Integrating Toxicity Data to Gain a Systems-Level Understanding for Safety Evaluation


The goal of the Symposium Session “Systems Toxicology Approaches to the Science of Safety Evaluation” during the SOT 58th Annual Meeting and ToxExpo was to review recent advances in the application of systems toxicology in safety evaluation. The chairs of the symposium defined systems toxicology as “a transformational sub-discipline within toxicology that applies approaches that have evolved from modeling of complex biological systems to toxicology-related questions.” Over the course of the discussion with the four speakers, this definition became more fluid, with the more inclusive definition described by Dr. Jason Papin as simply: studying more than the individual parts.

Dr. Ed Lobenhofer from Amgen described examples of using existing knowledge to create target liability assessments that can help predict safety endpoints of interest. He called on Stephen Covey’s mantra, “begin with the end in mind,” to add endpoints of safety concern earlier in the process of drug discovery. He gave two examples of molecular targets identified from genomic literature that were associated with osteoarthritis and a different protein associated with dyslipidemia. His team was able to create a liability assessment and assign values of concern and safety risk associated with drugs that target these proteins, based on literature searches. By testing and looking for these outcomes of interest, they were able to screen out these chemicals and avoid wasting resources in the phase I trial of molecules of interest.

Dr. Wendy Freebern from Bristol-Myers Squibb shared their methodology in making immunotoxicity predictions for a drug candidate. She described how gene expression profiles from antigen-presenting dendritic cells can be used to predict whether a candidate drug is a sensitizer, by iteratively testing known sensitizing agents. Signatures associated with cytokines and chemokines also find utility in understanding T cell–dependent antibody response. By integrating data from gene expression profiles from innate immune cells, cytokines/chemokines, proteomic profiles, complement cascade activation, and histopathologic changes, their group has become better at predicting skin sensitizers sooner.

At the University of Pittsburgh, Dr. D. Lansing Taylor applies the human vascularized liver acinus microphysiology system (vLAMPS) that embodies microenvironments of the liver representing real biology, by recreating the three zones of a liver acinus with different oxygen tensions, making it an attractive model to use for liver toxicity studies. In vitro studies often ignore the real oxygen tension found in vivo when recreating models. vLAMPS uses fluorescent beads that are quenched in the presence of oxygen and produce a gradient based on oxygen tension. Many redox toxicologists use tri-gas chambers with oxygen tension that represent in vivo levels to study redox signaling in the cell. Some have suggested that atmospheric levels of oxygen over cell cultures increase the expression of nrf-2 and other redox signaling pathways, making it harder to detect effects of a redox stressinducing toxicant compared to controls.

The last speaker, Dr. Jason Papin from University of Virginia, took us through a stoichiometric matrix that they use to create metabolic networks that are species specific. The ability to abstract biochemical data into a matrix makes it amenable to mathematical manipulation and allows toxicologists to include fluxes and add restraints in models to predict changes in biochemistry due to an insult. By recreating species-specific maps, their model was able to make predictions of metabolic changes in the liver that accompany exposure to chemicals or drugs pretty well. There were changes that were specific to rats and humans, which were predicted and validated in the lab. While really cool data that shows how systems biology approaches can be used in toxicology, it also throws light on underlying biochemical differences between rats and humans, where metabolically, the liver of the two species responds differently to the same chemical/drug exposure.

As a member of a lab that studies metabolomic changes that accompany exposure to environmental toxicants, this finding reminds me that crucial biochemical information needs curation. There should be increased focus in recreating species-specific maps that are complete, with regular updates following those made by metabolomic consortia in the US and Europe.

This blog was prepared by an SOT Reporter. SOT Reporters are SOT members who volunteer to write about sessions and events they attend during the SOT Annual Meeting and ToxExpo. If you are interested in participating in the SOT Reporter program in the future, please email Giuliana Macaluso.

Recent Stories
Volunteers Make the Undergraduate Diversity Program Possible—Thank You!

Toxicology History: Reference Materials

Reflections on the 2019 Domestic ToxScholar Outreach Program