The researchers combined information on the genome of almost 150,000 people to find this result, which they say could lead to new treatments.
'This discovery underscores what can be accomplished when human genetics experts on both sides of the Atlantic come together to apply their craft to founder populations, enabling us to find rare mutations with large effects on disease risk', said Dr Kári Stefánsson, one of the study's authors and CEO of deCODE Genetics.
Although the effect of the gene variants that were identified in this study was very large – a 65 percent reduction of diabetes risk – the mutations involved are very rare. Only people in west Finland and Iceland appear to have the protective gene variant in relatively high numbers.
The researchers started out by trying to find an effect in smaller-scale studies. In their first attempt, they included young people of normal weight who had type 2 diabetes, and compared them with elderly obese people who had healthy blood sugar levels. By comparing 750 people at these two extremes, they hoped to be able to identify gene variants that increased or protected from diabetes risk, but they did not find any definitive evidence.
Nevertheless, one of the mutations identified in a handful of people looked promising. It stopped a gene, ZnT8, from working. ZnT8 is involved in zinc transport in the same cells where insulin is produced. They included this gene in a larger study including just over 20,000 people.
The results looked more promising this time: mutations in the zinc-transporter gene lowered the risk of diabetes by 54 percent. Unfortunately, numbers were still too low for a definitive answer.
At this point, the team decided to branch out: they started looking at more mutations that had the same effect, a loss-of-function of the zinc transporter gene, and starting collaborating with deCODE to increase people they could include in their study to almost 150,000.
Even in this massive dataset, only 345 had the type of mutation of interest: but it was enough to show it had a protective effect on type 2 diabetes.
'Such genetic associations provide important new insights into the pathogenesis of diabetes, potentially leading to the discovery of drug targets, which may result in a novel medicine', said Tim Rolph, vice president of Pfizer, who was involved in the study. Pfizer and Amgen, owners of deCODE, are already planning to use these findings for drug development.
The team has dedicated the study, published in Nature Genetics, to David Cox, whose support helped drive the research and form the collaboration. David Cox was a leader of the Human Genome Project and died last year.