Antagonistic effects of probiotic Escherichia coli Nissle 1917 on EHEC strains of serotype O104:H4 and O157:H7

International Journal of Medical Microbiology 303 (2013) 1–8

Contents lists available at SciVerse ScienceDirect

International Journal of Medical Microbiology journal homepage: www.elsevier.com/locate/ijmm

Antagonistic effects of probiotic Escherichia coli Nissle 1917 on EHEC strains of serotype O104:H4 and O157:H7 Stefan A. Rund a , Holger Rohde b , Ulrich Sonnenborn c , Tobias A. Oelschlaeger a,∗ a b c

Institut für Molekulare Infektionsbiologie, Universität Würzburg, Würzburg, Germany Institute of Medical Microbiology, Virology and Hygiene, University Medical Center Hamburg–Eppendorf, Hamburg, Germany Ardeypharm GmbH, Herdecke, Germany

a r t i c l e

i n f o

Article history: Received 26 October 2012 Received in revised form 21 November 2012 Accepted 25 November 2012 Keywords: Probiotic E. coli Nissle 1917 EHEC O104:H4 Adhesion Shiga toxin

a b s t r a c t The largest EHEC outbreak up to now in Germany occurred in 2011. It was caused by the non-O157:H7 Shiga-toxinogenic enterohemorrhagic E. coli strain O104:H4. This strain encodes in addition to the Shiga toxin 2 (Stx2), responsible for the hemolytic uremic syndrome (HUS), several adhesins such as aggregative adherence fimbriae. Currently, there is no effective prophylaxis and treatment available for EHEC infections in humans. Especially antibiotics are not indicated for treatment as they may induce Stx production, thus worsening the symptoms. Alternative therapeutics are therefore desperately needed. We tested the probiotic Escherichia coli strain Nissle 1917 (EcN) for antagonistic effects on two O104:H4 EHEC isolates from the 2011 outbreak and on the classical O157:H7 EHEC strain EDL933. These tests included effects on adherence, growth, and Stx production in monoculture and co-culture together with EcN. The inoculum of each co-culture contained EcN and the respective EHEC strain either at a ratio of 1:1 or 10:1 (EcN:EHEC). Adhesion of EHEC strains to Caco-2 cells and to the mucin-producing LS-174T cells was reduced significantly in co-culture with EcN at the 1:1 ratio and very dramatically at the 10:1 ratio. This inhibitory effect of EcN on EHEC adherence was most likely not due to occupation of adhesion sites on the epithelial cells, because in monocultures EcN adhered with much lower bacterial numbers than the EHEC strains. Both EHEC strains of serotype O104:H4 showed reduced growth in the presence of EcN (10:1 ratio). EHEC strain EDL933 grew in co-culture with EcN only during the first 2 h of incubation. Thereafter, EHEC counts declined. At 24 h, the numbers of viable EDL933 was at or slightly below the numbers at the time of inoculation. The amount of Stx2 after 24 h co-incubation with EcN (EcN:EHEC ratio 10:1) was for all 3 EHEC strains tested significantly reduced in comparison to EHEC monocultures. Obviously, EcN shows very efficient antagonistic activity on the EHEC strains of serotype O104:H4 and O157:H7 tested here regarding adherence to human gut epithelial cells, bacterial growth, and Stx2 production in vitro. © 2012 Elsevier GmbH. All rights reserved.

Introduction In 2007, a review with the title “The non-O157 Shiga-toxigenic (verocytotoxigenic) Escherichia coli; under-rated pathogens” (Bettelheim, 2007) nicely outlined the problem of underestimated infections by non-O157 E. coli. In 2011, the non-O157 Shiga toxin-producing enterohemorrhagic E. coli strain O104:H4 caused the largest outbreak of EHEC recorded so far. A total of 2987 cases of gastroenteritis (18 deaths) and 855 cases of HUS (35 deaths) were reported (RKI, 2011). A single lot of Fenugreek seeds from Egypt was identified as the most likely source of the O104:H4

∗ Corresponding author at: Institut für Molekulare Infektionsbiologie, JosefSchneider-Str. 2 D15, D-97080 Würzburg, Germany. Tel.: +49 0931 31 82150; fax: +49 0931 31 82578. E-mail address: [email protected] (T.A. Oelschlaeger). 1438-4221/$ – see front matter © 2012 Elsevier GmbH. All rights reserved. http://dx.doi.org/10.1016/j.ijmm.2012.11.006

outbreak strain (Buchholz et al., 2011). However, contaminations with EHEC O104:H4 could not be detected in any analyzed sample (EFSA, 2011). Other studies found no signs of E. coli O104:H4 in the feces of cattle from northern Germany or France, indicating that ruminants are not the primary reservoir of this strain (Auvray et al., 2012; Wieler et al., 2011). So far, the most likely reservoir of E. coli O104:H4 is the human population (Harrington et al., 2006; Karch et al., 2012; Okeke et al., 2010). Regardless of the origin, the severity of the disease caused by EHEC O104:H4 with respect to HUS development (Frank et al., 2011) and neurological symptoms (Jansen and Kielstein, 2011) was alarming. This new E. coli strain, a combination of an enteroaggregative E. coli (EAEC) and an enterohemorrhagic E. coli (EHEC) strain (Mellmann et al., 2011; Rohde et al., 2011), exhibited a unique set of virulence factors, which contributed to its hazardousness. One of the most important virulence factors of E. coli O104:H4 is Shiga toxin 2 (Stx2) (Pacheco and Sperandio, 2012), which is closely

2

S.A. Rund et al. / International Journal of Medical Microbiology 303 (2013) 1–8

related to Stx2 from E. coli O111:H− strain JB1-95 (Laing et al., 2012). The two-component Shiga toxin (AB5 ) is released from dying bacteria, binds to globotriaosylceramide (Gb3) on eukaryotic cells via subunit B, and ultimately inhibits protein synthesis (subunit A), which leads to cell death by apoptosis (Karch et al., 2012; Pacheco and Sperandio, 2012). Small blood vessels, which are found in the lungs, digestive tract or kidneys, are the main targets for Shiga toxin. Bloody diarrhea, renal dysfunction and kidney failure, damage of lungs, and severe neurological symptoms can be the result (Jansen and Kielstein, 2011; Karch et al., 2012). Another important virulence factor of EHEC O104:H4 is a plasmid encoding an extended-spectrum class A beta lactamase (CTX-M-15), which is responsible for resistance against a wide range of antibiotics (Bielaszewska et al., 2011). Whilst some antibiotics might be useful in the treatment of patients with EHEC O104:H4 (Bielaszewska et al., 2012), antibiotic treatment of EHEC is not generally recommended due to apprehension of increased Shiga toxin production (Smith et al., 2012; Wong et al., 2000). Furthermore, E. coli O104:H4 shows an aggregative adherence pattern (stacked-brick) like enteroaggregative E. coli (EAEC) (Bielaszewska et al., 2011). In contrast to its close relative EAEC 55989, EHEC O104:H4 harbors genes for aggregative adherence fimbriae I (AAF/1) and not AAF/III (Mellmann et al., 2011). The most fearsome aspect of the EHEC outbreak in 2011 is that it might happen again. Different ‘under-rated pathogens’ (Bettelheim, 2007) may arise and prove current treatment options as insufficient. Highly antibiotic-resistant pathogens are on the rise while development of new antibiotics is stagnating. Treatment of gastrointestinal diseases with probiotics as an alternative to antibiotics has been widely discussed over the last few years (Carey et al., 2008; Eaton et al., 2011; Fukuda et al., 2012; Muniesa et al., 2012; Oelschlaeger, 2010; Sonnenborn and Schulze, 2009). Probiotic bacterial strains are mostly members of the Gram-positive genera Lactobacillus and Bifidobacterium, but also Gram-negative bacteria such as the probiotic E. coli strain Nissle 1917 (EcN) are used. EcN has been in use in medicine as a probiotic drug (Mutaflor® ) since 1917 (Nissle, 1918) and can be applied for the treatment of various dysfunctions and diseases of the intestinal tract (Montrose and Floch, 2005; Schultz, 2008; Sonnenborn and Schulze, 2009). EcN is an effective alternative to the ‘gold standard’ mesalazin (5-aminosalicylic acid) in maintaining remission in patients with ulcerative colitis (Kruis, 2004; Rembacken et al., 1999). Other indications are chronic habitual constipation (Bär et al., 2009; Möllenbrink and Bruckschen, 1994) and diarrhea in young children (Henker et al., 2007, 2008). Means and molecular mechanisms responsible for EcN’s clinical efficacy are only partially understood. It has been shown that EcN inhibits invasion of intestinal epithelial cells by enteroinvasive bacteria (Altenhoefer et al., 2004). Moreover, EcN induces human beta-defensin 2 in epithelial cells via its flagellin (Schlee et al., 2007; Wehkamp et al., 2004). Antagonistic effects of EcN on colonization of the mouse intestine by EHEC O157:H7 strain EDL933 (Leatham et al., 2009) and inhibition of Shiga toxin production of several STEC strains (Reissbrodt et al., 2009) have been reported. Some of the fitness factors thought to contribute to the probiotic properties of E. coli Nissle 1917 are multiple iron acquisition systems (ent, iro, iuc/aer, ybt, chu, cit), different adhesins (F1C fimbriae, F1A fimbriae, Curli fimbriae, H1 flagella), and 2 different microcins (M, H47) (Patzer et al., 2003; Schlee et al., 2007; Valdebenito et al., 2006; Vassiliadis et al., 2010). An additional safety aspect in comparison to probiotic lactobacilli is EcN’s serum sensitivity, due to a mutation in the wzy gene encoding the O6 antigen polymerase (Grozdanov et al., 2002). This results in a truncated O6 carbohydrate side chain of its lipopolysaccharide, leading to the

remarkable feature that the EcN strain is immunogenic without being immunotoxic. The objective of the present study was to investigate potential antagonistic effects of EcN on adhesion, growth, and Shiga toxin production of EHEC strains, such as the O157:H7 strain EDL933 or EHEC O104:H4 isolates from the 2011 outbreak in Germany. Materials and methods Bacterial strains Bacterial strains used in this study are listed in Table 1. Bacteria were grown in TY medium (10 g/l tryptone, 5 g/l yeast extract, and 5 g/l NaCl) or DMEM medium containing 10% fetal bovine serum (PAA, Cölbe, Germany) at 37 ◦ C. Caco-2 and LS-174T cell cultures The human colonic epithelial cell lines Caco-2 (mucin-negative) and LS-174T (mucin-positive) were obtained from CLS (Eppelheim, Germany). The cell lines were cultured in DMEM medium containing 10% fetal bovine serum (FBS) and maintained at 37 ◦ C and 5% CO2 . Passages of Caco-2 cells were performed every 3–4 days with a split ratio of 1:4. LS-174T cells were cultured for 2–3 days and then subcultured with a split ratio of 1:2.5. All cell culture media and supplements were purchased from PAA. Adhesion to Caco-2 and LS-174T cells The cells were grown until confluency in cell culture flasks and seeded in 24-well microtiter plates. Caco-2 and LS-174T cells were seeded at 4.0 × 105 and 1.0 × 106 cells/well, respectively, and incubated for 24 h at 37 ◦ C and 5% CO2 to reach confluency (6.0 × 105 and 1.4 × 106 for Caco-2 and LS-174T, respectively). The DMEM medium was replaced with fresh medium 30 min before incubation with the bacterial inoculum. Bacterial overnight cultures in TY medium were used to inoculate cell cultures in DMEM medium, which were incubated for 2 h at 37 ◦ C under shaking conditions. Monocultures with EcN (∼1.5 × 107 CFUs/well) or pathogenic E. coli (pEc) (∼1.5 × 107 CFUs/well) and co-cultures at a ratio of 1:1 (EcN:pEc; ∼1.5 × 107 :1.5 × 107 CFUs/well) or 10:1 (EcN:pEc; ∼1.5 × 108 :1.5 × 107 CFUs/well) were made with these 2-h cultures. The CFUs/well of the inocula were confirmed plating serial dilutions on agar plates. 24-well microtiter plates with Caco2 or LS-174T cells and the bacteria were incubated for 2 h at 37 ◦ C and 5% CO2 . This was followed by removal of the supernatant containing non-adhering bacteria and 3 times washing of the epithelial cell layer with 1 ml PBS. Epithelial cells were resuspended in 1 ml Trypsin-EDTA (1:250, PAA) and lysed during 15 min of extensive shaking. Serial dilutions of the resulting bacterial suspension were placed on agar plates. To differentiate between the bacterial strains, the CFUs/ml on LB-agar, and LB-agar + antibiotic (TY3730, TY3456: 50 ␮g/ml Ampicillin; 55989: 6 ␮g/ml doxycyclin) were determined. Chromogenic ECC agar plates (Medco, München, Germany) were used to distinguish between EDL933 (pink color) and EcN (mauve color). Determination of bacterial growth in monoculture and co-culture The CFUs/ml of inocula (t = 0 h) and supernatants (t = 2 h) from the adhesion experiments were determined, and the supernatants were further incubated at 37 ◦ C and 0% CO2 to assess the CFUs at t = 5 h and t = 24 h. CFUs/ml of individual strains in co-culture were determined as described in the paragraph above.

S.A. Rund et al. / International Journal of Medical Microbiology 303 (2013) 1–8

3

Table 1 Bacterial strains used in this study. E. coli strain

Serotype

Important properties

Source

EAEC (55989) EHEC (EDL933) EHEC (TY3730) EHEC (TY3456) E. coli Nissle 1917 (EcN)

O104:H4 O157:H7 O104:H4 O104:H4 O6:K5:H1

No Stx production, DoxR High Stx production Clinical isolate from gastroenteritis patient, low Stx production, AmpR Clinical isolate from HUS patient, low Stx production, AmpR No Stx production

Ulrich Dobrindt, Münster Ulrich Dobrindt, Münster Holger Rohde, Hamburg Holger Rohde, Hamburg Ardeypharm, Herdecke

Shiga toxin assays The Shiga toxin production of all EHEC strains in monocultures and in co-cultures was determined after 24 h with the Ridascreen® Verotoxin ELISA (R-biopharm, Darmstadt, Germany). Bacteria were incubated in DMEM medium and the samples filtered with a lowprotein-binding filter (Millex-GV, Millipore, Tullagreen, Ireland). The ELISA readings (OD450 nm ) of monocultures and co-cultures were compared to assess the degree of inhibition of toxin production. The positive control was provided with the Ridascreen® Verotoxin ELISA. As negative control, DMEM medium containing 10% FBS was used. Statistical analysis Experiments were independently repeated at least 3 times and performed in quadruplicates (adhesion) or duplicates (growth/toxin production). Data are expressed as means ± SD. The statistical analyses were performed with the Student’s ttest. Differences were considered statistically significant with a p-value < 0.05. Results Influence of EcN on adhesion of pathogenic E. coli strains to epithelial cell lines Adherence to host cells is viewed as an important initial step for the establishment of an infection by pathogenic bacteria as well as for starting colonization by probiotic bacteria. Therefore, we first determined the adhesion efficiency of EcN and pathogenic E. coli strains in monocultures. The number of bacteria adhering to Caco-2 and LS-174T cells was lowest for EcN (Caco-2: 53.2 × 105 ± 2.2 × 105 CFUs/well; LS-174T: 44.5 × 105 ± 2.0 × 105 CFUs/well) and comparable for both cell lines. The highest numbers of adhering bacteria were observed for both EHEC O104:H4 isolates (TY3730: Caco-2: 401.7 × 105 ± 93.1 × 105 CFUs/well; LS-174T: 588.3 × 105 ± 140.3 × 105 CFUs/well; TY3456: 486.7 × 105 ± 103.4 × 105 CFUs/well; LS-174T: Caco-2: 593.3 × 105 ± 108.7 × 105 CFUs/well). The close relative of these isolates (EAEC 55989) was adhering less efficiently (Caco-2: 348.7 × 105 ± 82.5 × 105 CFUs/well; LS-174T: 225.7 × 105 ± 67.6 × 105 CFUs/well). Adhesion of EHEC strain EDL933 was even less pronounced than adhesion of the EAEC strain, although stronger than that of LS-174T: EcN (Caco-2: 179.2 × 105 ± 36.8 × 105 CFUs/well; 166.9 × 105 ± 28.0 × 105 CFUs/well) (Fig. 1). The adhesion efficiency to the 2 cell lines was significantly different for both EHEC O104:H4 isolates and the EAEC strain, but not for EDL933 and EcN (Fig. 1). In a second set of experiments, we investigated whether EcN might have an antagonistic effect on the adhesion efficiency of the pathogenic E. coli strains to Caco-2 and to the mucus-producing LS-174T cells. First, adhesion of EcN to epithelial cells in co-culture with pEc was determined. The number of adhering EcN bacteria

was static or slightly increased at a ratio of 10:1 (TY3456, Caco-2: 99.1 × 105 ± 32.5 × 105 CFUs/well; data not shown). Secondly, the adhesion efficiency for all tested pathogenic E. coli strains was enumerated. The number of adhering pEc was reduced in the presence of EcN. Adherence of both EHEC O104:H4 isolates to Caco-2 cells was reduced by 56% and 72% after co-incubation with EcN at a ratio of 1:1. At a ratio of 10:1 (EcN:EHEC O104:H4), adherence of these EHEC strains dropped to 5% and 13%, respectively (Fig. 2A). EcN inhibited the adherence of the EAEC strain by 45% and >99% at a ratio of 1:1 and 10:1 (EcN:EAEC), respectively (Fig. 2A). Also the adherence of EDL933 was lowered in the presence of EcN. At a ratio of 1:1, a residual adherence of 39% and at the 10:1 ratio a residual adherence of only 0.8% was found (Fig. 2A). Compared to the strong reduction of adhesion of both EHEC O104:H4 strains to Caco-2 cells by the presence of EcN, the reduction of adherence to the mucin-producing LS-174T cells was less pronounced. At a 1:1 ratio, adherence of these strains was reduced by 44% and 50% and at a 10:1 ratio by 79% and 87% (Fig. 2B). In contrast, in the presence of EcN, the reduction of adherence of the EAEC strain to LS-174T cells was more pronounced than the reduction of adhesion to Caco-2 cells only at the 1:1 ratio. Adherence of EAEC bacteria dropped to 26% (1:1 ratio) and 1.1% (10:1 ratio) (Fig. 2B). The adherence of EDL933 to the LS-174T cells was reduced by EcN to a comparably low level as with Caco-2 cells (36% and 1.7% residual adherence at the 1:1 and the 10:1 ratio) (Fig. 2B). EcN was able to inhibit adhesion of the tested pathogenic bacteria to both cell lines employed in a dose-dependent manner. This reduction of adherence was for both cell lines in all cases significant (p < 0.0001). Influence of EcN on growth of pathogenic bacteria Another feature frequently attributed to probiotic bacteria is their ability to inhibit the growth of pathogenic microorganisms. We wondered, whether EcN would also be able to interfere with the growth of different EHEC strains and the EAEC strain used in this study. For this purpose, we determined the bacterial cell counts (CFUs) at the start of the experiment (0 h) and at 2 h, 5 h, and 24 h post inoculation for monocultures of single strains as well as for

Fig. 1. Adhesion of E. coli strains to Caco-2 or LS-174T cells after 2 h of incubation at 37 ◦ C in a 5% CO2 atmosphere (ns: not significant, *p < 0.05, ***p < 0.001).

4

S.A. Rund et al. / International Journal of Medical Microbiology 303 (2013) 1–8

Fig. 2. Relative adhesion of pathogenic E. coli in co-culture with EcN to confluent monolayers of (A) Caco-2 cells and (B) LS-174T cells after 2 h of incubation at 37 ◦ C in a 5% CO2 atmosphere. Mean value of adherent pathogenic E. coli in monocultures (Fig. 1) served as reference (100%). All results were significant with p < 0.0001.

co-cultures of EcN and one of the pathogenic strains. In this approach, the bacteria were added to 1 ml DMEM on top of a confluent Caco-2 or LS-174T cell monolayer. The kind of epithelial cell line had no influence on the outcome of the experiments. Therefore, only the results obtained with the LS-174T cells are shown. In monocultures, both EHEC O104:H4 isolates and the EAEC strain had slightly lower cell counts compared to EcN at the same time points (Fig. 3A, D, and G). Cell counts of the EHEC O157:H7 strain EDL933 were almost identical at all time points of determination to those of the EcN culture (Fig. 3K). In co-culture experiments, EcN and the pathogenic E. coli strains were inoculated at a ratio of 1:1 and 10:1 (EcN:pEc). At the 1:1 ratio, growth of E. coli O104:H4 (TY3730) was not reduced by coincubation with EcN (Fig. 3B). However, both strains showed lower cell counts compared to their monocultures at 5 h and 24 h. Starting with a 10-fold higher EcN inoculum, growth of TY3730 was retarded, and the strain did not reach the same cell density at the respective time points as in the monoculture (compare Fig. 3C with Fig. 3A). In contrast to E. coli strain TY3730, growth of the other E. coli O104:H4 isolate (TY3456) was already slightly reduced at a 1:1 ratio of EcN and pEc (Fig. 3E). At 5 h and 24 h, the number of TY3456 bacteria in the co-culture was significantly lower than in the monoculture. However, this was also true for EcN at these time points (Fig. 3E). As could be expected, the inhibitory effect of EcN against TY3456 was much more pronounced at the 10:1 ratio (Fig. 3F). Cell counts for TY3456 in the co-culture with the 10:1 ratio were not only lower in comparison to the monoculture, but also to the culture with the 1:1 ratio. After 24 h, the number of TY3456 bacteria in the co-culture experiments had increased from 107 to 108 CFU only (Fig. 3F). The EAEC strain 55989 showed a slightly reduced growth in the monoculture as compared to the EcN monoculture (Fig. 3G). The presence of EcN in the 1:1 co-culture did not influence growth of the strain during the first 2 h. However, at 5 h, the number of EAEC bacteria had dropped dramatically, and cell counts were already below those at time point zero (t0 ). Thereafter, bacterial growth slowly returned, but the initial cell counts of t0 were hardly reached at 24 h (Fig. 3H). The decrease of EAEC counts was even more pronounced in the 10:1 co-culture (Fig. 3I). Here, already at 2 h, the number of EAEC bacteria was below that of the inoculum and dropped further, down to about 10% of the inoculum at 24 h. In monoculture, EHEC strain EDL933 grew as well as the EcN strain (Fig. 3K). In co-culture with EcN (1:1 and 10:1), strain EDL933

multiplied only during the first 2 h. At later time points, however, bacterial growth ceased, and the cell counts at time point 24 h were about those at the start of co-incubation (Fig. 3L and M). Obviously, the strains 55989 (EAEC) and EDL933 (EHEC O157:H7) were at least partially killed by EcN after 2 h. In contrast, both EHEC O104:H4 strains showed only reduced growth in the presence of EcN. Effects of EcN on Shiga toxin production by EHEC strains Very likely the most important virulence factor of EHEC strains is Shiga toxin. For EHEC O104:H4, it was shown that they produce Shiga toxin 2 (Stx2) (Bielaszewska et al., 2011) whereas the classical O157:H7 EHEC strain EDL933 produces Stx1 and Stx2 (Strockbine et al., 1986). EcN was able to reduce Stx2 production of Shiga toxin-producing E. coli (STEC) O157:H7 (97-10085) and Stx1 production of STEC O26:H11 (03-03231) and STEC O91:H− (03-06891) (Reissbrodt et al., 2009). Therefore, we were interested to investigate, whether EcN would be able to inhibit Shiga toxin production not only in STEC, but also in the classical EHEC EDL933 or in the EHEC O104:H4 isolates from the German outbreak in 2011. For EHEC O104:H4 (TY3730), the ELISA used to determine the Stx2 level showed a reduction of Stx2 after 24 h co-incubation with EcN in a dose-dependent manner (Fig. 4A). EHEC O104:H4 isolate TY3456 showed a significantly reduced Stx2 amount after co-incubation with EcN for 24 h, but only if EcN was present in a 10-fold excess (Fig. 4B). The amount of Shiga toxins 1 and 2 detectable in the EDL933 monoculture was about 5 times higher than the one produced by each of the O104:H4 isolates (compare Fig. 4C with Fig. 4A and B). In the co-culture inoculated with EcN and EDL933 at a ratio of 1:1, we could not observe a decrease of Shiga toxin production after 24 h. However, if the inoculum of EcN was 10-fold increased, the amount of Stx1/2 after 24 h co-incubation was reduced to about 15% of the amount detected in the EDL933 monoculture (Fig. 4C). EcN was able to reduce the amount of Shiga toxin produced by both EHEC types, O104:H4 and O157:H7, as long as the inocula for the co-culture contained 10-fold more EcN than EHEC bacteria. Discussion The main goal of this study was to investigate presumed antagonistic effects of the probiotic E. coli strain Nissle 1917 (EcN) on 2 clinical isolates of EHEC O104:H4 from the German EHEC outbreak in 2011. In addition, we also tested for inhibitory effects of EcN on

S.A. Rund et al. / International Journal of Medical Microbiology 303 (2013) 1–8

5

Fig. 3. Growth curves of E. coli Nissle 1917 (䊉) or EHEC/EAEC () strains in monoculture or in co-culture. (A) EcN and EHEC (O104:H4, TY3730) in monoculture. (B) EcN and EHEC (O104:H4, TY3730) in co-culture at a ratio of 1:1. (C) EcN and EHEC (O104:H4, TY3730) in co-culture at a ratio of 10:1. (D) EcN and EHEC (O104:H4, TY3456) in monoculture. (E) EcN and EHEC (O104:H4, TY3456) in co-culture at a ratio of 1:1. (F) EcN and EHEC (O104:H4, TY3456) in co-culture at a ratio of 10:1. (G) EcN and EAEC (O104:H4, 55989) in monoculture. (H) EcN and EAEC (O104:H4, 55989) in co-culture at a ratio of 1:1. (I) EcN and EAEC (O104:H4, 55989) in co-culture at a ratio of 10:1. (K) EcN and EHEC (O157:H7, EDL933) in monoculture. (L) EcN and EHEC (O157:H7, EDL933) in co-culture at a ratio of 1:1. (M) EcN and EHEC (O157:H7, EDL933) in co-culture at a ratio of 10:1. The differences between monocultures and co-cultures regarding CFUs of EHEC and EAEC at 24 h were always significant (p < 0.0001).

6

S.A. Rund et al. / International Journal of Medical Microbiology 303 (2013) 1–8

the classical EHEC O157:H7 strain EDL933. As EAEC strain 55989 of serotype O104:H4 is most likely the closest non-Stx producing relative of the 2011 German EHEC outbreak strain, we included this strain in our test systems (Mellmann et al., 2011). Adherence is viewed as an early step in establishing an infection by pathogenic bacteria. Therefore, we employed 2 human gut epithelial cell lines to check for interference of EcN with adhesion of pathogenic E. coli. First, adhesion efficiencies of bacteria from monocultures of different E. coli strains were determined. The least adhering strain to both cell lines was EcN. This was surprising, since probiotics are expected to adhere more efficiently to epithelial cells than pathogens. On the contrary, both O104:H4 isolates adhered 7.5–13.3-fold, EAEC strain 55989 adhered 5.1–6.5-fold, and EDL933 3.4–3.7-fold more efficiently than EcN. The very strong adherence phenotype of the O104:H4 isolates may be due to the fact, that EHEC O104:H4 has a new composition of adhesion factors such as the aggregative adherence fimbriae I (AAF/I), the IrgA homolog adhesin (Iha), long polar fimbriae (LPF), and an absent intimin (Eae) (Bielaszewska et al., 2011). Adherence of EHEC O104:H4 to the mucus-producing cell line LS-174T was significantly more efficient than to the non-mucus-producing cell line Caco-2. This implies that the ability to adhere to the mucin layer of the gut might be increased in this new EHEC strain. Therefore, infection of the human host might be easier for EHEC O104:H4 than for a classic EHEC or EAEC strain. Still, in co-culture, EcN was able to reduce the adherence efficiency of all pathogens employed in a dose-dependent fashion. Obviously, one often suggested mode of action of probiotics, i.e. interference with the adhesion of pathogens by occupying receptors and thus preventing adherence of pathogens (Ohland and Macnaughton, 2010; Sherman et al., 2009) is not the way EcN achieves inhibition of pathogen adherence in our test system. This assumption is also supported by the fact, that we performed coincubation and not pre-incubation experiments with EcN. Since adherence efficiency was calculated by enumerating living bacteria by determination of CFUs, the lowered number of living adhering bacteria might have reflected killing of the pathogenic bacteria by EcN. However, we can exclude this explanation for both O104:H4 isolates due to the fact that after 2 h of co-incubation with EcN the number of these bacteria was increased and not reduced. However, growth of EHEC strains is slowed down to some extent in the presence of EcN compared to O104:H4 monocultures. We cannot rule out inhibition of adhesin expression in the pathogens mediated by EcN. If EcN blocks adhesin expression at the start of the co-culture, then the adhesins present on the surface of EHEC and EAEC cells would be diluted 1:64 after the 6 duplications that could have occurred in this time period. Further investigations are under way to test this hypothesis. Killing by EcN might in fact at least be part of the reason for the reduced cell counts of EDL933, because this strain showed a reduced number of living bacteria after a 2h co-culture, if the incubation had been started at a ratio of 10:1 (EcN:EDL933) (Fig. 3L). This killing effect is even more pronounced for EAEC in the 10:1 (EcN:EAEC) co-culture. Another reason for the observed reduction of bacterial counts might have been the induction of human beta-defensin 2 expression in the epithelial cells by EcN, mediated by its flagella (Schlee et al., 2007; Wehkamp et al., 2004). However, this host-derived killing mechanism can probably be excluded due to the fact that EcN is also sensitive to this and other defensins, whereas the amount of adherent EcN is not reduced in co-cultures and even slightly increased at the EcN:pEc 10:1 ratio. Moreover, the Fig. 4. Shiga toxin production by EHEC O104:H4 or EHEC O157:H7 in monoculture or co-culture with EcN in DMEM medium after incubation for 24 h at 37 ◦ C was measured via Verotoxin ELISA (OD450 ). (A) Shiga toxin levels of EcN and EHEC (O104:H4, TY3730) in monoculture or co-culture. (B) Shiga toxin levels of EcN and EHEC (O104:H4, TY3456) in monoculture or co-culture. (C) Shiga toxin levels of EcN

and EHEC (O157:H7, EDL933) in monoculture or co-culture. The positive control was provided with the Ridascreen® Verotoxin ELISA. As a negative control, DMEM medium containing 10% FBS was used. *Samples were diluted (1:10) before the Verotoxin ELISA was performed.

S.A. Rund et al. / International Journal of Medical Microbiology 303 (2013) 1–8

bacterial density and the time of exposure to EcN is probably too low to induce enough human beta-defensin 2 to cause the effect seen in our experiments (Wehkamp et al., 2004). It is therefore likely that several distinct mechanisms are involved in the EcNmediated inhibition of adhering EHEC and EAEC bacteria. EcN is able to secrete several effectors which might kill other E. coli strains. These are the 2 microcins M and H47 (Patzer et al., 2003) and a polyketide (Nougayrede et al., 2006). In a second approach, the influence of EcN on replication and survival of EHEC and EAEC in co-culture was investigated. Growth of both O104:H4 EHEC isolates, determined as numbers of living bacterial cells (CFUs) at various time points during co-culture, was reduced. As expected, growth reduction was stronger in the 10:1 (EcN:EHEC) co-cultures than in the 1:1 co-cultures. In contrast, the closely related EAEC strain at 2 h showed in the 1:1 co-culture almost identical bacterial counts to those of the monoculture. However, thereafter, the number of EAEC bacteria decreased dramatically. In the EcN:EAEC 10:1 co-culture, the numbers of living EAEC bacteria were already lowered at 2 h and dropped further at the later time points. Obviously, EcN kills this strain in a time- and dose-dependent manner. It may be speculated that the 2 microcins (H47 and M) encoded by EcN (Patzer et al., 2003) might have been responsible for this killing. The microcins might have also been responsible for the killing observed in co-cultures of EcN with EHEC strain EDL933, although this effect was not as pronounced as with the EAEC strain. The catecholate receptors Fiu, Cir, and FepA of E. coli act also as receptors for these microcins (Patzer et al., 2003). A possibly lower expression of these receptors could explain the lower susceptibility of EHEC O104:H4 in co-culture with EcN. The most crucial factor in EHEC infections is the production of Shiga toxin(s). EHEC O104:H4 produces Stx2, whereas EDL933 produces Stx1 and Stx2. Shiga toxin produced by EHEC in the gut is transported to the site of pathological lesions where it inactivates the protein machinery via the Stx A subunit, which ultimately results in cell death (Muniesa et al., 2012). An extraordinarily high frequency of hemolytic uremic syndrome (HUS) and of neurological symptoms such as paresis, epileptic seizures, delirium, and coma was observed with the new EHEC strain of serotype O104:H4 (Jansen and Kielstein, 2011; Karch et al., 2012). The use of antibiotics for the treatment of EHEC infections is controversially discussed and not recommended, due to the resistance of EHEC O104:H4 against a wide range of antibiotics and a possibly increased toxin release induced by these agents (Proulx et al., 1992; Tarr et al., 2005; Wong et al., 2000). A previous study has shown EcN to reduce the production of Stx1 in STEC O26:H11 and STEC O91:H− as well as Stx2 production in STEC O157:H7 (Reissbrodt et al., 2009). Since effective treatment options for EHEC infections are still lacking, although they are urgently needed, we investigated the effect of EcN on toxin production by these EHEC strains. For that purpose, an ELISA was employed which detects Stx1 as well as Stx2. In accordance with earlier reports, we observed a much higher Stx production by EHEC strain EDL933 compared to the O104:H4 EHEC isolates (Laing et al., 2012). A reduction of Stx production was observed in all 10:1 (EcN:EHEC) co-cultures after 24 h of cocultivation. Interestingly, the inhibitory effect on toxin production exerted by EcN was greatest for the most efficient producer, EDL933 (85% inhibition). For the O104:H4 isolate TY3730, the reduction was almost equally pronounced (84% inhibition). Although isolate TY3456 was also isolated from a patient during the 2011 EHEC outbreak in Germany, the Stx2 production of this isolate was far less repressed by EcN compared to isolate TY3730. These 2 isolates, both of serotype O104:H4, differ also in their resistance to Trimethroprim/Cotrimoxazol. The different behavior of the isolates observed by us might well be the result of (micro)evolution in different hosts.

7

The mechanisms responsible for the reduced Stx titers after 24 h co-incubation with EcN are not known. It can only be speculated that the dramatically reduced Stx amount in the co-culture of EDL933 with EcN (ratio 10:1) reflects the fast killing of this strain. A dramatic reduction of the cell counts of EDL933 in these co-cultures was documented in the experiments for the determination of growth (=CFUs at various time points) and again in the co-cultures for the determination of Stx (data not shown). Preliminary results indicate that EcN microcins are not responsible for the reduction of Stx production in EDL933. A previous study showed that EcN does not inhibit EDL933 in colicin and microcin assays (Leatham et al., 2009), which is consistent with our preliminary results. The situation for the EHEC O104:H4 isolates is different. These pathogens are not killed by EcN. Nevertheless, Stx2 levels are clearly reduced in the presence of a 10-fold higher EcN inoculum compared to the O104:H4 inoculum, although to different degrees for these 2 strains. The reduction might be the result of repressed Stx expression in the O104:H4 isolates. Reduced Stx expression as a result of lowering the pH through organic acid production by lactobacilli or bifidobacteria without affecting viability of EHEC has been reported (Carey et al., 2008). In DMEM medium, which was the medium mainly used in these experiments, a transient pH reduction was observed. At time point 24 h, however, the pH was again at about 7. Furthermore, Stx production in LB medium was also determined (data not shown). There, reduction of Stx production in co-culture with EcN (10:1) was observed for all the EHEC strains investigated to a similar degree as in DMEM, but no lowering of pH was observed. A simple mechanism, such as the production of (low-molecularweight) metabolites could also be one way for the down-regulation of Shiga toxin production. Acetate is also the main end product of carbohydrate degradation by E. coli Nissle 1917. Fukuda et al. have shown that high amounts of acetate produced by bifidobacteria can prevent a lethal infection induced by the enterohemorrhagic E. coli O157:H7 in the mouse model (Fukuda et al., 2011, 2012). As Pacheco and Sperandio (2012) already discussed in an article this year, Shiga toxin expression is connected to the phage cycle and the SOS response of the EHEC bacteria. Here, the protein RecA plays a key role, produced and activated upon triggering of the SOS response (Mühldorfer et al., 1996). Decreasing the amount of activated RecA via various ways could be a general mechanism responsible for the reduced level of Shiga toxin production. Alternatively, only the secretion/release of Stx could be inhibited by EcN. This explanation is, however, rather unlikely because in a test with O104:H4 strain TY3730 the amount of Stx detected was the same in the cell-free culture supernatant and in the lysate of the whole co-culture with EcN (data not shown). We are already planning experiments to elucidate the mechanism(s) responsible for this important anti-EHEC effect of EcN as well as for the effects on adherence and growth of these pathogens. Acknowledgements We are grateful to Klaus Hantke and Ulrich Dobrindt for the supply of E. coli strains and to Ulrich Dobrindt for helpful discussions and critical reading of the manuscript. This work was kindly supported by Ardeypharm GmbH. References Altenhoefer, A., Oswald, S., Sonnenborn, U., Enders, C., Schulze, J., Hacker, J., Oelschlaeger, T.A., 2004. The probiotic Escherichia coli strain Nissle 1917 interferes with invasion of human intestinal epithelial cells by different enteroinvasive bacterial pathogens. FEMS Immunol. Med. Microbiol. 40, 223–229. Auvray, F., Dilasser, F., Bibbal, D., Kerouredan, M., Oswald, E., Brugere, H., 2012. French cattle is not a reservoir of the highly virulent enteroaggregative Shiga

8

S.A. Rund et al. / International Journal of Medical Microbiology 303 (2013) 1–8

toxin-producing Escherichia coli of serotype O104:H4. Vet. Microbiol. 158, 443–445. Bär, F., von Koschitzky, H., Roblick, U., Bruch, H.P., Schulze, L., Sonnenborn, U., Böttner, M., Wedel, T., 2009. Cell-free supernatants of Escherichia coli Nissle 1917 modulate human colonic motility: evidence from an in vitro organ bath study. Neurogastroenterol. Motil. 21, 559–566, e516–e557. Bettelheim, K.A., 2007. The non-O157 shiga-toxigenic (verocytotoxigenic) Escherichia coli: under-rated pathogens. Crit. Rev. Microbiol. 33, 67–87. Bielaszewska, M., Idelevich, E.A., Zhang, W., Bauwens, A., Schaumburg, F., Mellmann, A., Peters, G., Karch, H., 2012. Effects of antibiotics on Shiga toxin 2 production and bacteriophage induction by epidemic Escherichia coli O104:H4 strain. Antimicrob. Agents Chemother. 56, 3277–3282. Bielaszewska, M., Mellmann, A., Zhang, W., Kock, R., Fruth, A., Bauwens, A., Peters, G., Karch, H., 2011. Characterisation of the Escherichia coli strain associated with an outbreak of haemolytic uraemic syndrome in Germany, 2011: a microbiological study. Lancet Infect. Dis. 11, 671–676. Buchholz, U., Bernard, H., Werber, D., Böhmer, M.M., Remschmidt, C., Wilking, H., Deleré, Y., an der Heiden, M., Adlhoch, C., Dreesman, J., Ehlers, J., Ethelberg, S., Faber, M., Frank, C., Fricke, G., Greiner, M., Höhle, M., Ivarsson, S., Jark, U., Kirchner, M., Koch, J., Krause, G., Luber, P., Rosner, B., Stark, K., Kühne, M., 2011. German outbreak of Escherichia coli O104:H4 associated with sprouts. N. Engl. J. Med. 365, 1763–1770. Carey, C.M., Kostrzynska, M., Ojha, S., Thompson, S., 2008. The effect of probiotics and organic acids on Shiga-toxin 2 gene expression in enterohemorrhagic Escherichia coli O157:H7. J. Microbiol. Methods 73, 125–132. Eaton, K.A., Honkala, A., Auchtung, T.A., Britton, R.A., 2011. Probiotic Lactobacillus reuteri ameliorates disease due to enterohemorrhagic Escherichia coli in germfree mice. Infect. Immun. 79, 185–191. EFSA, 2011. Tracing seeds, in particular fenugreek (Trigonella foenum-graecum) seeds, in relation to the Shiga toxin-producing E. coli (STEC) O104:H4 2011 outbreaks in Germany and France. Technical Report (European Food Safety Authority). Frank, C., Werber, D., Cramer, J.P., Askar, M., Faber, M., an der Heiden, M., Bernard, H., Fruth, A., Prager, R., Spode, A., Wadl, M., Zoufaly, A., Jordan, S., Kemper, M.J., Follin, P., Müller, L., King, L.A., Rosner, B., Buchholz, U., Stark, K., Krause, G., HUS Investigation Team, 2011. Epidemic profile of Shiga-toxin-producing Escherichia coli O104:H4 outbreak in Germany. N. Engl. J. Med. 365, 1771–1780. Fukuda, S., Toh, H., Hase, K., Oshima, K., Nakanishi, Y., Yoshimura, K., Tobe, T., Clarke, J.M., Topping, D.L., Suzuki, T., Taylor, T.D., Itoh, K., Kikuchi, J., Morita, H., Hattori, M., Ohno, H., 2011. Bifidobacteria can protect from enteropathogenic infection through production of acetate. Nature 469, 543–547. Fukuda, S., Toh, H., Taylor, T.D., Ohno, H., Hattori, M., 2012. Acetate-producing bifidobacteria protect the host from enteropathogenic infection via carbohydrate transporters. Gut Microbes 3, 449–454. Grozdanov, L., Zähringer, U., Blum-Oehler, G., Brade, L., Henne, A., Knirel, Y.A., Schombel, U., Schulze, J., Sonnenborn, U., Gottschalk, G., Hacker, J., Rietschel, E.T., Dobrindt, U., 2002. A single nucleotide exchange in the wzy gene is responsible for the semirough O6 lipopolysaccharide phenotype and serum sensitivity of Escherichia coli strain Nissle 1917. J. Bacteriol. 184, 5912–5925. Harrington, S.M., Dudley, E.G., Nataro, J.P., 2006. Pathogenesis of enteroaggregative Escherichia coli infection. FEMS Microbiol. Lett. 254, 12–18. Henker, J., Laass, M., Blokhin, B.M., Bolbot, Y.K., Maydannik, V.G., Elze, M., Wolff, C., Schulze, J., 2007. The probiotic Escherichia coli strain Nissle 1917 (EcN) stops acute diarrhoea in infants and toddlers. Eur. J. Pediatr. 166, 311–318. Henker, J., Laass, M.W., Blokhin, B.M., Maydannik, V.G., Bolbot, Y.K., Elze, M., Wolff, C., Schreiner, A., Schulze, J., 2008. Probiotic Escherichia coli Nissle 1917 versus placebo for treating diarrhea of greater than 4 days duration in infants and toddlers. Pediatr. Infect. Dis. J. 27, 494–499. Jansen, A., Kielstein, J.T., 2011. The new face of enterohaemorrhagic Escherichia coli infections. Euro Surveill. 16 (25), pii=19898. Karch, H., Denamur, E., Dobrindt, U., Finlay, B.B., Hengge, R., Johannes, L., Ron, E.Z., Tonjum, T., Sansonetti, P.J., Vicente, M., 2012. The enemy within us: lessons from the 2011 European Escherichia coli O104:H4 outbreak. EMBO Mol. Med. 4, 841–848. Kruis, W., 2004. Review article: antibiotics and probiotics in inflammatory bowel disease. Aliment. Pharmacol. Ther. 20 (Suppl. 4), 75–78. Laing, C.R., Zhang, Y., Gilmour, M.W., Allen, V., Johnson, R., Thomas, J.E., Gannon, V.P., 2012. A comparison of Shiga-toxin 2 bacteriophage from classical enterohemorrhagic Escherichia coli serotypes and the German E. coli O104:H4 outbreak strain. PLoS ONE 7, e37362. Leatham, M.P., Banerjee, S., Autieri, S.M., Mercado-Lubo, R., Conway, T., Cohen, P.S., 2009. Precolonized human commensal Escherichia coli strains serve as a barrier to E. coli O157:H7 growth in the streptomycin-treated mouse intestine. Infect. Immun. 77, 2876–2886. Mellmann, A., Harmsen, D., Cummings, C.A., Zentz, E.B., Leopold, S.R., Rico, A., Prior, K., Szczepanowski, R., Ji, Y., Zhang, W., McLaughlin, S.F., Henkhaus, J.K., Leopold, B., Bielaszewska, M., Prager, R., Brzoska, P.M., Moore, R.L., Guenther, S., Rothberg, J.M., Karch, H., 2011. Prospective genomic characterization of the German enterohemorrhagic Escherichia coli O104:H4 outbreak by rapid next generation sequencing technology. PLoS ONE 6, e22751. Möllenbrink, M., Bruckschen, E., 1994. Treatment of chronic constipation with physiological Escherichia coli bacteria. Results of a clinical study of the effectiveness and tolerance of microbiological therapy with the E. coli Nissle 1917 strain (Mutaflor). Med. Klin. (Munich) 89, 587–593 (in German).

Montrose, D.C., Floch, M.H., 2005. Probiotics used in human studies. J. Clin. Gastroenterol. 39, 469–484. Mühldorfer, I., Hacker, J., Keusch, G.T., Acheson, D.W., Tschäpe, H., Kane, A.V., Ritter, A., Oelschläger, T., Donohue-Rolfe, A., 1996. Regulation of the Shiga-like toxin II operon in Escherichia coli. Infect. Immun. 64, 495–502. Muniesa, M., Hammerl, J.A., Hertwig, S., Appel, B., Brussow, H., 2012. Shiga toxinproducing Escherichia coli O104:H4: a new challenge for microbiology. Appl. Environ. Microbiol. 78, 4065–4073. Nissle, 1918. Die antagonistische Behandlung chronischer Darmstörungen mit Kolibakterien. Med. Klein 2, 29–30 (in German). Nougayrede, J.P., Homburg, S., Taieb, F., Boury, M., Brzuszkiewicz, E., Gottschalk, G., Buchrieser, C., Hacker, J., Dobrindt, U., Oswald, E., 2006. Escherichia coli induces DNA double-strand breaks in eukaryotic cells. Science 313, 848–851. Oelschlaeger, T.A., 2010. Mechanisms of probiotic actions – a review. Int. J. Med. Microbiol. 300, 57–62. Ohland, C.L., Macnaughton, W.K., 2010. Probiotic bacteria and intestinal epithelial barrier function. Am. J. Physiol. Gastrointest. Liver Physiol. 298, G807–G819. Okeke, I.N., Wallace-Gadsden, F., Simons, H.R., Matthews, N., Labar, A.S., Hwang, J., Wain, J., 2010. Multi-locus sequence typing of enteroaggregative Escherichia coli isolates from Nigerian children uncovers multiple lineages. PLoS ONE 5, e14093. Pacheco, A.R., Sperandio, V., 2012. Shiga toxin in enterohemorrhagic E. coli: regulation and novel anti-virulence strategies. Front. Cell. Infect. Microbiol. 2, 81. Patzer, S.I., Baquero, M.R., Bravo, D., Moreno, F., Hantke, K., 2003. The colicin G, H and X determinants encode microcins M and H47, which might utilize the catecholate siderophore receptors FepA, Cir, Fiu and IroN. Microbiology 149, 2557–2570. Proulx, F., Turgeon, J.P., Delage, G., Lafleur, L., Chicoine, L., 1992. Randomized, controlled trial of antibiotic therapy for Escherichia coli O157:H7 enteritis. J. Pediatr. 121, 299–303. Reissbrodt, R., Hammes, W.P., dal Bello, F., Prager, R., Fruth, A., Hantke, K., Rakin, A., Starcic-Erjavec, M., Williams, P.H., 2009. Inhibition of growth of Shiga toxinproducing Escherichia coli by nonpathogenic Escherichia coli. FEMS Microbiol. Lett. 290, 62–69. Rembacken, B.J., Snelling, A.M., Hawkey, P.M., Chalmers, D.M., Axon, A.T., 1999. Nonpathogenic Escherichia coli versus mesalazine for the treatment of ulcerative colitis: a randomised trial. Lancet 354, 635–639. RKI, 2011. Abschließende Darstellung und Bewertung der epidemiologischen Erkenntnisse im EHEC O104:H4 Ausbruch (Berlin), Abschlussbericht. Rohde, H., Qin, J., Cui, Y., Li, D., Loman, N.J., Hentschke, M., Chen, W., Pu, F., Peng, Y., Li, J., Xi, F., Li, S., Li, Y., Zhang, Z., Yang, X., Zhao, M., Wang, P., Guan, Y., Cen, Z., Zhao, X., Christner, M., Kobbe, R., Loos, S., Oh, J., Yang, L., Danchin, A., Gao, G.F., Song, Y., Yang, H., Wang, J., Xu, J., Pallen, M.J., Aepfelbacher, M., Yang, R., 2011. Open-source genomic analysis of Shiga-toxin-producing E. coli O104:H4. N. Engl. J. Med. 365, 718–724. Schlee, M., Wehkamp, J., Altenhoefer, A., Oelschlaeger, T.A., Stange, E.F., Fellermann, K., 2007. Induction of human beta-defensin 2 by the probiotic Escherichia coli Nissle 1917 is mediated through flagellin. Infect. Immun. 75, 2399–2407. Schultz, M., 2008. Clinical use of E. coli Nissle 1917 in inflammatory bowel disease. Inflamm. Bowel Dis. 14, 1012–1018. Sherman, P.M., Ossa, J.C., Johnson-Henry, K., 2009. Unraveling mechanisms of action of probiotics. Nutr. Clin. Pract. 24, 10–14. Smith, K.E., Wilker, P.R., Reiter, P.L., Hedican, E.B., Bender, J.B., Hedberg, C.W., 2012. Antibiotic treatment of Escherichia coli O157 infection and the risk of hemolytic uremic syndrome, Minnesota. Pediatr. Infect. Dis. J. 31, 37–41. Sonnenborn, U., Schulze, J., 2009. The non-pathogenic Escherichia coli strain Nissle 1917 – features of a versatile probiotic. Microb. Ecol. Health Dis. 21, 122–158. Strockbine, N.A., Marques, L.R., Newland, J.W., Smith, H.W., Holmes, R.K., O’Brien, A.D., 1986. Two toxin-converting phages from Escherichia coli O157:H7 strain 933 encode antigenically distinct toxins with similar biologic activities. Infect. Immun. 53, 135–140. Tarr, P.I., Gordon, C.A., Chandler, W.L., 2005. Shiga-toxin-producing Escherichia coli and haemolytic uraemic syndrome. Lancet 365, 1073–1086. Valdebenito, M., Crumbliss, A.L., Winkelmann, G., Hantke, K., 2006. Environmental factors influence the production of enterobactin, salmochelin, aerobactin, and yersiniabactin in Escherichia coli strain Nissle 1917. Int. J. Med. Microbiol. 296, 513–520. Vassiliadis, G., Destoumieux-Garzon, D., Lombard, C., Rebuffat, S., Peduzzi, J., 2010. Isolation and characterization of two members of the siderophoremicrocin family, microcins M and H47. Antimicrob. Agents Chemother. 54, 288–297. Wehkamp, J., Harder, J., Wehkamp, K., Wehkamp-von Meissner, B., Schlee, M., Enders, C., Sonnenborn, U., Nuding, S., Bengmark, S., Fellermann, K., Schröder, J.M., Stange, E.F., 2004. NF-kappaB- and AP-1-mediated induction of human beta defensin-2 in intestinal epithelial cells by Escherichia coli Nissle 1917: a novel effect of a probiotic bacterium. Infect. Immun. 72, 5750–5758. Wieler, L.H., Semmler, T., Eichhorn, I., Antao, E.M., Kinnemann, B., Geue, L., Karch, H., Guenther, S., Bethe, A., 2011. No evidence of the Shiga toxin-producing E. coli O104:H4 outbreak strain or enteroaggregative E. coli (EAEC) found in cattle faeces in northern Germany, the hotspot of the 2011 HUS outbreak area. Gut Pathog. 3, 17. Wong, C.S., Jelacic, S., Habeeb, R.L., Watkins, S.L., Tarr, P.I., 2000. The risk of the hemolytic-uremic syndrome after antibiotic treatment of Escherichia coli O157:H7 infections. N. Engl. J. Med. 342, 1930–1936.