(556g) High-Throughput Toxicity Testing of Chemicals and Mixtures in Organotypic Multi-Cellular Cultures of Primary Human Hepatic Cells

Authors: 
Orbach, S., Virginia Tech
Ehrich, M., Virginia-Maryland College of Veterinary Medicine
Rajagopalan, P., Virginia Tech
There are thousands of chemicals whose effects on human and animal populations are unknown or only partially understood. Specifically, combinatorial effects of these chemicals to the liver are unknown. Traditional in vitro and in vivo studies, while informative, can be laborious and time-consuming. High-throughput screening (HTS) methodologies can rapidly test individual and cocktails of chemicals across a wide range of concentrations. To date, HTS assays on liver cells have primarily been conducted on hepatocyte monolayers. When hepatocytes are cultured in monolayer, they begin to lose liver-specific functions within 24 hours. Moreover, these models fail to recapitulate the complex interplay between hepatic cells and their in vivo microenvironment.

We have assembled a novel micro liver organotypic culture model assembled in a 96-well plate (denoted as μOCM) that can be used to rapidly screen individual and mixtures of chemicals. A novel aspect of the μOCMs is the use of automated procedures. Liver sinusoidal endothelial cells (LSECs) and Kupffer cells (KCs) were encapsulated in a type I collagen/fibronectin gel to mimic the Space of Disse. The encapsulated cells were seeded on top of the hepatocytes.

Acetaminophen (APAP), ethanol (EtOH), isoniazid (INH) and perfluorooctanoic acid (PFOA) were administered to μOCMs assembled with rodent and human primary cells. The collagen sandwich model served as a 2D comparison. The μOCMs assembled with rodent cells exhibited between 39% and 49% cell death in response to LC50 for all four chemicals. In contrast, there was no more than 25% cell death in the collagen sandwich cultures when treated with APAP, EtOH or INH. A similar trend was found in the models assembled with human cells. These μOCMs exhibited 41-48% cell death at LC50, while the collagen sandwich models demonstrated no more than 33% cell death. Apoptosis increased at least 51% in response to each chemical in the human μOCMs. Glutathione was depleted 52-74% in response to APAP, EtOH and INH but was unchanged upon PFOA treatment. The mitochondrial membrane integrity of the human μOCMs decreased approximately 35% in response upon treatment with APAP, INH and PFOA. CYP2E1 activity increased 2.6-fold relative to the untreated samples upon treatment with EtOH. In contrast, CYP2E1 activity decreased 1.8-fold and 1.6-fold in response to APAP and INH, respectively. These results are consistent with in vivo findings.

Chemical mixtures of APAP, EtOH and INH were added to the cultures, where each chemical was added at ½ LC50. In the human μOCMs, all mixtures resulted in at least 40% cell death for each mixture. In the collagen sandwich models, 50% cell death only occurred in response to the mixture of all three chemicals and EtOH + INH. Changes in apoptosis and GSH in the μOCMs in response to the chemical mixtures were non-additive. In contrast, the mixtures were additive in the human CS models indicating that the LSECs and KCs were necessary for the chemical interactions. These findings suggest that not only are the μOCMs more sensitive to chemical mixtures, but that the toxicity occurs through mechanisms similar to those reported in vivo.

Ongoing efforts are focused on the use of HTS assays to investigate changes in hepatic cellular functions, protein expression and functional markers in the presence of individual and cocktails of chemicals. The combination of HTS methodologies and human μOCMs has the potential to transform current testing of pharmaceuticals and environmental toxicants.