An international research team has identified 11 novel locations in the human genome where common variations appear to influence cholesterol or triglyceride levels, bringing the total number of lipid-associated genes to 30. While major mutations in some of these genes have been known to underlie rare lipid metabolism disorders, it is becoming apparent that common changes in the same genes that have modest effects can combine with risk-associated variants in other genes to significantly influence blood lipid levels.
“Over the past 2 years, our work has identified at least a dozen gene regions that previously had not been suspected of playing a role in human cholesterol metabolism,” says Sekar Kathiresan, MD, director of Preventive Cardiology at Massachusetts General Hospital (MGH), the paper’s corresponding author. “Finding new gene regions involved with lipid levels could improve our understanding of cholesterol regulation, help identify individuals who are at greater risk for heart disease, and point out new targets for drugs to control cholesterol levels. For example, the statin class of cholesterol-lowering drugs acts by blocking HMGCR, one of these 30 genes. We hope that some of these newly discovered genes will be as beneficial drug targets as HMGCR is.”
Building on previous work that associated common variants at 19 locations with lipid levels, the researchers first conducted a meta-analysis of 7 genome-wide association studies of blood lipid levels. Potential lipid-associated sites identified in that first-phase analysis of data from almost 20,000 individuals – including participants in the Framingham Heart Study, which first established blood cholesterol as a risk factor for heart disease in 1961 – were then genotyped using DNA samples from another 20,600 participants in 5 different studies. This validation phase found 30 gene sites, including the 19 previously identified, that were associated with levels of either high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, or triglycerides.
Rare mutations in some of the 11 newly identified genes have been known to cause serious cholesterol disorders or conditions like type 2 diabetes, but this study is the first to associate common variations in those genes with cholesterol levels. Others of the novel genes had no previous link to lipid metabolism. Some of them appear to be transcription factors, which control how other genes work, or enzymes that modify the molecules that transport cholesterol in the blood, but the function of others is still unknown.
“We’re finding that many of the variants that affect blood lipids seem to do so by influencing how much of a gene is expressed in the liver,” Kathiresan explains. “With rare mutations, gene function is usually completely abolished, but these common variants appear to affect the quantity but not necessarily the quality of protein being produced.”
It is estimated that the 30 currently known genes probably represent about 20 percent of the genetic contribution to lipid levels, so additional studies in larger populations – including participants from a broader range of ethnic backgrounds – are required to find more risk variants. Detailed sequencing of risk genes, already underway, could define the precise variants that influence lipid levels, allowing development of screening tests to identify individuals at elevated risk for cardiovascular disease because of the genes they have inherited.
“We are currently designing studies to test whether individuals inheriting several of these lipid risk genes really are at higher risk for heart attack and whether they are more likely to benefit from cholesterol-lowering treatments like statins,” Kathiresan says. A member of the MGH Cardiovascular Research Center and Center for Human Genetic Research, he is also affiliated with the Broad Institute of MIT and Harvard and is an assistant professor of Medicine at Harvard Medical School.
Illustration: Massachusetts General Hospital.
Massachusetts General Hospital News Release (12/07/08)
Washington Post (12/08/08)
Abstract (Nature Genetics; 41(1) 56-65 (12/07/08))