INTRODUCTION
Enteropathogenic
Escherichia coli (EPEC) frequently causes acute and persistent diarrhea in animals and humans. EPEC infection leads to serious acute diarrhea in weaned pigs (
1,
2). In humans, EPEC infections are particularly serious for infants and toddlers and are often accompanied by high lethality in developing countries (
3,
4). However, it is now recognized that atypical EPEC (aEPEC) strains are more frequent in humans than typical EPEC (tEPEC; expression of bundle-forming pili) strains in both developing and developed countries (
5,
6). In animals, aEPEC strains are much more prevalent than EPEC strains, and aEPEC strains isolated from cattle, sheep, and pigs belong to the same serotypes found in humans (
7,
8). Furthermore, recent studies showed a close clonal relationship between human and animal aEPEC isolates, suggesting a possible zoonotic potential of animal aEPEC strains that could serve as a reservoir for human infections (
9).
Infection of intestinal epithelial cells by EPEC is a complex multistage process. Initially, EPEC loosely adheres to epithelial cells by diverse adhesins and subsequently translocates effector molecules, including the translocated intimin receptor (Tir), into host cells using a type 3 secretion system (T3SS). After integration of Tir into the host cell membrane, EPEC binds tightly through the adhesin intimin with Tir. EPEC is able to intimately adhere to epithelial cells and to form microcolonies with resulting typically associated histopathological alterations of the host cell surface known as attaching and effacing (AE) lesions. Rearrangement and massive accumulation of actin and other cytoskeletal proteins beneath the site of bacterial attachment lead to the formation of pedestal structures and destruction of microvilli (effacement). The pathogenesis is further characterized by loss of tight-junction integrity and barrier functions of the gut epithelium and destruction of microvilli (effacement) and the brush border that leads to diarrhea (
4,
10–13).
The nonpathogenic
E. coli strain Nissle 1917 (EcN) is a widely employed probiotic strain, and several
in vivo studies have demonstrated its promising probiotic activity in humans and animals, including the treatment of acute, chronic, or frequent recurring diarrhea and inflammatory bowel disease (
14–19). Proposed probiotic actions of EcN include effects on pathogens, host epithelial cells, host smooth muscle cell activity, and the host immune system (
20–28).
In vitro, EcN has been shown to inhibit invasion of host cells by several enteric pathogens, including
Salmonella,
Yersinia,
Shigella,
Legionella,
Listeria, and adherent-invasive
E. coli (
29,
30). However, the underlying molecular mechanisms remain largely unknown. Here, we characterize the effects of EcN on aEPEC infection of IPEC-J2 cells by means of confocal laser scanning microscopy and scanning electron microscopy, as well as molecular and protein biochemical methods. Our data provide new insights into host-bacterium and interbacterial interactions and show that EcN might be a promising tool in prophylactic defense against EPEC infections.
MATERIALS AND METHODS
Cell line and bacterial strains.
The porcine intestinal epithelial cell line IPEC-J2 (
31) was grown to confluence in Dulbecco's modified Eagle medium (DMEM)–Ham's F-12 (1:1) (Biochrom, Berlin, Germany) supplemented with 5% fetal calf serum (FCS) and maintained in an atmosphere of 5% CO
2 at 37°C. The bacterial strains used in this study are listed in
Table 1. EcN was kindly provided by G. Breves (Hannover, Germany). The EcN mutant EcN Δ
fliA (this study) was generated using the method of Datsenko and Wanner (
32). The
fliA gene was replaced by a kanamycin resistance (
kan) antibiotic cassette generated using the plasmid pKD4 as the template and primer pair fliAH1P1 (5′-GTGAATTCACTCTATACCGCTGAAGGTGTAATGGATAAACAGTGTAGGCTGGAGCTGCTTC-3′) and fliAH2P2 (5′-ACTTACCCAGTTTAGTGCGTAACCGTTTAATGCCTGGCTGTGCATATGAATATCCTCCTTAG-3′). EcN Δ
fliA was complemented using the plasmid pACYC177 harboring the sequence of
fliA (strain EcN Δ
fliA +
fliA).
E. coli MG1655 was a laboratory strain originally obtained from C. A. Cross (San Francisco, CA, USA).
E. coli IMT13962 was from the collection of the Institute of Microbiology and Epizootics (Berlin, Germany) and was originally isolated from the colon of a clinical healthy piglet and chosen for its strong adherence to IPEC-J2 cells. Strain IMT13962(pCosF1C6) was generated from strain IMT13962 by complementation with the
foc operon cloned into the pSuperCos1 vector (Stratagene, Heidelberg, Germany). Strain P2005/03 (kindly provided by R. Bauerfeind, Gießen, Germany) was isolated from a piglet with diarrhea and classified as aEPEC. Human EPEC E2348/69 was kindly provided by J. B. Kaper (Baltimore, MD, USA).
E. coli strain H5316 is a microcin-sensitive indicator strain kindly provided by K. Hantke (Tübingen, Germany). Uropathogenic
E. coli (UPEC) strain RZ525 was kindly provided by U. Dobrindt (Würzburg, Germany). Unless otherwise indicated, bacterial strains were grown in LB broth at 37°C with agitation at 200 rpm. For cultivation of strains P2005/03 and E2348/69, LB broth was supplemented with 20 μg/ml tetracycline and 30 μg/ml nalidixic acid, respectively.
Preparation of bacterial supernatants.
Bacterial strains were grown in LB broth at 37°C with agitation at 200 rpm to an optical density at 600 nm (OD600) of 1.0, diluted 1:100 in DMEM–Ham's F-12 cell culture medium containing 5% fetal calf serum, and grown again to an OD600 of 1.0. The bacterial cultures were centrifuged at 8,000 × g at 4°C for 15 min. The supernatants (SN) were sterile filtered with 0.22-μm filters and kept until use at −20°C.
Infection assay.
aEPEC P2005/03 and EPEC E2348/69 were grown to an OD600 of 1.0, washed by centrifugation, resuspended in cell culture medium, and adjusted by dilution to provide a multiplicity of infection (MOI) of 100:1 (bacteria to host cells) in wells of 12- or 24-well cell culture plates. Confluent monolayers of IPEC-J2 cells were infected with P2005/03 or E2348/69 and incubated at 37°C. After 3 h, nonadherent bacteria were removed by three washes with phosphate-buffered saline (PBS). For P2005/03, incubation was continued for an additional 3 h. Efficiencies of infection of epithelial cells were determined by washing and lysing the cells with 0.1% Triton X-100 in double-distilled H2O (ddH2O) and plating serial dilutions on LB agar plates containing 5 μg/ml tetracycline or 30 μg/ml nalidixic acid, which allowed the selective growth of P2005/03 and E2348/69, respectively. The plates were incubated overnight at 37°C, and the resulting numbers of CFU were determined. For preincubation experiments, epithelial cells were first incubated with the strains indicated for 2 h and washed three times with PBS prior to EPEC infections. For co- and postincubation experiments, strains were added simultaneously or 1 h after aEPEC infection. For growth kinetics of adherent aEPEC on epithelial cells, the cell culture medium was changed every 30 min 2 h after the beginning of aEPEC infection to remove nonadherent bacteria and to exchange exhausted cell culture medium.
Adhesion assay.
Bacterial strains were prepared and added to IPEC-J2 cell cultures as described for infection assays. At the indicated time points, nonadherent bacteria were removed by washing the cells three times with PBS. Numbers of adherent bacteria were determined by lysing the cells with 0.1% Triton X-100 and plating serial dilutions on LB agar plates.
Fluorescence microscopy.
For fluorescence microscopy, epithelial cells were grown to confluence on glass coverslips and fixed with acetone for 2 min at −20°C (confocal laser scanning microscopy) or 4% paraformaldehyde for 30 min at 4°C (epifluorescence microscopy) after performing infection or adhesion experiments. All incubation steps during staining with fluorescent dyes were performed in the dark. Strain P2005/03 was detected by immunohistochemical staining. Samples were incubated with polyclonal antibodies against serotype O108 raised in rabbit (Federal Institute for Risk Assessment, Berlin, Germany; diluted 1:50 in PBS with 0.5% bovine serum albumin [BSA]) for 30 min at room temperature, followed by incubation with goat anti-rabbit tetramethyl rhodamine isocyanate (TRITC)-labeled secondary monoclonal antibodies (Sigma-Aldrich, Munich, Germany; diluted 1:200 in PBS with 0.5% BSA) for 30 min at room temperature. Samples were washed three times with PBS after each antibody labeling.
The ability to produce AE lesions was indirectly examined by fluorescent actin staining (FAS) according to the method of Knutton et al. (
33). F-actin was stained with fluorescein isothiocyanate (FITC)-phalloidin (5 μg/ml; Invitrogen, Karlsruhe, Germany) for 30 min at room temperature. Samples were washed three times with PBS. Cell nuclei and bacteria in general were visualized by staining DNA with 0.3 μg/ml propidium iodide (Invitrogen, Karlsruhe, Germany) for 3 min at room temperature. Before staining, paraformaldehyde (PFA)-fixed cells were first incubated with 1 ml ice-cold 0.1% Triton X-100 in PBS for 4 min on ice and washed three times with PBS in order to permeabilize the cells. Images were acquired with the confocal laser scanning microscope DMIRE 2 TCS SP2 or the epifluorescence microscope DMBL (both from Leica, Wetzlar, Germany).
EcN flagella were stained with polyclonal antibodies against flagellum H1, kindly provided by L. Beutin (National Reference Laboratory for E. coli, Federal Institute for Risk Assessment, Berlin, Germany). Antibodies were diluted 1:50 in PBS with 0.5% BSA and applied for 30 min at room temperature, followed by incubation with FITC-labeled secondary monoclonal antibodies against rabbit (Sigma-Aldrich, Munich, Germany) for 30 min at room temperature. Samples were washed three times with PBS after each antibody labeling.
Scanning electron microscopy.
Infected cells grown on glass coverslips were fixed with 2% glutaraldehyde and 0.05% calcium chloride in 0.1 M sodium cacodylate buffer, pH 7.4, for 24 h at 4°C. Then, the samples were rinsed three times with ice-cold 0.1 M sodium cacodylate buffer, pH 7.4; fixed a second time with 1% osmium tetroxide in 0.1 M sodium cacodylate buffer for 3 h at 4°C; and subsequently rinsed again. For raster preparation, samples were dehydrated in a graduated series of ethanol solutions and finally in 100% acetone, critical-point dried with liquid carbon dioxide using the point dryer CPD 030 (Bal-Tec, Witten, Germany), and sputtered with 20-nm gold particles using the sputter coater SCD 005 (Bal-Tec, Witten, Germany). Samples were examined with a Leo 1430 scanning electron microscope (Leo Elektronenmikroskopie, Oberkochen, Germany).
Microcin test.
The expression, as well as the sensitivity, of
E. coli strains versus microcins was tested as described by Kleta et al. (
34).
Isolation of secreted and intracellular EPEC proteins.
The human EPEC strain E2348/69 was grown in 6 ml DMEM–Ham's F-12 cell culture medium to an OD600 of approximately 1.0. The bacteria were centrifuged for 5 min at 8,000 × g at room temperature, and the pellet was resuspended in 6 ml DMEM–Ham's F-12 cell culture medium. The bacterial suspension was then diluted 1:100 in 50 ml DMEM–Ham's F-12 in a 250-ml Erlenmeyer flask and grown again to an OD600 of 1.0 with shaking at 200 rpm at 37°C. The bacterial culture (45 ml) was centrifuged (15 min; 8,000 × g; 4°C), and the resulting supernatants were sterile filtered using 0.22-μm filters. Secreted proteins in the supernatants were precipitated with trichloroacetic acid (TCA) (final concentration, 10%) overnight at 4°C. The next day, the TCA precipitates were centrifuged for 30 min at 10,000 × g at 4°C. The resulting protein pellet was dried for 5 min at room temperature; resuspended in 1 ml 0.2% SDS in 25 mM Tris-HCl, pH 8.0; mixed with ice-cold acetone (1:4); and incubated for 1 h at −20°C. After centrifugation (30 min; 10,000 × g; 4°C), the protein pellet was dried (5 min; room temperature) and diluted in 25 mM Tris-HCl (pH 8.0) and 4 M urea. The secreted proteins (1,000-fold concentrated) were stored at −20°C.
In order to extract proteins from EPEC bacteria, 1 ml of the 50-ml bacterial suspension in the 250-ml Erlenmeyer flask was centrifuged for 5 min at 8,000 × g at room temperature. The bacterial pellet was resuspended in 4× sample buffer (1 M Tris-HCl, pH 6.8, 4% SDS, 8% β-mercaptoethanol, 20% glycerin, 0.025% bromophenol blue). The proteins were denatured by heating at 99°C for 5 min. The tubes were incubated on ice for 5 min, centrifuged for 5 s, and kept at −20°C until further use.
To determine the effects of EcN and MG1655 on protein secretion by EPEC, E2348/69 was grown in culture supernatants of EcN and MG1655, respectively. Bacterial supernatants were obtained from DMEM–Ham's F-12 medium as described above but without added FCS. For dilution of bacterial supernatants, DMEM–Ham's F-12 cell culture medium was used.
Western blot analysis.
Proteins were separated by SDS-PAGE according to the method of Laemmli (
35). For nonspecific protein detection, SDS-PAGE gels were stained with the Silver Stain Plus Kit (Bio-Rad, Munich, Germany) according to the manufacturer's instructions. The proteins EspA, EspB, and Tir were detected using Western blot analysis according to the method of Towbin et al. (
36). Briefly, proteins were transferred to nitrocellulose membranes (Sartorius, Göttingen, Germany) using a tank blot apparatus (Bio-Rad Laboratories, Munich, Germany). The membrane was blocked in 5% skim milk powder in Tris-buffered saline (TBS) buffer for 1 h. Primary antibodies against EspA, EspB, and Tir were kindly provided by J. B. Kaper (University of Maryland, Baltimore, MD, USA). The membranes were incubated with primary antibodies diluted 1:5,000 in blocking buffer overnight at 4°C with shaking, washed three times with TBS for 10 min each time, incubated with 1:2,000-diluted secondary antibodies labeled with horseradish peroxidase (Sigma-Aldrich, Munich, Germany) in blocking buffer for 1 h, and washed as described above. The proteins were visualized by enhanced chemiluminescence.
Statistical methods.
Statistical calculations were performed using SPSS 11.5 (SPSS, Chicago, IL, USA). Significances (P values) were calculated using Student's t test for normally distributed data or the Mann-Whitney U test for nonnormally distributed data. A P value of less than 0.05 was considered to indicate a statistically significant difference.
DISCUSSION
Probiotic bacteria might affect or interfere with aEPEC infections due to direct effects against pathogens or due to stabilizing effects on the intestinal microbiota, mucosa, or mucosal immunity. EcN has been successfully applied against different porcine and human intestinal disorders, including infection by enterotoxigenic
E. coli, acute nonspecific diarrhea, or chronic inflammatory bowel diseases (
14–18).
In vitro, EcN has been shown to inhibit the invasion of porcine and human intestinal epithelial cell lines by enteropathogens (
29,
30,
42), although the underlying mechanisms have rarely been defined. As there is currently no information about probiotic effects of EcN and other probiotics against aEPEC infection—either in animals or in humans—we determined the effects of EcN on aEPEC infection using a porcine intestinal-epithelial
in vitro model.
In the present study, preincubation of IPEC-J2 cells with EcN resulted in a drastic reduction of cell-adherent EPEC (aEPEC, as well as tEPEC). EcN affected the attachment and growth of cell-adherent aEPEC and therefore the growth of microcolonies, but not the formation of attaching and effacing lesions of adherent EPEC. We conclude that EcN interferes with the EPEC infection process very early and that this interference is a complex multistage process. Furthermore, culture supernatants of EcN were also found to be very effective against aEPEC infection. This is consistent with the observations of Altenhoefer et al. (
29) and Schierack et al. (
42), which showed that supernatants are effective against invasion of enteropathogens. The bacterial supernatants appear to act directly on aEPEC bacteria and not on host epithelial cells, as aEPEC infection was not inhibited by preincubation of IPEC-J2 cells with bacterial supernatants; only coincubation resulted in reduced infection efficiencies of aEPEC. However, we cannot exclude the possibility that other bacterial factors from the supernatant may have effects on host cells. Furthermore, the effects of supernatants on aEPEC infection were not EcN specific, as the supernatants of control
E. coli strains also inhibited aEPEC infection. This is in accordance with a recent study on the effects of EcN against
Salmonella invasion (
42) but in contrast to another study that showed a specific effect of EcN supernatants against enteroinvasive pathogens (
29). The discrepancy might be based on the use of
E. coli strain DH5α in the latter study. The strain grows very slowly in LB, as well as in cell culture medium (our unpublished observation), suggesting metabolic deficiencies, and might not secrete sufficient amounts of inhibitory supernatant factor(s). In addition, these prior studies used different intestinal epithelial cell lines and experimental conditions, which could also have been responsible for divergent results. Although there are several reports on the effects of bacterial supernatants on infection by enteropathogens, to date, a clearly defined inhibitor has not been identified.
The probiotic effect of EcN is likely to depend in part on the very high adhesion rate of EcN compared to the control strains. Only EcN was still present in high numbers after numerous washing steps of IPEC-J2 cell monolayers. In contrast, poorly adherent control bacteria were inevitably removed. As a result, only EcN was still able to grow to sufficient densities at host cell surfaces and possibly secrete sufficient amounts of putative inhibitors. Successful probiotic action of a bacterial strain is often associated with its colonization of the intestine. The colonization of hosts by EcN can be very successful but is likely to be specific for individual hosts (
34,
43).
Our results suggest that the inhibitory factor secreted by EcN is effective only when present at sufficient concentrations prior to aEPEC adhesion. Inhibitory effects were observed only in preincubation experiments where sufficient EcN bacteria (approximately 20 adherent bacteria per epithelial cell) were present close to the host cell surface, and thus the interaction site of aEPEC with epithelial cells. This suggestion is also supported by our observations that dilution of EcN supernatants reduced the inhibitory effect on secretion of virulence-associated effector proteins by EPEC.
EcN expresses several fimbrial-gene clusters, including (i) type 1 fimbriae (
66), expressed by most
E. coli strains and involved in the infection processes of many pathogens, such as UPEC or adherent invasive
E. coli (AIEC) (
44,
45); (ii) F1C fimbriae, which have usually been associated with uropathogenic
E. coli; and (iii) curli fimbriae (
46,
47). The type 1 fimbriae do not appear to play a predominant role in EcN adhesion or in mediating an inhibitory effect of EcN. An EcN Δ
fim mutant adhered as well as the EcN wild-type strain to IPEC-J2 cells, and adhesion of EcN to IPEC-J2 cells was not inhibited by α-
d-mannose (data not shown), indicating that type 1 fimbriae play only a subordinate role in EcN adhesion.
Recently, it was demonstrated that F1C fimbriae play an important role in EcN biofilm formation, adherence to intestinal epithelial cells
in vitro, and intestinal colonization of mice and a probiotic effect against
Salmonella invasion (
42,
48,
49). In the present study, we showed that F1C fimbriae were expressed and appeared to play a prominent role in adhesion of EcN to IPEC-J2 epithelial cells and contributed to the inhibitory effects on aEPEC infection. The notion that F1C fimbriae were indeed an important adhesin involved in these effects was further strengthened by complementation studies of
E. coli strain IMT13962 with the
foc operon and adhesion assays with the UPEC strain RZ525 carrying the F1C fimbriae. Both strains adhered at levels comparable to those of the EcN Δ
fliA mutant, and both inhibited aEPEC infection.
A second important EcN adhesion determinant we identified was the flagellum. We showed for the first time that EcN expresses H1 flagella during the adhesion process at epithelial cell surfaces. Time course experiments showed elongation of flagella after adhesion and the formation of a tight network on epithelial cell monolayers. Flagella contributed to the strong adherence of EcN to IPEC-J2 cells through interactions between single bacteria, as well as apparent anchorage to the host cell surface. We suggest that flagella may also prevent detachment of EcN after bacterial cell division, as well as facilitate adhesion of further EcN bacteria from the cell culture medium. Flagella, therefore, may contribute to the inhibitory effects on pathogen infection by enhancing EcN adhesion to host cell surfaces.
The role of flagella in adhesion might be of much greater importance
in vivo, as flagella enable bacteria to direct their movement toward the host cell by chemotaxis and thus support subsequent adhesion (
50). Moreover, flagella can also serve as direct adhesion determinants, as shown for enteropathogens, including typical EPEC strains (
51). The cross-linking of EPEC bacteria on the surfaces of human epithelial HeLa cells via flagella contributed to a three-dimensional structure but required a eukaryotic signal, similar to the expression of flagella by EcN in our present study. As scanning electron microscopy did not show expression of flagella by EcN on human intestinal Caco-2 cells in another study (
52), expression of EcN flagella might depend on the epithelial cell line, suggesting host or tissue specificity.
A prerequisite for successful infection of epithelial cells by enteropathogens is their adhesion to receptors on host cell surfaces. Receptors are predominantly carbohydrates of transmembrane or transmembrane-associated glycolipids and glycoproteins, as well as proteins or glycoproteins of extracellular matrices. The specificity of binding is determined by the sequence of carbohydrate chains (
53). The blocking of host cell surface receptors by probiotics (competitive exclusion), due to direct binding either to similar epitopes or to adjacent structures, has been proposed to be a potential mechanism for inhibiting pathogenic infections (
54,
55).
In the present study, aEPEC infection was inhibited by EcN adhesion, as well as by EcN supernatants that were isolated from shaking cultures and thus could also contain fimbrial components serving as potential ligands for host cell receptors. It might be possible that EcN and aEPEC use similar host cell receptors for adhesion, supporting competitive exclusion as one mechanism for the inhibitory effects. F1C fimbriae mediating adhesion of EcN to IPEC-J2 cells exhibit a high binding affinity to glycosphingolipids (e.g., asialo-GM1 and asialo-GM2), with the disaccharide sequence GalNacβ1-4Galβ as a minimal binding epitope and lower affinity for other ceramides (
56,
57). In contrast, EPEC host cell receptors have only rarely been defined, but in
in vitro studies of EPEC bacteria showing a localized adherent phenotype, they were found to bind to glycosphingolipids, suggesting that both EcN and EPEC might use the same host cell receptor (
58). Our experiments in which IPEC-J2 receptors were blocked with antibodies directed against asialo-GM2 indicated the expression of asialo-GM2 by IPEC-J2 and the binding of aEPEC to this receptor. In contrast to aEPEC, EcN adhesion levels were not affected by blocking asialo-GM2 (data not shown). In order to verify whether both EcN and EPEC bind to the same host cell receptors, more detailed studies identifying receptors present on IPEC-J2 cell surfaces and the binding of EcN and EPEC are necessary.
The results of the present study also show that the inhibitory factor might not affect aEPEC infection indirectly via epithelial cells, as preincubation of supernatants had no effect on aEPEC infection. The binding of ligands to receptors is in general very stable, and thus, the blocking of receptor structures used by pathogens to bind to cell surfaces would also result in inhibition of pathogen adhesion in preincubation experiments with supernatants, as shown for probiotic bifidobacteria (
59,
60). Our results, therefore, do not appear to support competitive exclusion as a predominant mechanism for the inhibitory effects of EcN.
EPEC infection of intestinal epithelial cells is mediated by a number of virulence-associated proteins that are secreted or translocated by the T3SS. Inhibition of protein expression, secretion, or translocation results in decreased EPEC virulence (
4,
61). In the present study, we tested the effects of EcN supernatants on protein expression and secretion. As the porcine aEPEC strain secreted protein concentrations insufficient for extensive SDS-PAGE and Western blot analysis, we examined the effects of EcN on secretion of virulence proteins in the human tEPEC strain E2348/69. The well-characterized tEPEC strain E2348/69 secreted detectable amounts of proteins common for EPEC adherence factor (EAF) plasmid-positive (tEPEC) compared to EAF plasmid-negative (aEPEC) EPEC strains (
41,
62). Protein secretion of strain E2348/69 was almost completely inhibited by EcN supernatants, including inhibition of secretion of EspA, EspB, and Tir. Supernatants of the control strain
E. coli MG1655 inhibited protein secretion to a much lesser extent. A reduction in secretion of all three proteins indicated that the T3SS or expression of locus of enterocyte effacement (LEE) genes encoding the secreted proteins was affected. T3SSs are used by a number of pathogenic Gram-negative bacteria, including
E. coli,
Yersinia,
Shigella,
Salmonella, and
Chlamydia. Inhibition of the T3SS would therefore be an effective mechanism for prevention of infection. It has previously been shown that a low-molecular-weight polyketide compound isolated from cultures of
Streptomyces sp. was able to inhibit the T3SS (
63). However, there are no data available on whether and how probiotics might block the T3SS. As intracellular cytosolic protein concentrations of EPEC were only slightly reduced, we conclude that protein secretion via the T3SS was the major target of inhibition, rather than bacterial protein expression. Further studies with respect to the expression of virulence genes are needed to confirm this.
In conclusion, a number of previous studies have indicated that EcN is effective against a range of pathogenic bacteria that possess numerous different infection strategies with diverse mechanisms to adhere to host cells. Our data give insights into potential mechanisms by which EcN could affect infection due to one of these pathogens, aEPEC. However, it appears unlikely that EcN will be found to inhibit infections by different specific mechanisms. The probiotic mode of action might rather be based on interfering with global mechanisms of bacterial pathogenesis, as suggested by this study. Although this mode of action might not be specifically limited to EcN, we assume that the strong colonization capabilities of EcN, both in vitro and in vivo, together with its fitness characteristics, support its probiotic effect.