GIANT makes short work of height gene discovery

Don't you just love it when an acronym comes together? In the pages of last week's edition of Nature (see BioNews 887), a consortium that styles itself with the moniker 'GIANT' (short for Genetic Investigation of Anthropometric Traits – I’ll try to keep further puns to a minimum) describes 83 novel but rare genetic variants that are related to human height. The study involved more than 700,000 people and the dark art of genome-wide association studies, or GWAS. Basically, GWAS use DNA arrays or 'chips' – collections of known DNA variations – to search the genomes of a large number of individuals for these variations, and so link them to traits. The more people are analysed, the more 'power' the study is said to have, in that it can accurately link less common variations to traits. Height is a complex trait involving many genes, and this study brings the total known to be involved to around 700.

It comes as no surprise to anyone to learn that height is heritable. Tall people tend to have tall parents, and the converse applies for their more vertically challenged counterparts. A tall mum and a short dad (or the other way around) often produce a child of intermediate height. So many genes are involved in human height that classic Mendelian genetics does not seem to apply. In fact, it was probably observation of traits like height (which appear at first glance to involve 'blending') that slowed acceptance of Mendelian genetics in the first place. Early height research involved simply examining inherited disease genes that lead to short stature, but these were only ever going to be the tip of the iceberg.

So did the GIANT study deliver the goods? In many ways, it did. The main selling point of this paper is the technique they used which allowed them to detect some new, but rare genetic variants. The authors focused solely on the 'exome' – those bits of the genome which are exons – whereas standard GWAS use chips containing DNA variations from both exons and introns. Yet exons make up just 1–2 percent of the human genome. So in other GWAS studies, frequently variations found in introns are associated with traits, but we don't really know what those pesky introns do. The great advantage of the 'exome' approach, then, is that it can detect rare variants which would not be picked up by standard GWAS.

Recent large-scale sequencing studies suggest that the rapid expansion in the human population has introduced a whole set of new rare variants. Indeed, this fact was behind the invention of the grandly named 'Illumina HumanExome BeadChip' ('Exome Chip' for short) used in the GIANT study, which detects rare variants more efficiently than its predecessors. In a prior study, the investigators used this chip to identify rare variants that influence glycaemic traits and type 2 diabetes, in a significant advance towards improved drug targets and better treatment. An Exome Chip analysis can also back up previous standard GWAS studies, where an association may have been found but the genetic function poorly understood.

While the variants found in this study were rare, some of them made a difference of nearly an inch. The gene stars of the show were linked to signaling molecules for growth processes, hormone receptors and insulin growth factors, and also cellular processes that have not previously been much associated with growth. The authors also found genes that had already been associated with growth restriction disorders. One of the nice things about reading this paper and the commentaries that surrounded it is that, for once, there was no talk of 'making people taller', no prospects for genetic manipulation or designer babies, just pure and simple scientific enquiry. By standing on the shoulders of GIANT (sorry, couldn't resist) we now think we know about a little more than a quarter of the genetic variation associated with height, a trait that also is affected by our environment.

So what are the applications? Well, both the authors and the commentaries that surround the manuscript point out that, through understanding height, we can understand those diseases that affect it. In turn, this could lead to new drug treatments for those most severely affected. Mostly, however, the study is an advert for the mantra 'the harder you look, the more you'll find', and the groups are already embarking on a larger study.

Where else can these studies go? No matter how comprehensive the analysis of an Exome Chip might be, the design still relies on genes that have already been discovered for one reason or another. It's a little hard to measure, but estimates suggest that any new individual's genome contains 2–5 percent which have not yet appeared on a public database. The other problem is that some DNA variants, which control genome processes such as inserting, duplicating or deleting DNA, can't be studied by this approach. Finally, some authors have suggested that introns are responsible for the majority of the variation underlying complex phenotypes. This still requires much investigation.

The solution to these problems lies, in part, with whole-genome sequencing new individuals each time they are presented for testing – an approach which is becoming more and more cost-effective by the day. It is particularly useful for studying diseases where the genetic architecture of a specific trait is poorly understood. Whereas a GWAS search might miss interesting genetic variations – because we don't know about them yet and have not put them on the DNA chip – when it comes to height, understanding a little over 25 percent of the variation means that we have a long way to go.

The press release refers to height as the 'poster child' of complex genetic traits, because it's easy to measure. When interviewed by the International Business Times, co-author Panagiotis Deloukas from Queen Mary University of London said: 'Compared to other traits, height receives a very high contribution of genes and this is why it can be a model for large genetic studies.' Dear me – I thought my puns were bad!