Fewer animal experiments due to optimal animal model selection

20-Jun-2016 - Germany

Are the results of animal experiments transferrable to humans? This question has time and again been the subject of controversial discussions. Using a novel approach, scientists at the Federal Institute for Risk Assessment (BfR) have analysed data sets on inflammation processes in humans and mice and systematically compared them to each other. Certain mice models correlate very well with most studies conducted on humans but others don’t. "The results show that targeted research is required in order to identify the optimal animal model for a given research question", says BfR President Professor Dr. Dr. Andreas Hensel. "However, the rule still applies: where there are reliable alternative methods, these must be used instead of animal experiments. Through the German Centre for the Protection of laboratory animals (Bf3R), we support the d evelopment, validation, and use of such alternative methods to animal experiments."

In the year 2013, an international research consortium concluded on the basis of extensive genome data that inflammatory reactions of humans and mice cannot be compared. The scientists investigated the consequences of different triggers of inflammation for gene activity. As part of the study, they also compared the information of ribonucleic acid (RNA) of white blood cells of humans and mice from several scientific studies. For the first time it was concluded on the basis of extensive genome data that results from animal experiments are not necessarily transferrable to humans.

One year later, another research group made the opposite inference based on the same genome data: at the molecular level, the reaction of mice is very similar to that of humans. It was concluded that mice were therefore very useful as an animal model when researching human diseases.

It is not uncommon in science that study results are interpreted differently. It is, however, rare that diametrically opposed inferences are made from the same data. This has particularly far-reaching consequences, since given a specific question animals may be used in an experiment unnecessarily, because the chosen animal model may not be suitable in terms of the validity of the results for humans.

For this reason, the objective can only be to develop standardised approaches for a systematic data analysis of extensive genome data to ensure that such consequences are avoided. At the German Centre for the Protection of Laboratory Animals (Bf3R), these genome data were therefore studied again using the gene set enrichment analysis (GSEA) method. By means of GSEA, a direct connection between the disease and the potentially responsible biological process is established at the expression level. The study uses the statistically significant combination of those genes that are crucial for the biological process to be investigated. For example, using special computer programmes, comparisons are made to establish which genes are activated during the inflammation process and to which processes these genes can be assigned. If these molecular signal paths are changed in the same way in humans and mice, this would mean that the mouse is a good model for the study of inflammation processes in the human body.

Using the GSEA method, the BfR scientists also analysed all the changes in gene activities that are available for inflammation in humans and mice. In contrast to the usual approach, not only one part of the data was considered but the vast majority of only slightly changed gene products was also analysed systematically. The analysis of the BfR showed that in case of inflammations the reactions of some mouse models correlated very well with the data collected for humans. However, this did not apply to other mouse models. Such a data analysis is thus suitable for choosing the most appropriate animal model for the question to be investigated. Unnecessary animal experiments can thereby be avoided in future. The study published in the scientific journal EMBO Molecular Medicine further shows that the GSEA method is a very good instrument for a standardised interpretation of genome data.

Original publication

Christopher Weidner, Matthias Steinfath, Elisa Opitz, Michael Oelgeschläger, Gilbert Schönfelder; "Defining the optimal animal model for translational research using gene set enrichment analysis"; EMBO Molecular Medicine; 2016

https://www.bionity.com/en/news/158573/fewer-animal-experiments-due-to-optimal-animal-model-selection.html