Probiotic Lactobacillus spp. diminish Helicobacter hepaticus-induced inflammatory bowel disease in interleukin-10-deficient mice

Abstract

Clinical and experimental evidence has demonstrated the potential role of probiotics in the prevention or treatment of inflammatory bowel disease. Probiotic clones with direct immunomodulatory activity may have anti-inflammatory effects in the intestine. We investigated the roles of tumor necrosis factor alpha (TNF-alpha)-inhibitory Lactobacillus clones with a pathogen-induced murine colitis model. Murine-derived probiotic lactobacilli were selected in vitro for their ability to inhibit TNF-alpha secretion by Helicobacter hepaticus-stimulated macrophages. Interleukin-10 (IL-10)-deficient mice were treated with probiotic Lactobacillus reuteri in combination with Lactobacillus paracasei and then challenged with H. hepaticus. Ten weeks postinoculation, the severity of typhlocolitis was assessed by histologic examination of the cecocolic region. Intestinal proinflammatory cytokine responses were evaluated by real-time quantitative reverse transcriptase PCR and immunoassays, and the quantities of intestinal H. hepaticus were evaluated by real-time PCR. Intestinal colonization by TNF-alpha-inhibitory lactobacilli reduced intestinal inflammation in H. hepaticus-challenged IL-10-deficient mice despite similar quantities of H. hepaticus in cocolonized animals. Proinflammatory colonic cytokine (TNF-alpha and IL-12) levels were lowered in Lactobacillus-treated animals. In this H. hepaticus-challenged IL-10-deficient murine colitis model, lactobacilli demonstrated probiotic effects by direct modulation of mucosal inflammatory responses.

FIG. 1.

Timeline of H. hepaticus infection and Lactobacillus cocolonization study. Bacteria were administered to mice by orogastric gavage. An equal mixture (10⁹ cells per dose) of L. paracasei 1602 and L. reuteri 6798 (Lacto or Lb) was used to antagonize colitis induced by H. hepaticus (Hh) (10⁷ cells per dose).

FIG. 2.

In vitro selection of probiotic lactobacilli for studies in IL-10-deficient mice. L. paracasei 1602 (Lp1602) and L. reuteri 6798 (Lr6798) blocked TNF-α production by murine RAW 264.7 macrophages activated with bacterium-free stationary-phase culture medium of H. hepaticus (Hh) 3B1. Briefly, cells were coincubated with H. hepaticus-conditioned medium and Lactobacillus-conditioned medium (CM) derived from cultures of Lp1602 or Lr6798. TNF-α production was measured in picograms per milliliter by quantitative enzyme-linked immunosorbent assay. *, Lactobacillus-CM significantly decreased TNF-α output by Hh-stimulated macrophages (P < 0.05).

FIG. 3.

Lactobacillus decreases typhlocolitis in female mice but not in male mice. The cecocolic region of mouse intestines was evaluated by hematoxylin-and-eosin staining and scored blindly by a board-certified veterinary pathologist. Independent scores (scale, 0 to 4) were obtained for degree of mucosal inflammation (INFLAMM), hyperplasia (HYPER), and dysplasia (DYSPL). (a) In a pilot study utilizing both genders, prophylaxis with lactobacilli failed to diminish visible disease by histologic examination. (b) When only female mice were analyzed, lactobacilli exerted significant protective effects. Mice were uninfected (not colonized or infected with exogenous bacteria), cocolonized with L. paracasei 1602/L. reuteri 6798 (Lp1602/Lr6798), infected with H. hepaticus (Hh), or cocolonized with L. paracasei 1602/L. reuteri 6798 and H. hepaticus (Lp1602/Lr6798 + Hh). Data are presented as box plots with median (line inside box), interquartile range (shaded box), and range (error bars). *, pretreatment with lactobacilli prior to H. hepaticus challenge (Lp1602/Lr6798 + Hh) significantly decreased all lesion types compared to results for animals receiving H. hepaticus alone (Hh) (P < 0.05).

FIG. 4.

Lactobacillus ameliorates typhlocolitis in H. hepaticus-infected IL-10-deficient mice. (a) Histopathology of cecocolic junction from an uninfected mouse demonstrates normal mucosal architecture. (b) Tissue from a mouse monoinfected with H. hepaticus exhibits moderately severe inflammation with reactive epithelial changes including goblet cell loss and crypt hyperplasia and dysplasia. (c) Tissue from a mouse administered a combination of L. paracasei 1602/L. reuteri 6798 prior to H. hepaticus infection exhibits a moderate reduction in intestinal inflammation and hyperplasia. Bar = 200 μm.

FIG. 5.

Levels of H. hepaticus colonization in the ceca of IL-10-deficient mice. Cecal colonization by H. hepaticus (Hh) was evaluated by TaqMan probe-based real-time quantitative PCR, using the cytolethal distending toxin B-subunit gene (cdtB) as the species-specific DNA target. CdtB gene copy numbers were normalized by mouse DNA content, which was quantified by using murine 18S rRNA gene-based primers and probe. Quantitative results are expressed as copies of Hh cdtB per ng mouse 18S rRNA gene. Mice were infected with H. hepaticus (Hh) or precolonized with L. paracasei 1602/L. reuteri 6798 prior to H. hepaticus infection (Lp1602/Lr6798 + Hh). Data are presented as box plots with median (line inside box), interquartile range (shaded box), and range (error bars).

FIG. 6.

Colonization with probiotic Lactobacillus affects levels of cytokine mRNA transcripts in the intestine. The mid-ceca of mice were homogenized, and total RNA was extracted. mRNA levels for (A) TNF-α, (B) IL-12 p40 subunit, (C) IFN-γ, and (D) IL-4 were quantitated by TaqMan probe-based real-time RT-PCR and normalized by mouse glyceraldehyde-3-phosphate dehydrogenase coamplification, using commercially available primer and probe sets for each target mRNA. IL-12 is presented as n-fold change from levels for uninfected control animals due to absence of a quantitation standard. Mice were uninfected (not colonized with probiotic lactobacilli), colonized with L. paracasei 1602/L. reuteri 6798 (Lp1602/Lr6798), infected with H. hepaticus (Hh), or precolonized with L. paracasei 1602/L. reuteri 6798 prior to H. hepaticus infection (Lp1602/Lr6798 + Hh). Data are presented as box plots with median (line inside box), interquartile range (shaded box), and range (error bars). *, mice monoinfected with H. hepaticus have IL-12 levels significantly greater than those for mice given Lactobacillus prior to H. hepaticus challenge (P < 0.05).

FIG. 7.

Proinflammatory cytokine responses are modulated by probiotic lactobacilli in IL-10-deficient mice. Proinflammatory cytokine profiles were generated by using cecal levels of IL-12 and TNF-α mRNA transcripts. Mean quantities of IL-12 and TNF-α for all groups receiving exogenous bacteria were divided by means for uninfected mice. Bars represent n-fold change relative to results with uninfected sham-dosed controls. Note that the magnitudes of inflammatory profiles generated are proportional to histologic findings. Mice were uninfected (not colonized with probiotic lactobacilli), colonized with L. paracasei 1602/L. reuteri 6798 (Lp1602/Lr6798), infected with H. hepaticus (Hh), or precolonized with L. paracasei 1602/L. reuteri 6798 prior to H. hepaticus infection (Lp1602/Lr6798 + Hh).

FIG. 8.

Reduction of LPS-induced TNF-α production by colonic explants from mice treated with Lactobacillus. TNF-α output was corrected by the wet weight (measured at necropsy, prior to culture) and total DNA content of each explant. DNA content was determined by extraction of total explant DNA after culture and subsequent amplification of murine 18S rRNA gene by real-time quantitative PCR. TNF-α output in response to LPS stimulation is decreased in mice treated with L. paracasei 1602 and L. reuteri 6798 (Lp1602/Lr6798), indicating that Lp1602/Lr6798 may have direct anti-inflammatory effects. Mice were uninfected (not colonized with probiotic lactobacilli), colonized with L. paracasei 1602/L. reuteri 6798 (Lp1602/Lr6798), infected with H. hepaticus (Hh), or precolonized with L. paracasei 1602/L. reuteri 6798 prior to H. hepaticus infection (Lp1602/Lr6798 + Hh). Data corrected by weight are shown in this figure; data corrected by DNA (data not shown) were similar. Data are presented as box plots with median (line inside box), interquartile range (shaded box), and range (error bars).