8 ± 0.5, 6.4 ± 0.4 and 6.5 ± 0.3 log10

MCN/ml in R1, R2 and R3, respectively (Bif; Figure 2A). Addition of B. thermophilum RBL67 beads FRAX597 in vitro increased Salmonella counts in R1 compared to the previous E. coli L1000 treatment (Ecol II, Figure 2C). However, Salmonella invasion efficiency did not change for any of the reactors and the invasion ratio measured with JSH-23 nmr transverse reactor samples significantly decreased during Bif compared to Ecol II periods (Figure 2B). B. thermophilum RBL67 addition (Bif) significantly (P = 0.0001) increased the mean TER measured across HT29-MTX cell monolayers applied with effluents of all reactors by 58 ± 17% compared to previous E. coli L1000 period (Ecol II, Figure 2D). Mean TER measured after 24 h of incubation with effluents from proximal reactors (130 ± 47 Ω cm2) was similar (P > 0.05) to initial model stabilization periods (Stab) before Salmonella infection (127 ± 23 Ω cm2; Table 1). Confocal microscopy analysis revealed high integrity of

intracellular junctions upon application of distal colon reactor effluents of F1 after addition of B. thermophilum RBL67 (Figure 4D) despite high Salmonella counts (6.4 ± 0.6 log10 cfu/ml). Inulin stimulates B. thermophilum RBL67 growth but increases Salmonella invasion in proximal colon environments Addition of inulin induced a significant (P = 0.022) increase in Salmonella counts (Figure 2A) in R3 NCT-501 concentration compared to previous B. thermophilum RBL67 periods (Bif). Furthermore a pronounced enhancement of B. thermophilum RBL67 growth (Figure 2A) and an increase in SCFA concentrations and butyrate ratios (Table 1) occurred in all reactors. Inulin supplementation in R1 was accompanied by a significant (P = 0.024) increase in the efficiency of Salmonella to invade HT29-MTX cells compared to the previous B. thermophilum RBL67 period (Bif). This effect was not significant for transverse and distal reactor samples. Inulin treatment also induced a 25%-decrease (P = 0.088) in TER after 1-3

h of incubation for effluents of R1 compared to the previous B. thermophilum RBL67 periods (Table 1), while a similar but less pronounced tendency was observed for transverse and distal reactors. Discussion next Accurate assessment of probiotic-mediated anti-Salmonella activities is complicated by the fact that mechanisms involved in enteric protection are the function of many probiotic features. Various interactions take place in complex gut environments, including competition for substrates, direct antagonism by the production of inhibitory substances (e.g. SCFA or bacteriocins), competitive exclusion, and potentially host-mediated effects such as improved barrier function and altered immune response [5, 28, 29]. It is therefore crucial to consider microbe-microbe as well as host-microbe interactions for the development of probiotics with targeted efficacy.