A toxicogenomics-based parallelogram approach to evaluate the relevance of coumarin-induced responses in primary human hepatocytes in vitro for humans in vivo Anne S. Kienhuis a,b, * , Marcel C.G. van de Poll c , Cornelis H.C. Dejong c , Ralph Gottschalk a , Marcel van Herwijnen a , André Boorsma b , Jos C.S. Kleinjans a , Rob H. Stierum b , Joost H.M. van Delft a a Department of Health Risk Analysis and Toxicology, Maastricht University, Maastricht, The Netherlands b Physiological Genomics, TNO Quality of Life, Zeist, The Netherlands c Department of Surgery, University Hospital Maastricht and Nutrition and Toxicology Research Institute (NUTRIM), Maastricht University, Maastricht, The Netherlands article info Article history: Received 15 December 2008 Accepted 9 June 2009 Available online 13 June 2009 Keywords: Gene expression profiling Toxicogenomics Parallelogram approach Coumarin Primary human hepatocytes Species difference Hepatotoxicity abstract A compound for which marked species differences have been reported in laboratory animals and humans is coumarin. In rats, metabolites of coumarin are highly toxic, whereas in humans, the compound is mainly metabolized to non-toxic metabolites. In the present study, a toxicogenomics-based parallelo- gram approach was used to compare effects of coumarin on gene expression in human hepatocytes rel- evant for the situation in vivo. To this purpose, gene expression profiling was performed on human hepatocytes treated with coumarin in a pharmacological relevant and proposed toxic concentration and results were compared to a previously performed coumarin in vivo and in vitro rat toxicogenomics study. No cytotoxicity was observed in human hepatocytes at both concentrations, whereas rats showed clear toxic effects in vitro as well as in vivo. In all three systems, coumarin affected genes involved in the blood coagulation pathway; this indicates relevant responses in cases of human exposure. However, no pathways and processes related to hepatotoxicity in rats were observed in human hepatocytes. Still, repression of energy-consuming biochemical pathways and impairment of mitochondrial function were observed in human hepatocytes treated with the highest concentration of coumarin, possibly indicating toxicity. In conclusion, although species differences in response to coumarin are evident in the present results, the toxicogenomics-based parallelogram approach enables clear discrimination between phar- macological responses at pharmacological doses and proposed toxic responses at high (toxic) doses rel- evant for humans in vivo. Ó 2009 Elsevier Ltd. All rights reserved. 1. Introduction Liver-based in vitro systems, such as primary hepatocytes, pre- cision cut liver slices, and hepatic cell lines are valuable tools for studying hepatotoxicity. Many studies use liver material isolated from laboratory animals, especially rats. However, due to com- pound-related metabolism-mediated species differences in re- sponse to toxicants, these in vitro models are not always suitable to predict effects in human liver in vivo. Therefore, human hepato- cytes in culture are considered as the closest model to predict hep- atotoxicity actually occurring in livers of humans in vivo (Maurel, 1996; LeCluyse, 2001). The toxicology of coumarin merits special attention with regard to its marked species differences in both metabolism and toxicity (Cohen, 1979; Lake, 1999; Lewis et al., 2006). The use of coumarin was banned in the USA in 1954 based on reports of hepatotoxicity in rats and was recommended for withdrawal from use in the UK in 1965 (Cohen, 1979; Egan et al., 1990; Lake, 1999). Consecutive re- search revealed that coumarin metabolism in rats and some other species is quite different compared to metabolism in man (Cohen, 1979). The severe hepatotoxicity caused by coumarin in rats depends on conversion of coumarin to the toxic metabolites 0887-2333/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.tiv.2009.06.005 Abbreviations: CYP450, cytochrome P450; HBSS, Ca 2+ - and Mg 2+ -free Hank’s buffered salt solution; DMEM, Dulbecco’s minimal essential medium; FCS, foetal calf serum; BSA, bovine serum albumin; DEX, dexamethasone; MTT, 3-(4,5- dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; ITS, insulin–transferrin– selenium; CTP, cyanine 5-cytosine triphosphate; ECVAM, European centre for the validation of alternative methods; DMSO, dimethylsulfoxide; MMLV-RT, molony murine leukemia virus-reverse transcriptase; GO, gene ontology; KEGG, Kyoto encyclopedia of genes and genomes; DAVID, database for annotation, visualisation and integrated discovery. * Corresponding author. Address: Department of Health Risk Analysis and Toxicology, Maastricht University, P.O. Box 616, 6200 MD, Maastricht, The Netherlands. Tel.: +31 43 388 1097; fax: +31 43 388 4146. E-mail address: anne.kienhuis@rivm.nl (A.S. Kienhuis). Toxicology in Vitro 23 (2009) 1163–1169 Contents lists available at ScienceDirect Toxicology in Vitro journal homepage: www.elsevier.com/locate/toxinvit