Evaluation of a battery of bacterial and human stress gene tests for their application in ecotoxicological and food testing

Transgenic cells - be it prokaryotic or eukaryotic - were introduced in environmental toxicology thanks to biotechnology and genetic engineering. Such cellular systems typically combine a promoter::operator, which acts as the sensing element, with a reporter gene coding for an easily detectable protein. In the absence of toxicant the repressor is bound to the operator and expression of the reporter is blocked. Under stress conditions, the repressor is released from the operator and reporter expression is switched on either through direct interaction with the repressor or through a cascade of connected responses (signal transduction). Important aspects in the different transgenic models are a sensitive promoter, which is reliable and relevant for a certain stress, and a reporter gene allowing efficient detection of the gene product thanks to a specific substrate or, ideally detectable as such. Most, if not all, reporter proteins can be detected using optical, immunological or electrochemical methods. Many different sensing strains and cell lines of different genera have already been constructed for the detection of specific analytes such as metals, aromatic compounds and genotoxic compounds.

In this study a battery of Escherichia coli (14) strains and 13 transgenic human (hepG2) cell lines each containing a different stress gene promoter fused to a reporter gene, are evaluated for their possible added value in ecotoxicological test programs. The promoters belong to various endpoint classes e.g. osmotic stress, oxidative damage, DNA repair induction and protein perturbation, ensuring sensitivity to a broad range of stressors.

In order to position stress gene profiling among a number of whole organism tests such as fish and daphnia acute toxicity testing, an array of reference toxicants is assayed on both types of tests and the results compared. Additionally the response profiles obtained with single toxicants and more or less complex mixtures are modelled in order to assess the power of gene profiling in predicting mode of action and toxicity of mixtures from the profile of the single compounds.