Science - March 9, 2017

Animal testing going down

Roelof Kleis

Wageningen researchers are using fewer and fewer lab animals. The number of animals used for research purposes has been halved in five years, mainly due to stricter rules and research budget cuts. And WUR is working on new technologies which can replace animal testing.

photo Guy Ackermans; shutterstock; Rob de Winter; Rikilt.

The latest figures, for 2015, show that a total of 33,631 animals were used for Wageningen research and teaching (see figure 1). Roughly one quarter of these were used at the university; the rest were used by Wageningen Research institutes.

Besides these animals, 20,526 fish were caught by Wageningen Marine Research in order to study fish stocks and the make-up of fish populations in the North Sea. This figure is reported separately because this fisheries monitoring has only been included in Dutch animal testing legislation since 2014, and these high numbers could hide the downward trend in animal testing.

These figures do not include every animal used by scientists for research. The law only counts vertebrates and cephalopods as lab animals, explains university animal testing expert Rob Steenmans. ‘So using insects does not count as animal testing. It also only counts if the animal is being used for particular research goals. Consumption does not count, for instance. And there has to be a certain level of distress for the animal in question. A lot of nutrition experiments do not come under the law because the animal is not disturbed by them.’

Downward trend

At both Wageningen Research and the university, according to annual reports, animal testing was halved between 2011 and 2015. A closer analysis gives rather a different picture, though. Much the biggest drop in numbers took place in 2015, the last year included in the report. According to Steenmans and animal testing expert Jan van der Meulen of Wageningen Research, this ‘dip’ is a result of the new animal testing legislation which came into force at the end of 2014. This legislation made the licensing process lengthier, and more complicated and expensive. Researchers had to get used to the new way of working. These changes distort the picture considerably, says Steenmans. ‘The figures for 2016 should be submitted to the ministry next month. I am pretty sure the numbers will have gone up again.’ Van der Meulen expects the same. So the ‘dip’ in 2015 will remain an exception in what is nevertheless a clear downward trend.

That downward trend is partly the result of the wish of the general public and the government to reduce the use of animals in research. The government’s aim is for legally required safety tests on chemicals and allergy tests for new products, food ingredients, pesticides and vaccinations to be free of animal testing by 2025. Applications to use animals for experiments are therefore being assessed increasingly critically.

But the financial side of this is at least as important a factor, says Van der Meulen. ‘There is simply less money for testing on animals. The Ministry of Economic Affairs, in particular, is commissioning less research in the areas of animal welfare and health. Animal experiments are expensive so that is one of the first things to be cut when savings have to be made. The abolition of the product boards has also led to a drop in assignments.’ Steenmans also points out that scientists are going about their research differently. ‘There is a shift from experiments in animal testing facilities to observations in the field. That is a very different way of doing research, which doesn’t usually involve anything that counts as an animal experiment in the eyes of the law.’

Testing mussels without rats

Mussels can contain toxic substances from algae which can give people diarrhoea. During the mussel season, Rikilt gets weekly deliveries of mussels for testing. Until 2011 this was done on rats or mice. Then a PhD candidate Arjen Gerssen came up with an analytical method which did not use any animals. Instead it made use of LC-MS, a combination of liquid chromatography and mass spectrometry. The first technique separates the various substances in a sample, and the second then identifies them by comparing them with known substances.


Gerssen’s method was officially approved by the European Union in 2011. Rikilt researcher Toine Bovee estimates that this means 300,000 fewer lab mice. But that can be improved on. The Gerssen method only works for known toxins, so a number of EU countries are still testing toxins on mice. Bovee thinks he has found a solution to this by developing a new bioassay. ‘An assay with neurons, which reveals unknown toxics as well through an effect that is easy to spot with coloration.’ The method will be published shortly. Bovee has high hopes that it will become standard in Europe.

Digital human replaces lab animal

If you want to know how the human gut works, you shouldn’t really settle for results based on animal testing. And thanks to the latest technology, you soon won’t need to. With their gut-on-a-chip project, scientists from Wageningen and Twente are seeking to simulate the microworld of intestinal cells.

12-ACH foto kader Darm op chip.foto.jpg

In vitro research on gut cells is not new in itself,’ explains Hans Bouwmeester of the Toxicology chair group. ‘The classic approach is to grow cells on a membrane with a medium on both sides. With this static apparatus you can measure the permeability of the intestinal epithelium. What is new about the gut-on-a-chip is that it is a dynamic model. The contents of the gut flow over the top of the gut cells, and the bloodstream over the bottom. This produces a much more realistic approach to the gut.’

The gut-on-a-chip looks simple enough: two glass slides with a layer of intestinal cells between them which are almost invisible to the naked eye. There are connections on the chip for the inflow and outflow of the simulated gut content and bloodstream. So does a bunch of cells like this work like a real intestine? ‘No, of course not,’ responds Bouwmeester. ‘But in some cases that makes no difference at all. It just depends on which question you want to answer. You can make the system as complex and realistic as you like. For example, you can add mucus cells which simulate the slimy layer on the gut side. Or you can add gut microbiota.’

At the end of last year Bouwmeester and his colleagues gained NWO funding to elaborate on the system. Rikilt is involved in the project too.


The bare figures on animal testing only tell part of the story, however. More important, perhaps, is the suffering caused by an animal experiment. Taking blood from a mouse is a very different thing to inserting a rumen cannula – a hole with a lid on it – into a cow’s stomach. So experiments are rated according to the level of distress – fear or pain – caused to the animal. The scale ranges from terminal through light and moderate to severe distress. Terminal is seen as the mildest form of distress. This may sound illogical but it is not, says Van der Meulen. ‘This category covers animals which are experimented on under general anaesthetic and never regain consciousness. So they do not suffer.’

The distress scores show that far and away the most Wageningen lab animals suffer slight distress (see figure 2). This may include killing an animal for its organs, taking limited amounts of blood or injecting it with substances that only affect it mildly. Inserting a rumen cannula in a cow comes under moderate distress, while emptying the stomach through the cannula only counts as slight distress. Two examples of severe distress are toxicity testing which ends in death, and total isolation of a sociable animal.

Digital human replaces lab animal


WUR toxicologist Jochem Louisse developed a ‘digital lab animal’ for his PhD in 2012. This is a computer model which describes what happens to a substance in the body. The model can be used to forecast precisely where the substance collects and the dilution rate in the blood or in an organ. It works the other way round as well. Once you know how much a cell can cope with, you can calculate what dose has had a particular effect. Louisse is now refining the digital lab animal, with the help of a group of PhD researchers. But his real aim is to go one step further and develop a digital test human: a computer model of human physiology combined with in vitro tests on human material. Louisse: ‘The digital lab animal predicts a safe dose in animals based on in vitro tests on animal cells. A safe dose for humans can be worked out on that basis, but it give a false sense of security. Lab animals are not humans.’ One of the things Louisse is now working on is human heart cells, so as to predict the toxic effects of substances on the heart.

Meanwhile, the digital lab animal is proving its worth. ‘We are working with BASF, for example, on using our model as a toxicological test prior to developing new substances. This gives them a cheap way of ruling out substances at an early stage because they are high-risk toxicologically.’

Number of tests

The number of animals used for experiments tells us nothing about the number of animal experiments carried out by WUR. The research institutes between them conducted 90 animal experiments in 2015. The university conducted 60 animal experiments, with the smallest involving just two animals and the largest almost 1000. And some animals are used in several different experiments.

Another striking point is that besides university research, teaching at the university makes use of a significant number of animals – 1567, one sixth of the total. These are tests which only cause slight distress, says Steenmans. ‘For example, students learn to measure the recirculation system in a fish tank. Officially that is an animal experiment. In theory the water quality could deteriorate so much that the fish suffer from it.’

Whether the drop in animal testing will continue in future depends partly on the new WUR policy on animal testing. This new vision was announced in the annual report on animal testing in 2015. The executive board wants to ‘go further than the legislative framework’. Quite how the board wants to do that is not clear, however. But there is certainly innovative research going on in various WUR departments on alternatives to animal testing (see boxes).

Reactions 1

  • onlinelezer

    Waar vind ik figuur 1 uit de eerste alinea? "De jongste cijfers laten zien dat er in 2015 in totaal 33.631 dieren zijn gebruikt voor Wagenings onderzoek en onderwijs (zie figuur 1)."