You are here:
Leveraging toxicogenomics data to build predictive biomarkers supporting AOP assessment
Citation:
Corton, C. Leveraging toxicogenomics data to build predictive biomarkers supporting AOP assessment. HESI Genomics Meeting, Montreal, Canada, CANADA, May 25 - 26, 2017.
Impact/Purpose:
Chemicals induce liver cancer in rodents through well characterized adverse outcome pathways (AOPs) that include molecular initiating events (MIEs). In addition to genotoxicity, these include nongenotoxic mechanisms of cytotoxicity and receptor activation (aryl hydrocarbon receptor (AhR), constitutive activated/androstane receptor (CAR), estrogen receptor (ER), peroxisome proliferator-activated receptor α (PPARα)). We hypothesized that measurement of MIEs and downstream key events (KEs) in short-term assays will allow the identification of chemicals that are tumorigenic in the liver in two-year bioassays.
Description:
Chemicals induce liver cancer in rodents through well characterized adverse outcome pathways (AOPs) that include molecular initiating events (MIEs). In addition to genotoxicity, these include nongenotoxic mechanisms of cytotoxicity and receptor activation (aryl hydrocarbon receptor (AhR), constitutive activated/androstane receptor (CAR), estrogen receptor (ER), peroxisome proliferator-activated receptor α (PPARα)). We hypothesized that measurement of MIEs and downstream key events (KEs) in short-term assays will allow the identification of chemicals that are tumorigenic in the liver in two-year bioassays. We tested this hypothesis using the TG-GATES study to measure 6 MIEs and 3 KEs (oxidative stress, cell proliferation, liver to body weights) across 133 chemicals at three doses and 8 time points (up to 29d) for a total of 3127 chemical-dose-time comparisons including 77 chemicals linked to doses with known effects on liver tumor induction. AhR, CAR, ER, and PPARα gene expression biomarkers were built using microarray comparisons from the livers of rats treated with prototypical activators of the receptor. The genotoxicity biomarker was comprised of 7 p53-responsive genes that are induced upon DNA damage. Predictive accuracies of the biomarkers ranged from 91% to 98%. Examination of the MIEs/KEs across time showed that sustained activation was most clearly associated with liver tumorigenesis. The Toxicological Priority Index (ToxPi) was used to rank chemicals based on their ability to activate MIEs/KEs at all 8 time points. The tumorigenic chemicals clearly gave the highest ranked scores. Using a ToxPi threshold 75 out of 77 chemicals (97%) were correctly classified. Using only the 8, 15, or 29d data, the 29d time point gave the most accurate ranking of chemicals. Our AOP-directed approach could be used to identify chemicals at low risk of causing liver tumors that have human relevance.