• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of iaiPermissionsJournals.ASM.orgJournalIAI ArticleJournal InfoAuthorsReviewers
Infect Immun. Apr 2008; 76(4): 1340–1348.
Published online Jan 28, 2008. doi:  10.1128/IAI.00778-07
PMCID: PMC2292865

Lactobacillus rhamnosus Strain GG Prevents Enterohemorrhagic Escherichia coli O157:H7-Induced Changes in Epithelial Barrier Function[down-pointing small open triangle]

Abstract

Enterohemorrhagic Escherichia coli (EHEC) O157:H7 intimately attaches to intestinal epithelial monolayers and produces attaching and effacing (A/E) lesions. In addition, EHEC infection causes disruptions of intercellular tight junctions, leading to clinical sequelae that include acute diarrhea, hemorrhagic colitis, and the hemolytic-uremic syndrome. Current therapy remains supportive since antibiotic therapy increases the risk of systemic complications. This study focused on the potential therapeutic effect of an alternative form of therapy, probiotic Lactobacillus rhamnosus strain GG, to attenuate EHEC-induced changes in paracellular permeability in polarized MDCK-I and T84 epithelial cell monolayers. Changes in epithelial cell morphology, electrical resistance, dextran permeability, and distribution and expression of claudin-1 and ZO-1 were assessed using phase-contrast, immunofluorescence, and transmission electron microscopy and macromolecular flux. This study demonstrated that pretreatment of polarized MDCK-I and T84 cells with the probiotic L. rhamnosus GG reduced morphological changes and diminished the number of A/E lesions induced in response to EHEC O157:H7 infection. With probiotic pretreatment there was corresponding attenuation of the EHEC-induced drop in electrical resistance and the increase in barrier permeability assays. In addition, L. rhamnosus GG protected epithelial monolayers against EHEC-induced redistribution of the claudin-1 and ZO-1 tight junction proteins. In contrast to the effects seen with the live probiotic, heat-inactivated L. rhamnosus GG had no effect on EHEC binding and A/E lesion formation or on disruption of the barrier function. Collectively, these findings provide in vitro evidence that treatment with the probiotic L. rhamnosus strain GG could prove to be an effective management treatment for preventing injury of the epithelial cell barrier induced by A/E bacterial enteropathogens.

Enterohemorrhagic Escherichia coli (EHEC) O157:H7 infection is a common cause of intestinal disease in humans (43). Intestinal infections are known to mediate multiple changes in host cellular signal transduction responses, which impair epithelial transport and the barrier function (9). The current treatment remains predominantly supportive because antibiotics appear to increase the risk of systemic complications (e.g., the hemolytic-uremic syndrome), perhaps by releasing Shiga-like toxins from the bacterial cell periplasm into the gut lumen (43).

The intestinal tract is comprised of a mucosal epithelial cell barrier, which is critical in providing the first line of defense against external insults (24). Tight junctions represent the luminal-most portion of a broader “apical junction complex” that circumscribes polarized epithelial cells (44). Tight junction domains are comprised of integral membrane proteins, including the junction adhesion molecule, occludin, and members of the claudin superfamily, which differentially modulate the transport of ions and macromolecules and prevent toxin, antigen, and microflora interactions with subepithelial tissues (28, 44). The primary purpose of claudins is to regulate the paracellular flow of ions and small molecules mediated through extracellular loops that protrude into the spaces between adjacent cells (44). These integral proteins are connected to cytoplasmic plaque proteins, including ZO-1, which anchor tight junctions to the F-actin cytoskeleton for mechanical regulation (28). Disruption of epithelial tight junctions due to E. coli O157:H7 infection depends on bacterial attachment and the deployment of a type III secretion system (encoded by the locus of enterocyte effacement), which acts as a molecular syringe to transmit bacterial effector proteins into the cytoplasm of host cells (32, 45). As a result, cascades of signal transduction events cause F-actin pedestal formation (mediated by actin bridging proteins, including α-actinin), the loss of microvilli, and attaching and effacing (A/E) lesions (22).

Probiotics, including Lactobacillus species, are defined as live microorganisms which have beneficial effects for the host (8). Increasing evidence indicates that probiotics are effective in preventing and treating a variety of intestinal disorders, including pouchitis, traveler's diarrhea, antibiotic-associated diarrhea, and acute infectious diarrhea (33). Probiotics exert their beneficial effects by a variety of complementary mechanisms, including an ability to modulate host innate immune responses, such as secretion of mucins from goblet cells, adaptive immune responses, protection from the damaging effects of aspirin, and competitive exclusion of pathogens (36). In addition, we have previously shown that probiotics have the ability to decrease Citrobacter rodentium-induced mucosal inflammation and disruption of paracellular junctions (20).

Lactobacillus rhamnosus strain GG is a gram-positive, lactic acid-producing bacterium that was first isolated from the stools of a healthy human (7, 10). L. rhamnosus GG has been employed experimentally as a supplement to the human neonatal intestinal microflora and has been studied to determine its effects on the enhancement of immunoglobulin (Ig) secretion. This probiotic prevents rotavirus-induced diarrhea, Salmonella infection in vitro, and the recurrence of colitis (7, 10), and it protects against indomethacin-induced changes in barrier function both in humans and in polarized intestinal cell monolayers grown in tissue culture (11). A notable characteristic of this bacterium is its ability to adhere to epithelial cells in tissue culture and displace intestinal pathogens, including E. coli (23).

The maintenance of the cytoskeleton and tight junction integrity of polarized epithelial monolayers plays an integral role in modulating paracellular diffusion. This epithelial barrier function can be weakened by infection with bacteria, including pathogenic E. coli strains (39). Therefore, the aim of this study was to determine the potential ability of L. rhamnosus GG to protect the epithelial cell barrier in response to challenge with the enteric pathogen E. coli O157:H7 and to delineate the mechanistic aspects by which L. rhamnosus GG exerts its effects.

[Part of this research was presented at the 6th International Symposium on Shiga Toxin (Verocytotoxin)-Producing Escherichia coli Infections (VTEC 2006), Melbourne, Australia.]

MATERIALS AND METHODS

Bacterial strains, cell cultures, and growth conditions.

L. rhamnosus strain GG (ATCC 53103) was purchased frozen from the American Type Culture Collection (Manassas, VA), was cultured on blood agar (Difco, Detroit, MI), and was grown in de Mann-Rogosa-Sharpe (MRS) broth (Difco) aerobically at 37°C for 24 h. E. coli O157:H7 strain CL56 was grown overnight at 37°C in static, nonaerated Penassay broth (Difco, Detroit, MI).

As previously described, Madin-Darby canine kidney (MDCK-I) and T84 epithelial cells were used as model epithelia to study the dynamics of barrier function due to their ability to form polarized monolayers with high resistance (47). MDCK-I epithelial cells were kindly provided by Roger Worrell (The Vontz Center for Molecular Studies, University of Cincinnati, Cincinnati, OH), and were cultured until they were confluent in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (Invitrogen) and 2% penicillin-streptomycin (Invitrogen), as previously described (47). T84 epithelial cells were purchased from the American Type Culture Collection and grown in tissue culture flasks using established methods (37). Briefly, T84 cells were cultured using DMEM and Ham's F-12 medium (Invitrogen Canada Inc., Burlington, Canada) at a 1:1 ratio supplemented with 10% fetal bovine serum (Invitrogen), 0.6% glutamine, 1.9% sodium bicarbonate, and 2% penicillin-streptomycin (Invitrogen) until they were confluent. The cell culture medium was changed to fresh medium without antibiotics prior to treatment of the cells with bacteria.

Cells were grown in 25-cm2 flasks (Corning, Corning, NY) until they were confluent or in Lab-Tek chamber slides (VWR International Ltd., Mississauga, Canada). In addition, T84 and MDCK-I polarized monolayers were grown on 6.5- or 12-mm-diameter Transwells (Corning) at 37°C in the presence of 5% CO2 until the transepithelial electrical resistance (TER), as measured with a Millicell ERS voltmeter (Millipore, Bedford, MA), was greater than 1,000 Ω·cm2.

Effects of viable probiotics and their products on pathogen viability.

The effects of probiotics on the viability of the pathogen were assessed using coincubation methods, as previously described (20). Briefly, L. rhamnosus GG (108 CFU/ml) was coincubated at a ratio of 1:1 (vol/vol) with E. coli O157:H7 (107 CFU/ml) in Penassay broth (10 ml) for 3 and 18 h at 37°C. Serial dilutions were prepared, and 0.1-ml aliquots of 1:100,000 dilutions were spread onto MacConkey agar plates (PML Microbiologicals, Mississauga, Ontario, Canada) and incubated overnight at 37°C. Colonies were then counted, and the results were expressed as absolute numbers of CFU.

To determine if secreted products affected pathogen viability, culture supernatants were employed. Culture supernatants were prepared as previously described (20). Briefly, supernatants resulting from 4 h of growth of L. rhamnosus GG in MRS broth were centrifuged at 1,600 × g for 5 min and filtered twice through a 0.2-μm filter (Millipore). E. coli O157:H7 (108 CFU/ml) was incubated in an equal volume of L. rhamnosus GG culture supernatant for 18 h at 37°C and then plated onto MacConkey agar plates. Following overnight incubation at 37°C, visible bacteria were enumerated to determine the number of viable CFU.

Transmission electron microscopy.

To assess epithelial cell monolayer structures, polarized epithelial cells were prepared as previously described, with minor modifications (20). Briefly, MDCK-I cells were grown on 60-mm plates at 37°C in the presence of 5% CO2 to confluence. Confluent polarized monolayers were either infected with EHEC O157:H7 (107 CFU/ml) for 4 h, treated with L. rhamnosus GG (108 CFU/ml) for 4 h, pretreated for 1 h with L. rhamnosus GG prior to EHEC infection (3 h), or left untreated. Epithelial cell monolayers were then washed twice with phosphate-buffered saline (PBS). Fixative (2.5% glutaraldehyde in phosphate buffer) was then added to the cells and incubated at room temperature for 15 min. Cells were then scraped from the dishes and centrifuged at 41 × g for 10 min. The cells were subsequently fixed in osmium tetroxide for 1 h, dehydrated in a graded acetone series (50 to 100% acetone), and then embedded in epoxy resin. Ultrathin sections (80 nm) were cut with a Reichert Ultracat E (Leica Inc., Richmond Hill, Ontario, Canada). Samples were viewed with a JEM-1230 (JOEL USA Corp., Peabody, MA) transmission electron microscope operated at 80 kV. Digital images were acquired with a charge-coupled device camera (AMT Advantage HR camera system; AMT, Massachusetts) attached to the electron microscope.

Detection of E. coli O157:H7 adhesion and A/E lesions by alternating phase-contrast and immunofluorescence microscopy.

As described previously (15, 17), indirect immunofluorescence using a murine monoclonal antibody against the F-actin bridging protein α-actinin was employed to detect E. coli O157:H7-induced A/E lesions. Briefly, T84 and MDCK-I cells grown overnight on chamber slides in the presence of 5% CO2 at 37°C were washed with sterile PBS. Cells were then either infected with E. coli O157:H7 (107 CFU/ml) (3 h), treated with viable or heat-killed (100°C for 1 h) L. rhamnosus GG (108 CFU/ml) (3 h), pretreated with viable L. rhamnosus GG or heat-killed L. rhamnosus GG for 1 h at 37°C prior to EHEC infection (3 h), or left untreated. Cells were fixed in 100% cold methanol for 10 min (Caledon Laboratories, Georgetown, Ontario, Canada), and slides were examined by alternating phase-contrast and immunofluorescence microscopy at a magnification of ×40 (Leitz Dialuz 22; Leica Canada Inc., Willowdale, Ontario, Canada). A/E lesions were quantified, and the results were expressed as the number of α-actinin foci per cell. A total of ≥100 cells were counted in each experiment.

TER and dextran permeability as measurements of changes in the barrier function of polarized epithelial cell monolayers.

As previously described (18), polarized MDCK-I and T84 cells were grown on 6.5-mm (pore size, 0.4 μm) or 12-mm (pore size, 0.4 μm) Transwells (Corning) and cultured until the TER reached a minimum of 1,000 Ω·cm2. Cells were infected with EHEC (107 CFU/ml), treated with L. rhamnosus GG (108 CFU/ml), or pretreated with L. rhamnosus GG for 3 h prior to pathogenic infection. Two complementary assays were used to measure the epithelial barrier function.

First, TER was employed as a marker of intercellular tight junction integrity, because it provides an electrical measurement of barrier function toward passive ion flow and the measurements are inversely related to the permeability of a polarized epithelium to macromolecules, such as mannitol and 51Cr-labeled EDTA (3). TER was measured after 18 h of incubation in the presence of 5% CO2 at 37°C with a Millicell probe (Millipore Corporation, Bedford, MA), and changes were expressed as percentages of untreated control measurements.

Second, the movement of macromolecules across polarized epithelial cell monolayers was assayed using a macromolecular conjugate probe, Alexa Fluor 647 dextran (10 kDa; Molecular Probes, Eugene, OR) (1). Briefly, 0.2 ml of conjugated dextran suspended in DMEM (Invitrogen) was added to the apical compartment of Transwells, and 0.4 ml of DMEM alone added to the basolateral compartment. After incubation for 5 h at 37°C, samples (0.5 ml) from the basolateral compartment were placed into a 96-well plate (Corning) and analyzed to determine their fluorescent intensity using the Odyssey infrared imaging system (LI-COR Biosciences, Lincoln, NE) at a wavelength of 700 nm. Integrated intensities were expressed relative to the integrated intensity of medium samples from untreated controls.

Analysis of ZO-1 and claudin-1 distribution by confocal microscopy.

MDCK-I cells were seeded into 6.5-mm Transwells and infected under the experimental conditions used for TER experiments, as described above. The immunofluorescence staining protocol was adapted from the protocol of Zareie et al. (47). Briefly, MDCK-I cell monolayers were rinsed in PBS, and this was followed by fixation and permeabilization in 100% methanol (−20°C) (Caledon Laboratories) for 10 min. Cells were incubated in 5% (vol/vol) normal goat serum (Jackson Immunoresearch, West Grove, PA) in PBS for 1 h at room temperature and then incubated with primary rabbit anti-ZO-1 or rabbit anti-claudin-1 (Zymed, San Francisco, CA) for 1 h at 37°C. After unbound primary antibodies were rinsed away with PBS, cells were incubated with secondary Cy2-conjugated goat anti-rabbit IgG (1:200 dilution; Zymed) for 1 h at room temperature. Host cell nuclei were then counterstained with 300 nM 4′,6-diamidino-2-phenylindole dilactate (DAPI) (Molecular Probes) in PBS for 5 min. Monolayers were thoroughly rinsed with PBS, excised from the wells, mounted on slides, and examined with a confocal laser scanning microscope at 1-μm intervals (Zeiss LSM510; Zeiss, Frankfurt, Germany).

Western blotting for tight junction proteins. (i) Protein extraction.

Western blotting for tight junction proteins was performed as previously described (47) using MDCK-I whole-cell protein extracts. In addition, MDCK-I cells grown on 6.5-mm Transwells were processed for protein extraction using a Triton X-100-containing lysis buffer, as previously described (26), to separate detergent-soluble protein fractions from detergent-insoluble (cytoskeletal) protein fractions. The initial studies were performed with 12-mm Transwells to increase the protein yield. There were substantial functional differences between cells grown on 6.5-mm Transwells and cells grown on 12-mm Transwells; therefore, subsequent experiments were carried out using 6.5-mm Transwells. To ensure equal protein loading for electrophoresis, protein concentrations were determined using the standard Bradford assay (Bio-Rad Laboratories, Hercules, CA).

(ii) Electrophoresis and immunoblotting.

As described by Zareie et al. (47), samples were run on a 12% polyacrylamide Tris-HCl gel (Ready Gel; Bio-Rad Laboratories, Hercules, CA) at 175 V for 45 min (for claudin-1) or on a 7.5% polyacrylamide gel at 111 V for 80 min (for ZO-1), transferred onto a nitrocellulose membrane (BioTrace NT; Pall Corporation, Ann Arbor, MI), blocked for 1 h with Odyssey blocking buffer (LI-COR Biosciences), and probed overnight with either rabbit anti-claudin-1 (1:1,000) or rabbit anti-ZO-1 (1:2,500) together with mouse anti-β-actin (1:5,000; Sigma, St. Louis, MO). Membranes were then probed with IRDye 800 goat anti-rabbit IgG (1:10,000; Rockland Immunochemicals, Gilbertsville, PA) and Alexa Fluor 680 goat anti-mouse IgG (1:20,000; Molecular Probes). Blots were imaged using an infrared imaging system (Odyssey) using both 700- and 800-nm channels at 169-μm resolution. Quantification was performed with commercial analysis software provided by LI-COR. Western blot intensity measurements for whole-cell proteins were determined from the ratio of the integrated intensity of the claudin-1 or ZO-1 band to the integrated intensity of the β-actin band in the same sample. The resulting ratios were normalized using untreated samples included in each experiment and were expressed as percentages (47). For Triton X-100 fractionation extraction, a ratio of the integrated intensity of the claudin-1 band in the soluble fraction to the total claudin-1 band intensity (sum of the signals from soluble and insoluble fractions) was calculated.

Statistics.

The results were expressed as means ± standard errors of the means. Statistical differences between multiple groups were calculated using analysis of variance (ANOVA). To examine differences between two experimental groups, the unpaired Student t test was employed. A P value of <0.05 was considered statistically significant.

RESULTS

Coincubation of E. coli O157:H7 with viable L. rhamnosus GG or culture supernatants does not affect the growth of the pathogen.

There was no difference in growth when E. coli O157:H7 was grown alone in Penassay broth (147 ± 91.1 CFU) and when the pathogen was coincubated with the probiotic (183.6 ± 156.2 CFU) (P = 0.85) for 18 h (comparable results were obtained with a coincubation time of 3 h [data not shown]). Similarly, incubation of E. coli O157:H7 in L. rhamnosus GG culture supernatant did not affect the growth of the pathogen (97.5 ± 92.5 CFU, compared to 66.0 ± 32 CFU for E. coli O157:H7 grown in PBS alone [P = 0.78]).

L. rhamnosus GG pretreatment prevents E. coli O157:H7-induced ultrastructural changes in polarized epithelial monolayers.

Transmission electron microscopy of untreated MDCK-I epithelial cells revealed normal cell morphology architecture with intact microvilli and nuclei. In addition, electron-dense tight junctions demonstrated that there were intact intercellular membrane appositions (Fig. (Fig.1a).1a). Comparable ultrastructural architecture was seen in epithelial cell monolayers treated with L. rhamnosus GG alone (Fig. (Fig.1b).1b). In contrast, epithelial cells infected with E. coli O157:H7 (Fig. (Fig.1c)1c) had intracellular vacuoles, and there were distinct gaps in the continuity of intercellular membrane contacts. Probiotic pretreatment prior to EHEC infection (Fig. (Fig.1d)1d) preserved the intercellular contacts but did not prevent cytoplasmic vacuolization.

FIG. 1.
L. rhamnosus GG inhibits E. coli O157:H7-induced ultrastructural changes of host cells. (a) Transmission electron photomicrograph of untreated MDCK-I cells showing intact microvilli, normal cellular morphology, and tight junctions (dashed arrows). (b) ...

L. rhamnosus GG prevents E. coli O157:H7-induced rearrangements of the epithelial cell cytoskeleton.

Phase-contrast and indirect immunofluorescence imaging of MDCK-I and T84 cells infected for 3 h with E. coli O157:H7 (107 CFU/ml) revealed pathogen binding (Fig. (Fig.2a)2a) and bacterium-induced A/E lesions, as indicated by the formation of α-actinin foci (Fig. (Fig.2b).2b). Adhesion of L. rhamnosus GG (108 CFU/ml) to MDCK-I cells did not result in the formation of α-actinin foci (data not shown). Preincubation of the epithelial cells for 1 h at 37°C with L. rhamnosus GG prior to E. coli O157:H7 infection inhibited pathogen attachment (Fig. (Fig.2c)2c) and attenuated bacterium-induced A/E lesion formation (Fig. (Fig.2d).2d). By contrast, preincubation of epithelial cells with heat-killed L. rhamnosus GG prior to EHEC infection did not prevent pathogen attachment and A/E lesion formation (data not shown).

FIG. 2.
L. rhamnosus GG inhibits E. coli O157:H7-induced rearrangements of α-actinin in MDCK-I and T84 epithelial cells. (a) Phase-contrast photomicrographs showing binding of E. coli O157:H7 (arrows) to epithelial cells following infection for 3 h at ...

Semiquantitative analysis of the number of A/E foci in MDCK-I cells pretreated with probiotics prior to E. coli O157:H7 showed that there was a decrease in A/E lesions (1.8 ± 1.1 foci/cell) compared with cells infected with the pathogen alone (7.5 ± 1.9 foci/cell) (P = 0.01). Probiotic-pretreated T84 cells also had decreased numbers of E. coli O157:H7-induced A/E lesions (2.6 ± 0.8 foci/cell with EHEC infection alone and 0.2 ± 0.6 foci/cell with L. rhamnosus GG pretreatment and EHEC infection; P = 0.04). By contrast, there was no decrease in the number of foci of α-actinin when heat-killed L. rhamnosus GG was added to MDCK-I or T84 cell monolayers for 1 h at 37°C prior to challenge with the bacterial pathogen (for MDCK-I, 6.6 ± 2.3 foci/cell; and for T84, 7.4 ± 3.0 foci/cell).

L. rhamnosus GG attenuates E. coli O157:H7-induced decreases in TER.

Two complementary polarized epithelial cell lines (MDCK-I and T84) were used to assess barrier function in response to EHEC infection in the absence or presence of probiotics. Treatment of polarized epithelial cells with L. rhamnosus GG alone did not alter the TER of polarized T84 cell monolayers (which was 92.1% ± 5.9% of the TER of uninfected control monolayers; n = 12) (Fig. (Fig.3a)3a) and MDCK-I cell monolayers (95.5% ± 4.5%; n = 5) (Fig. (Fig.3b).3b). Similarly, heat-killed probiotic L. rhamnosus GG did not alter the TER of T84 cell monolayers (96.8% ± 0.8%; n = 5) or MDCK-I cell monolayers (100%; n = 3). As observed previously (47), E. coli O157:H7 infection resulted in reduced TER of both T84 cells (19.0% ± 3.0%; n = 12) and MDCK-I cells (27.9% ± 11.0%; n = 5) cells. Pretreatment of polarized monolayers with L. rhamnosus GG for 3 h prior to infection attenuated the EHEC-induced decrease in TER (T84 cells, 59.21% ± 9.6% [n = 12]; MDCK-I cells, 86.3% ± 8.1% [n = 5]). By contrast, amelioration of the reduced TER in response to EHEC infection was not observed when T84 epithelial cell monolayers (18.37% ± 4.4% of the control value; n = 5) and MDCK-I epithelial cell monolayers (10.5% ± 3.3%; n = 3) were pretreated with heat-killed L. rhamnosus GG.

FIG. 3.
L. rhamnosus GG attenuates E. coli O157:H7-induced decreases in TER in T84 and MDCK-I cells. Pretreatment of polarized T84 (a) and MDCK-I (b) monolayers with L. rhamnosus GG (108 CFU in 0.2 ml) for 3 h prior to infection with E. coli O157:H7 (107 CFU ...

Increases in macromolecular permeability of T84 and MDCK-I cells in response to EHEC infection are attenuated by pretreatment with L. rhamnosus GG.

Macromolecular permeability assays with T84 cell monolayers using an infrared-sensitive dextran (10-kDa) probe (Fig. (Fig.4a)4a) demonstrated that L. rhamnosus GG treatment alone did not increase the diffusion of the probe (as measured by the signal intensity for basal medium samples) from apical to basolateral Transwell compartments (relative integrated intensity [RI] compared to untreated monolayers, 1.4 ± 0.4; P > 0.05; n = 4). In addition, E. coli O157:H7-infected monolayers did not exhibit a marked increase in the permeability to the dextran probe (RI, 3.9 ± 2.4; P > 0.05; n = 4). Likewise, pretreatment of T84 cell monolayers with viable probiotic for 3 h prior to pathogen infection did not affect the permeability compared to that of untreated cells (RI, 1.5 ± 0.4; P > 0.05; n = 4), nor did pretreatment of T84 cell monolayers with heat-killed probiotic (RI = 2.96 ± 1.2, P > 0.05; n = 4).

FIG. 4.
L. rhamnosus GG decreases EHEC-induced increases in macromolecular permeability. Integrated intensities of dextran diffused into the basolateral compartments of Transwells containing T84 cells (a) and MDCK-I cells (b) 5 h after addition to the apical ...

Pathogen-induced increases in the dextran permeability of MDCK-I cell monolayers (RI compared to uninfected controls, 11.9 ± 2.6; P < 0.01; n = 3) (Fig. (Fig.4b)4b) were reduced when epithelial cells were also pretreated with L. rhamnosus GG (RI, 3.42 ± 2.1, P < 0.05; n = 3). Attenuation of increased macromolecular permeability was not observed when monolayers were pretreated with heat-killed L. rhamnosus GG (RI, 7.35 ± 3.5, P > 0.05; n = 3). Treatment of MDCK-I cell monolayers with viable L. rhamnosus GG alone did not affect the dextran permeability compared to that of untreated cells (RI, 1.0 ± 0.1; P > 0.05; n = 3), nor did treatment with heat-killed L. rhamnosus GG alone (RI, 0.9 ± 0.03; P > 0.05; n = 3).

L. rhamnosus GG prevents pathogen-induced redistribution or expression of ZO-1 and claudin-1.

Untreated MDCK-I cells and monolayers treated with L. rhamnosus GG alone had intact tight junctions, as demonstrated by continuous and circumferential ZO-1 distribution visualized by confocal immunofluorescence microscopy (Fig. 5a and c, respectively). In contrast, fragmented ZO-1 staining was observed in cells infected with E. coli O157:H7 alone (Fig. (Fig.5b).5b). Pretreatment with L. rhamnosus GG for 3 h prior to EHEC infection prevented these changes (Fig. (Fig.5d5d).

FIG. 5.
L. rhamnosus GG prevents ZO-1 redistribution and decreased protein expression in E. coli O157:H7-infected polarized monolayers. (a) Polarized untreated MDCK-I cell monolayer labeled with rabbit anti-ZO-1 (green), showing continuous, intact circumferential ...

Corresponding z-line stacks (xz plane) of the en face images (Fig. (Fig.5e)5e) demonstrated that there was apically polarized distribution of ZO-1 in untreated cells and monolayers treated with L. rhamnosus GG alone. In contrast, punctuate ZO-1 staining was scattered throughout the depth of the cell monolayers infected with O157:H7 alone, indicating a loss of polarization (29). Pretreatment of MDCK-I cells with L. rhamnosus GG prior to EHEC infection resulted in prevention of EHEC-induced disruption of epithelial cell polarization.

Western blotting of epithelial whole-cell protein extracts (Fig. (Fig.5f)5f) showed that there was a reduction in ZO-1 expression (doublet seen in the 220-kDa region of control samples) in E. coli O157:H7-infected cells (band intensity, 0.65 ± 0.13; P < 0.05, as determined by ANOVA,; n = 4) compared to the expression in untreated monolayers (band intensity, 1). Treatment of MDCK-I cells with probiotics alone maintained the intensity at levels similar to that of the control (band intensity, 0.85 ± 0.14). Epithelial cell monolayers treated with probiotics prior to E. coli O157:H7 infection maintained higher levels of ZO-1 expression than monolayers infected with the pathogen alone (band intensity, 0.86 ± 0.09; P < 0.05, as determined by ANOVA; n = 4). Western blots of Triton X-100-soluble and Triton X-100-insoluble fractionated protein extracts from untreated, EHEC O157:H7-infected, L. rhamnosus GG-treated, and L. rhamnosus GG-pretreated EHEC O157:H7-infected cells demonstrated that ZO-1 was found only in the Triton X-100-insoluble fractions and not in the soluble fractions (n = 4) (Fig. (Fig.5g5g).

The claudin-1 distribution in untreated MDCK-I cells (Fig. (Fig.6a)6a) and probiotic-treated cells (Fig. (Fig.6c)6c) showed that there was continuous protein distribution circumscribing the epithelial cells, compared to the more diffuse and fragmented pattern of staining in cells infected with EHEC O157:H7 (Fig. (Fig.6b).6b). Pretreatment of MDCK-I cells with L. rhamnosus GG prior to pathogen infection prevented the changes in claudin-1 redistribution caused by E. coli O157:H7 (Fig. (Fig.6d).6d). Claudin-1 polarization was disrupted when eukaryotic cells were infected with E. coli O157:H7 alone, but these effects were ameliorated by L. rhamnosus GG pretreatment (Fig. (Fig.6e).6e). As shown in Fig. Fig.6f,6f, Western blotting to determine levels of whole-cell claudin-1 expression demonstrated that there was no change in monolayers under any of the experimental conditions (control band intensity, 1.00; band intensity with E. coli O157:H7 alone, 0.78 ± 0.01; band intensity with L. rhamnosus alone, 0.70 ± 0.03; band intensity for cells pretreated with probiotic prior to pathogen infection, 0.72 ± 0.03; P = 0.10, as determined by ANOVA; n = 4). Similarly, Western blots (Fig. (Fig.6g)6g) of Triton X-100-soluble and Triton X-100-insoluble protein extracts from E. coli O157:H7-infected MDCK-I cells did not show that there was a statistically significant shift in the distribution of claudin-1 from the soluble fraction to the insoluble fraction compared to untreated cells, monolayers treated with probiotics alone, or cells pretreated with probiotics prior to E. coli O157:H7 infection (the ratios of the soluble fraction band intensity to the sum of the band intensities for both the soluble and insoluble fractions were 0.78 ± 0.05 for uninfected cells, 0.55 ± 0.06 for EHEC-infected cells, 0.70 ± 0.05 for L. rhamnosus GG-treated cells, and 0.66 ± 0.06 for L. rhamnosus GG-treated cells prior to EHEC infection; P = 0.10, as determined by ANOVA; n = 3).

FIG. 6.
L. rhamnosus GG prevents claudin-1 redistribution in EHEC-infected epithelial monolayers. (a) Untreated MDCK-I cell monolayer labeled with rabbit anti-claudin-1 (green), showing intact intercellular tight junction distribution. (b) E. coli O157:H7-infected ...

DISCUSSION

Although many clinical studies have reported that probiotics, such as L. rhamnosus GG, have beneficial health effects (11, 13, 27, 33), it is still difficult to ascertain their direct mechanism(s) of action. Therefore, the current trend in research in this field is to determine the mechanisms by which particular strains of probiotic bacteria are efficacious in treating specific gut abnormalities or protect against defined microbial infections (35).

We have previously shown that in vivo use of a probiotic mixture containing L. rhamnosus R0011 and Lactobacillus helveticus R0052 in mice is beneficial in ameliorating intestinal injury in response to C. rodentium (19, 20), gastric injury in response to Helicobacter pylori infection (21), and stress-induced barrier dysfunction in epithelial ultrastructure in the colon of stressed rats (48). The response to a nonspecific injury, such as stress, leads to defective mucosal barrier function, enhanced luminal bacterial adherence, and alterations in cellular morphology (40, 41). Other investigators have demonstrated that probiotics, such as Lactobacillus casei strain DN-114 001, preserve barrier function in polarized T84 and Caco-2 cells infected with adherent-invasive E. coli and enteropathogenic E. coli strain E2348/69 (14, 31). Isolauri et al. previously demonstrated that suckling rats treated with L. casei GG have fewer gastrointestinal permeability disorders induced by cow's milk (16). Interestingly, Yan et al. (46) demonstrated that secreted proteins of L. rhamnosus GG prevent tumor necrosis factor alpha-induced apoptosis in human and mouse epithelial cells. Taken together, these studies support the protective effect that probiotics have on the host against pathogenic bacteria.

Probiotics are reported to exert their beneficial effects by producing bacteriostatic or bactericidal agents (6, 42), competitively excluding pathogenic bacteria (37), or regulating immunomodulatory effects (25, 34). Silva et al. (38) previously described the ability of L. rhamnosus GG to exert bactericidal effects against a variety of pathogens, such as Pseudomonas, Salmonella, Clostridium, and E. coli B-44. However, in the present study, such effects were not observed when pathogenic E. coli O157:H7 was incubated in the presence of either viable probiotic L. rhamnosus GG culture supernatants or L. rhamnosus GG-conditioned medium. These findings emphasize differences in individual probiotic species and indicate that the beneficial effects of L. rhamnosus GG observed in this study were likely not due to a reduction in the absolute number of pathogenic bacteria present.

This study broadens our current understanding of how probiotics exert their beneficial effects and emphasizes the ability of L. rhamnosus GG (= ATCC 53101) to protect polarized epithelial cells against the effects of E. coli O157:H7-induced changes in barrier function at several different levels. Using high-resistance in vitro models of barrier function, we demonstrated that exposure of polarized epithelial cells to the viable probiotic before infection with E. coli O157:H7 attenuated pathogen-induced alterations in epithelial barrier function. Differences in the effectiveness of L. rhamnosus GG pretreatment for the two cell lines used could be attributed to differences in cell physiology with respect to the absorptive (MDCK-I) and secretory (T84) phenotypes (5, 12).

Similar to previous studies (45, 47), this study demonstrated that EHEC O157:H7 disrupts epithelial tight junction structure, including ZO-1 and claudin-1 distribution, in both MDCK-I and T84 tissue culture cells, resulting in decreased TER and increased permeability to macromolecules. Infection models used by other investigators demonstrated that both probiotic mixtures (such as VSL#3) and additional single strains (e.g., E. coli Nissle 1917 and L. casei DN-114 001) prevent ZO-1 redistribution in response to Salmonella enterica serovar Dublin and enteropathogenic E. coli infections in vitro (30, 31). In our study, L. rhamnosus GG pretreatment prior to E. coli O157:H7 infection ameliorated the pathogen-induced redistribution of ZO-1. We also demonstrated, for the first time, using immunofluorescence microscopy, that L. rhamnosus GG pretreatment stabilizes cellular tight junctions, thereby preventing E. coli O157:H7-induced redistribution of the integral tight junction protein claudin-1.

To support microscopy observations, we also employed Western blotting techniques to determine levels of ZO-1 and claudin-1. The infrared-based detection system yielded greater resolution than conventional immunoblotting, and there was a doublet in the 220-kDa range. Previous studies demonstrated that this doublet corresponds to α+ and α isoforms (the amounts of the isoforms may differ in various cell lines) (2). Despite observing a redistribution of ZO-1 by immunofluorescence microscopy, we did not observe a change in the characteristics of detergent solubility (an indicator of protein localization between detergent-insoluble regions of the cell membrane and the rest of the cell [4]) in the control, probiotic-treated, or infected cells. Although this study shows that the immunofluorescence staining pattern of claudin-1 was redistributed in E. coli O157:H7-infected epithelia, a statistically significant change in the detergent solubility characteristics of this protein was not observed. Therefore, visible changes in tight junction morphology, as observed by microscopy, does not necessarily indicate a change in detergent solubility characteristics. This conclusion is supported by findings of Boyle et al. (4), who noted that tight junction-associated proteins can be mobilized within a particular fraction rather than between detergent-soluble and -insoluble fractions (4). E. coli O157:H7 decreases whole-cell expression of ZO-1 and thereby contributes to decreased barrier function (47). The present study demonstrated that this effect can be ameliorated when L. rhamnosus GG is added to a polarized monolayer prior to E. coli O157:H7 infection.

Taken together, the results of this study indicate that L. rhamnosus GG has the ability to protect against E. coli O157:H7-induced damage of the epithelial monolayer barrier function by preventing changes in host cell morphology, A/E lesion formation, monolayer resistance, and macromolecular permeability. In addition, L. rhamnosus GG pretreatment prevents E. coli O157:H7-induced morphological redistribution of intercellular tight junction proteins and a decrease in the expression of ZO-1.

We expanded findings of previous investigators by demonstrating that L. rhamnosus GG pretreatment interrupts the infectious processes of E. coli O157:H7, without bactericidal activity. By demonstrating the mode of action of this probiotic strain in attenuating E. coli O157:H7 infection, we expanded our knowledge regarding the unique protective contributions of this specific probiotic bacterium when it is cultured with epithelial cells. It is increasingly recognized that the effects of probiotics are both species and strain specific (33). Accordingly, it is important to better define how individual probiotics elicit their beneficial effects as biotherapeutic agents against pathogen-induced disorders of the gastrointestinal tract.

Acknowledgments

We thank Yew Meng Heng (Pathology Department, Hospital for Sick Children, Toronto, Canada) for technical assistance with the electron microscopy.

This study was supported by grants from the Canadian Institutes of Health Research (CIHR) and the Crohn's and Colitis Foundation of Canada (Fay Shapiro Cutler Grant in Aid of Research). P.M.S. is the recipient of a Canada Research Chair in Gastrointestinal Disease. K.A.D. and G.S.-T. are recipients of CIHR Canada Graduate Scholarship Master's Awards. K.A.D. received a Dutkevich Foundation and Laboratory Medicine and Pathobiology travel award.

Notes

Editor: V. J. DiRita

Footnotes

[down-pointing small open triangle]Published ahead of print on 28 January 2008.

REFERENCES

1. Amieva, M., and R. Vogelmann. 2004. Epithelial cells and pathogens—the Odyssey System brings light into the darkness. Tight junction barrier function in epithelial cells. http://www.licor.com/bio/PDF/EpithelialCells.pdf. Accessed, 24 August 2006.
2. Balda, M. S., and J. M. Anderson. 1993. Two classes of tight junctions are revealed by ZO-1 isoforms. Am. J. Physiol. 264C918-C24. [PubMed]
3. Berkes, J., V. K. Viswanathan, S. D. Savkovic, and G. Hecht. 2003. Intestinal epithelial responses to enteric pathogens: effects on tight junction barrier, ion transport, and inflammation. Gut 52439-451. [PMC free article] [PubMed]
4. Boyle, E. C., N. F. Brown, and B. B. Finlay. 2006. Salmonella enterica serovar Typhimurium effectors SopB, SopE, SopE2 and SipA disrupt tight junction structure and function. Cell. Microbiol. 81946-1957. [PubMed]
5. Bradbury, N. A. 2000. Protein kinase-A-mediated secretion of mucin from human colonic epithelial cells. J. Cell Physiol. 185408-415. [PubMed]
6. Corr, S. C., Y. Li, C. U. Riedel, P. W. O'Toole, C. Hill, and C. G. M. Gahan. 2007. Bacteriocin production as a mechanism for the antiinfective activity of Lactobacillus salivarius UCC118. Proc. Natl. Acad. Sci. USA 1047617-7621. [PMC free article] [PubMed]
7. Doron, S., D. R. Snydman, and S. L. Gorbach. 2005. Lactobacillus GG: bacteriology and clinical applications. Gastroenterol. Clin. N. Am. 34483-498. [PubMed]
8. FAO/WHO. 2001. Joint FAO/WHO expert consultation on evaluation of health and nutritional properties of probiotics in food including powder milk with live lactic acid bacteria, Cordoba, Argentina, 1 to 4 October 2001. http://www.fao.org/es/esn/food/foodandfood_probio_en.stm.
9. Finlay, B. B., and M. Caparon. 2005. Bacterial adherence to cell surfaces and extracellular matrix, p. 105-120. In P. Cossart, P. Boquet, S. Normark, and R. Rappuoli (ed.), Cellular microbiology, 2nd ed. ASM Press, Washington, DC.
10. Gorbach, S. L. 2000. Probiotics and gastrointestinal health. Am. J. Gastroenterol. 95S2-S4. [PubMed]
11. Gotteland, M., S. Cruchet, and S. Verbeke. 2001. Effect of Lactobacillus ingestion on the gastrointestinal mucosal barrier alterations induced by indomethacin in humans. Aliment. Pharmacol. Ther. 1511-17. [PubMed]
12. Hansson, G. C. 1988. A blood group B-like pentaglycosylceramide is the major complex glycosphingolipid of the Madin-Darby canine kidney epithelial cell line I (MDCK I). Biochim. Biophys. Acta 96787-91. [PubMed]
13. Huebner, E. S., and C. M. Surawicz. 2006. Probiotics in the prevention and treatment of gastrointestinal infections. Gastroenterol. Clin. North. Am. 35355-365. [PubMed]
14. Ingrassia, I., A. Leplingard, and A. Darfeuille-Michaud. 2005. Lactobacillus casei DN-114 001 inhibits the ability of adherent-invasive Escherichia coli isolated from Crohn's disease patients to adhere to and to invade intestinal epithelial cells. Appl. Environ. Microbiol. 712880-2887. [PMC free article] [PubMed]
15. Ismaili, A., D. J. Philpott, M. T. Dytoc, R. Soni, S. Ratnam, and P. M. Sherman. 1995. Alpha-actinin accumulation in epithelial cells infected with attaching and effacing gastrointestinal pathogens. J. Infect. Dis. 1721393-1396. [PubMed]
16. Isolauri, E., H. Majamaa, T. Arvola, I. Rantala, E. Virtanen, and H. Arvilommi. 1993. Lactobacillus casei strain GG reverses increased intestinal permeability induced by cow milk in suckling rats. Gastroenterology 1051643-1650. [PubMed]
17. Johnson-Henry, K., J. L. Wallace, N. S. Basappa, R. Soni, G. K. P. Wu, and P. M. Sherman. 2001. Inhibition of attaching and effacing lesion formation following enteropathogenic Escherichia coli and Shiga toxin-producing E. coli infection. Infect. Immun. 697152-7158. [PMC free article] [PubMed]
18. Johnson-Henry, K. C., K. E. Hagen, M. Gordonpour, T. A. Tompkins, and P. M. Sherman. 2007. Surface layer extracts from Lactobacillus helveticus protect against the effects of enterohemorrhagic Escherichia coli O157:H7 infection. Cell. Microbiol. 9356-367. [PubMed]
19. Johnson-Henry, K. C., J. D. Riff, M. Gordanpour, and P. M. Sherman. 2005. Probiotics in models of gastrointestinal bacterial infections. Nutrafoods 429-36.
20. Johnson-Henry, K. C., M. Nadjafi, Y. Avitzur, D. J. Mitchell, B.-Y. Ngan, E. Galindo-Mata, N. L. Jones, and P. M. Sherman. 2005. Amelioration of the effects of Citrobacter rodentium infection in mice by pretreatment with probiotics. J. Infect. Dis. 1912106-2117. [PubMed]
21. Johnson-Henry, K. C., D. J. Mitchell, Y. Avitzur, E. Galindo-Mata, N. L. Jones, and P. M. Sherman. 2004. Probiotics reduce bacterial colonization and gastric inflammation in H. pylori-infected mice. Dig. Dis. Sci. 491095-1102. [PubMed]
22. Kaper, J. B., J. P. Nataro, and H. L. Mobley. 2004. Pathogenic Escherichia coli. Nat. Rev. Microbiol. 2123-140. [PubMed]
23. Lee, Y.-K., K.-Y. Puong, A. C. Ouwehand, and S. Salminen. 2003. Displacement of bacterial pathogens from mucus and Caco-2 cell surface by lactobacilli. J. Med. Microbiol. 52925-930. [PubMed]
24. Liévin-Le Moal, V., and A. L. Servin. 2006. The front line of enteric host defense against unwelcome intrusion of harmful microorganisms: mucins, antimicrobial peptides, and microbiota. Clin. Microbiol. Rev. 19315-337. [PMC free article] [PubMed]
25. Madsen, K., A. Cornish, P. Soper, C. McKaigney, H. Jijon, C. Yachimec, J. Doyle, L. Jewell, and C. De Simone. 2001. Probiotic bacteria enhance murine and human intestinal epithelial barrier function. Gastroenterology 121580-591. [PubMed]
26. Mally, A., M. Decker, M. Bekteshi, and W. Dekant. 2006. Ochratoxin A alters cell adhesion and gap junction intercellular communication in MDCK cells. Toxicology 22315-25. [PubMed]
27. Mastretta, E., P. Longo, A. Laccisaglia, L. Balbo, R. Russo, A. Mazzaccara, and P. Gianino. 2002. Effect of Lactobacillus GG and breast-feeding in the prevention of rotavirus nosocomial infection. J. Pediatr. Gastroenterol. Nutr. 351046-1049. [PubMed]
28. McCormick, B. A. 2003. The use of transepithelial models to examine host-pathogen interactions. Curr. Opin. Microbiol. 677-81. [PubMed]
29. Muza-Moons, M., E. E. Schneeberger, and G. A. Hecht. 2004. Enteropathogenic Escherichia coli infection leads to appearance of aberrant tight junctions strands in the lateral membrane of intestinal epithelial cells. Cell. Microbiol. 6783-793. [PubMed]
30. Otte, J.-M., and D. K. Podolsky. 2004. Functional modulation of enterocytes by gram-positive and gram-negative microorganisms. Am. J. Physiol. 286G613-G626. [PubMed]
31. Parassol, N., M. Freitas, K. Thoreux, G. Dalmasso, R. Bourdet-Sicard, and P. Rampal. 2005. Lactobacillus casei DN-114 001 inhibits the increase in paracellular permeability of enteropathogenic Escherichia coli-infected T84 cells. Res. Microbiol. 156256-262. [PubMed]
32. Pizarro-Cerdá, J., and P. Cossart. 2006. Bacterial adhesion and entry into host cells. Cell 124715-727. [PubMed]
33. Reid, G., J. Jass, M. T. Sebulsky, and J. K. McCormick. 2003. Potential uses of probiotics in clinical practice. Clin. Microbiol. Rev. 16658-672. [PMC free article] [PubMed]
34. Resta-Lenert, S., and K. E. Barrett. 2006. Probiotics and commensals reverse TNF-alpha and IFN-gamma-induced dysfunction in human intestinal epithelial cells. Gastroenterology 130731-746. [PubMed]
35. Santosa, S., E. Farnworth, and P. J. Jones. 2006. Probiotics and their potential health claims. Nutr. Rev. 64265-274. [PubMed]
36. Servin, A. L. 2004. Antagonistic activities of lactobacilli and bifidobacteria against microbial pathogens. FEMS Microbiol. Rev. 28405-440. [PubMed]
37. Sherman, P. M., K. C. Johnson-Henry, H. Yeung, P. Ngo, J. Goulet, and T. A. Tompkins. 2005. Probiotics reduce enterohemorrhagic Escherichia coli O157:H7 and enteropathogenic E. coli O127:H6-induced changes in polarized T84 epithelial cell monolayers by reducing bacterial adhesion and cytoskeletal rearrangements. Infect. Immun. 735183-5188. [PMC free article] [PubMed]
38. Silva, M., N. V. Jacobus, C. Deneke, and S. L. Gorbach. 1987. Antimicrobial substance from a human Lactobacillus strain. Antimicrob. Agents Chemother. 311231-1233. [PMC free article] [PubMed]
39. Simonovic, I., J. Rosenberg, A. Koutsouris, and G. Hecht. 2000. Enteropathogenic Escherichia coli dephosphorylates and dissociates occludin from intestinal epithelial tight junctions. Cell. Microbiol. 2305-315. [PubMed]
40. Söderholm, J. D., P. C. Yang, P. Ceponis, A. Vohara, R. Riddell, P. M. Sherman, and M. H. Perdue. 2002. Chronic stress induces mast cell-dependent bacterial adherence and initiates mucosal inflammation in rat intestine. Gastroenterology 1231099-1108. [PubMed]
41. Söderholm, J. D., and M. H. Perdue. 2001. Stress and the gastrointestinal tract. II. Stress and intestinal barrier function. Am. J. Physiol. Gastrointest. Liver. Physiol. 280G7-G13. [PubMed]
42. Takahashi, M., H. Taguchi, H. Yamaguchi, T. Osaki, A. Komatsu, and S. Kamiya. 2004. The effect of probiotic treatment with Clostridium butyricum on enterohemorrhagic Escherichia coli O157:H7 infection in mice. FEMS Immunol. Med. Microbiol. 41219-226. [PubMed]
43. Tarr, P. I., C. A. Gordon, and W. L. Chandler. 2005. Shiga-toxin-producing Escherichia coli and haemolytic uraemic syndrome. Lancet 3651073-1086. [PubMed]
44. Van Itallie, C. M., and J. M. Anderson. 2006. Claudins and epithelial paracellular transport. Annu. Rev. Physiol. 68403-429. [PubMed]
45. Viswanathan, V. K., A. Koutsouris, S. Lukic, M. Pilkinton, I. Simonovic, M. Simonovic, and G. Hecht. 2004. Comparative analysis of EspF from enteropathogenic and enterohemorrhagic Escherichia coli in alteration of epithelial barrier function. Infect. Immun. 723218-3227. [PMC free article] [PubMed]
46. Yan, F., H. Cao, T. L. Cover, R. Whitehead, M. K. Washington, and D. B. Polk. 2007. Soluble proteins produced by probiotic bacteria regulate intestinal epithelial cell survival and growth. Gastroenterology 132562-575. [PMC free article] [PubMed]
47. Zareie, M., J. D. Riff, K. Donato, D. M. McKay, M. H. Perdue, J. D. Soderholm, M. Karmali, M. B. Cohen, J. Hawkins, and P. M. Sherman. 2005. Novel effects of the prototype translocating Escherichia coli, strain C25 on intestinal epithelial structure and barrier function. Cell. Microbiol. 71782-1797. [PubMed]
48. Zareie, M., K. C. Johnson-Henry, J. Jury, P.-C. Yang, B.-Y. Ngan, D. M. McKay, M. H. Perdue, J. D. Soderholm, and P. M. Sherman. 2006. Probiotics prevent bacterial translocation and improve intestinal barrier function in rats following chronic psychological stress. Gut 551553-1560. [PMC free article] [PubMed]

Articles from Infection and Immunity are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...