Applying mixture toxicity modelling to assess the contribution of detected chemicals to in vitro effects : An example from the Danube River
Abstract
Surface water can contain a diverse range of organic micropollutants stemming from a number of sources. Targeted chemical analysis alone is unable to detect all micropollutants present and in vitro bioassays can be applied complementary to assess the biological effects of complex environmental mixtures. The aim of this study was to test the hypothesis that the in vitro effects in large human-impacted rivers are primarily driven by the mixture effects of many chemicals, rather than being dominated by a particular chemical. In this study a suite of bioassays indicative of activation of the aryl hydrocarbon receptor (AhR), activation of the pregnane X receptor (PXR), activation of the estrogen receptor (ER), adaptive stress responses to oxidative stress (Nrf2), genotoxicity (p53) and inflammation (NF-κB) and fish embryo toxicity were applied to large volume solid phase extracted (LVSPE) samples from the Danube River. This was complemented with targeted chemical analysis of 272 chemicals, with available effect concentrations for the detected chemicals collected from the literature or the US EPA ToxCast database. The contribution of detected chemicals to the in vitro effects was assessed using the bioanalytical equivalent concentration (BEQ) concept. Available effect concentrations for between 0 to 13 detected chemicals were found for the different assays. Up to 80% of ER activation was explained by five detected chemicals, with the hormone estrone and the phytoestrogen genistein contributing significantly, while up to 71% of AhR activation could be explained by three chemicals. In contrast, less than 0.2% of the effect was explained by the detected chemicals for PXR activation, oxidative stress response, p53 response and fish embryo toxicity. As a significant fraction of the effect could not be explained by detected chemicals, this study emphasises the importance of applying bioanalytical tools along with chemical analysis for water quality monitoring. The study was limited by the availability of effect concentrations for the detected chemicals and consequently further fingerprinting of water pollutants is recommended.