A key role of dendritic cells in probiotic functionality

Abstract

Background: Disruption of the intestinal homeostasis and tolerance towards the resident microbiota is a major mechanism involved in the development of inflammatory bowel disease. While some bacteria are inducers of disease, others, known as probiotics, are able to reduce inflammation. Because dendritic cells (DCs) play a central role in regulating immune responses and in inducing tolerance, we investigated their role in the anti-inflammatory potential of probiotic lactic acid bacteria.

Methodology/principal findings: Selected LAB strains, while efficiently taken up by DCs in vitro, induced a partial maturation of the cells. Transfer of probiotic-treated DCs conferred protection against 2, 4, 6-trinitrobenzenesulfonic acid (TNBS)-induced colitis. Protection was associated with a reduction of inflammatory scores and colonic expression of pro-inflammatory genes, while a high local expression of the immunoregulatory enzyme indolamine 2, 3 dioxgenase (IDO) was observed. The preventive effect of probiotic-pulsed DCs required not only MyD88-, TLR2- and NOD2-dependent signaling but also the induction of CD4+ CD25+ regulatory cells in an IL-10-independent pathway.

Conclusions/significance: Altogether, these results suggest that selected probiotics can stimulate DC regulatory functions by targeting specific pattern-recognition receptors and pathways. The results not only emphasize the role of DCs in probiotic immune interactions, but indicate a possible role in immune-intervention therapy for IBD.

Figure 1

Internalization of L. rhamnosus by BMDCs. BMDCs were incubated for 18 h with Lr32 strain at a bacteria-to-DC ratio of 10 and then processed for transmission electron microscopy. Bacteria are visible within membrane-bound phagosomes at various stages of degradation (arrows). Bar, 2 µm.

Figure 2

Cytokine (A) and chemokine (B) response of murine BMDCs (0.5×10⁶ cells/ml) derived from BALB/c mice to stimulation with strains L. rhamnosus Lr32, L. salivarius Ls33, L. acidophilus NCFM, L. lactis MG1363 and E. coli TG1. Bacteria were collected after overnight culture and the stimulation of BMDC was done at a ratio of 10∶1 (bacteria/DC). Results represent the mean±SEM of 7 independent experiments.

Figure 3

Phenotypic characteristics of non-treated (control) or bacteria-pulsed BMDCs from BALB/c mice. Plots show flow cytometric profiles of MHC class II, CD86 and CD40 expression (thick line) in comparison to an isotypic control (thin line). The data are representative of 4 independent experiments.

Figure 4

Protective effect of intra-peritoneal administration of LAB-treated BMDCs on acute TNBS-induced colitis in BALB/c mice. Wallace inflammation scores were calculated after a TNBS challenge in mice either not treated (None) or intraperitoneally injected by untreated BMDCs (DC) or BMDCs treated with (A) Lr32 (Lr32-DC) or (B) Ls33 (Ls33-DC) strains. The comparison between the TNBS-control groups and the groups that received the corresponding untreated BMDCs (DC) was calculated using the Mann-Whitney U test (<0.01, **; p<0.001, ***). Percentage mortality and protection (reduction of the mean Wallace scores of mice treated with BMDCs, in relation to the mean score of TNBS control group) are also indicated. In A, the effect of Lr32-treated BMDCs was also compared to the intra-peritoneal administration of 10⁸ CFU of Lr32 bacteria (Lr32-IP) and in B, an additional group of mice was included, pre-treated with 3 intra-peritoneal administrations of prednisone (10 mg/kg), representing a clinically relevant standard treatment for Crohn's disease (36). Data represent the mean±SEM of two representative experiments (number of mice n = 10).

Figure 5

Representative histology of May-Grunwald stained sections (×10) of the distal colon from healthy control mice (A), mice with acute TNBS-induced colitis that have been or not (B) administered with non-treated DC (C) and Lr32-treated DC (D) and corresponding Ameho scores (E), mean±SEM (number of mice n = 10).

Figure 6

Colonic myeloperoxydase (MPO) activity (A) and Serum Amyloid A protein levels obtained 48 h after TNBS-induced colitis. Mice were left untreated (None) or IP administered with untreated BMDC (DC) or Lr32-treated BMDCs (Lr32-DC). The values are expressed as the mean±SEM (number of mice n = 10).

Figure 7

Quantitative real time PCR analysis of mRNA expression of pro-inflammatory or regulatory mediators in colons obtained 48 h after TNBS-induced colitis. Mice were left untreated (None) or IP administered with untreated BMDC (DC) or Lr32-treated BMDCs (Lr32-DC). The values are expressed as the mean ratio±SEM of mRNA levels after TNBS challenge compared with non challenged healthy mice.

Figure 8

Result of the in vivo administration of anti-CD25 mAb (PC61) on the protective effect of probiotic-pulsed BMDCs in TNBS-induced colitis. Ten mice per group were challenged with TNBS alone (CTL) or after Lr32-pulsed BMDC transfer (Lr32-DC). The effect of a pre-treatment by either 200 µg of anti-CD25 (α-CD25) or control rat IgG (rat IgG) mAb, 24 h before colitis induction was analyzed in mice treated or not by Lr32-pulsed BMDC (Lr32-DC) transfer. Wallace inflammation scores (mean ± SEM) were calculated in each group and compared to each other by a Mann-Whitney U test (<0.01, **; p<0.001, ***). Percent of mortality and protection (reduction of the mean Wallace scores of mice treated with BMDCs, in relation to the mean score of TNBS control group) are also reported.

Figure 9

Protective effect of intra-peritoneal administration of untreated (DC) or Lr32-treated BMDCs (Lr32-DC) derived from C57Bl/6 WT or MyD88^−/−, TLR2^−/−, NOD2 ^−/− or IL-10^−/− KO mice on TNBS-induced colitis in C57Bl/6 mice. For technical reason, the protective effect of all KO BMDC's have not been analyzed in the same experiment. For that reason, results are only expressed as the reduction (in %) of the mean macroscopic inflammation scores of mice treated with BMDCs, in relation to the mean scores of non-treated mice (corresponding control TNBS). Significance values p<0.05 (*) or p<0.01 (**), as compared to the corresponding TNBS-control group or the groups that received the corresponding non-treated BMDCs, were calculated by the Mann-Whitney U test (n = 10). The dashed line indicates the 25% threshold of the uncertain statistical significance as previously described (48).