Use of An In-Vitro Gut-on-Chip System to Investigate the Effect of X-Ray Irradiation on the Human GI Tract and Gut Microbiome

The investigation of radiation’s effects on the GI tract, the identification of dosimetry biomarkers, and the testing of new radiation countermeasure drugs are all limited in humans due to the restricted access to human samples or the use of animal models that are not representative of human physiology. Thus, there is a need for improved in vitro models to elucidate the effects of x-ray radiation on the human GI system that mimic the in vivo physiological environment and interaction between human GI epithelium and gut microbiome. Here, we developed a Gut-on-Chip system (HuMiX) to reproduce multiple in vivo parameters classically associated with human GI following acute irradiation. The HuMiX device is a co-culture system consisting of three chambers separated by two porous polycarbonate membranes, delimiting the microbial, intestinal, and vascular compartments. Previous design iterations of HuMiX have demonstrated that the system can imitate the in vivo immunologic, metabolic, and transcriptional responses to commensal gut bacteria [Shah et al, 2016]. In the newly designed system, ~16,000 Caco-2 cells/mm2 were seeded on a collagen-treated membrane in the middle chamber and incubated under normal cell culture incubator conditions. The middle and bottom chambers were then flowed with Caco-2 cell medium at 67 ul/min for 6 days, when 8 x 107/mL of mixed bacterial flora, isolated from human stool samples from 6 different donors, were injected into the top chamber. Post 12 hours co-culture, the devices were sham- or 8 Gy-irradiated at 1 Gy/min using 320 keV x-ray beam with 2 mm Al filter. Twenty-four hours after irradiation, the supernatant was collected for inflammatory cytokine detection, the bacterial cells were collected from the top chamber for microbiome profiling and Caco2 cells were stained. Preliminary microbiome profiling identified the presence of three phyla: Bacteroides, FAFV (Firmicutes, Actinobacteria, Fusobacteria, and Verrucomicrobia), and Proteobacteria. DNA concentration, species richness, and species diversity obtained from the devices decreased as compared to those profiles isolated from direct fecal bacteria isolation and/or cultured in bacterial culture broths, but still maintained highly diverse profiles. Current analyses are on-going to identify if radiation modified the Firmicutes/Bacteriodetes ratio as shown by in vivo studies. The Caco2 phenotype revealed the formation of villi, the presence of cell-cell tight junctions and the production of mucus, demonstrating the establishment of physiological intestinal hallmarks. Irradiation showed ability to disrupt cell junctions and to permeabilize the intestinal barrier as suggested by dextran assay. Citrulline, an amino acid whose plasma level decrease following intestinal radiation-induced damage, was also quantified. This study is an ongoing project and the injection of immune cells will be performed in the future to assess if the HuMiX can recapitulate the expression change of well-known radiation dosimetry biomarkers in blood. Overall, our data suggest that the HuMiX offers a promising tool to study radiation effect on human gut and test radiation countermeasure approaches.