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

Highlights

• An in vitro/ex vivo gastrointestinal model of Salmonella invasivity is proposed.

• Physiological conditions were simulated in vitro and in silico.

• Histological analysis from ex vivo tissue revealed Salmonella damage.

• In silico interactions displayed Salmonella-food matrix affinity.

• Salmonella probability of infection and severity can be evaluated using this model.

Abstract

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_inf) were found for intra-experimental repeatability. However, the P_inf 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.

Introduction

Salmonellosis is a worldwide common cause for morbidity and mortality in humans (Schlundt, 2017), which is caused by the non-typhoidal Salmonella, mainly due to the consumption of contaminated food with the pathogen. Depending on the host and strain characteristics, the minimum Salmonella infective dose can be one cell (FDA, 2012). The lack of information about the inter-strain variability generates uncertainty in the risk estimations (Haddad et al., 2018; McLauchlin et al., 2004). Few researchers have included the strain virulence variability into quantitative risk assessment models, being Listeria monocytogenes the most studied foodborne pathogen (Chen et al., 2006; FDA/USDA, 2003Fritsch et al., 2018; Pouillot et al., 2015 ). Improvements in quantitative risk assessment models are important for assessing the virulence variability of the pathogen's strains to include it into the hazard characterization.

The evaluation of virulence variability could be done using in vitro and in vivo approaches. In this sense, in vitro experiments are preferred since specific factors can be set under controlled environmental conditions (FAO/WHO, 2003). For instance, in vitro gastrointestinal models can be used for such purpose, but mimicking human digestion is still a challenge. The digestive tract comprises of the mouth, esophagus, stomach, small intestine, large intestine, and the anus (McDonald et al., 2018), presenting abiotic and biotic barriers against the pathogens (Ceuppens et al., 2012). For this reason, all of these factors should be considered in the design of gastrointestinal models.

Reductions in the particle size take place in the mouth, and the pathogens might be protected by the food matrix (Guerra et al., 2012). At the gastric stage, low pH is a critical barrier against pathogenic microorganisms (Smith, 2003). However, gastric pH depends on the acid secretion and the buffer capacity of the food (Bellmann et al., 2016). The pH from the empty stomach usually ranges between 1 and 3, but the food arrival increases its value, followed by a gradual decrease (Gardner et al., 2002; Koziolek et al., 2015; Malagelada et al., 1976). The third stage is the intestine, where intestinal enzymes, the bile, and the commensal microbiota influence the survival and colonization of the pathogen. The intestinal microbiota protects the host not only for the competitiveness of nutrients and space but also for its antimicrobial compounds (Ceuppens et al., 2012).

Regarding Salmonella, most in vitro gastrointestinal simulations have evaluated the behavior of the pathogen in simulated fluids of each stage. The two-compartment system (stomach and intestine) is the most common (Aviles et al., 2013; Kim et al., 2014; Seixas et al., 2014). On the other hand, only a few studies have assessed the pathogen behavior using a three-compartment system (mouth, stomach, and intestine) with simulated fluids for each stage (De Sales et al., 2018; Ferreira de Melo et al., 2017).

Other studies have also evaluated Salmonella ability to adheres and invades polarized Caco-2 cells (Oliveira et al., 2011; Wijnands et al., 2017), HEp-2, HT-29, and the mucus-secreting HT29-MTX cells (Chakroun et al., 2017; Gagnon et al., 2013; Solano et al., 2001; Yoon et al., 2013). Nonetheless, cell culture assays lack of intestinal tissue complexity such as mucus production, intestinal transporters, tissue conformation, enzymatic production, and the interplay between immune cells and gut microbiota. In comparison, the everted gut sac displays all of these complexities, including multiple cell types, protein transporters, and enzymes (Mateer et al., 2016; McDonald et al., 2018), being a more realistic alternative to simulate what is happening during the intestinal phase. For this reason, this study aimed to propose and evaluate a dynamic and integrated in vitro/ex vivo gastrointestinal model to determine the damage severity and infection probability of S. Typhimurium vehiculated in food matrices. This model could be used as a tool to evaluate virulence variability among different pathogen strains.