Citation:

Corton, C. AND J. Rooney. Integrating toxicogenomics data into cancer adverse outcome pathways. Society of Toxicology, Baltimore, Maryland, March 13 - 16, 2017.

Impact/Purpose:

As the toxicology field continues to move towards a new paradigm in toxicity testing and safety assessment, there is the expectation that models will be increasingly developed and refined to enable better prediction of cancer risk from short-term tests in animals or even cultured cells. The growing number of publically available genomic profiling studies from chemically-treated animals and cells provides opportunities to construct models for prediction of molecular initiating events (MIEs) and key events (KEs), and assess linkages between the events as KE relationships (KERs) in cancer adverse outcome pathways (AOPs). In this study we use a number of computational approaches to build biomarkers and integrate findings into the AOP framework.

Description:

Integrating toxicogenomics data into adverse outcome pathways for cancer.J. Christopher CortonNHEERL/ORD, EPA, Research Triangle Park, NCAs the toxicology field continues to move towards a new paradigm in toxicity testing and safety assessment, there is the expectation that models will be increasingly developed and refined to enable better prediction of cancer risk from short-term tests in animals or even cultured cells. The growing number of publically available genomic profiling studies from chemically-treated animals and cells provides opportunities to construct models for prediction of molecular initiating events (MIEs) and key events (KEs), and assess linkages between the events as KE relationships (KERs) in cancer adverse outcome pathways (AOPs). Using a number of computational approaches including the comparison of wild-type and nullizygous mice as well as annotated conditions in which key events in known cancer pathways occur, we derived gene expression biomarkers that accurately predict events, including activation of transcription factors (TFs) (e.g., aryl hydrocarbon receptor (AhR), constitutive activated receptor (CAR), peroxisome proliferator-activated receptor alpha (PPARalpha), and sex steroid receptors) and increases in cytotoxicity and downstream KEs including increases in oxidative stress (using Nrf2 activation as a surrogate) and inflammation. These biomarkers were used to comprehensively assess MIE/KE modulation by ~200 chemicals linked to known incidences of hepatocellular adenomas and carcinomas in rodents under chronic exposure conditions. The analysis highlighted a number of features of chemically-induced mouse liver tumor induction: 1) most carcinogenic chemicals activated multiple MIEs; 2) simultaneous assessment of known AOPs uncovered unexpected KERs between TFs (e.g., activation of CAR and Nrf2); and 3) in vitro models evaluated were generally poor surrogates of the intact liver in prediction of events. Because chemical-independent (and potential modulating) factors that regulate the MIEs/KEs are simultaneously evaluated in the database (e.g., diet, genetic background, and life stage), the integrated predictions may inform assessment of cumulative risk using AOPs. (This abstract does not represent EPA policy.)

Record Details: