3Rs-Centre Utrecht Life Sciences

Faculteit Dierengeneeskunde
September 2016
New ontology model could replace and reduce animal use in developmental and reproductive toxicity testing

A lot of laboratory animals are used for the screening of potential toxic substances in developmental and reproductive toxicity testing. To replace and reduce animal use in these research areas, Prof. Dr Piersma wants to make a change by introducing a computer model which integrates chemical and toxicological data with information about prenatal development, called an ontology.

Toxic substances have to be tested for their safety and efficacy, before they are allowed for human use in many ways (for example drugs, industrial chemicals, pesticides, food ingredients or contact substances). This is not only relevant for adults, but also for the unborn or embryo. All kinds of substances can be potentially toxic during embryonic development, addresses Prof. Dr. Aldert Piersma, senior scientist developmental toxicity at the National Institute for Public Health and the Environment (RIVM) and Professor by Special Appointment in Reproductive and Developmental Toxicology at the Institute for Risk Assessment Sciences (IRAS) of Utrecht University. Unfortunately, animal testing is still the golden standard in testing potential hazardous substances for harmful effects during human embryonic development.

Since 2007, the European REACH regulation (Registration, Evaluation, Authorisation and Restriction of Chemicals) program1 demands manufacturers to provide safety information on their potential toxic substances, also for predicting developmental toxicity. This accounts for an increased use of laboratory animals, because current available in vitro non-animal test methods do not suffice in determining the hazardous properties of chemicals. Around 65% of all laboratory animals are used in reproductive and developmental toxicity testing under REACH. Piersma wants to make a change towards non-animal testing in this research field, by integrating two research areas that have been relatively isolated so far: developmental biology and toxicology.  

New Collaboration “Substances and Prenatal Health Monitoring”

To foster the relevant integration of the two isolated research areas, a collaboration has been organized within a public-private partnership between RIVM, IRAS and the Dutch Society for the Replacement of Animal Testing (Stichting Proefdiervrij). The latter also provides funding by means of a crowdfunding initiative for Pierma's research project2. Within this new collaboration, Piersma wants to improve human reproductive and developmental hazard and risk assessments, by especially looking for innovative non-animal research methods in this field of research.

Foto: Stichting Proefdiervrij

Developmental toxicity ontology database

Currently, Piersma is working on the development of a new computational analysis method, called “developmental ontology”. In the ontology, large datasets of existing publicly available literature will be combined and analysed for a better human risk assessment. Specifically, the dataset includes several phenotypic and genetic effects of chemical substances on one hand (the chemical data), and information about substances causing reduced fertility or developmental effects of the embryo (the developmental toxicity data), on the other hand. Also, information about all processes of normal embryonic development is added. So the database will entail information from three different perspectives: chemical, toxicological and biological. The new database will be based on the ontology method as initiated by the US Environmental Protection Agency (EPA), a collaborator of Piersma. Funding is provided by CEFIC-LRI3 (Long- Research Initiative Programme of the European Chemical Industry Council (CEFIC)), because the project generates a better understanding of the impact of chemicals and translates research outcomes to a better product safety.

Better translation, fewer animals
One of the main expected success factors of this non-animal research method is the integration of multiple research areas in the new ontology. Existing ontologies do not include a combination of developmental processes and toxicities, but are only focusing on either one of them. Like the DevTox Database of the International Life Sciences Institute (ILSI), which only includes published data of fetal malformations at term pregnancy4.

Piersma is happy to introduce this new model and plans to come up with a proof-of-principle within two years. By using this new computer model, new relations between substances and developmental effects can be discovered and new hypotheses can be formulated and tested, without the use of animals. The model will be updated as soon as new discoveries are made, as an on-going-process. By choosing the right battery of tests, the predictive capacity of potential hazardous substances will increase. This will improve research, because a better translation towards the human situation can be made. And, more importantly, this will tremendously decrease the number of animals used in developmental toxicity testing.

1: Read more about REACH and animal testing.
2: Read more about the crowdfunding initiative of the Dutch Society for the Replacement of Animal Testing.
3: More information can be found on the CEFIC-LRI Project website.
4: Read more about the DevTox Database.