Reverse engineering adverse outcome pathways

Abstract

The toxicological effects of many stressors are mediated through unknown, or incompletely characterized, mechanisms of action. The application of reverse engineering complex interaction networks from high dimensional omics data (gene, protein, metabolic, signaling) can be used to overcome these limitations. This approach was used to characterize adverse outcome pathways (AOPs) for chemicals that disrupt the hypothalamus-pituitary-gonadal endocrine axis in fathead minnows (FHM, Pimephales promelas). Gene expression changes in FHM ovaries in response to seven different chemicals, over different times, doses, and in vivo versus in vitro conditions, were captured in a large data set of 868 arrays. Potential AOPs of the antiandrogen flutamide were examined using two mutual information-based methods to infer gene regulatory networks and potential AOPs. Representative networks from these studies were used to predict network paths from stressor to adverse outcome as candidate AOPs. The relationship of individual chemicals to an adverse outcome can be determined by following perturbations through the network in response to chemical treatment, thus leading to the nodes associated with the adverse outcome. Identification of candidate pathways allows for formation of testable hypotheses about key biological processes, biomarkers, or alternative endpoints that can be used to monitor an AOP. Finally, the unique challenges facing the application of this approach in ecotoxicology were identified and a road map for the utilization of these tools presented.

Pathway-based toxicity: history, current approaches and liver fibrosis and steatosis as prototypes

“…AOPs serve the purpose of linking observations at all levels of biological complexity and can identify alternative endpoints for use in hazard assessment. During the Pellston workshop, two complementary approaches to AOP development were demonstrated by case examples: using domoic acid as a case study, an AOP was created from existing information in the literature (Watanabe et al, 2011); using disruption of the hypothalamus-pituitary-gonadal endocrine axis in fathead minnows as a case study, omics data were used to reverse-engineer pathways (Perkins et al, 2011). two other publications from that workshop explore quantitative prediction models that begin to address consideration of the complexities of predicting outcomes at higher-level biological organization from information at relatively less complex levels,…”

Species extrapolation for the 21st century

“…Furthermore, many compounds elicit their toxic response through multiple parallel interactions. Building models for such networks is extremely challenging, although harnessing the networks constructed from transcriptomic data generated through temporal and dose-dependent studies and anchored through comprehensive physiological and cellular measurements may well enable even these complex scenarios to be addressed eventually [16]. Substantial experimental data will be necessary to validate the key concepts and current limitations outlined in this report to address the challenge of extrapolating models at the omic, life history, and toxicokinetic levels.…”

“…Several comprehensive pathway resources are available [15], and some of these resources contain comprehensive biochemical data for metabolic pathways relating to pathway kinetics. Regulatory networks, however, are less comprehensively represented, and many are still being inferred using reverse engineering [16]. However, computational formats have been developed (Systems Biology Markup Language, www.sbml.org; and Biological Pathways eXchange, www.biopax.org), allowing novel pathways to be relatively straightforwardly integrated into predictive software.…”

Adverse outcome pathways and ecological risk assessment: Bridging to population‐level effects

“…In the absence of knowledge about the functional role of transcripts that happen to respond to a treatment, a more generalized analysis could still be used to identify those transcripts that are shown to be predictive of toxic effects. Such informative patterns can be ascertained by using a variety of network analysis tools [70]. Although outside of the field of toxicology, a very complete approach to identifying regulatory networks in yeast was employed by Zhu et al [71].…”