Transcriptomic profile indicative of immunotoxic exposure; in vitro studies in peripheral blood mononuclear cells

Investigating the immunotoxic effects of exposure to chemicals usually comprises evaluation of weight and histopathology of lymphoid tissues, various lymphocyte parameters in the circulation and immune function. Immunotoxicity assessment is time consuming in humans or requires a high number of animals, making it expensive. Furthermore, reducing the use of animals in research is an important ethical and political issue. Immunotoxicogenomics represents a novel approach to investigate immunotoxicity able of overcoming these limitations. The current research, embedded in the EU project NewGeneris, aimed to retrieve gene expression profiles that are indicative of exposure to immunotoxicants. To this end, whole genome gene expression was investigated in human peripheral blood mononuclear cells (PBMC) in response to in vitro exposure to a range of immunotoxic chemicals (4-hydroxy-2-nonenal, aflatoxin B1, benzo[a]pyrene, deoxynivalenol, ethanol, malondialdehyde, polychlorinated biphenyl 153, 2,3,7,8-tetrachlorodibenzo-p-dioxin) and non-immunotoxic chemicals (acrylamide, dimethylnitrosamine, 2-amino-3-methyl-3H-imidazo[4,5-F]quinoline, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine). Using Agilent oligonucleotide microarrays, whole genome gene expression profiles were generated, which were analysed using Genedata’s Expressionist® software. Using Recursive Feature Elimination and Support Vector Machine, a set of 48 genes was identified that distinguishes the immunotoxic from the non-immunotoxic compounds. Analysis for enrichment of biological processes showed the gene set to be highly biologically and immunologically relevant. We conclude that we have identified a transcriptomic profile indicative of immunotoxic exposure. Keywords: Genome wide gene expression analysis, Transcriptomic profile indicative of immunotoxic exposure For analysis of whole genome gene expression by microarray, PBMC from five independent donors per compound were exposed for 20 hours to three concentrations, i.e. the 100% and two serial ten-fold dilutions (10% and 1%), and a DMSO or PBS vehicle control. Exposed samples were always labelled with Cy5, whereas the vehicle control samples were labelled with Cy3, and were competetively hybridized on 4x44K Agilent microarrays.