Development and Application of an Integrated Addition and Interaction Model of Mixture Toxicity

Abstract

The focus of this work was to elucidate general principles of chemical interactions among mixture constituents using mathematical modeling approaches. Mixtures of diverse acting chemicals displaying both toxicokinetic and toxicodynamic interactions were assessed. Endpoints measured in these assessments ranged from acute toxicity to molecular adaptive responses in Daphnia magna. An Integrated Addition and Interaction (IAI) model of mixture toxicity was constructed and validated using mixtures of organophosphate pesticides (malathion and parathion), chlorophenols (2-chlorophenol and 4-chlorophenol), and the P450 inhibitor piperonyl butoxide. Modeled data was compared to experimentally derived data from Daphnia magna acute toxicity assays. Toxicokinetic interactions between the organophosphates and piperonyl butoxide were characterized and incorporated into the model to accurately predict mixture toxicity. Unexpected toxicodynamic interactions between organophosphates and chlorophenols were assessed.

Next, we developed a molecular endpoint for use in mixture modeling. We explored endocrine-mediated stress signaling in daphnids towards this end. We determined that juvenoid hormones induce hemoglobin mRNA and protein production. Hemoglobin expression in daphnids provided a promising molecular endpoint for use in mixture toxicity assessments. We identified the cis-acting juvenoid response element (JRE) that is likely responsible for activating the hemoglobin gene in response to juvenoids. The JRE was used to capture a specific protein that may prove to be the juvenoid receptor. Hemoglobin 2 (hb2) was identified as a highly inducible molecular endpoint which responds to multiple signaling pathways.

The hb2 gene was used to model molecular interactions of chemical mixtures. We demonstrated that the ubiquitous herbicide atrazine induced hb2 expression, while having no obvious relatedness to juvenoid hormones. Additive effects of binary mixtures of atrazine and the juvenoid pyriproxyfen were modeled and experimentally assessed to determine whether atrazine induced the hb2 gene via the juvenoid signaling pathway. Results indicated that atrazine was not inducing hb2 through this endocrine pathway.

In summary, mathematical models of mixture toxicity can be used to predict the joint effects of diverse chemicals displaying interactions, identify unexpected interactions, and elucidate potential mechanisms of interaction.