Investigating mechanisms behind invasive Salmonella infections using an intestinal organ-on-a-chip model

Typhoidal Salmonellae establish disease by bypassing innate immune defences in the human gut. Unlike localised gut infections by non-typhoidal Salmonellae, typhoidal Salmonellae disseminate from the intestinal epithelium to sterile sites, causing a systemic infection that results in acute enteric fever. It is hypothesised that the typhoid toxin encoded by typhoidal Salmonellae activate host DNA damage responses (DDR) to manipulate innate gut defences, thereby facilitating Salmonella dissemination, yet the mechanisms are unresolved. Previous studies on the typhoid toxin have been performed on 2D non-polarised human cell cultures, animal models and human challenge models. Human challenge models are subject to stringent regulations, which limit the extent of studies that can be performed and subsequent interpretations. On the other hand, typhoidal Salmonella are strict human pathogens and make animal models difficult to interpret. Using cultured human cells remains the most faithful way of studying host-pathogen interactions. However, 2D cell cultures do not account for the 3D polarised microenvironment in vivo. This study sought to establish a 3D intestinal organ-on-a-chip model to elucidate host interactions with purified recombinant typhoid toxin. Multiple 3D gut-on-chips were developed using polarized intestinal epithelial cells such as Caco-2 and DLD-1 cells, and human primary intestinal cells from biopsy-derived colon organoids. Of these, the Caco-2 gut-on-chip was the most sensitive to the typhoid toxin. Toxin nuclease activity is known to activate DDRs that mediate cell-cycle arrest to enable repair. Indeed, toxin-induced DDR marked by γH2AX was observed in both nonpolarized 2D Caco-2 cells and the 3D Caco-2 gut-on-chip system. Further investigation through RNA sequencing revealed a divergence in the transcriptomes of 2D Caco-2 cells and the 3D Caco-2 gut-on-chip in response to the toxin, which was dependent on polarisation in the 3D Caco-2 gut-on-a-chip. Interestingly, enrichment analyses of genes differentially regulated by the toxin in 2D Caco-2 cells showed cell cycle checkpoint and response to type I interferon as the two most significant biological processes associated with them. On the other hand, the two most significant processes associated with genes differentially regulated by the toxin in the 3D Caco-2 gut-on-chip were sterol biosynthetic process and cholesterol biosynthetic process. Nevertheless, toxin-induced DDRs were observed to reduce intracellular NTS Salmonella burden in both 3D and 2D Caco-2 models. Overall, findings in this PhD thesis reveal that the 3D environment of a system plays an important role in the transcriptional programs activated by the typhoid toxin, which may influence the ability of typhoidal Salmonellae to disseminate and establish systemic infections in vivo.