The 'gene deserts' unravelling the mysteries of disease

Ever since the Human Genome Project was declared complete in 2003, scientists have sought to pinpoint new regions among the three billion letters of our genetic code which may play a critical role in disease.

With the help of technologies which can analyse whole genome samples faster and more cheaply than ever before, vast numbers of genome-wide association studies – dubbed GWAS – have been published, identifying genetic variants linked to different chronic illnesses.

Frustratingly for many geneticists, this has turned out to be the easy bit. The much harder part is understanding how they are relevant. For example, while GWAS have identified segments of DNA associated with inflammatory bowel disease at 215 different chromosomal sites, scientists have only been able to pinpoint the exact mechanisms involved for four of them.  

One of the biggest challenges is that many of these pieces of DNA lie in so-called gene deserts, swathes of the genome that initially appeared to contain nothing of relevance – genetic "junk" that could be disregarded. After all, less than 2% percent of the human genome is dedicated to coding for genes which produce proteins, while much of the remaining 98% has no obvious meaning or purpose.

"You'll go, 'Oh here's a really important association and it increases your risk of many different diseases'," says James Lee, a clinician-scientist who runs a research group at the Francis Crick Institute in London. "But when you actually go and look at that bit of DNA, there's just nothing there."

For many years, gene deserts have been one of the most perplexing areas of medical science, but scientists are slowly managing to accrue information about their apparent purpose and why they exist.

Recently, Lee and colleagues at the Crick Institute published a new investigation into a particular gene desert known as chr21q22. Geneticists have known about this gene desert for more than a decade, because it is associated with at least five different inflammatory diseases from inflammatory bowel disease (IBD) to a form of spinal arthritis known as ankylosing spondylitis. Yet deciphering its function has always proven elusive.

However, for the first time, the Crick scientists were able to show that chr21q22 contains an enhancer, a segment of DNA which can regulate nearby or distant genes, capable of cranking up the amount of proteins they make. Lee refers to this behaviour as "a volume dial". Delving deeper, they found that this enhancer is only active in white blood cells called macrophages where it can ramp up the activity of a previously little-known gene called ETS2.

While macrophages play a vital role in clearing dead cells or fighting off harmful micro-organisms, when the body produces too many they can wreak havoc in inflammatory or autoimmune diseases, flooding into affected tissues and secreting damaging chemicals which attack them. The new study demonstrated that when ETS2 is boosted in macrophages, it heightens virtually all their inflammatory functions.

Lee describes it as "the central orchestrator of inflammation". "We've known for a while that there must be something at the top of the pyramid that is telling the macrophages to behave like this," he says. "But we've never known what it was. The most exciting bit of this, is if we can target it in some way, we might have a new way to treat these diseases."

But if gene deserts are capable of causing us so much harm, why are they in our DNA?

Tracing back in time, Lee's colleagues at the Crick's Ancient Genomics Laboratory were able to show that the disease-causing mutation in chr21q22 first entered the human genome somewhere between 500,000 and one million years ago. This particular DNA change is so ancient that it was even present in the genomes of Neanderthals as well as some ancestors of Homo sapiens.

It turns out that its original purpose was to help the body fight off foreign pathogens. After all, before antibiotics were invented, being able to rapidly switch on a heightened inflammatory response through ETS2 was extremely useful. "Within the first couple of hours of seeing bacteria, it ramps up your macrophage responses," says Lee.

As a result, blocking ETS2 completely could leave IBD patients vulnerable to future infections. However, Lee says when its activity is turned down by between 25 to 50%, it seems to be capable of eliciting a profound anti-inflammatory effect, without risking making the patient too immunosuppressed. While this theory has yet to be tested in clinical trials, the researchers showed that MEK inhibitors – a class of cancer drugs which can dampen ETS2 signalling – were capable of reducing inflammation not just in macrophages but in gut samples taken from people with IBD.

This appears to represent a new pathway to a completely novel class of treatments for IBD patients. "Some of these MEK inhibitor drugs do have side effects, and what we're trying to do now is to make them more targeted and safer, so that for lifelong diseases like IBD, we would actually be able to offer patients a drug that could switch off the inflammatory process and actually make them a lot better," says Lee. 

Now the Crick's researchers are turning their attention to the other four diseases which have been linked to the chr21q22 gene desert, to see whether altering ETS2 activity can also help alleviate the excess inflammation which seems to be driving the condition.

"One of the most significant ones is an inflammatory liver disease called primary sclerosing cholangitis," says Lee. "It's a particularly nasty disease because it can cause liver failure leaving people needing transplants. It can also have a much higher risk of causing liver cancers, and this can happen in young people. And at the moment, there's not a single drug that has been shown to work, there's very little to offer patients," he says.

From cancer to lupus

Scientists also predict that studying gene deserts will yield vital information which will help to improve our understanding of the various pathways involved in tumour development.

As an example, cancer researchers have pinpointed a gene desert called 8q24.21 which is known to contribute to cervical cancer as the human papilloma virus, the main cause of the disease, embeds itself in this part of the genome. In doing so, the virus enhances a gene called Myc which is a well-known driver of cancer. Studies are suggesting that the connection between 8q24.21 and Myc may also play a role in a number of ovarian, breast, prostate and colorectal cancers. 

Richard Houlston, of the Institute of Cancer Research in London, says that various genetic variants which have been identified as contributing to the heritable risk of many common cancers have been found in gene deserts. Knowledge of these target genes will provide opportunities for drug discovery as well as for cancer prevention.

However Houlston points out it is harder to translate this knowledge into new therapeutics for cancer compared to IBD, because tumours are not static targets, but continuously evolve over time. "This is the challenge, whereas with something like Crohn's disease and other bowel conditions, it's not evolving," he says.

Lee is optimistic that the Crick's work on IBD will provide a template for how researchers can find new ways of understanding the pathways involved in all kinds of autoimmune and inflammatory diseases. The institute's scientists are now investigating other gene deserts which have been associated with conditions such as lupus, a disease in which the immune system damages the body's tissues, leading to symptoms such as skin rashes and tiredness.

Other research centres around the world such as the University of Basel in Switzerland are also examining how single inherited mutations in gene deserts could lead to some rare genetic diseases. Three years ago, Basel scientists discovered how one of these mutations could lead to babies being born with limb malformation due to its regulatory effects on a nearby gene.

Lee predicts that understanding the roles of gene deserts will ultimately help improve the notoriously inefficient drug development process. "Making new drugs for these diseases is terribly unsuccessful," he says. "Only about 10% of the drugs going into clinical studies are ever approved at the end, so 90% of them fail because they don't make people better. But if you know that your drug going into development is actually targeting a pathway supported by genetics, the chances of that drug actually being approved is at least somewhere between three- and five-fold higher."

-bbc