Search
Close this search box.
Search
Close this search box.

Computer-assisted drug discovery for rare epilepsies

Dr Jonathan Lippiat

Rare Epilepsies - Computer-assisted drug discovery for rare epilepsies

- Endeavour Project Awardee
- University of Leeds 

Dr Jonathan Lippiat is studying a mutation in a gene called KCNT1 which causes people to experience severe seizures, several times a day. We know that existing anti-epilepsy drugs often can’t control these seizures. So instead, Jonathan is testing drugs which specifically counteract the KCNT1 mutation, potentially reversing the problem at its source. In this blog, Jonathan shares his progress on finding new treatments for this form of epilepsy, and the potential of using computers to help design new drugs.

In 2012 we discovered that some children with severe forms of uncontrolled epilepsy had variations in the KCNT1 gene. We were studying the protein encoded by this gene which allows potassium ions to pass through nerve cell membranes in response to certain conditions. We initially thought it could be a new target for painkillers, but on hearing about its role in epilepsy, our research changed direction. The problem is that genetic variations make the KCNT1 potassium channel overactive. Electrical activity in the brain involves billions of tiny proteins that allow charged ions, such as potassium or sodium, to flow across nerve cell membranes in the right place and at the right time. These are similar to the electrical components that control the flow of electrons in the device that you are looking at right now. When these components fail, the device usually stops working correctly. The brain is no different, which is why genetic variations in several genes that encode ion channel proteins cause epilepsy. As pharmacologists faced with this medical problem, we thought about how this form of epilepsy could be treated and whether there are any existing medicines that could work. We reasoned that if the protein is overactive then a chemical that reduces the flow of potassium ions through this type of protein, like putting a plug in the hole of a sink, could be the answer. We were aware of three drugs that could do this, but two had been withdrawn from clinical use as they are unsafe. The third, quinidine, was being tested with patients with KCNT1 epilepsy, but doses high enough to be effective could not be achieved. Quinidine is a drug that affects the heart and is given to patients with irregular heartbeats. It works by altering the electrical activity in the heart by interacting with several types of ion channel. Unfortunately, it does this much better than interacting with potassium channels in the brain, which puts a stop to its use in this type of epilepsy as there is a risk of a heart attack. What we needed to do was to find new chemicals that better controlled the KCNT1 potassium channel without affecting the heartbeat. Traditional drug discovery involves testing many thousands of chemicals individually to see if they have the desired effect. This is very expensive, time-consuming, and requires specialist equipment and all the chemicals themselves. Thankfully, we have colleagues in our Chemistry department who are experts at using computational tools to predict how potential drugs interact with their target proteins. A computer programme “tested” 100,000 chemicals in a virtual experiment which provided us with a shortlist for chemicals to test in the lab. We really didn’t know if this would work so we were pleasantly surprised when one third of the chemicals predicted by the computer actually inhibited KCNT1 potassium channels in our experiments. The next question that we need to address is “do these chemicals work on real brain cells and could they treat epilepsy?” To help our research, we have obtained a mouse model of KCNT1 epilepsy, where the mouse’s own KCNT1 gene has been altered to replicate a variant found in patients. With the generous support of Epilepsy Research UK we will be able to embark on the next chapter in this story and answer this important question. Many laboratory techniques have remained unchanged for decades, but computers and their applications have become ever more powerful and sophisticated over time. In the future, it is possible that computers will discover more drugs than laboratory scientists will.