Behavioural fingerprints predict drug mode of action

 25 May 2021   Research News

By Shannon Quinney

Pesticides are used globally by farmers to protect their crops and livestock from pests. They work by killing invertebrates such as worms or flies. In the same way invertebrates can cause harm to plants, they can also cause harm to humans. Nematodes, also known as roundworms, infect over 1 billion people worldwide. Worm-combatting drugs, known as anthelmintics, are used to battle these parasites.

Widespread use of anthelmintics and insecticides can lead to resistance, rendering these treatments less effective. This has created the need for novel treatments, but to discover new drugs we need to work out what exactly each potential drug is doing to animals.

Most drugs for human infection and crop protection are discovered by looking at the ability of a compound to kill or impair a target species. The aim is to find out which substances alter the phenotype (observable traits) of a cell or an organism in a desired manner. This is called phenotypic screening.

Both insecticides and anthelmintics tend to target the organism’s neuromuscular system, this includes all the muscles in the body and the nerves serving them. Therefore, it is possible to simply observe the behaviour, or symptoms of the subject organism to figure out how the drug is affecting the pest/parasite. However, this process of ‘watching’ the behaviour is time consuming and doesn’t always capture the small details.

Research published today in Molecular Systems Biology, involving the Behavioural Phenomics group at the LMS presents a more automated and measurable method to predict the mode of action of insecticides and other drugs, using the worm C. elegans. The team are using megapixel camera arrays to record the behavioural responses of worms to a library of 110 insecticides and anthelmintics. They have extracted a ‘pose’ of each worm and a behavioural fingerprint capturing posture, motion, and path.

They have also developed a machine learning approach to accurately predict the mode of action of previously unseen compounds, they can now predict the mode of action of a compound with an accuracy of 88%.

Andre Brown, Head of the Behavioural Phenomics group at the LMS expands on the suitability of C. elegans in these studies:

“Many people think worms are very simple, they don’t do much, they go forwards and backwards and they pause and so you’re not going to be able to distinguish a large number of different compounds. However, it turns out that there’s enough richness in their behaviour that we cannot just tell what these compound classes are, but we can even distinguish compounds within those classes.”

This new method has shown to be a promising approach to the discovery of novel drugs and insecticides.

“If you’re sitting in a company trying to discover new drugs in this way, this screening method will just be one piece of evidence you would use and try to decide on further experiments and narrow down the target.”

The Behavioural Phenomics group are working with Syngenta, a global provider of agriculture technology, to bring this automated method of screening in house. This will help them to develop novel insecticides.

As well as playing a role in plant health, this new method will contribute to the discovery of novel drugs combatting diseases associated with nematodes.

Behavioural fingerprints predict insecticide and anthelmintic mode of action was published on 25th May 2021 in Molecular Systems Biology. Read the full article here.