Stefanie Hoffmann ^1* Philip Hewitt ¹ Isabel Koscielski ¹ Dorota Kurek ² Wouter Strijker ² Kinga Kosim ²

• ¹ Merck (Germany), Darmstadt, Germany
• ² Mimetas BV, Leiden, Netherlands

The potential for drug-induced gastrointestinal (GI) toxicity is significant, since the GI tract is one of the first barriers which come in to contact with oral drugs. In pharmaceutical research, the complex behavior of human intestinal cells is traditionally investigated using 2D cultures, in which one cell type grows under static conditions. With the development of advanced microphysiological systems (MPS) more in vivo like conditions can be generated which increase the physiological nature and also the predictive validity of these models.Caco-2 cells are known for their capability to build tight junctions. These connections are responsible for the maintenance of intestinal homeostasis and can be used as a specific safety endpoint, by measuring the Trans Epithelial Electrical Resistance (TEER), for the investigation of drug-induced GI toxicity.Compared to a widely used Caco-2 cell 2D Transwell model, the advanced MPS model (Mimetas OrganoPlate®) allows for the recapitulation of the enterocyte cell layer of the intestinal barrier as the Caco-2 cells grow in a tubular structure through which the medium continuously flows.The OrganoPlate® intestinal model was qualified to be implemented as a routine test system for the early prediction of drug-induced GI toxicity based on the measurement of the tightness of the cell layer by measuring changes in the TEER. For this qualification 23 well known compounds as well as a positive, negative and solvent control were selected. The compounds were selected based on their known effect on the GI system (chemotherapeutics, tight junction disruptor, liver toxins, controls, NSAIDs and a mixed group of drugs).The TEER values were measured 4h and 24h after treatment. In parallel the cell viability was determined after 24h to be able to distinguish between an unspecific cytotoxic effect or direct tight junction damage. Overall, from the 23 tested compounds, 15 showed the expected outcome, i.e.,the compound led to a decrease of the TEER for the positive control compounds, or the TEER value remained stable after treatment with non-GI-toxic compounds. In summary, this MPS model allowed the recapitulation of the human intestinal GI barrier and will enable a faster and more robust assessment of drug-induced damage in the GI tract.