Scientists identify 7 genes linked to diabetes

Researchers said yesterday that they have identified seven new genes connected to the most common form of diabetes - the latest result of an intensifying race between university researchers and private companies to find genes linked to a range of diseases.

The findings, presented in three reports by university scientists and one report by a private company, offer novel insights into the biology of a disease that affects 170 million people worldwide.


And the sudden spate of new results marks an acceleration, and perhaps a turning point, in the ability to find disease genes, the long-promised payoff from the Human Genome Project that began in 1989.

Yesterday's reports bring the number of well-attested genes involved in adult onset, or Type 2, diabetes up to 10 from the three known previously. The new genes do not immediately suggest any new therapy but could point to a new biological basis for the disease, from which effective treatments could emerge.


DeCode Genetics, an Icelandic company responsible for one of the reports, has dominated the gene-finding field for the past five years. The other papers come from three academic consortiums led by David Altshuler of the Broad Institute, Michael Boehnke of the University of Michigan and Mark McCarthy of the University of Oxford in England.

DeCode Genetics has been the first to find several major disease genes through skillful analysis of the Icelandic patient population and its well-documented health care system.

Its academic competitors, with the support of the National Institutes of Health, have had to describe and analyze the much wider range of genetic variation in the American and European populations.

This has required analyzing 10 million SNPs (pronounced snips), or sites of common genetic variation on the human genome, and cataloging them in a project known as the HapMap.

The academic approach has resulted in claims of disease genes that have proven inaccurate.

"It has not been a terribly productive field until the last two or three months," McCarthy said.

But with improved technology and better understanding of the sources of error, the academic groups seem at last to have gotten their technique working. Known as whole genome association, the technique has now helped them identify several genes not yet found by DeCode.

"The DeCode paper only found one new gene, and we found three," McCarthy said.


Kari Stefansson, DeCode's chief executive, said that the new diabetes genes found by the university consortiums contributed only a small degree of extra risk and that he did not have the time or resources to go after them. But he acknowledged that his academic competitors were breathing harder on his heels.

Having conducted neurological research at the University of Chicago and Harvard, Stefansson is no stranger to the sharp elbows sometimes wielded in biomedical research, and he is not without a good pair of his own.

He said his report last year on an important new diabetes gene, known as TCF7L2, had been rejected by both Nature and Science, two leading scientific journals, because of negative reports from his competitors, and he chided Science for now publishing reports on genes of lesser effect from the same people.

Stefansson's paper on TCF7L2 was published in March 2006 in Nature Genetics.

A spokeswoman for Science declined to comment.