Disease Risk Depends on Which Parent a DNA Variant Is Inherited From

Icelandic biologists have discovered that the genetic risk of several common diseases, like Type 2 diabetes and cancer, can depend on which parent a DNA variant is inherited from.
The finding may help explain part of a serious gap in understanding the genetics of common diseases.

Using an extensive genealogy that includes almost all the present population of Iceland and many in previous generations, the Reykjavik company DeCode Genetics managed to distinguish which chromosomes came from the father and which from the mother in some 40,000 people.

The company then ran standard tests, known as genome wide association studies, the tool that scientists have hoped would track down the roots of common diseases and fulfill the promise of the Human Genome Project.

But with most of these common diseases, the tests have so far identified genetic variants that account for only a small percentage of the risk. This is in contrast to simple diseases, most of them rare, where a single gene is the cause and the disease has a clear family pedigree.

The failure has left biologists puzzled about the missing heritability, which some have jokingly attributed to “dark matter” within the human genome, an analogy with the dark matter invoked by cosmologists to explain the missing mass of the universe.

Some experts believe the missing risk might be carried in a large number of rare DNA variants not included in current scans of patients’ genomes. Because it costs too much to decode all three billion units of DNA in a patient’s genome, biologists use chips that scan just the 500,000 sites where variation is most common. These sites were expected to explain the genetic component of common diseases like cancer or schizophrenia, but mostly do not do so.

Another explanation is that the missing heritability lies in aspects of cell biology that are not yet understood.

Decode scientists have found one such instance. They report in Friday’s Nature that a DNA variant increases a person’s risk of Type 2 diabetes by 30 percent if inherited from the father, but reduces the risk by 10 percent if comes from the mother.

Because the two effects tend to cancel each other out, they have not been picked up by the standard tests that do not identify the parental origin of each section of DNA.

DeCode found that five of seven variants tested made different contributions to disease depending on the parent of origin. In most cases the effect was of differing degrees of severity, depending on the parent involved.

To increase the chances of seeing such an effect the researchers tested known disease-associated variants that lay close to so-called imprinted genes, ones where the copy from one parent is suppressed. They now plan to survey the rest of the genome.

Kari Stefansson, DeCode’s chief executive, said he could not predict how large a percentage of the missing heritability the parent-of-origin effect might account for. “But I think it’s going to be substantial,” he said.

DeCode Genetics recently filed for bankruptcy, but the company plans to continue operations in another form. Dr. Stefansson said that its access to the Icelandic genealogical database would give it an advantage in searching for disease genes in circumstances where it is essential to know the genetic structure of a population.

Mark Daly, a medical geneticist at Massachusetts General Hospital, said that DeCode’s result was “a significant finding” and that it confirmed the idea that effects of this nature would account for some of the missing heritability.