Metformin was originally developed to control blood glucose but is now drawing attention for a very different reason. New research suggests that it can energise human cells that rebuild myelin – the protective insulation around nerves – raising cautious optimism that it could slow, or even reverse, damage in multiple sclerosis.
No treatments can restore lost nerve function for the estimated 3 million people worldwide living with multiple sclerosis, a degenerative disease that destroys the protective coating around nerves and disrupts movement, sensation and vision. But metformin, a drug originally developed to treat high blood glucose, is showing unexpected promise for slowing – and potentially reversing – symptoms of the disease.
Experiments in rats with multiple sclerosis were so encouraging that scientists began clinical trials involving people even before fully understanding how metformin acts at the cellular level.
“Metformin seems to be really good in a rat – but we are not rats,” says Anna Williams, a neuroscientist and neurologist at the University of Edinburgh, United Kingdom. To reveal how metformin interacts with human cells, Williams and colleagues turned to a range of experimental systems, including miniature petri-dish brains and chimeras – hybrids of human and animal tissue – that enable human cells to mature and function in living brains.
Their findings, published in Nature Communications, provide a mechanistic framework that supports – but does not yet predict – the outcome of ongoing clinical trials, says Nina-Lydia Kazakou, lead author of the study and a neuroscientist at the Novo Nordisk Foundation Center for Stem Cell Medicine at the University of Copenhagen in Denmark.
“This reassures us that what we are seeing in rats, in human cells grown in the laboratory and in postmortem human brain tissue is likely to translate to people,” Williams adds.
Insulating the nervous system
Multiple sclerosis is a disease of myelin – a structure made of proteins and fats that wraps around our nerve cells and acts like wiring insulation, helping the electrical signals we use to control our movements travel quickly and over great distances. A single nerve cell can stretch up to a metre long, Williams says, and this myelin sheath also provides food to the most distant parts of the nerve.
In the early stages of multiple sclerosis, the myelin sheath is damaged by attacks from the immune system. “But later on, it is very much that the nerves themselves are dying back,” Williams says. “If we can put the myelin back on, then we hope that we can preserve those nerves, because it re-establishes the food supply.”
In a healthy body, cells called oligodendrocytes build these myelin sheaths. Researchers often use rats and mice to study oligodendrocytes and myelination, but these cells differ crucially between rodents and humans. Human oligodendrocytes rely on some gene programmes that are absent in rodents – including genes involved in lipid metabolism and late-stage maturation – meaning that human cells build and maintain myelin differently and on a slower timetable than their rodent counterparts.
Oligodendrocytes that Kazakou and her team grew in two laboratory systems – petri-dish monocultures, in which a single cell type grows in isolation, and organoids, in which multiple cell types interact in a three-dimensional community – did not act like those found in adult human brains. Instead, the cells resembled oligodendrocytes from the second or third trimester of fetal development.
“Mostly we myelinate after birth,” Williams says. “That is why babies are so limited in function. This is not because they do not have nerves – it is because they do not have myelin” to help carry signals from the brain to the muscles.
Letting human myelin cells mature – inside a living brain
To give the oligodendrocytes time to mature, Kazakou and her team used chimeras – living organisms that contain cells from multiple individuals, or in this case, two different species – to observe human oligodendrocytes producing myelin in a living brain environment.
Importantly, this approach enables researchers to study human oligodendrocyte-intrinsic behaviour in vivo – effectively watching human myelin cells at work inside a functioning brain – reducing but not eliminating the need to assume that findings in animal models will translate directly to people.
The researchers used laboratory mice “as incubators for these human cells,” Williams explains. The team started with Shiverer mice, a strain of laboratory mice that have a slight tremor because they are incapable of producing functional myelin proteins. Because Shiverer mice produce no myelin of their own, “anything that looks like myelin is human myelin,” Williams says.
The researchers treated five of the chimeras with metformin for 10 days to determine how metformin affects oligodendrocytes. Then, they analysed brain RNA – which acts as a record of what genes are actively being used – and used powerful microscopes to look inside the oligodendrocytes directly in slices of each mouse brain.
How metformin boosts the cell’s energy system
Gene activity linked to mitochondria – the cell’s energy-producing structures – was elevated in oligodendrocytes treated with metformin. The researchers examined brain slices from animals carrying the human cells and found physical changes in the mitochondria themselves. “The mitochondria were larger and much more obvious” Kazakou says.
“What we think metformin does is increase the components a cell needs to make more mitochondria and likely encourages them to link up into networks,” Williams explains. “That is beneficial for cells, because it enables energy to be shared very efficiently – which myelin-producing cells need to keep long nerve fibres alive and functioning.”
For oligodendrocytes, that extra energy would be expected to support more efficient myelin production – a mechanism that could help to explain how metformin lessens the symptoms of multiple sclerosis in rodents. In simple terms, metformin appears to help the cells that repair nerve insulation do this more effectively.
But Williams says that the impact could be much broader. “What I found really exciting is that it is not just doing it in the oligodendrocytes, but it is in the nerves, in the neurons and in all of the other glia as well,” Williams says.
Promise for the future but not a panacea for today
The findings raise the possibility of broader benefit from metformin, Williams says. “It may well be helping the energetics of other cells as well, which makes me cautiously optimistic that the trials will show something beneficial.”
Nevertheless, the researchers stress that this does not mean that metformin is ready for widespread use in multiple sclerosis. “Metformin is not specific for oligodendrocytes – you swallow it, and it goes everywhere,” Williams explains, and how it affects other cell types and organ systems is not yet fully understood.
Pinpointing how metformin acts at the molecular level could pave the way for more targeted therapies. Future work will need to “drill down into precisely how metformin works – which parts are beneficial for the brain, and which potential side-effects we want to avoid,” Williams says.
Currently, Kazakou, Williams and their co-authors are looking even more closely at mitochondria, this time under electron microscopes, to determine whether they are shaped differently after metformin treatment and whether they really are forming networks.
For now, attention is turning to the Octopus trial – a clinical trial in the United Kingdom and Australia testing multiple drugs that might slow the progression of multiple sclerosis, including metformin. “Its first analysis point, based on MRI scans of disease progression, is expected in 2026,” Williams says.
