Cracking the mystery of human variation, at last

Why does one person suffer from schizophrenia and not another? We know the answer lies in tiny changes in their DNA code. But which code changes, and how do such glitches cause the disease?

The Human Genome Project promised us answers to such questions, not just for schizophrenia but many common illnesses such as diabetes and heart disease, where DNA has a lot to say about whether we will succumb. But 16 years on, few answers have been forthcoming. Now an epic new study promises to forge the elusive link between DNA code glitches and how it affects information flowing from genes.

The catalogue of gene activity known as GTEx – which stands for Genotype-Tissue Expression project  — tracked how genes are dialled up or down in the tissues of 449 organ donors. It is published in Nature and a suite of other journals this week.

“It’s extremely valuable; the most comprehensive catalogue of gene regulation we’ve ever had,” says geneticist Yoav Gilad at the University of Chicago, who was not a member of the project.

In 2001 when we finished reading all three billion letters of the human DNA genome, many predictions followed. Foremost was that the era of personalised medicine was nigh. The strategy was to read the DNA of a large group of people who all suffered from the same illness — say schizophrenia — and compare their DNA to those who did not have it.

In theory, the DNA glitches associated with schizophrenia should just pop out. But, by and large, so-called ‘genome wide association studies’, have failed to deliver. In the case of schizophrenia even though a person’s DNA is 80 to 90% responsible for the illness, finding the glitches has been difficult. And even when those glitches have been convincingly associated with schizophrenia, it hasn’t been all that helpful.

That’s because most of them lie in that part of the DNA we don’t know how to decode, so-called “non-coding DNA”. It accounts for 98.8% of the total. Unlike the tiny 1.2% of our DNA that codes for proteins, we have no cipher for non-coding DNA.

To remedy the situation, researchers needed a way to see how these glitches actually affected the information flow from genes. Funded mostly by the US National Institutes of Health, as well as other government bodies, in 2010 a group of researchers, led by Kristin Ardlie at the Broad Institute of the Massachusetts Institute of Technology, started GTEx.

The team sourced their material from organ donors. Some 960 people consented to have their remaining tissues used once the needed organ had been donated. The GTEx collectors had to be fast, harvesting tissues samples within an hour of death because the messenger RNA, which carries information transcribed from genes, is extremely fragile.

So far, Stephen Montgomery, a GTEx member at Stanford University is amazed by how much individuals vary in their levels of gene activity. In some cases, genes are completely switched off as a result of glitches in nearby non-coding DNA.

“We may have hypothesised it, but we’ve never had an opportunity to look across so many tissues before,” he says. “This allows us to drill down into the mechanisms that underpin DNA variation.”

One fruitful area is that it’s now possible for the first time is to see how DNA glitches play out in different issues. For instance, a particular code glitch may have been associated with schizophrenia. The expectation is that would influence the activity of genes in the brain. But what if instead researchers discover an effect in the immune system?  

“This helps drug developers decide which tissue to target,” says Montgomery.

Other researchers have already been using the catalogue to get a sense of what sort of gene output is normal in different tissues, says Ardlie. For instance, Daniel McArthur’s group at the Broad Institute has used it to as a reference to try to identify which genes are abnormally expressed in people with collagen VI dystrophy, a rare muscle disease.

The use of the data, McArthur and colleagues report, “highlights its utility for the detection and interpretation of variants missed by current standard diagnostic approaches”.

In a News & Views article that accompanies the research paper in Nature, Gilad observes, “the extensive catalogue generated by the GTEx consortium takes us one step closer to decoding the regulatory code of the genome.”

As researchers begin diving into the catalogue in earnest, we may at last be on the cusp of decoding human variation.