A new, more life-like physical model of microscopic nerve fibres called axons could speed up the discovery of medicines for multiple sclerosis and other degenerative brain diseases, suggests a new study led by UCL researchers.
Testing drugs on an artificial replica that better matches the physical properties of human axons can reduce the risk of drugs that appear to perform well in the laboratory but ultimately fail in human trials, the researchers said.
In multiple sclerosis (MS), the immune system mistakenly attacks myelin, the protective sheath or insulation around axons - the long nerve fibres modelled in this study and responsible for carrying electrical signals in the brain and spinal cord.
The new model, described in a paper in Nature Methods, consists of tiny pillars tens of times thinner than a human hair that replicate these axons. Unlike previous models made of hard materials such as plastics, this new one is made of a water-filled gel (hydrogel) that can be adjusted to be as soft as real axons.
The research team grew myelin from human and animal cells around these pillars and added drugs proposed as therapies to restore or repair damaged myelin to see whether they promoted myelin growth.
However, they found that drugs performed less well when the model axons were more life-like, that is, were adjusted to be as soft as the real-life versions - pointing to one reason why past drugs, tested on less life-like models, may have succeeded in the laboratory and then failed in humans.
The study also marks the first time that myelin has been successfully grown from human cells in the laboratory.
Senior author Professor Emad Moeendarbary, based at UCL Mechanical Engineering, said: "To stop MS, we need therapies that repair myelin. Promising drug candidates in the past have failed when tested in human patients. One factor might be that laboratory models do not replicate the basic physical properties of the human brain. Another is that animal brains differ from human brains in so-called white matter, regions densely packed with myelinated axons.
"Our work suggests that commonly used rigid models, hundreds of times stiffer than real axons, can generate misleading drug hits. We believe that our more life-like model can be used as a more robust early test of drug candidates and as a platform to discover new drugs."
