Live probiotics protect intestinal epithelial cells from the effects of infection with enteroinvasive Escherichia coli (EIEC)

Abstract

Background: The colonic epithelium maintains a life long reciprocally beneficial interaction with the colonic microbiota. Disruption is associated with mucosal injury.

Aims: We hypothesised that probiotics may limit epithelial damage induced by enteroinvasive pathogens, and promote restitution.

Methods: Human intestinal epithelial cell lines (HT29/cl.19A and Caco-2) were exposed to enteroinvasive Escherichia coli (EIEC 029:NM), and/or probiotics (Streptococcus thermophilus (ST), ATCC19258, and Lactobacillus acidophilus (LA), ATCC4356). Infected cells and controls were assessed for transepithelial resistance, chloride secretory responses, alterations in cytoskeletal and tight junctional proteins, and responses to epidermal growth factor (EGF) stimulation.

Results: Exposure of cell monolayers to live ST/LA, but not to heat inactivated ST/LA, significantly limited adhesion, invasion, and physiological dysfunction induced by EIEC. Antibiotic killed ST/LA reduced adhesion somewhat but were less effective in limiting the consequences of EIEC invasion of cell monolayers. Furthermore, live ST/LA alone increased transepithelial resistance, contrasting markedly with the fall in resistance evoked by EIEC infection, which could also be blocked by live ST/LA. The effect of ST/LA on resistance was accompanied by maintenance (actin, ZO-1) or enhancement (actinin, occludin) of cytoskeletal and tight junctional protein phosphorylation. ST/LA had no effect on chloride secretion by themselves but reversed the increase in basal secretion evoked by EIEC. EIEC also reduced the ability of EGF to activate its receptor, which was reversed by ST/LA.

Conclusions: Live ST/LA interact with intestinal epithelial cells to protect them from the deleterious effect of EIEC via mechanisms that include, but are not limited to, interference with pathogen adhesion and invasion. Probiotics likely also enhance the barrier function of naïve epithelial cells not exposed to any pathogen.

Figures

Figure 1

Probiotic colonisation potentiates barrier function in Caco-2 monolayers. Caco-2 cells were exposed to living (A) or antibiotic killed (a) (B) Streptococcus thermophilus (ST) and Lactobacillus acidophilus (LA), alone or in combination. Transepithelial resistance (TER) was measured at the time points indicated. Resistance values of cell free inserts were subtracted from each measurement. Data are means (SEM), n=10. Where error bars are not shown, they are obscured by the symbols. *p<0.05, **p<0.01, ***p<0.001 compared with uninfected controls (by ANOVA). SM, spent medium.

Figure 2

Pretreatment with probiotics prevents enteroinvasive Escherichia coli (EIEC) induced decrease in transepithelial resistance (TER) of polarised HT29/cl.19A cell monolayers. HT29/cl.19A cells were exposed to medium, EIEC followed by Streptococcus thermophilus and Lactobacillus acidophilus (EIEC+ST/LA), EIEC and ST/LA given simultaneously (EIEC/ST/LA), EIEC alone, ST/LA followed by EIEC (ST/LA+EIEC), or ST/LA alone on the apical side. Preparations of living (A), antibiotic killed (a) (B), heat inactivated (h) (B), or spent medium (SM) from ST and LA cultures (B) were tested in these experiments. TER was measured at the times indicated. Resistance values of cell free inserts were subtracted from each measurement. Data are means (SEM), n=8. *p<0.05, **p<0.01, ***p<0.001 compared with uninfected controls (by ANOVA). Similar findings were observed in Caco-2 cells.

Figure 3

Pretreatment with probiotics significantly reduces enteroinvasive Escherichia coli (EIEC) invasion of infected Caco-2 cell monolayers. (A) Caco-2 cells were exposed to EIEC alone, Streptococcus thermophilus and Lactobacillus acidophilus followed by EIEC (ST/LA+EIEC), ST/LA and EIEC simultaneously (ST/LA/EIEC), or EIEC followed by ST/LA (EIEC+ST/LA), and treated with gentamicin one hour after EIEC infection to kill extracellular bacteria. Living, antibiotic killed (a), and heat inactivated (h) ST and LA were tested in these experiments. Cell lysates were plated on agar for colony forming unit (CFU) counts in the experiments shown. (B) (pretreatment) and (C) (simultaneous treatment) summarise invasion data obtained at various ST/LA multiplicity of infection (MOI) (EIEC 50:1 MOI). (D) Results of invasion in cell monolayers pretreated with inactivated probiotics plus their spent medium (SM) or spent medium alone. Data are means (SEM), n=6. *p<0.05, **p<0.01, ***p<0.001 compared with cells infected with EIEC in the absence of probiotics (by ANOVA).

Figure 4

Infection with enteroinvasive Escherichia coli (EIEC) alters phosphorylation of the tight junction proteins occludin and ZO-1. HT29/Cl.19A cell monolayers were exposed to medium, Streptococcus thermophilus and Lactobacillus acidophilus (ST/LA) (multiplicity of infection=50:1), EIEC, or ST/LA followed by EIEC. Both living and antibiotic killed (a) ST and LA were tested in these experiments. Cells were harvested at 12 hours after infection. Whole cell lysates were immunoprecipitated with specific antibodies against occludin (A) or ZO-1 (B), then blotted with the same antibody or with an antiphosphoserine (A) or antiphosphotyrosine (B) antibody. The upper panels are representative blots whereas the lower panels show means (SEM) for densitometric analyses, n=6. **p<0.01; ***p<0.001 compared with uninfected controls (by ANOVA).

Figure 5

Pretreatment with probiotics prevents actinin dephosphorylation in enteroinvasive Escherichia coli (EIEC) infected Caco-2 cell monolayers. Caco-2 cell monolayers were exposed to medium, Streptococcus thermophilus and Lactobacillus acidophilus (ST/LA) (multiplicity of infection=50:1), EIEC, ST/LA and EIEC simultaneously (ST/LA/EIEC), EIEC followed by ST/LA (EIEC+ST/LA), or ST/LA followed by EIEC (ST/LA+EIEC) and cells were harvested at 12 hours after infection. Both living and antibiotic killed (a) ST and LA were tested in these experiments. Whole cell lysates were immunoprecipitated with specific antibodies against α-actinin, then blotted with the same antibody or an antiphosphotyrosine antibody. The upper panel shows representative blots whereas the lower panel shows means (SEM) for densitometric analyses, n=6. ***p<0.001 compared with uninfected controls (by ANOVA)

Figure 6

Pretreatment with probiotics limits enteroinvasive Escherichia coli (EIEC) induced increases in chloride secretion. HT29/Cl.19A cells were exposed to medium, EIEC, Streptococcus thermophilus and Lactobacillus acidophilus (ST/LA)+EIEC, or ST/LA alone on the apical side. Preparations of living or antibiotic killed (a) ST/LA, or spent medium (SM) from ST and LA cultures were tested in these experiments. Short circuit current (Isc), an indicator of chloride secretion, was measured at the times indicated. Data are means (SEM), n=6. ***p<0.001 compared with uninfected controls (by ANOVA).

Figure 7

Pretreatment with probiotics prevents galanin stimulated chloride secretory responses in enteroinvasive Escherichia coli (EIEC) infected HT29/cl.19A cell monolayers. HT29/cl.19A cell monolayers were exposed to medium alone, Streptococcus thermophilus and Lactobacillus acidophilus (ST/LA) alone, EIEC alone, ST/LA followed by EIEC (ST/LA+EIEC), ST/LA and EIEC simultaneously (ST/LA/EIEC), or EIEC followed by ST/LA (EIEC+ST/LA), or spent medium (SM) from ST and LA cultures with or without EIEC. Preparations of living or antibiotic killed (a) ST/LA were used. Monolayers were mounted in Ussing chambers at the times indicated and chloride secretory responses to galanin (100 nM) assessed as maximal increases in short circuit current (ΔIsc). *p<0.05, ***p<0.001 compared with uninfected controls (by ANOVA).

Figure 8

Pretreatment with probiotics prevents epidermal growth factor receptor (EGFr) inactivation by enteroinvasive Escherichia coli (EIEC). HT29/cl.19A cell monolayers were exposed to medium, Streptococcus thermophilus and Lactobacillus acidophilus (ST/LA), EIEC, or ST/LA followed by EIEC. Both living and antibiotic killed (a) ST and LA were tested in these experiments. Before harvest at three hours after infection, cells were treated with EGF (50 nM, two minutes) or medium alone. After lysis and immunoprecipitation with anti-EGFr, proteins were electrophoresed and blotted with the same antibody or with antiphosphotyrosine. The top panel shows representative blots whereas the bottom panel shows means (SEM) for densitometric analyses, n=6. ***p<0.001 compared with uninfected cells stimulated with EGF (by ANOVA) (n=4, means (SEM)).

Figure 9

Effect of probiotics and enteroinvasive Escherichia coli (EIEC) on basal and epidermal growth factor (EGF) stimulated ERK 1 and 2 activation. HT29/cl.19A cell monolayers were exposed to medium, Streptococcus thermophilus and Lactobacillus acidophilus (ST/LA), EIEC, or ST/LA followed by EIEC. Both living and antibiotic killed (a) ST and LA were tested in these experiments. ERK 1 and 2 were also activated under the various conditions studied following EGF stimulation. The top panel shows two representative blots whereas the bottom panel shows means (SEM) for densitometric analyses. ***p<0.001 compared with uninfected cells (by ANOVA, n=6).