Three Groups Find Genetic Links to Diabetes

ANN ARBOR, Mich., April 26 -- Several genetic regions appear to contribute to the risk of developing type 2 diabetes, three research groups have found in a major collaborative effort.

The three groups -- one based in the United Kingdom and two in the U.S. -- used variations on a method called genome-wide association analysis to look for genes that were more common in those with type 2 diabetes than in healthy controls.

"This represents sort of a landmark moment in type II diabetes research," said Michael Boehnke, Ph.D., of the University of Michigan at Ann Arbor, who led the so-called FUSION study, which included investigators from several American sites, the National Human Genome Research Institute, and the Finnish National Public Health Institute.

He said the "very compelling results" -- published online today by Science -- show the power of genome-wide association studies in finding genetic variations associated with disease.

Each of the three groups began with a different group of participants with and without type 2 diabetes and used microchip array techniques to scan and compare their genomes.

Once they found regions of interest that differed between patients and healthy controls, the groups collaborated to confirm or refute the findings in larger cohorts.

All told, the three groups scanned the genomes of 32,544 people with and without type 2 diabetes.

Dr. Boehnke said the findings differ slightly from group to group "mainly in the way we present our data" although there are also differences that arise from variations in their methods.

But all three groups agreed that type 2 diabetes is associated with genetic variations in:

• A region on chromosome three, near the gene IGF2BP2.
• A region on chromosome six, near the gene CDKAL1.
• And a region on chromosome nine, near the genes CDKN2A and CDKN2B.

While those regions are near known genes, Dr. Boehnke said, whether the genes themselves are involved in diabetes is not yet known.

In addition, all of the groups confirmed previously reported links to the genes HHEX, PPAR-gamma, KCNJ11, SLC30A8, and TCF7L2.

Dr. Boehnke and colleagues also reported a link to a region on chromosome 11, which appears not to be inhabited by any known genes, although neither of the other groups did so.

Dr. Boehnke and colleagues also reported a link to the gene FTO, on chromosome 16, but the British group, led by Mark McCarthy, M.B., of Oxford University, did not.

That's because Dr. McCarthy and colleagues reported within the past two weeks that FTO is linked to increased risk of obesity, which is itself a risk factor for diabetes. (Gene Variant Increases Obesity Risk)

In their study -- the Wellcome Trust Case Control Consortium -- the FTO link to diabetes vanished when the researchers adjusted for body mass index, so they did not report it as a diabetes gene.

"Essentially," Dr. Boehnke said, "they think of it as an obesity gene."

The third group -- the Diabetes Genetic Initiative -- had similar results, but reported an additional link to a region near the gene for a protein that regulates glucokinase, an enzyme that breaks down sugars, according to David Altshuler, M.D., Ph.D., of the Broad Institute of Harvard University.

The immediate clinical significance of the findings remains unclear. Dr. Altshuler said in a statement that the results "open surprising new avenues for disease research, treatment and prevention."

But Dr. Boehnke said in an interview that diabetes is such a complex disease that the step from finding genes to developing drugs or improving diagnosis is likely to be a long one.

The pro-diabetic variations that were found tended to increase the risk of the disease by between 10% and 25% per allele.

In other words, Dr. Boehnke said, a person with two copies of one of the risky variations would have between a 20% and 50% increase in diabetes risk.

On the other hand, he said, the picture is clouded by the environmental and behavioral factors that also contribute to the disease - such as lack of exercise or over-eating.

Because of that complexity, he said, "for type II diabetes, we can't really say yet what the impact (of these genetic variants) is on the disease risk."