The genetics of human height

A variant in a gene called HMGA2 is associated with differences in adult and childhood height in the general population, reports a study to be published online this week in Nature Genetics. This is the first such common variant to be reproducibly associated with human stature, and likely heralds a new wave of studies associating genetic variation with non-disease-related complex traits in humans.

Studies of twins have shown that up to 90% of the normal variation in human height is due to genetic variation, with such variants likely to be distributed among a large number of genes. A large consortium of investigators led by Timothy Frayling, Joel Hirschhorn, and Mark McCarthy carried out a genome-wide association study of nearly 5,000 individuals, and found two variants very close to HMGA2 to be associated with variation in height. A follow-up study of more then 19,000 individuals confirmed the association.

HMGA2 is an excellent candidate to be associated with height, as rare, severe mutations in the gene cause dramatic alterations of body size in mice and humans. Although the effect of these common HMGA2 variants is small—explaining approximately 0.3% of population variation in height, or approximately 0.4 cm increased height per copy in an individual—the success of the study suggests that additional genes will soon be associated with height and other visible traits in humans.

Diet shapes the human genome

Human populations that have high starch diets have an increase in the number of copies of a gene whose product breaks down starch, reports a study to be published online this week in Nature Genetics. Although copy number variation has attracted a lot of recent attention, this is one of the first documented examples of positive selection on gene copy number in humans

Starch has become a prominent component of the human diet. It is metabolized in part by salivary amylase, and the gene encoding it, AMY1, shows extensive variation in copy number. George Perry and colleagues estimated AMY1 copy number in 50 European Americans and showed that the levels of salivary amylase protein are positively correlated with gene copy number. They went on to show that individuals from three populations with high-starch diets tend to have more copies of AMY1 than individuals from populations with low-starch diets. Finally, the authors compared the extent of variation across the genome between two Asian populations -- Japanese (high-starch diet) and Yakut pastoralists (low-starch diet) -- and found that variation at AMY1 exceeds that of more than 97% of the other sites in the genome that were assessed.

The authors conclude from this that natural selection favored increased AMY1 copy number in at least some populations with high-starch diets. Interestingly, humans have significantly more copies of AMY1 than chimpanzees, which ingest relatively little starch. Increased AMY1 expression would probably improve the digestion of starchy foods, and possibly maintain energy absorption in the face of intestinal disease.

Author contact:
Nathaniel Dominy (University of California, Santa Cruz, CA, USA)
Tel: +1 831 459 2541; E-mail: njdominy@ucsc.edu

Muscle metabolism and human evolution

A variant of a gene associated with elite athletic performance has been subject to strong, recent positive selection in humans, according to a study published online this week in Nature Genetics. Experiments on mice suggest that this variant may promote more efficient muscle metabolism.

The gene ACTN3, encoding alpha-actinin-3, is specifically expressed in the fast-twitch muscle fibers that are responsible for generating force at high velocity. ACTN3 exists in a non-functional truncated form in more than a billion people worldwide, and is overrepresented in endurance athletes. By contrast, the functional form is overrepresented in elite sprinters. Kathryn North and colleagues examined the extent of genetic variation in the vicinity of ACTN3 in individuals with the truncated version. They found very little variation, which is consistent with the truncated version of the gene being under positive selection.

To understand better the effect of the truncation on muscle function, the authors generated a line of mice lacking ACTN3, and found that there is a shift in muscle metabolism toward the more efficient aerobic pathway. These mice were also able to run on average 33% further before reaching exhaustion than mice with normal ACTN3 function. The authors conclude that this increased metabolic efficiency could explain the overrepresentation of the truncated form of ACTN3 in endurance athletes