A dynamic and integrated in vitro/ex vivo gastrointestinal model for the evaluation of the probability and severity of infection in humans by Salmonella spp. vehiculated in different matrices.

The lack of proper gastrointestinal models assessing the inter-strain virulence variability of foodborne pathogens and the effect of the vehicle (food matrix) affects the risk estimation. This research aimed to propose a dynamic and integrated in vitro/ex vivo gastrointestinal model to evaluate the probability and severity of infection of foodborne pathogens at different matrices. An everted gut sac was used to determine the adhesion and invasion of Salmonella enterica and tissue damage. S. Typhimurium ATCC 14028 was used as a representative bacterium, and two matrices (water and cheese) were used as vehicles. No differences (p > 0.05) in the probability of infection (P) were found for intra-experimental repeatability. However, the P of cheese-vehiculated S. Typhimurium was different compared to water- vehiculated S. Typhimurium, 7.2-fold higher. The histological analysis revealed Salmonella-induced tissue damage, compared with the control (p < 0.05). In silico proposed interactions between two major Salmonella outer membrane proteins (OmpA and Rck) and digested peptides from cheese casein showed high binding affinity and stability, suggesting a potential protective function from the food matrix. The results showed that the everted gut sac model is suitable to evaluate the inter-strain virulence variability, considering both physiological conditions and the effect of the food matrix.