Subsequently, we found that some DAEC strains that cause epithelial cells to secrete high levels of IL-8 are likely to be diarrheagenic, particularly in pediatric patients (19, 20, 27).

In the present study, we investigated how some Afa/Dr DAEC strains can induce high amounts of IL-8 secretion from cultured human epithelial cells to the same extent as EAggEC. The mechanisms suggest characteristics of Afa/Dr DAEC that may be useful in helping clinical microbiologists predict which strains of Afa/Dr DAEC are diarrheagenic. We studied the role of flagella and TLR5 because motile Afa/Dr DAEC strains cause prominent IL-8 induction but non-motile strains show weaker induction of IL-8 (19). Furthermore, we evaluated the effects of Afa/Dr DAEC infection on TJ in tissue culture epithelial cells. The flagella must interact with their innate receptors (TLR5), which are located on the basolateral side of the epithelia, so that Afa/Dr DAEC can cause inflammation. We suggest that Afa/Dr-possessing motile DAEC organisms swarm and adhere to epithelial cells, loosen the TJ, and then flagella reach TLR5 and trigger signaling that results in the induction of IL-8 at high levels.


  1. Top of page

Bacterial strains

A total of 19 DAEC strains expressing the Afa gene as reported previously were used in this study (Table 1) (19, 20). The strains were isolated as causal agents from the stool specimens of 924 patients with diarrhea throughout a surveillance study (28), although they were not included as etiological agents in the report because the enteropathogenicity of DAEC was still controversial. EAggEC strain V546 and E. coli laboratory strain DH5α were used as positive and negative controls in the IL-8 induction assays, respectively. Salmonella Enteritidis (V215) and Listeria monocytogenes were used as representative invasive pathogens. Bacterial inocula were prepared by culturing the strains in 1% Bacto-peptone (Becton Dickinson and Co., Sparks, MD) in water containing 0.5% NaCl at 37°C for 20 hr. LIM was used to examine the motility of each strain.

Table 1.  Bacterial isolates used in the present study
StrainSerotypeRelevant propertiesMotilityaSwarming diameter (mm)IL-8 secretionb (pg/ml)satpetaatETT2 gene regionc
  1. aMotility was determined in lysine-indole-motility medium at 37°C.

  2. bSecretion of IL-8 by Caco-2 cells after exposure to strains for 22 hr.

  3. cETT2 gene region was regarded as positive when three genes, prgEC, InvH, spaS, were expressed.

  4. NE, not examined.

V36OUT:H18AfaE1+69.7 ± 9.3940 ± 311+
V289UTAfaE14.7 ± 0.389 ± 30++
V561OUT:H5AfaE1+851070 ± 42
V582O1AfaE155 ± 15NE
V599O1AfaE118.7 ± 2.3112 ± 19+
V720O1AfaE13.3 ± 0.3106 ± 55
V242UTAfaE34.0 ± 0.698 ± 4+
V547UTAfaE35.3 ± 0.3109 ± 23
V550UTAfaE33.7 ± 0.389 ± 23
V679UTAfaE33.7 ± 0.7160 ± 141
V725UTAfaE33.7 ± 0.777 ± 10+
V880UTAfaE33134 ± 65
V922-1UTAfaE34.3 ± 0.396 ± 6
V1OUT:H4AfaEX+85930 ± 495++
V19OUT:H4AfaEX+80 ± 51220 ± 28+
V64O1:H4AfaEX+85630 ± 71+
V554O1:H28AfaEX+NE610 ± 71+
V827O86a:H18AfaEX+76.7 ± 8.3680 ± 255+
V205O143:H4AfaEX31 ± 20.2305 ± 219
V546OUTEAggEC+852920 ± 962++
DH5α  48.7 ± 9.8NENENENENE

Swarming and motility

Bacterial strains were precultured on trypticase soy agar at 37°C for 20 hr. The fresh colonies were picked, inoculated into the center of EMEM plates containing 0.3% agar, and grown at 37°C. Swarming diameters were measured at 6 hr and then every 24 hr for 120 hr.


PCR was used to examine the expression of virulence genes (29–32), which had not been examined in our previous study (19). The primer sets and amplification conditions used are shown in Table 2.

Table 2.  PCR parameters used in the present study
Target genePrimer nameSequence (5' to 3')Denaturing/Annealing/Extension/CyclesProduct size (bp)Reference
satSat1GCAGCAAATATTGATATATCA60 s, 94°C/60 s, 57°C/120 s, 72°C/30306131
aatAaatA-FATGTTACCAGATATAAATATAG30 s, 94°C/40 s, 50°C/60 s, 72°C/30106430
petpet1TTTCCAGCACTTCCTGTTCC60 s, 94°C/45 s, 55°C/60 s, 72°C/3029632
prgECprgEC C1TACAAAACTTCCGCTAAACTGA30 s, 94°C/60 s, 52°C/60 s, 72°C/30112629
InvHInvH H1CTTCTTCCTAACGAAACTATCATTA30 s, 94°C/60 s, 57°C/60 s, 72°C/3091329
spaSspaS S1GTTGGACAATGTTATCAAA30 s, 94°C/60 s, 52°C/60 s, 72°C/3085229

Eukaryotic cell lines

Caco-2 and HEp-2 cells were grown in EMEM containing 2 mM l-glutamine, 0.15% NaHCO3, and 10% FBS at 37°C in a 5% CO2 incubator. Cells were seeded at high density in 25-cm2 polystyrene tissue culture flasks. For each experiment, HEp-2 cells were cultured until they became semi-confluent, and Caco-2 cells were cultured for at least 2 weeks until they reached the polarized mature state.

Preparation of flagella

Each strain of Afa/Dr DAEC was grown on LB agar at 37°C for 24 hr. The bacteria were harvested from each plate with 2 ml PBS. Organisms recovered from 10 plates were pooled and vortexed for 1 min. Bacterial cells were removed by centrifugation at 14 000 ×g at 4°C for 30 min. The supernatant containing the sheared flagella was ultracentrifuged at 100 000 ×g at 4°C for 90 min. The supernatant was discarded, and the flagellar pellet was resuspended in 1 ml PBS. Total protein concentration of the flagella suspension was determined using a protein assay (Bio-Rad, Hercules, CA) according to the manufacturer's instructions. Purified flagella were identified by western blotting using polyclonal E. coli H antibody (Escherichia coli Antisera set 2, 1:800; Denka Seiken, Tokyo, Japan). In some experiments, the flagella preparation was incubated in EMEM with or without 100 μg/ml disialoganglioside-GD1a (Sigma, St Louis, MO), which is a competitor of endogenous gangliosides at the TLR5 coreceptor, for 1 hr at 37°C before the inoculation onto epithelial cells (33, 34).

Establishment of Caco-2 cells that were stably transfected with TLR5 siRNA

siRNA targeting human TLR5 in a mammalian cell expression plasmid (psiRNA-hTLR5) was purchased from Invivogen (San Diego, CA). A plasmid containing scrambled nucleotides (psiRNA-h7SKgzScr) was used as a control. Both Caco-2 and HEp-2 cells were transfected with either psiRNA-hTLR5 or psiRNA-h7SKgzScr using the Polyfect Transfection Reagent (Qiagen, Valencia, CA) according to the manufacturer's instructions. Transfected cells were selected and maintained with zeocin (200 μg/ml; Invitrogen, Carlsbad, CA). Transfection efficiency was monitored by the expression of jellyfish GFP fusion protein.

Adhesion and invasion assays

The adhesion and invasion assays were carried out as described (20). Briefly, confluent monolayers of cells grown in 24-well plates or on 24-well Transwell filters (Corning, NY, NY) (Coaster, 3.0-μm porosity, 0.33 cm2) were washed three times with PBS. Then, 1 ml EMEM containing d-mannose (1%, w/v) and FBS (0.5%, v/v) without antibiotics was added to each well. The cells were infected at a bacteria-to-cell ratio of 100:1 for 3 hr at 37°C. After removing medium, the epithelial cells were washed three times with PBS to remove extracellular bacteria. A 100-μL aliquot of culture medium was used for an ELISA for IL-8 using the ELISA kit Human interleukin-8 [IL-8] (ELISA Biotrak™ System; Amersham Biosciences, Buckinghamshire, UK). Then, cells were lysed with 1% (v/v) Triton X-100, and the CFU of the adhering bacteria was determined. The number of intracellular bacteria was determined in the same manner as for the adhesion test, but the epithelial cells were incubated in EMEM containing 100 μg/ml gentamycin (Sigma) for 1 hr before lysis to kill extracellular bacteria. The medium containing gentamycin was removed, and internalized bacteria were counted. The level of IL-8 in the supernatants was determined.

Transepithelial electrical resistance measurements

Transepithelial electrical resistance (TEER) of Caco-2 monolayers grown on 24-well Transwell filters was measured using a Millicell Electrical Resistance System (Millipore Corporation, Bedford, MA). TEER was calculated as Ω cm2 using the measured electrical resistance of the surface area of the filter. The background reading of a cell-free control filter was subtracted.

Permeability tests

The permeability of Caco-2 cell monolayers was determined by measuring the paracellular passage of uranin (Wako, Osaka, Japan) from apical to basolateral compartments of the Transwell filter culture. Uranin (0.2 mg/ml) was dissolved in phenol red-free EMEM. Bacterial inoculum was prepared in this solution, and the 150-μl mixture was loaded onto the apical side of the monolayer. The uranin concentration in the basolateral compartment was determined longitudinally on the basis of fluorescence intensity analyzed using a spectrofluorophotometer (Wallac 1420 ARVOSX; Perkin Elmer Life Sciences, Boston, MA) at an excitation wavelength of 485 nm and an emission wavelength of 535 nm.


  1. Top of page

Swarming and IL-8 induction by the DAEC strains

A swarming test was carried out to investigate whether the motility of Afa/Dr DAEC was implicated in the high levels of IL-8 induction. The DAEC strains were assigned to three groups on the basis of swarming: (i) high motility; (ii) low motility; and (iii) no motility (Fig. 1). Low-motility strains were defined as organisms having a weak ability to swarm in the EMEM soft agar plates; their largest diameter of swarming area on the EMEM soft agar was approximately 34.9 ± 10.7 mm in contrast to the 80.2 ± 2.3-mm diameter seen with highly motile strains after 5 days. The maximum diameter of the swarming area of non-motile strains was 4.0 ± 0.2 mm. Low-motility strains did not show any motility in LIM medium after 24 hr. Neither non-motile nor low-motility strains induced strong IL-8 production in epithelial cells, as we reported previously (19): the strains induced IL-8 production of approximately 77–160 pg/ml on Caco-2 cells, which was as low as the IL-8 induced by the negative control E. coli laboratory strain HB101. The high IL-8-inducing strains were all highly motile in the swarming assay. The amount of flagella protein was compared between nearly the same number (8.7–9.8 × 1011 CFU) of highly motile and low-motility organisms recovered from 10 culture plates using western blotting. The amounts of flagellin from flagella partially purified from organisms of the highly motile strain V19 and EAggEC V546 were 88.7 and 82.7 μg, respectively. In contrast, the low-motility strain V205 produced only 32.0 μg flagellin in the flagella fraction. Although a few background protein bands were observed (Fig. 1c), western blotting showed that the flagellin band that was detected from low-motility strains was faint compared to bands from highly motile strains (Fig. 1d). The swarming ability, which correlated with the amount of flagella, was coincident with high levels of IL-8 induction.


Figure 1. Motility of Afa/Dr DAEC strains in EMEM soft agar plates. (a) Motility of the DAEC strains was determined by measuring the diameters of the swarming area formed by the organisms inoculated onto EMEM soft agar plates. Red and blue lines show IL-8 induction from highly-inductive DAEC strains (V1, V19, V36, V64, V554, V561, and V827) and low-inducing strains (V205, V582, V599, V720, V242, V289, V547, V550, V679, V725, V880, and V922), respectively. Black lines are controls. Salmonella enteritidis (V215) and EAggEC strain V546 were used as positive controls. Non-pathogenic E. coli, DH5α was used as a negative control. Values are the mean ± standard deviation (n= 3). DAEC strain V544 was excluded because the organisms did not grow on EMEM soft agar plates. (b) Swarming on EMEM soft agar plates by the highly motile strain V64, the low-motility strain V205, and the non-motile strain V679. (c) SDS-polyacrylamide gel electrophoresis (7.5%) of flagella protein recovered from strain V64. Lanes 1–4 are markers (Precision Plus Protein All Blue Standards; Bio-Rad, Hercules, CA), the first supernatant of cultured broth, precipitant obtained from ultra-high speed centrifugation, and the supernatant, respectively. The arrow indicates flagellin. (d) Western blotting of partially purified flagella with anti-H4 serum. Lanes 1–3 show flagellin bands from the highly motile strains V1, V19, and V64, respectively; lane 4 is from the low-motility strain V205. Lanes 5–8 are negative controls for the antiserum and include flagella from the highly motile serotypes H5 (strain V561), H18 (strain V36), H18 (strain V827), and H28 (strain V554).

Download figure to PowerPoint

IL-8 induction by flagella

It has been reported that flagella purified from EAggEC O42 (26) or enteropathogenic E. coli (35, 36) induce IL-8 release from Caco-2 cells. We investigated this possibility using Afa/Dr DAEC strains because swarming correlated with induction of IL-8. Flagella from the motile DAEC strain V561, serotype H4, induced IL-8 secretion from Caco-2 cells in a dose-dependent manner to an equal extent as live bacteria (Fig. 2). It was recently reported that signal transduction due to flagellin acting via TLR5 was inhibited by gangliosides (33, 34). Indeed, the induction of IL-8 by the DAEC flagella and bacteria was inhibited by disialoganglioside-GD1a (Table 3). The specificity of the inhibitory effect against TLR5 by the gangliosides could not be examined because other TLR systems such as TLR4 were not active in Caco-2 cells (37). In HEp-2 cells, however, the gangliosides inhibited IL-8 induction by flagella without inhibiting IL-8 induction by lipopolysaccharide or CpG-containing oligonucleotides (data not shown). These results suggested that flagella play a key role in the high IL-8 induction by DAEC.


Figure 2. IL-8 secretion from Caco-2 cells induced by flagella from the high IL-8 inducer V561. Partially purified flagella were inoculated on Caco-2 cells grown in 24-well plates and were incubated for 6 hr. Then, the amount of IL-8 in the tissue culture medium was assayed. Values are the mean ± standard deviation (n= 3).

Download figure to PowerPoint

Table 3.  Effects of disialoganglioside treatment of flagella on induction of IL-8 in Caco-2 cells
Reagent added to culturesIL-8 (pg/ml after 18 hra)
  1. aData are the mean ± SEM (n= 3).

  2. bCaco-2 cells were treated with flagella serotype H4 (5 μg/ml) purified from V64.

  3. cH4 (5 μg/ml) was incubated with disialoganglioside-GD1a (100 μg/ml) at 37°C for 1 hr prior to stimulation.

  4. dLive bacteria (1.6 × 107 CFU/well) were inoculated onto Caco-2 cells.

H4b290 ± 46
H4 + Disialoganglioside-GD1ac15 ± 1
V64d333 ± 64
V64 + Disialoganglioside-GD1a10
Medium control17 ± 2

IL-8 induction in cultured epithelial cells expressing TLR5-siRNA

To clarify whether expression of IL-8 in epithelial cells was induced via TLR5 by DAEC flagella, we examined IL-8 induction using cultured epithelial cells expressing TLR5-siRNA. Motile DAEC strains (V64, V554) induced high levels of IL-8, and EAEC (V546) as a positive control induced IL-8 in Caco-2 and HEp-2 cells expressing control vector containing scrambled nucleotides (psiRNA-h7SKgzScr). However, IL-8 induction was clearly inhibited in epithelial cells expressing TLR5-siRNA (Fig. 3a, b). The level of IL-8 secreted from transfected Caco-2 cells was exceptionally high compared with normal Caco-2 cells. This is presumably because their TJ were nascent or fragile, as transfected Caco-2 cells grown in the presence of zeocin were morphologically undifferentiated.


Figure 3. Decrease in IL-8 secreted from human cultured epithelial cells (Caco-2 and HEp-2) following silencing of TLR5 with siRNA. (a) Caco-2 and (b) HEp-2 cells treated with siRNA specific for TLR5 showed lower IL-8 production after infection by motile strains (DAEC, V64, V554, EAEC, V546) and flagella from DAEC (serotype H4) compared with vehicle-treated control epithelial cells. The level of IL-8 production in epithelial cells infected with a non-motile DAEC strain (V679) was as low as non-infected cells regardless of whether the cells were transfected with siRNA. In panel (a), values are the mean ± standard error (n= 3). Experiments in panel (b) were done twice, and the values are the mean.

Download figure to PowerPoint

Influence of TJ on IL-8 induction by DAEC

It has been reported that TLR5 is expressed on the basolateral side of intestinal epithelial cells (38). IL-8 induction by Afa/Dr DAEC in differentiated Caco-2 cells was lower than in unpolarized HEp-2 cells, as we reported previously (19). We hypothesized that flagella contact TLR5 on unpolarized cells expressing TLR5 over the entire surface more readily than on Caco-2 cells, which express TLR5 on the basolateral side only. To examine whether the inhibition of the passage of bacteria or flagella onto the basolateral side inhibits IL-8 induction, we measured the level of IL-8 induced by Afa/Dr DAEC in Caco-2 cells with different TEER. The level of IL-8 in the culture medium of Caco-2 cells with TEER over 150 Ω cm2 remained approximately 100 pg/ml. However, cells with TEER below 150 Ω cm2 induced IL-8 up to a concentration of 800 pg/ml (Fig. 4). This result showed that the strength of the TJ was related to IL-8 induction.


Figure 4. IL-8 induction by DAEC in Caco-2 cell monolayers with different TEER. Caco-2 monolayers were grown on a 3.0-μm Pore Polycarbonate Membrane Insert (Corning, NY, NY) for 2–3 weeks. After measuring the TEER, 1 × 107/ml DAEC strain (V561, n= 20) was inoculated into the culture medium on the apical side, and IL-8 was measured in the medium 6 hr later. IL-8 production in bacteria-free medium was 9.8 ± 0.8 pg/ml (n= 3).

Download figure to PowerPoint

Translocation of flagella to TLR5

Finally, we examined how Afa/Dr DAEC flagella reached TLR5 when crossing a polarized epithelial monolayer. Partially purified flagella did not decrease the TEER (Fig. 5a). To investigate the effect of the DAEC strain on epithelial barrier function, we longitudinally measured the TEER (Fig. 5a) and transepithelial flux of uranin (MW = 362) (Fig. 5b) on Caco-2 epithelial monolayers grown on filters after inoculation of Afa/Dr DAEC strain V64 or non-pathogenic E. coli DH5α onto the apical bathing medium. The V64 strain decreased TEER and accelerated transepithelial flux of uranin more rapidly than DH5α.


Figure 5. Change in gate functions at tight junctions of DAEC-infected Caco-2 cell monolayers. (a) TEER of Caco-2 monolayers infected with DAEC strain V64 was compared with that of cells inoculated with flagella, non-pathogenic E. coli DH5α, or uninfected cells. V64 caused a marked decrease in TEER at 9 hr. (b) Paracellular passage of uranin in Caco-2 cell monolayers increased when cells were infected with V64 compared with uranin passage through a monolayer inoculated with non-pathogenic E. coli DH5α or uninfected cells. Uranin diffused from the upper to the lower chamber beyond the 3.0-μm Pore Polycarbonate Membrane Insert (Corning, NY, NY) quickly unless the monolayer was present (cell-free membrane). Values are the mean (n= 3).

Download figure to PowerPoint

Next, the DAEC strains were examined for known virulence genes using PCR to evaluate their contribution to the reduction of TEER. Although sat has reportedly been implicated in lesions of TJ (39), only one of the seven high IL-8 inducers expressed the gene (Table 1).

We further determined whether an intracellular route was involved in the delivery of flagella to TLR5. Listeria monocytogenes and Yersinia spp. invade cells via endocytosis, and the activation of PI3K is involved in the signaling pathway (40). Wortmannin, an inhibitor of PI3K activation (41), did not affect the number of adherent bacteria, whereas the number of internal bacteria was decreased by wortmannin in a dose-dependent manner. IL-8 was produced abundantly regardless of the decrease in invasive bacteria (Fig. 6), similar to Yersinia (41).


Figure 6. Effect of the PI3K inhibitor wortmannin on adhesion, invasion, and IL-8 induction by Afa/Dr DAEC. Wortmannin decreased the number of invading V64 (DAEC) organisms into polarized Caco-2 cells in a dose-dependent manner, but the drug did not affect IL-8 induction. Values are the mean (n= 3). Error bars show the standard deviation.

Download figure to PowerPoint


  1. Top of page

Although DAEC is often isolated from fecal specimens (6), it is unknown whether all such strains are enteropathogenic. If only some strains of DAEC are diarrheagenic, methods for discrimination between diarrheagenic strains and other strains are clearly required. Previously, we attempted to determine whether all Afa-positive DAEC strains isolated from diarrheal stools induced IL-8 secretion in cultured human epithelial cells to the same extent as EAggEC organisms (19). Some Afa/Dr DAEC strains induced IL-8 secretion from cultured human intestinal cells, whereas others only caused a weak reaction despite their diffuse adhesion. Thus, diffuse adhesion itself does not necessarily indicate that the strain is inflammatory in the intestine. Our data suggested that the motility (i.e. the amount of flagella in DAEC) may be more important than invasion for Afa-positive DAEC strains to induce inflammatory cytokines. Thus, only a subset of Afa/Dr DAEC strains would be enteropathogenic if chemokine induction was an essential step for full virulence (19). This hypothesis may explain why epidemiological studies have so far failed to yield conclusive answers regarding the role of DAEC in diarrheal disease. In fact, we found that high IL-8-inducing strains were significantly prevalent among 1- to 4-year-old diarrheal patients who are reportedly prone to infection by DAEC (27).

Because motility was suggested to be involved in Afa/Dr DAEC virulence (19), in the present study, we looked at the role of flagella in the induction of high levels of IL-8. The strongly inducing strains in this study were all highly motile in the swarming tests (Fig. 1). The highly motile strains had more flagella than low-motility or non-motile strains. Bacterial flagella trigger inflammatory responses via Toll-like receptors (26, 35). Disialoganglioside-GD1a, an inhibitor of TLR5, completely blocked IL-8 induction by the highly motile DAEC strains (Table 3). Transfection of siRNA for TLR5 also suppressed IL-8 secretion (Fig. 3). The flagella appeared to induce IL-8 secretion via TLR5 because partially purified flagella by themselves induced IL-8 secretion (Figs 2, 3). However, large amounts of flagella were required compared to the amount of flagella that was estimated to be associated with inoculated DAEC organisms (2.5 × 107 CFU).

Flagella and adhesins mediate opposite actions: flagella propel bacteria along mucus layers, whereas adhesins allow bacteria to adhere to specific receptors present on epithelial cells. Pathogens must coordinately regulate motility and adhesion (42, 43). Upon adhesion to surfaces, Pseudomonas aeruginosa alters its outer membrane composition, thereby shedding flagellin from the flagella (44). Likewise, Afa/Dr DAEC may have unknown additional mechanisms for shedding flagella, allowing both stable colonization and efficient delivery of flagella or flagellin to TLR5 in addition to the close apposition associated with diffuse adhesion to epithelial cells.

TLR5 is reportedly located on the basolateral side of intestinal epithelia (38), and flagella do not reach from the apical side to the basolateral side when the barrier function of TJ is normal. Indeed, lower TEER led to increased IL-8 production (Fig. 4). Immature unpolarized Caco-2 cells grown under the selective pressure of zeocin produced more IL-8 than polarized normal Caco-2 cells, indicating the importance of TJ as a barrier (Fig. 3). Because the barrier function of the epithelia decreased in 6–9 hr after Afa/Dr DAEC infection (Fig. 5), deterioration of epithelial TJ following adhesion of the DAEC would make the epithelia more vulnerable to the transfer of flagella to the basolateral side. Invasion of Afa/Dr DAEC was unlikely to be involved in the IL-8 secretion in terms of delivery of flagellin to TLR5 or to cytosolic NOD-like receptor C4 (45) (Fig. 6). Guignot et al. reported that fully differentiated Caco-2 cells do not allow the entry of DAEC strain IH11128 unless the TJ are disrupted by EGTA treatment and the basolateral surface is exposed to the bacteria (46). The present findings likely support the idea that the DAEC organisms weaken TJ. The mechanisms of how Afa/Dr DAEC organisms alter conductivity of epithelial TJ remain to be clarified (39, 47).

In conclusion, possession of sufficient amounts of flagella to stimulate TLR5 is a candidate second criterion for predicting the ability of Afa/Dr DAEC isolates to induce high amounts of IL-8 in clinical laboratories. This phenomenon can be observed by bacterial spreading into LIM medium in 24 hr rather than a swarming test, although diffuse adhesion is a prerequisite condition. When the ability of DAEC strains to trigger IL-8 secretion was compared between Afa/Dr DAEC strains isolated from healthy specimens and diarrheal specimens, most Afa-possessing motile DAEC strains from healthy adults induced less IL-8 than the strains from diarrheal patients (48). In the present study, we suggest that loosening of TJ by DAEC is important for flagella to stimulate TLR5. In future studies, changes in the TEER following infection with motile Afa-possessing DAEC strains from healthy individuals should be compared with TEER changes induced by strains from diarrheal patients. The factors contributing to the deterioration of TJ may be a third criterion used to assess the diarrheagenicity of Afa/Dr DAEC strains.


  1. Top of page
  • 1
    Black R.E., Morris S.S., Bryce J. (2003) Where and why are 10 million children dying every year Lancet 361: 222634.
  • 2
    Kosek M., Bern C., Guerrant R.L. (2003) The global burden of diarrhoeal disease, as estimated from studies published between 1992 and 2000. Bull World Health Organ 81: 197204.
  • 3
    Nataro J.P., Kaper J.B. (1998) Diarrheagenic Escherichia coli. Clin Microbiol Rev 11: 142201.
  • 4
    Scaletsky I.C., Silva M.L., Trabulsi L.R. (1984) Distinctive patterns of adherence of enteropathogenic Escherichia coli to HeLa cells. Infect Immun 45: 5346.
  • 5
    Giron J.A., Jones T., Millan Velasco F., Castro Munoz E., Zarate L., Fry J., Frankel G., Moseley S.L., Baudry B., Kaper J.B., Schoolnik G.K., Riley L.W. (1991) Diffuse-adhering Escherichia coli (DAEC) as a putative cause of diarrhea in Mayan children in Mexico. J Infect Dis 163: 50713.
  • 6
    Jallat C., Livrelli V., Darfeuille Michaud A., Rich C., Joly B. (1993) Escherichia coli strains involved in diarrhea in France: high prevalence and heterogeneity of diffusely adhering strains. J Clin Microbiol 31: 20317.
  • 7
    Echeverria P., Serichantalerg O., Changchawalit S., Baudry B., Levine M.M., Orskov F., Orskov I. (1992) Tissue culture-adherent Escherichia coli in infantile diarrhea. J Infect Dis 165: 1413.
  • 8
    Scaletsky I.C., Fabbricotti S.H., Carvalho R.L. B., Nunes C.R., Maranhao H.S., Morais M.B., Fagundes Neto U. (2002) Diffusely adherent Escherichia coli as a cause of acute diarrhea in young children in northeast Brazil: a case-control study. J Clin Microbiol 40: 6458.
  • 9
    Levine M.M., Ferreccio C., Prado V., Cayazzo M., Abrego P., Martinez J., Maggi L., Baldini M.M., Martin W., Maneval D., Kay B., Guers L., Lior H., Wasserman S.S., Nataro J.P. (1993) Epidemiologic studies of Escherichia coli diarrheal infections in a low socioeconomic level peri-urban community in Santiago, Chile. Am J Epidemiol 138: 84969.
  • 10
    Gunzburg S.T., Chang B.J., Elliott S.J., Burke V., Gracey M. (1993) Diffuse and enteroaggregative patterns of adherence of enteric Escherichia coli isolated from aboriginal children from the Kimberley region of Western Australia. J Infect Dis 167: 7558.
  • 11
    Baqui A.H., Sack R.B., Black R.E., Haider K., Hossain A., Alim A.R., Yunus M., Chowdhury H.R., Siddique A.K. (1992) Enteropathogens associated with acute and persistent diarrhea in Bangladeshi children less than 5 years of age. J Infect Dis 166: 7926.
  • 12
    Poitrineau P., Forestier C., Meyer M., Jallat C., Rich C., Malpuech G., De Champs C. (1995) Retrospective case-control study of diffusely adhering Escherichia coli and clinical features in children with diarrhea. J Clin Microbiol 33: 19612.
  • 13
    Kallenius G., Mollby R., Svenson S.B., Windberg J., Lundblad A., Svenson S., Cedergren B. (1980) The Pk antigen as receptor for the haemagglutination of pyelonephritic Escherichia coli. FEMS Microbiol Lett 7: 297302.
  • 14
    Labigne Roussel A.F., Lark D., Schoolnik G., Falkow S. (1984) Cloning and expression of an afimbrial adhesin (AFA-I) responsible for P blood group-independent, mannose-resistant hemagglutination from a pyelonephritic Escherichia coli strain. Infect Immun 46: 2519.
  • 15
    Telen M.J., Hall S.E., Green A.M., Moulds J.J., Rosse W.F. (1988) Identification of human erythrocyte blood group antigens on decay-accelerating factor (DAF) and an erythrocyte phenotype negative for DAF. J Exp Med 167: 19938.
  • 16
    Bilge S.S., Clausen C.R., Lau W., Moseley S.L. (1989) Molecular characterization of a fimbrial adhesin, F1845, mediating diffuse adherence of diarrhea-associated Escherichia coli to HEp-2 cells. J Bacteriol 171: 42819.
  • 17
    Le Bouguénec C., Lalioui L., Du Merle L., Jouve M., Courcoux P., Bouzari S., Selvarangan R., Nowicki B.J., Germani Y., Andremont A., Gounon P., Garcia M.I. (2001) Characterization of AfaE adhesins produced by extraintestinal and intestinal human Escherichia coli isolates: PCR assays for detection of afa adhesins that do or do not recognize Dr blood group antigens. J Clin Microbiol 39: 173845.
  • 18
    Servin A.L. (2005) Pathogenesis of Afa/Dr diffusely adhering Escherichia coli. Clin Microbiol Rev 18: 26492.
  • 19
    Arikawa K., Meraz I.M., Nishikawa Y., Ogasawara J., Hase A. (2005) Interleukin-8 secretion by epithelial cells infected with diffusely adherent Escherichia coli possessing Afa adhesin-coding genes. Microbiol Immunol 49: 493503.
  • 20
    Meraz I.M., Arikawa K., Ogasawara J., Hase A., Nishikawa Y. (2006) Epithelial cells secrete interleukin-8 in response to adhesion and invasion of diffusely adhering Escherichia coli lacking Afa/Dr genes. Microbiol Immunol 50: 15969.
  • 21
    Germani Y., Begaud E., Duval P., Le Bouguénec C. (1996) Prevalence of enteropathogenic, enteroaggregative, and diffusely adherent Escherichia coli among isolates from children with diarrhea in New Caledonia. J Infect Dis 174: 11246.
  • 22
    Bétis F., Brest P., Hofman V., Guignot J., Kansau I., Rossi B., Servin A., Hofman P. (2003) Afa/Dr diffusely adhering Escherichia coli infection in T84 cell monolayers induces increased neutrophil transepithelial migration, which in turn promotes cytokine-dependent upregulation of decay-accelerating factor (CD55), the receptor for Afa/Dr adhesins. Infect Immun 71: 177483.
  • 23
    Bétis F., Brest P., Hofman V., Guignot J., Bernet-Camard M.F., Rossi B., Servin A., Hofman P. (2003) The Afa/Dr adhesins of diffusely adhering Escherichia coli stimulate interleukin-8 secretion, activate mitogen-activated protein kinases, and promote polymorphonuclear transepithelial migration in T84 polarized epithelial cells. Infect Immun 71: 106874.
  • 24
    Steiner T.S., Lima A.A., Nataro J.P., Guerrant R.L. (1998) Enteroaggregative Escherichia coli produce intestinal inflammation and growth impairment and cause interleukin-8 release from intestinal epithelial cells. J Infect Dis 177: 8896.
  • 25
    Jiang Z.D., Greenberg D., Nataro J.P., Steffen R., Dupont H.L. (2002) Rate of occurrence and pathogenic effect of enteroaggregative Escherichia coli virulence factors in international travelers. J Clin Microbiol 40: 418590.
  • 26
    Steiner T.S., Nataro J.P., Poteet Smith C.E., Smith J.A., Guerrant R.L. (2000) Enteroaggregative Escherichia coli expresses a novel flagellin that causes IL-8 release from intestinal epithelial cells. J Clin Invest 105: 176977.
  • 27
    Meraz I.M., Arikawa K., Nakamura H., Ogasawara J., Hase A., Nishikawa Y. (2007) Association of IL-8-inducing strains of diffusely adherent Escherichia coli with sporadic diarrheal patients with less than 5 years of age. Braz J Infect Dis 11: 449.
  • 28
    Nishikawa Y., Zhou Z., Hase A., Ogasawara J., Kitase T., Abe N., Nakamura N., Wada T., Ishii E., Haruki K., Team T.S. (2003) Diarrheagenic Escherichia coli isolated from stools of sporadic cases of diarrheal illness in Osaka City, Japan between 1997 and 2000: prevalence of enteroaggregative E. coli heat-stable enterotoxin 1 gene-possessing E. coli. Jpn J Infect Dis 55: 18390.
  • 29
    Hartleib S., Prager R., Hedenstrom I., Lofdahl S., Tschape H. (2003) Prevalence of the new, SPI1-like, pathogenicity island ETT2 among Escherichia coli. Int J Med Microbiol 292: 48793.
  • 30
    Nishi J., Sheikh J., Mizuguchi K., Luisi B., Burland V., Boutin A., Rose D.J., Blattner F.R., Nataro J.P. (2003) The export of coat protein from enteroaggregative Escherichia coli by a specific ATP-binding cassette transporter system. J Biol Chem 278: 456809.
  • 31
    Taddei C.R., Moreno A.C., Fernandes Filho A., Montemor L.P., Martinez M.B. (2003) Prevalence of secreted autotransporter toxin gene among diffusely adhering Escherichia coli isolated from stools of children. FEMS Microbiol Lett 227: 24953.
  • 32
    Yamazaki M., Inuzuka K., Matsui H., Sakae K., Suzuki Y., Miyazaki Y., Ito K. (2000) Plasmid encoded enterotoxin (Pet) gene in enteroaggregative Escherichia coli isolated from sporadic diarrhea cases. Jpn J Infect Dis 53: 2489.
  • 33
    Ogushi K., Wada A., Niidome T., Okuda T., Llanes R., Nakayama M., Nishi Y., Kurazono H., Smith K.D., Aderem A., Moss J., Hirayama T. (2004) Gangliosides act as co-receptors for Salmonella enteritidis FliC and promote FliC induction of human beta-defensin-2 expression in Caco-2 cells. J Biol Chem 279: 122139.
  • 34
    West A.P., Dancho B.A., Mizel S.B. (2005) Gangliosides inhibit flagellin signaling in the absence of an effect on flagellin binding to toll-like receptor 5. J Biol Chem 280: 94828.
  • 35
    Zhou X., Giron J.A., Torres A.G., Crawford J.A., Negrete E., Vogel S.N., Kaper J.B. (2003) Flagellin of enteropathogenic Escherichia coli stimulates interleukin-8 production in T84 cells. Infect Immun 71: 21209.
  • 36
    Sampaio S.C., Gomes T.A., Pichon C., Du Merle L., Guadagnini S., Abe C.M., Sampaio J.L., Le Bouguénec C. (2009) The flagella of an atypical enteropathogenic Escherichia coli are required for efficient interaction with and stimulation of IL-8 production by enterocytes in vitro. Infect Immun. 77: 440613.
  • 37
    Takahashi K., Sugi Y., Hosono A., Kaminogawa S. (2009) Epigenetic regulation of TLR4 gene expression in intestinal epithelial cells for the maintenance of intestinal homeostasis. J Immunol 183: 65229.
  • 38
    Gewirtz A.T., Navas T.A., Lyons S., Godowski P.J., Madara J.L. (2001) Cutting edge: bacterial flagellin activates basolaterally expressed TLR5 to induce epithelial proinflammatory gene expression. J Immunol 167: 18825.
  • 39
    Guignot J., Chaplais C., Coconnier-Polter M.H., Servin A.L. (2007) The secreted autotransporter toxin, Sat, functions as a virulence factor in Afa/Dr diffusely adhering Escherichia coli by promoting lesions in tight junction of polarized epithelial cells. Cell Microbiol 9: 20421.
  • 40
    Mecsas J., Raupach B., Falkow S. (1998) The Yersinia Yops inhibit invasion of Listeria, Shigella and Edwardsiella but not Salmonella into epithelial cells. Mol Microbiol 28: 126981.
  • 41
    Schulte R., Zumbihl R., Kampik D., Fauconnier A., Autenrieth I.B. (1998) Wortmannin blocks Yersinia invasin-triggered internalization, but not interleukin-8 production by epithelial cells. Med Microbiol Immunol 187: 5360.
  • 42
    Pesavento C., Becker G., Sommerfeldt N., Possling A., Tschowri N., Mehlis A., Hengge R. (2008) Inverse regulatory coordination of motility and curli-mediated adhesion in Escherichia coli. Genes Dev 22: 243446.
  • 43
    Simms A.N., Mobley H.L. (2008) Multiple genes repress motility in uropathogenic Escherichia coli constitutively expressing type 1 fimbriae. J Bacteriol 190: 374756.
  • 44
    Gerstel U., Czapp M., Bartels J., Schroder J.M. (2009) Rhamnolipid-induced shedding of flagellin from Pseudomonas aeruginosa provokes hBD-2 and IL-8 response in human keratinocytes. Cell Microbiol. 11:84253.
  • 45
    Miao E.A., Andersen-Nissen E., Warren S.E., Aderem A. (2007) TLR5 and Ipaf: dual sensors of bacterial flagellin in the innate immune system. Semin Immunopathol 29: 27588.
  • 46
    Guignot J., Bernet Camard M.F., Pous C., Plancon L., Le Bouguénec C., Servin A.L. (2001) Polarized entry of uropathogenic Afa/Dr diffusely adhering Escherichia coli strain IH11128 into human epithelial cells: evidence for a5b1 integrin recognition and subsequent internalization through a pathway involving caveolae and dynamic unstable microtubules. Infect Immun 69: 185668.
  • 47
    Peiffer I., Blanc-Potard A.B., Bernet-Camard M.F., Guignot J., Barbat A., Servin A.L. (2000) Afa/Dr diffusely adhering Escherichia coli C1845 infection promotes selective injuries in the junctional domain of polarized human intestinal Caco-2/TC7 cells. Infect Immun 68: 343142.
  • 48
    Fujihara S., Arikawa K., Aota T., Tanaka H., Nakamura H., Wada T., Hase A., Nishikawa Y. (2009) Prevalence and properties of diarrheagenic Escherichia coli among healthy individuals in Osaka City, Japan. Jpn J Infect Dis 62: 31823.