Junk DNA holds clues to heart disease
Deleting a non-coding region leads to narrowing of arteries in mice.
Researchers have made headway in working out why a section of junk DNA — the 98% or so of the genome that does not code for proteins — raises the risk of at least one form of heart disease.
About one in five deaths in the United States results from excessive build-up of fatty plaques inside arteries supplying blood to the heart — known as coronary artery disease (CAD). In 2007, genome-wide association studies1,2 on thousands of participants linked a non-coding stretch of chromosome 9p21 with the disease, and showed that people who carry certain single nucleotide mutations in this stretch of DNA have an increased chance of developing CAD.
The latest work, published online today in Nature3, builds on these studies by knocking out this area of the equivalent chromosome in mice. "We were really interested in understanding how this purely non-coding interval leads to CAD, so we thought, 'Let's delete it and see what happens'," says geneticist Len Pennacchio of the Lawrence Berkeley National Laboratory in Berkeley, California, who led the study.
"We did, and found that the expression of two genes nearly 100,000 base pairs away from the deletion dramatically decreased in mice," Pennacchio explains. In addition, many of the mice without the non-coding DNA died earlier than normal and some developed tumours.
On the basis of this genetic discovery, the team uncovered a potential mechanism for how the region of non-coding DNA might increase the risk of heart disease. The two genes that showed decreased expression, called Cdkn2a and Cdkn2b, are cell-cycle inhibitors that control cell proliferation in the heart and other tissues. In mice lacking the non-coding region, muscle cells taken from the aorta multiplied much faster than normal — potentially obstructing blood flow to the heart.
"How this translates into humans, we don't know yet," Pennacchio says. He suggests that humans with abnormalities in cell-cycle gene expression are susceptible to increased cell division in coronary arteries. The extra cells could build up and restrict blood flow to the heart, ultimately causing a heart attack, he proposes.
The association between chromosome 9p21 and CAD has remained unclear because the single-nucleotide variants in the non-coding region are not associated with established risk factors for heart disease such as diabetes, hypertension or high cholesterol levels.
"This study brings the understanding of 9p21 and CAD risk to another level," says cardiovascular specialist Ruth McPherson of the University of Ottawa Heart Institute in Ontario, Canada, who led a genome-wide association study on CAD1. But she cautions that more work is needed before the mechanistic link between 9p21 and Cdkn2a/b regulation is established. For instance, it's a mystery why the mice in the study did not get fatty plaques building up in their arteries when humans with single-nucleotide variants in chromosome 9p21 do get a build-up of plaque. And whereas some of the mice developed tumours, that is not associated with CAD in humans.
"The fact that this non-coding region works on genes over 100,000 base pairs away goes to show that non-coding DNA can play important roles in common human disorders," Pennacchio says.
"We want to globally understand what fraction of human diseases are due to variation in the coding versus non-coding regions," he adds. "That's a huge unanswered question as we continue into the post-genomic era."
- McPherson, R. et al. Science 316, 1488-1491 (2007).
- Helgadottir, A. et al. Science 316, 1491-1493 (2007).
- Visel, A. et al. Nature advance online publication doi: 10.1038/nature08801 (2010).