The differences between people result from differences in less than 0.1% of human DNA, and determining which tiny variations influence whether people get ill or stay healthy is extremely difficult. Now, in massively parallel experiments, researchers tested 9,000 variants of an enzyme associated with the risk of developing type 2 diabetes. The results show that the genetic changes that reduce the stability of the enzyme’s structure can lead to either decreased or increased activity and can help to identify people with a type of diabetes who may not need medication. The researchers are currently testing this method on diseases of the nervous system.
When you eat carbohydrates, they are broken down and released into the bloodstream as glucose and other types of sugar. To ensure that blood glucose levels do not become too high, the glucokinase (GCK) enzyme acts as a sensor protein that regulates the release of insulin in the pancreas. This key role means that small genetic changes to GCK can greatly affect blood glucose and the development of diabetes. Now researchers have analysed more than 9,000 GCK variants, 97% of the possible genetic variants.
“What we have done is quite powerful. We have not only measured the effects of 9,000 gene variants but can also explain how many of these variants affect GCK. Probing how single amino-acid substitutions alter the stability and function of a protein is a key challenge, and we can now use our data to test and develop computer models, which we can then apply to other genes and diseases,” explains Kresten Lindorff-Larsen, Professor, Department of Biology and Director of the PRISM – Protein Interactions and Stability in Medicine and Genomics centre at the University of Copenhagen, in which researchers use biophysics, cell biology, genetics and machine learning to discover how mutations can lead to the development of hereditary diseases.
Helping to solve a widespread problem
The problem of linking genetic variants and disease has grown significantly as sequencing of the human genome has become a routine diagnostic tool. Unfortunately, determining which of the many tiny differences in DNA affect the risk of developing disease turns out to be far from simple. The GCK gene is a good example of this.
“GCK regulates the secretion of insulin in the pancreas, and GCK variants can therefore cause a form of hereditary diabetes. But although the link between GCK and diabetes has been known for many years, we have only known about the effect of a small percentage of the possible variants of GCK – until now,” says the leader of the study, Rasmus Hartmann-Petersen, Professor, Department of Biology, University of Copenhagen.
People are genetically very similar, and the usefulness of DNA sequencing is therefore often limited to the few cases in which researchers already know whether a specific gene variant can cause disease or whether it is a harmless example of the natural variation across the population. The new study shows one way to solve this widespread problem in genetic research.
“We used yeast cells to measure the activity of more than 9,000 GCK variants to generate a long list of the effects of the known variants and even more that people may have but have not yet been discovered. The new study thereby functions as a reference for future genetic diagnosis of GCK,” explains the first author, Sarah Gersing, PhD Fellow, Kaj Ulrik Linderstrøm‑Lang Centre for Protein Science, Department of Biology, University of Copenhagen.
Stability is important for activity
The researchers induced the yeast cells to produce 97% of all the missense and nonsense variants of GCK, and because of its special structure, not all were worse – hypoactive – than the original (wild-type) GCK. Many led to hyperactive GCK.
“Many of the less active variants have changes at sites buried deep inside GCK or near the active site, but also in an area of known importance for its structure and dynamics. On the other hand, many of the hyperactive variants locked the enzyme in the active state by destabilising the inactive state. GCK normally has both an accelerator and a brake. When you remove the brake, GCK runs at full speed all the time,” adds Sarah Gersing.
Now that the researchers have measured the effect of nearly all GCK variants, the next step will be to try to better understand why some variants have an effect and others do not. One key challenge in protein engineering is identifying amino acid substitutions that can improve both the stability and function of a protein.
“Scientists have long known that the protein’s active site was key to preserving and fine-tuning the function of the proteins and enzymes. However, the structure and stability of a protein turn out to be at least as important. However, separating the effects on function and stability is difficult, and some of the next important steps include collecting data and learning the rules that describe the different mechanisms,” says Kresten Lindorff-Larsen.
Maturity-onset diabetes of the young (MODY) and other diseases
The researchers therefore have high hopes that the new large-scale screening set-up can help to understand and create more and less stable and active proteins. This knowledge is also the key to trying to understand how apparently almost identical molecular mechanisms trigger many different diseases. GCK variants can trigger a form of hereditary diabetes – GCK maturity-onset diabetes of the young (GCK-MODY).
“Although people with GCK-MODY have elevated blood glucose, this is often not associated with complications, and unlike other types of diabetes, most should therefore not be given medication at all. However, missing or inaccurate genetic data mean that more than half the people with GCK-MODY are classified as having either type 1 or type 2 diabetes and may therefore unnecessarily be given medication,” explains Torben Hansen, Professor of Genetics, University of Copenhagen and part of the PRISM centre.
The researchers estimate that about 1% of the people in Denmark who have recently been diagnosed with type 2 diabetes have a GCK variant causing elevated blood glucose, which means that they do not need medication.
“Early diagnosis can limit unnecessary treatment and monitoring resulting from misdiagnosis. The new map of GCK variant activity can hopefully help to provide these people with a more accurate diagnosis,” adds Torben Hansen.
The new results thus greatly increase knowledge of GCK and this type of diabetes. However, the new method may also help in understanding the mechanisms of many other hereditary diseases.
“We have already started investigating a number of other genes involved in, for example, diseases of the nervous system and are trying to develop methods that are both precise and provide insight into the disease mechanisms,” concludes Rasmus Hartmann-Petersen.
