Tox21 Researchers Analyze Potential Negative Effects on Human Health of Thousands of Chemicals

Throughout our lives, we may be exposed to thousands of chemicals in our food, household cleaning products, medicines and environment. However, scientists know little about the potential for most of these substances to be hazardous to human health (i.e., their toxicity). Traditional toxicity testing often involves animals, which can be a slow and costly approach that often does not predict accurately how humans will respond to a chemical.

To address these challenges, researchers from the Toxicology in the 21st Century (Tox21) consortium have developed faster, cheaper and more effective toxicity testing methods. Tox21 is a collaboration among researchers from NIH’s NCATS and the National Institute of Environmental Health Sciences (NIEHS), the Environmental Protection Agency (EPA), and the Food and Drug Administration (FDA).

Tox21 scientists use a robotic system located in NCATS’ Rockville, Maryland, laboratories to perform automated tests (called assays) that expose cells and proteins to thousands of chemicals. Tox21 scientists currently are working on the second phase of the effort to screen a library of more than 10,000 chemicals (Tox21 10K) — the largest analysis of its kind. The library consists of substances to which people are regularly exposed. These include pesticides, industrial chemicals, food additives and drugs. The effort already has generated nearly 50 million data points that will provide information about the potential effects of the chemicals on human biological functions.

Tox21 scientists published the first results from screening the Tox21 10K library in the July 11, 2014, issue of Scientific Reports. The group developed an assay to test the effects of the approximately 10,000 chemicals on estrogen receptors, which are activated by the estrogen hormone and convey signals to the body that regulate reproductive functioning. Some chemicals may cause certain cancers or birth defects by mimicking estrogen and activating its receptors.

For example, bisphenol A (BPA), an industrial chemical used to make plastics and found in water bottles, has been linked to birth defects in infants. Animal studies suggest that BPA could cause these effects by mimicking estrogen. Many other compounds in food and the environment similarly may affect the estrogen receptor by turning it on or off and pose health hazards; therefore, performing the chemical screen with the estrogen receptor assay was a high priority for the Tox21 group. The screens reported in the paper identified chemical classes that were known to interact with estrogen receptors, as well as some not previously known to scientists.

Enter the NCATS Robots

The screen’s unprecedented scale was not its only distinguishing feature. The NCATS robotic system was developed to perform a technique called quantitative high-throughput screening. This innovative method enables the team to run each compound through the screen three times at 15 different concentrations. This approach is a key feature of the NCATS’ screening programs, and “it gives us higher confidence in the data and assures us that any given result is not just a trend, but real biological activity of a molecule,” explained Anton Simeonov, Ph.D., acting deputy director of the NCATS Division of Preclinical Innovation and an author on the paper.

Additionally, the assays included two versions of the estrogen receptor, one partial and one full length. The partial version of the receptor contains the ligand binding domain (LBD), which is the region that attaches to compounds, including the estrogen ligand. The full-length version of the receptor better mimics the biological environment, but chemicals could interact with it in areas other than the LBD, producing a different response than would be expected from estrogen and similar compounds.

Screening the chemicals against both versions of the receptor allowed the team to compare the results to get a more accurate idea of which compounds truly affect the receptor by interacting with it like estrogen.

“This study’s scale and comprehensive nature, with its exhaustive testing of compound activities in two different receptor versions, make it unique. It is the first of its kind to test so many chemicals against the estrogen receptor, which is an important focus of toxicology,” Simeonov said.

An Effective Federal Collaboration

The successful completion and publication of the estrogen receptor screen shows the ability of several federal agencies to pool resources and provide complementary expertise to carry out a large-scale project.

“This collaboration is probably one of the best-coordinated events in toxicology and human health assessment,” said Robert Kavlock, Ph.D., director of the National Center for Computational Toxicity at the EPA and an author on the paper. “It is a seamless organization — it’s often hard to tell who works for what agency.”

Raymond Tice, Ph.D., chief of the National Toxicology Program at NIEHS and another of the paper’s authors, added, “This project is something the government is uniquely capable of doing, and this initial publication of the estrogen receptor data demonstrates the success of this federal collaborative model.”

The Tox21 group submitted the data reported on in the paper to the National Library of Medicine’s PubChem website, making it available for the public and other scientists to freely access. The results will enable the Tox21 team — as well as other scientists — to prioritize chemicals for further in-depth studies to define their effects on human health.

In releasing the data publicly, the Tox21 group plans to harness the power of crowdsourcing to challenge the general scientific community to develop innovative and original ways to analyze the data. “The data provide scientists with the initial roadmap that will enable them to interpret the results and predict the chemicals’ effects on human biology,” Simeonov said.

In the near future, Tox21 experts plan to publish similar data sets from screens of the Tox21 10K library in assays of stress response pathways, which are designed to identify compounds that can damage or kill cells. The team will continue to test the chemical library and release screening data, moving Tox21 closer to its goal of transforming toxicity testing into a cheaper, more efficient, and more effective enterprise, bringing about the ultimate endpoint of improved human health.

In addition, NCATS recently launched the Tox21 Data Challenge 2014, which is a crowdsourcing competition to develop predictive models for chemical toxicity from data generated from 12 Tox21 assays, including the estrogen receptors and stress response pathways information. This public sharing of Tox21 data and methods reflects the NCATS mission to disseminate knowledge and information to arm scientists with tools to carry out more efficient and transformative translational research.


Posted August 2014