Genes linked to height no longer in short supply
Image by Nadav Kupiec and Bang Wong, Broad Communications
In the past several months, two genome-wide association studies by Broad researchers and others have found two common single-letter DNA changes (SNPs) that influence height in humans. Now, an expanded meta-analysis of nearly 16,000 people has uncovered ten additional SNPs that contribute to stature. Together with an accompanying manuscript from several groups in the United Kingdom, these discoveries bring the total number of height-related genes from two to twenty-six. The genes implicate new biological pathways that are important in human growth. The papers appear in the April 6 online version of Nature Genetics.
In an earlier study, a team of scientists led by Joel Hirschhorn (a senior associate member of the Broad Institute, a pediatric endocrinologist at Children’s Hospital Boston, and an associate professor of genetics at Harvard Medical School) and collaborators from the United Kingdom, helped uncover the first convincing genetic link to height — a single-letter change in the HMGA2 gene. The variation was found by combining genome-wide association data from the Diabetes Genetics Initiative (DGI) and a similar diabetes-related study from the Wellcome Trust Case Control Consortium (WTCCC). Adults with two copies of the height-increasing version of the HMGA2 variant are on average 0.8 centimeters taller than adults carrying two copies of the other version. Hirschhorn and Guillaume Lettre, postdoctoral researcher and first author on the new study, then collaborated on another whole genome association study published earlier this year that pointed to a second gene, GDF5, of which the average height difference between genotypes is 0.4 centimeters. The second study combined data from the SardiNIA and FUSION whole genome studies.
To identify genetic variations with smaller effects upon height, Hirschhorn led a team of researchers in a scaled-up search by merging results from the DGI, SardiNIA, and FUSION studies, in addition to three other genome-wide studies (KORA, Nurses’ Health Study and PLCO). By analyzing the genomes of over 16,000 people and verifying the results in over 10,000 additional people, they were able to spot ten new genetic variations that, along with the two original genes, account for a total of roughly 2% of the height variation in the general population. Because a person can have up to two copies of the height-increasing versions of each gene, those with less than eight of the height-increasing gene versions are on average 3.5 centimeters shorter than those with more than 16 of the variations.
While the meta-analysis highlighted some genes already known to play a role in height, like HMGA2 and GDF5, the analysis produced some biologically insightful surprises. Several of the identified genes are targeted by the microRNA let-7, suggesting a link between microRNA and growth, and several other SNPs may work by modifying the structure of chromatin. “This is a good example of the power of an unbiased genetic approach to provide new biological insights and hypotheses,” said Hirschhorn.
The work may also have relevance for patients with severe inherited problems with growth or bone formation, as several of the newly discovered genes also have rare mutations that cause genetic syndromes with abnormal skeletal growth as a feature. Only through further biological and genomic studies will researchers fully elucidate the mechanisms by which this growing list of genes influence human growth.
The work also demonstrates the power of collaboration and large data sets to find new genes that can lead to biological insights. “Our success in using large data sets to identify new variants with small effects suggests that a collaborative effort could work, not just for height, but also for common diseases for which genome-wide association studies are being performed,” said Hirschhorn.
Genetic factors are thought to explain 80-90% of the variability in human height. “These new findings only account for a small fraction of height variation among people, so we still have a lot of discovery left to do,” said Hirschhorn. The researchers plan to investigate even larger data sets to identify new height-related genes, in addition to searching for additional sequence variations within the genes already identified. “Our results also emphasize the need to consider other forms of genetic variation, such as rare variants or structural changes,” said Lettre. “This study is really only the second step in our quest to understand the genetics of human growth, and there are likely many other exciting discoveries yet to be made.”
Other Broad researchers contributing to the work include Benjamin Voight, Candace Guiducci, Rachel Hackett, Leena Peltonen, and David Hunter. The investigators on this study are part of a recently formed international consortium to study height and obesity-related traits (GIANT, Genetic Investigation of ANthropometric Traits).
Lettre G et al. Identification of ten loci associated with height and previously unknown biological pathways in human growth. 2008. Nature Genetics; DOI: 10.1038/ng.125
Weedon MN et al. Genome-wide association analysis identified 20 loci that influence adult height. 2008. Nature Genetics; DOI: 10.1038/ng.121
Sanna S et al. Common variants in the GDF5-UQCC region are associated with variation in human height. 2008. Nature Genetics; DOI: 10.1038/ng.74
Weedon MN et al. A common variant of HMGA2 is associated with adult and childhood height in the general population. 2007. Nature Genetics; DOI:10.1038/ng2121