September 03, 2007
Gene For Controlling Height Discovered

Some scientists have found a gene called HMGA2 that causes about 4 tenths of an inch difference in height.

Whilst we all know that tall parents are more likely to have tall children, scientists have been unable to identify any common genes that make people taller than others. Now, however, scientists have identified the first gene, known as HMGA2, a common variant of which directly influences height.

The difference in height between a person carrying two copies of the variant and a person carrying no copies is just under 1cm in height, so does not on its own explain the range of heights across the population. However, the researchers believe the findings may prove important.

Previous studies have suggested that, unlike conditions such as obesity, which is caused by a mix of genetic and environmental factors – so called "nature and nurture" – 90% of normal variation in human height is due to genetic factors rather than, for example, diet. However, other than very rare gene variants that affect height in only a small number of people, no common gene variants have until now been identified.

The research was led by Dr Tim Frayling from the Peninsula Medical School, Exeter, Professor Mark McCarthy from the University of Oxford and Dr Joel Hirschhorn from the Broad Institute of Harvard and MIT in Cambridge, US. Dr Frayling and Professor McCarthy were also part of a Wellcome Trust-funded study team that discovered the first common gene linked to obesity in April this year.

We are just now starting to see the results of big drops in the cost of DNA testing. The rate of discovery of the meaning of genetic variations is about to turn into a torrent. The discoveries will come so fast that only the most interesting ones will garner any press attention.

The scientists had to compare portions of the DNA of 35,000 people to find this connection between height and HMGA2.

The findings, published in the September 2 advance online edition of Nature Genetics, stem from a large-scale effort led by scientists at the Broad Institute of Harvard and MIT, Children’s Hospital Boston, the University of Oxford and Peninsula Medical School, Exeter. Using a new “genome-wide association” method, the research team searched the human genome for single letter differences in the genetic code that appear more often in tall individuals compared to shorter individuals. By analyzing DNA from nearly 35,000 people, the researchers zeroed in on a difference in the HMGA2 gene — a ‘C’ written in the DNA code instead of a ‘T’. Inheriting the ‘C’-containing copy of the gene often makes people taller: one copy can add about a half centimeter in height while two copies can add almost a full centimeter.

These scientists think this gene's variants account for just 0.3% of all variability in human height. So many more genes that contribute to height are waiting to be found.

The genomic find, though, is not the only indication that HMGA2 affects height. Previous studies in mice and humans revealed that a handful of rare stature disorders result from severe mutations in the gene. Taken together, the findings provide strong evidence for a role for HMGA2 in height. However, the identified SNP accounts for just 0.3% of the normal variability in human stature, which means there are probably many others yet to be found. To do this, researchers will need to study even larger groups of individuals.

Imagine what happens when we discover all the genetic variations that influence height. A couple gets their DNA sequenced. A genetic counselor tells them if they have a son he could be anywhere from 5' 6" to 5' 11". But if they produce a dozen in vitro fertilization (IVF) embryos they've got a very high chance of getting a couple of embryos on the tall end. Suddenly IVF becomes a lot more attractive.

How far are we away from the deluge of genetic sequencing discoveries and the beginning of the shift toward IVF as the preferred way to start pregnancies? Maybe 5 years.

Having brought the cost down by three orders of magnitude, the aim is to drop it by another three, to $1,000, and also to speed things up. To that end, the X Prize Foundation, an innovative American charity, is offering a $10m prize to the first team to decode the DNA of a hundred people within ten days. Dozens of groups from around the world have signed up, and the organisers expect a winner in less than five years. And it may not take that long. George Church of Harvard University recently started what he calls the Personal Genome Project. This aims to decode the genetic material of 100,000 people over the next year or so.

We haven't even seen the full discovery effects of the last few orders of magnitude in DNA sequencing costs. The costs have dropped too recently for researchers to have made much use of the new affordability of genetic sequencing. Once the costs drop a few more orders of magnitude full genome sequencing of millions of people will become affordable and then will come the deluge of discoveries of what all the genetic variations mean.

Share |      Randall Parker, 2007 September 03 08:13 PM  Human Population Genetics

rsilvetz said at September 3, 2007 9:51 PM:

Let me ask a stupid question. Are we really certain that we can take one cell away from an embryo without drastic consequences?

Randall Parker said at September 3, 2007 10:01 PM:


I've come across (and maybe even linked to) a study or two that has investigated that question and they didn't find problems. Not to say that there aren't problems. But so far so good.

But there's another way to solve that problem: Go back to an earlier stage pre-fertilization and figure out what DNA is going to come from the sperm and egg.

How to do that? David Friedman pointed out in the comments of a previous post that Robert Heinlein explained how to do this in his novel Beyond This Horizon:

The obvious problem was how to determine the genetics of a single egg or sperm without damaging it. The ingenious solution was to take advantage of the mechanism by which egg or sperm is produced, in which the full set of genes is divided between two cells (yes, I'm oversimplifying a little). Destructively analyze one of them, destructively analyze an ordinary cell, subtract the results of the first analysis from those of the second, and what remains is in the complement cell--which you haven't touched.

Of course, this probably requires arranging for the final division to occur in vitro.

That approach also reduces the size of the job when trying to come up with a desired combination and avoid all the less desired combinations.

rsilvetz said at September 4, 2007 7:20 AM:

Thanks Randall... Nifty thought from that Heinlein fellow... then again was always partial to Friday[1,2].

Hopefully Anonymous said at September 6, 2007 1:44 PM:

Too bad the market is in large part going to be for tall kids, and not more narrowly for kids that solve human aging and minimize existential risk.

rsilvetz said at September 6, 2007 7:10 PM:

The first few IvF screens might be for common desirable characteristics such as tallness, symmetry and strength. But as sophistication in the screening increases, other features, such as maximizing various flavors of intelligence and character, adrenalin responses, etc will become part of the equation. You will see multifactorial eugenics take place and there will be shifts, and hopefully these will end up being new Gaussian distributions with shifts of their means towards what we desire as optimal characteristics in our offspring.

Hopefully we will dodge GATTACA and Wrath_of_Khan....

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright ©