Abstract
In a previous study, we uncovered three immune-responsive patterns of gut microbes using an in vitro mesenteric lymph node cell suspension model, abbreviated as the MLN model hereafter. We used Akkermansia muciniphila and Clostridium butyricum as the first group directly inducing an immune response, Bifidobacterium sp. and Bacteroides sp. as the second group evoking an immune response with the help of stimuli (anti-CD3 and anti-CD28 antibodies), and Lactobacillus sp. as the third group blunting the immune response with or without stimuli. Our group previously clarified the immune-activation characteristics of A. muciniphila and linked its in vivo immune induction effect in GF and SPF mice under homeostasis. In the present study, we supplemented the characteristics of C. butyricum and B. bifidum in the in vitro MLN model and addressed the specific elements of the model. Finally, we used an in vivo TNBS-challenge model to show the functional differences between these species with different response patterns in vitro. The results showed that C. butyricum and B. bifidum evoked an immune response in vitro in a dose-dependent and strain-unique manner. Although TLR2, rather than TLR4, is indispensable for immune activation in the present in vitro model, it may not involve interaction between TLR2 and bacterial ligands. Like the PBMC model, the present in vitro MLN model is highly dependent on cell resources and should be given more attention when used to conduct a quantitative comparison. Finally, a mixture of two strong immunogenic strains, A. muciniphila and C. butyricum, significantly increased the mortality of TNBS-challenged (2,4,6-trinitrobenzene sulfonic acid, TNBS) mice, indicating a possible link between the in vitro MLN model and in vivo functional evaluation. However, more evidence is needed to clarify the associations and underlying mechanisms.