The aim of this thesis was to better understand the molecular mechanism of host res-ponses to probiotics. Probiotics can be used to stimulate or regulate immune responses in epithelial and immune cells of the intestinal mucosa and generate beneficial effects on the immune system. Carefully selected probiotics are able to steer the activity of the immune response in a predetermined manner by increasing or decreasing the activity of different aspects of the immune system (e.g. development and activity of T helper subsets). Beneficial effects of strains of probiotics have been established in the treatment and prevention of various intestinal disorders, including allergic diseases and diarrhea. However the precise molecular mechanisms and the strain dependent factors involved are poorly understood. Here in vitro molecular studies and in vivo mechanistic studies were combined in different mouse models to generate new insights into the beneficial mechanisms of selected lactobacilli and identify novel bacterial genes influencing the immune response. A further aim was to investigate the predictive value of in vitro immune assays for the effects of probiotics in vivo.
Chapter 1and chapter 2 describe the current knowledge and understanding of the immunomodulatory effects of different probiotic species and strains on mucosal immune system, dendritic cells (DCs) and the adaptive immune system. The relevance and the implications of in vitro studies for clinical trials or mechanistic research in animal mo-dels are discussed.
Chapter 3and chapter 4 present new insights gained from research on the strain-dependent factors involved in probiotic immune modulation. Extensive variation was observed in the immune responses to 42 L. plantarum strains. These results were used to identify genetic loci that correlated with levels of induced cytokines (such as IL-10 or IL-12) following co-culture with DCs (chapter 3) or peripheral blood mononuclear cells (PBMCs) (chapter 4). This in silico “gene-trait matching” approach led to the identification of several candidate genes in the L. plantarum genome that might modulate the immune cytokine response to L. plantarum. Selective gene deletions mutants were constructed for the candidate genes in L. plantarum WCFS1 and compared to the wild-type strain in immune assays with PBMCs and DCs. The predicted phenotype of the genetic knock-out was confirmed for most of the candidate loci including genes encoding an N-acetyl-glucosamine/galactosamine phosphotransferase system, the LamBDCA quorum sensing system, a predicted transcriptional regulator gene (lp_2991) and components of the plantaricin (bacteriocin) biosynthesis and transport pathway. Transcriptome analysis and qPCR data showed that transcript level of gtcA3, which is predicted to be involved in the glycosylation of cell wall teichoic acids, was substantially increased in the lp_2991 deletion mutant (44- and 29-fold respectively).
In vitroassays for pre-screening of candidate probiotics would benefit from standar-dized methods and cryopreservation techniques for immature DCs (iDCs) or precursor monocytes. Literature on the effects of cryopreservation and thawing of monocytes or monocyte-derived iDCs suggested that this strategy might be useful although bacteria had not been previously used as a stimulus. Thus in chapter 5 we investigated the effects of cryopreservation and thawing of precursor monocytes and iDCs on the maturation and immune response of DCs to potential probiotic strains and bacterial TLR agonists. Surface markers CD83 and CD86 were expressed at similar levels on iDCs generated from cryopreserved or freshly isolated monocytes. Cryopreservation of iDCs led to slightly decreased expression of CD86 and CD83 compared to freshly generated iDCs prepared from unfrozen cells but this did not affect the capacity of DCs to acquire fully mature characteristics after stimulation. In contrast the cytokine response to lipoteichoic acid and bacterial stimulation was altered by cryopreservation of monocytes or iDCs, particularly for IL-12 which was decreased up to 250 fold or even not detected at all. Cryopreservation also decreased TNF-α and IL-1β production in stimulated iDCs but to a lesser extent than for IL-12, depending on the maturation factors used. The amounts of IL-10 produced by stimulated iDCs were increased up to 3.6 fold when iDCs were cryopreserved, but decreased up to 90 fold when generated from cryopreserved monocytes. Immature DCs are often used to investigate the immunomodulatory properties of probiotics and here we showed for the first time that cryopreserved monocytes and cryopreserved iDCs have a skewed cytokine response to microbial stimulation. Therefore we consider that standardization of probiotic screening assays by the use of cryopreservation methods is currently not applicable. The detailed method for generating human monocyte derived DC described in chapter 5 may however be useful for developing standardized immune assays.
In chapter 6 we screened the immunomodulatory properties of 28 commercially available bacterial strains in vitro using human PBMCs and investigated selected strains for their in vivo immunomodulatory potential in an established mouse peanut allergy model. The 28 probiotic strains induced highly variable cytokine profiles in PBMCs. L. salivarius HMI001 (HMI001), L. casei Shirota (LCS) and L. plantarum WCFS1 (WCFS1) were selected for further investigation due to their distinct patterns of IL-10, IL-12 and IFN-γ induction. Prophylactic treatment with both HMI001 and LCS attenuated the Th2 phenotype in the mouse model (reduced mast cell responses and ex vivo IL-4 and/or IL-5 production). In contrast, WCFS1 augmented the Th2 phenotype (increased mast cell and antibody responses and ex vivo IL-4 production). In vitro PBMC screening was useful in selecting strains with anti-inflammatory and Th1 skewing properties. In the case of HMI001 (inducing a high IL-10/IL-12 ratio) and LCS (inducing high amounts of IFN-γ and IL-12) partial protection was seen in a mouse peanut allergy model. However, certain strains may worsen the allergic reaction as shown in the case of WCFS1. This approach indicated that pre-selection of candidate probiotics using in vitro immune assays is useful for selecting strains for translational research in humans.
Probiotics have been shown to increase the efficacy of different vaccines and can be easily consumed in food, and therefore probiotics might be useful in the improvement of current mucosal vaccines. In chapter 7 we have investigated the mechanisms behind the effect of lactobacilli on humoral responses to an intranasal vaccine. In addition to L. rhamnosus GG we selected 6 strains of Lactobacillus plantarum which have strikingly different immunomodulatory properties in vitro and TLR-2/6 activating properties. This selection was based on the approach outlined in chapter 3 and chapter 4 examining the in vitro immune responses of human monocyte derived DCs and PBMCs to 42 different L. plantarum strains. First we established an influenza vaccination model in Balb/c mice that would be sensitive to immunomodulation by lactobacilli, which allowed potential up- and down-regulation by the lactobacilli of the immune response. Strain WCFS1, that induced the lowest IL-10 to IL-12 cytokine ratio in DC co-culture significantly increased vaccine-specific antibody responses to the intranasal vaccine compared to the vaccine control group. Several Lactobacillus strains appeared to increase delayed-type hypersensitivity responses after vaccination compared to the vaccine control group indicating increased Th1-mediated vaccine responses. For strain LMG18021 this was also reflected in the significantly higher vaccine-specific IgG2a to IgG1 antibody ratio. LMG18021, CIP104448 and CIP104450 which have the highest IL-10 to IL-12 ratios of the strains tested, significantly enhanced the ex vivo vaccine-specific induction of IL-10, IL-17A, IL-6 and IL-4 in MLN cells. B1839 which was included as negative control, as it was a low cytokine inducer, did not enhance the vaccine-specific antibody or immune response indicating that the immune-stimulatory properties are important in mediating effects on the vaccine response. Further research is needed to demonstrate that these effects on the vaccine response impact on protection from influenza challenge and to validate the immunomodulatory mechanisms involved. Nevertheless, the in vivo studies described in this thesis support other publications proposing that in vitro immune assays can be useful for predicting which candidate probiotic strains will be most effective in vivo.
Chapter 8 completes this thesis with an overview of the most important findings of this thesis and discusses possible research limitations and future research perspectives. We stress the importance of proper strain selection using in vitro assays, and the use of strategies to identify novel immunomodulatory factors. The results described in this thesis support the rationale of using in vitro co-culture assays for selection of candidate probiotics for in vivo animal experiments or human trials.