Published online 14 March 2008 | Nature | doi:10.1038/news.2008.669


Gene hunters uncover networks behind disease

New technique offers different route to drug targets.

Networks of genes linked to obesity have been uncovered.Getty

Researchers have used a new technique to identify networks of genes linked to obesity in both mice and humans. The procedure is more comprehensive than the traditional method of hunting for genes associated with disease, and is already being used to identify new drug targets.

Over the past year, a flurry of studies have revealed genetic variations associated with disease. These ‘genome-wide association studies’ have been used to find variants associated with everything from heart disease to diabetes (See Genome studies: Genetics by numbers).

Traditionally, single genes are linked with particular diseases by locating genetic variants present in people who have the disease and identifying the part of a chromosome associated with that disease. Then researchers have to track down the gene on the chromosome, without knowing what it does or why it would be involved.

Eric Schadt of Rosetta Inpharmatics, a subsidiary of Merck Pharmaceuticals in Seattle, Washington, lead one of the research teams involved in the new work. He likens the traditional approach to finding a simple light switch for a disease: flipping this single gene switch on or off may produce a higher or lower risk of disease.

The new approach looks at changes in expression of already-known genes, and finds networks of genes associated with disease, rather than single switches. “Instead of the simple ‘turn the light on or off’ analogy, we would view this as a network of these switches,” says Schadt.

A weighty matter

To use this technique, the Seattle researchers teamed up with deCODE, a Reykjavik, Iceland-based genetics company, to collect blood and fat samples from hundreds of Icelanders. The deCODE team analyzed the expression of 23,720 known genes in each sample, and correlated these results with the body mass index of each participant1.

After controlling for confounding factors commonly associated with obesity, such as age and gender, the researchers found 2,000 genes whose expression is altered in blood cells in obese people. There were more than 17,000 such genes in fat tissue. They also found patterns in the gene expression data that allowed them to group genes into networks that are regulated together.

The researchers then searched the genome for additional DNA sequence variations associated with some of these changes in gene expression. This approach is conceptually similar to a traditional genetic association study, but starting with information about altered gene expression gives the researchers a hint about how these variants could be acting to affect obesity. Their approach yielded thousands of DNA variants, most of which were located near the gene whose expression was altered.

Mice and men

The results confirm that being heavy can be hereditary: 70% of the gene expression changes found in fat cells were heritable. And it confirms the importance of the immune system in metabolic disorders such as obesity: a similar study in mice led by Schadt determined that a group of genes related to obesity are also involved in inflammation2.


By going beyond a single gene association to studying networks of genes, the procedure could be used to identify new drug targets for other conditions or diseases. “Through the idea of a network, you increase your chances of finding genes responsible for the phenotypes you are interested in,” says Chiara Sabatti, a biostatistician at the University of California, Los Angeles.

Schadt says the gene-hunting technique is already being used by Merck. “We’re not just looking at one gene, that may or may not be druggable,” he says. “We’re looking at what are the best nodes, or information control points. What’s the best light switch to affect the network maximally?”

Several other labs are gearing up to use the technique as well, says Christian Dina, a biologist at the Institute of Biology in Lille, France, who has carried out genome-wide association studies of obesity and diabetes. “It’s really a great advance,” says Dina, who predicts that it could speed up the process of determining how genetic changes lead to disease. 

  • References

    1. Emilsson, V. et al. Nature doi:10.1038/nature06758 (2008).
    2. Chen, Y. et al. Nature doi:10.1038/nature06757 (2008).
Commenting is now closed.