Predicting chemical toxicity without animal testing

A study has just been published in the open-access journal Nature Communications which shows how predictions of chemical toxicity in humans can be made without animal testing.

If the research results are confirmed, albeit with reference to toxicity tests using animal subjects, the cell-based models could lead to better, more ethical methods for testing the toxicity of compounds such as pesticides, industrial chemicals, food additives and drugs.

Ruili Huang and her colleagues tested the activity of 10,000 chemicals compounds in various concentrations as they interacted with biological targets such as nuclear receptors and cellular pathways. With more than 50 million data points generated, and information on the structure of the tested chemicals, the researchers devised toxicity models that may be used to predict adverse effects of the compounds in animals or humans.

The results need further testing using additional pathways and targets, but, once that is done, the scientists propose that their in vitro methods could be employed in routine toxicity testing, and help prioritise the selection of compounds for more in-depth investigation.