A newly discovered molecular signal – found only in humans – may explain why some muscles stop responding to insulin, offering a promising lead in combatting type 2 diabetes.
Insulin resistance – when cells stop responding to insulin’s signal to store glucose – drives the global epidemic of type 2 diabetes, yet treatments are still lacking that target the body’s main glucose storage: muscle. No medicines can treat insulin resistance in skeletal muscle, the body’s main glucose warehouse.
Researchers from the University of Copenhagen in Denmark, the Karolinska Institutet in Stockholm, Sweden and Steno Diabetes Center Odense, Denmark examined samples of human skeletal muscle to tease out the origins of insulin resistance. They analysed the proteome – the full census of proteins, which are the cellular machines of life – and phosphoproteome, a molecular signalling system that gives proteins their marching orders.
In a recently published article, the team identified several potential targets to treat insulin resistance. But the biggest surprise was a molecular signal that exists only in humans – completely missed in decades of mouse research and invisible in even our closest relatives such as chimpanzees.
“Human muscle samples are enabling us to build the foundation for discovering new medicine, finding which pathway we need to tweak to improve the insulin response in the muscle,” says co-author Atul Deshmukh, an Associate Professor at the University of Copenhagen who studies the relationship between proteins and metabolic conditions.
How muscles manage glucose with tiny signals
Most muscles you can name – your delts, your quads, your biceps – are part of the skeletal muscle system, which performs the movements we voluntarily control. It is also where the body keeps much of its short-term energy reserves. “After a meal, almost 80% of the glucose extracted from your food goes into the muscle,” Deshmukh explains. “When it comes to glucose handling, muscle is probably the most important organ.”
To better understand what goes awry in type 2 diabetes, researchers studied 123 participants: 68 with type 2 diabetes and 55 with normal glucose tolerance.
First, the researchers assessed each participant’s insulin sensitivity. They found that the results did not cluster neatly into people with insulin-resistant diabetes and people who were insulin responsive and did not have diabetes. Instead, it was more of a continuum, says co-author Jeppe Kjærgaard, who recently moved to Broad Institute of MIT and Harvard for his postdoctoral research.
This suggests that there is a continuum of defects that may go wrong in responding to the insulin signal rather than a single point of failure somewhere along the complex pathway from DNA to cellular activity.
The mystery beyond DNA and proteins
Previous research has all but ruled out anything in the transcription from DNA to RNA as the culprit. “If you just look at the messenger RNA, there’s no difference between people with type 2 diabetes and ‘healthy’ people,” says co-author Anna Krook, Professor of Physiology and Pharmacology at the Karolinska Institutet. And the key protein– the surface receptors that enable cells to sense insulin, the gateways used to bring glucose inside – all seem to be assembled correctly and present in skeletal muscle from people with type 2 diabetes.
The researchers suspected that the answer might lie in the next layer governing cellular activity: the phosphoproteome.
A variety of molecular tags – tiny structures that can be attached and removed like sticky notes – give proteins their marching orders. “Once you have this tag on the protein on a particular place, the protein ‘knows’ it has to communicate with another protein, go to another location or execute a certain function,” Deshmukh explains. “That tag basically wakes that protein up for its activity.”
The best known of these tags is the phosphate group. If you take a snapshot of where these tags are placed on proteins at one moment in time, you get the phosphoproteome – the body’s molecular activity map. The researchers used samples of skeletal muscle taken from the 123 participants to chart their phosphoproteome at about 15,000 phosphosites across 3,000 proteins.
One molecule stands out among thousands
Comparing each patient’s insulin sensitivity with their phosphoproteome revealed differences at 184 phosphosites that may account for degrees of dysfunction along the continuum the researchers observed.
Many of the phosphosites were on genes associated with how cells convert materials – whether blood glucose or stores of fat – into useable energy. Any of these could be potential targets for therapies. But the researchers were most intrigued by one phosphorylation site in particular.
A protein called AMP-activated protein kinase (AMPK) “tells a cell whether it is hungry or not, whether it has enough fuel,” Krook explains. It is critical in regulating blood glucose: determining whether cells open their doors to glucose.
Creatures across the tree of life from yeasts to plants and animals use AMP-activated protein kinases, but the human AMPK have a different amino acid – the building block of proteins – at a site in the protein called position Ser65 on subunit AMPKγ3. “It is not present in any other species,” Kjærgaard says. “It must be a way for Homo sapiens to regulate glucose handling and insulin sensitivity better.”
Among more than 15,000 molecular sites, one stood out: a tiny tag at position Ser65 on the AMPKγ3 protein. Whether this site was tagged – or not – turned out to be the best predictor of how insulin sensitive a person was.
Humans are occasionally the odd man out in pathways like neuron development, since our structure and function diverge markedly from those of other animals. But the researchers say that it is vanishingly rare to find that humans differ from the animal consensus in something as fundamental as how muscles work.
This uniqueness becomes a major stumbling block for scientists trying to study AMPKγ3 in the laboratory. You do not even have this phosphorylation site in chimpanzees. Forget about the mouse model,” Deshmukh says.
The team says that this finding underscores the importance of studying human cells and tissues – and the risks of leaning too heavily on animal models. “Otherwise, we would never identify this,” Kjærgaard says.
Turning mice human to test new ideas
The next step in the team’s research on AMPKγ3 will be to “humanise mice,” Deshmukh says. Co-author Kei Sakamoto, a Professor at the Novo Nordisk Foundation Center for Basic Metabolic Research at the University of Copenhagen , has genetically modified a strain of mice by replacing the genes that code for AMPK with the version present in humans.
“In collaboration with Sakamoto group, we are going to challenge that mouse with different stimuli and see what happens to downstream parameters, such as glucose handling and glycogen synthesis,” Deshmukh says.
Scientists cannot afford to ignore these tiny molecular tags, which may hold the key to charting the “cascade of events after insulin has found its receptor in muscle tissue,” Krook adds.
“Show me your phosphoproteome and I can tell you how insulin sensitive you are,” she says. “That is stunning.”
