Science Inventory

The use of adverse outcome pathway-based toxicity predictions: A case study evaluating the effects of imazalil on fathead minnow reproduction


Jensen, K., G. Ankley, B. Blackwell, J. Cavallin, W. Cheng, R. Conolly, D. Feifarek, M. Kahl, S. Poole, E. Randolph, T. Saari, Dan Villeneuve, AND K. Watanabe. The use of adverse outcome pathway-based toxicity predictions: A case study evaluating the effects of imazalil on fathead minnow reproduction. SETAC North America, Minneapolis, MN, November 12 - 16, 2017.


The vision for toxicity testing in the 21st century is to enhance our ability to predict hazard and risk based on data that can be acquired more efficiently and cost effectively than is possible using traditional whole organism toxicity testing. The adverse outcome pathway (AOP) framework was developed to aid the translation of mechanistic data (e.g., effects of chemicals on enzyme activities, gene expression, receptor binding, etc.) into information that’s meaningful to risk managers (e.g., impacts on survival, growth, reproduction, disease, etc.). However, in order for regulators and others in the scientific community to accept these new/alternative approaches to chemical safety assessment, they need to develop confidence, through case studies, that alternative data and AOPs can provide reliable predictions of hazard and risk. The present case study is aimed at establishing that confidence for an important mechanism of endocrine disruption and its potential ecological consequences. The results show that AOP-based approaches were useful for predicting the effects of a previously untested chemical and that the accuracy of quantitative predictions was within the range of variability observed in whole organism studies.


Product Description: As a means to increase the efficiency of chemical safety assessment, there is an interest in using data from molecular and cellular bioassays, conducted in a highly automated fashion using modern robotics, to predict toxicity in humans and wildlife. The present study provides an example of how a series of biologically-based computer models could be used to predict the toxicity of a chemical previously untested in a fish reproduction assay. Experiments were then conducted to test the prediction. This research builds confidence in the ability to use these types of predictive approaches to aid decision-making concerning chemical safety. Adverse outcome pathways (AOPs) provide a framework supporting greater use of mechanistic data measured at lower levels of biological organization as a basis for regulatory decision-making. An AOP description linking inhibition of aromatase (cytochrome P450 [cyp] 19) to reproductive dysfunction was reviewed for scientific and technical quality and endorsed by the OECD ( High throughput screening in US EPA’s Toxcast program identified the azole fungicide, imazalil, as an endocrine active chemical capable of inhibiting mammalian cyp19 and likely 17á-hydroxylase/17,20lyase (cyp17) in vitro. Based on these results, imazalil was selected as a case study chemical to test an AOP-based hazard prediction. Twenty-four hour exposures with fathead minnow (Pimephales promelas), focused on effects on production and circulating concentrations of 17â-estradiol (E2), key events in the AOP, were conducted to verify in vivo activity. A computational model of the fish hypothalamic-pituitary-gonadal-liver axis and a statistically-based model of oocyte growth dynamics were used to predict impacts of different concentrations of imazalil on multiple key events along the AOP, assuming continuous exposure for 21 d. Results of the model simulations were used to select test concentrations and design a fathead minnow reproduction study in which fish were exposed to 20, 60, or 200 µg imazalil/L for durations of 2.5, 10, or 21d. Within 60 h of exposure, female fathead minnows showed significant reductions in ex vivo production of E2, circulating E2 concentrations, and significant increases in the ovarian expression of mRNA transcripts coding for cyp19a1a, cyp11a (cholesterol side-chain cleavage) , and cyp17. A concentration-dependent decrease in cumulative fecundity was also detected for fathead minnow pairs exposed continuously for 21 d. Overall, results of the study provide strong support for the qualitative relationships represented in the AOP and provide further evidence of concentration-response and temporal concordance among key events. The quantitative models were generally consistent with in vivo potency, supporting the use of this established AOP linking aromatase inhibition to reproductive impairment in fish as a means for predictive risk assessment. The contents of this presentation neither constitute nor necessarily reflect US EPA policy.

Record Details:

Product Published Date: 11/16/2017
Record Last Revised: 11/13/2017
OMB Category: Other
Record ID: 338281