When is a model sufficiently good to replace animal testing?
Aleksandra Rybacka shows in her dissertation how computational tools can be used in risk assessment of industrial chemicals and pharmaceuticals. Computational tools are planned to fully substitute animal testing in the future. Aleksandra Rybacka defends her dissertation at Umeå University on 3 June.
Aleksandra Rybacka comes from Gdańsk in Poland. She has a masterdegree in Chemoinformatics from the University of Gdańsk. Photo: Ingrid Söderbergh
Every chemical present in products we use in our daily life needs to be assessed with respect to our health and the environment. If a chemical is found to be toxic, the producers and importers of products that include such chemical need to ensure that we, the consumers, are safe. Identification of potential hazard, however, is not a simple task as chemicals can cause a variety of toxic effects, which are recognized by different tests. Additionally, mechanisms of some effects, such as disruption to the endocrine system, are not fully understood.
Computational tools are faster and cheaper alternatives to animal tests and can be applied to a big number of constantly emerging chemicals. The reliability of these tools, however, is questionable. When can we surely say that a model is sufficiently good to fully substitute animal testing? What are the missing links? For what toxic effects are models trustworthy? Aleksandra Rybacka tried to address these issues. Within her dissertation she evaluated available tools and developed new models that can be used for predicting adverse effects of industrial chemicals.
“By using mathematical equations we are able to correlate chemical structure with an adverse effect,” says Aleksandra Rybacka. “Such methodology is called quantitative structure activity relationships (QSARs) and at this moment the method is commonly used by industry to predict how chemicals accumulate in fish. But QSARs have big potential to predict other hazardous effects, such as carcinogenicity.”
The study focused on identifying substances of very high concern, that is carcinogens, mutagens and disruptors of reproductive and endocrine system. QSARs were satisfactory in predicting carcinogenic and mutagenic effects if used in so-called consensus approach that is if few models describing different aspects of the toxic response are combined at the same time. One of these aspects is biotransformation in liver, as some chemicals may originally not be toxic but can have toxic metabolites. Aleksandra Rybacka found that biotransformation was in particular important for effects related to endocrine disruption.
QSAR methodology can also aid understanding the phenomena behind a particular effect.
“We used QSARs to understand fate of pharmaceuticals in the environment by studying how they behave in waste water treatment processes. We were able to characterise main mechanisms of interacting with sludge, which can improve currently used strategies in waste water treatment,” says Aleksandra Rybacka.
About the public defence of the dissertation:
On Friday 3 June, Aleksandra Rybacka, Department of Chemistry at Umeå University, defends her dissertation entitled: A step forward in using QSARs for regulatory hazard and exposure assessment of chemicals.The public defence (in English) will take place at 13:00 KB3B1 in the KBC-building.External examiner is Dr Uko Maran, Institute of Chemistry, University of Tartu, Estonia. Main supervisor is Patrik Andersson.