It was quite a vision of our genetic future. Shortly before the first drafts
of the human genome were unveiled in 2001, Francis Collins spelt out in bold
terms what the project he had pioneered would ultimately mean for medicine.
By 2010, scientists would understand how genes contribute to at least a dozen
common illnesses such as diabetes and heart disease, the director of the US
National Human Genome Research Institute said. Preventive therapies would
soon be developed to match.
A new era of bespoke medicine lay ahead, in which drugs tailored to
individuals’ genetic profiles would replace the traditional
“one-size-fits-all” approach. Insights from humanity’s genetic code were
poised to transform healthcare.
Just a year ago, Dr Collins’s timetable looked like fantasy. Science had had a
version of the genome for five years, and parts of it for longer still, yet
it had added surprisingly little to knowledge of the genetic roots of common
diseases. The promised revolution had yet to begin.
It is now under way. This week’s announcement that four common genes have been
reliably linked to breast cancer is only the latest and most spectacular of
a series of remarkable results that have led cautious scientists to speak
openly of a tipping point in their understanding of how DNA affects our
health.
Scores of genes that cause or influence disease are now known to science, but
until recently it has been difficult to identify any that contribute to the
greatest causes of ill-health in the modern world – diabetes and heart
disease, mental illness and most cancers.
The successes of the first phase of genome research were limited largely to
genetic mutations that have devastating effects but for very few people,
such as those that cause Huntington’s disease and cystic fibrosis.
Four out of five women with mutated BRCA1 or BRCA2 genes will develop the
disease. Only one in 500, however, carries these defects in the first place.
Their discovery has transformed screening for a small minority, but it means
nothing for more than 95 per cent of the 44,000 British women whose breast
cancer is diagnosed each year.
Scientists have long known, from family and twin studies, that other genetic
factors affect women’s risk, but hitherto it has been impossible to pinpoint
what they are. The same has been true of heart disease, diabetes, and other
modern plagues.
All that is finally starting to change. After years of treading water,
geneticists have started to publish a cascade of data about very common
genetic variants that can subtly influence people’s risk of developing
common diseases.
The new breast cancer genes, discovered by a team assembled by Cancer Research
UK, have much less of an effect than the BRCA genes: the most damaging
genetic profile raises the lifetime risk to about 17 per cent. But they are
much, much more common: one variant is carried by one in six women, and the
rarest by one in 16.
That discovery followed the announcement in April of the first common gene
that has been shown reliably to contribute to obesity. Again, the risk is
low, but many are affected: one in six has inherited versions of the FTO
gene that raise it by 70 per cent.
Similar studies have identified common genetic variants that predispose to
heart disease and type-2 diabetes. Next week a Wellcome Trust consortium
will publish details of many more that are linked to seven major conditions,
including high blood pressure, bipolar disorder and rheumatoid arthritis.
Where genetics was once capable of pinning down only rare mutations with a
catastrophic impact, it is now tracing variants with smaller effects that
are much more widespread. You might call it the democratisation of the
genome.
“Democratisation is a good word for it,” said Mark Walport, director of the
Wellcome Trust, the British bio-medical charity that has funded much of the
key research. “What we are finding now is relevant to everyone, not just to
those rare families who happen to be exceptionally unfortunate.”
Professor Bruce Ponder of the University of Cambridge, whose team identified
the new breast cancer genes, said: “The genes we have seen so far are strong
genes for spectacular stuff, but this is the real genetics that underlies
variation in the population at large. These are the genes that affect how
people react to their environments, and that influences, however slightly,
their chances of developing common diseases.”
This step change has been brought about by new technology and a change of
approach. Traditional gene hunts have relied on a technique called linkage
analysis, in which scientists study families in which several members have
developed an inherited disorder. DNA samples from individuals with the
disease are compared with samples from those who are unaffected. About a
hundred subjects are usually enough, and no more than 300 DNA markers
generally need to be scanned.
While this is efficient, it works only for genes with a large impact, raising
risk by at least 200 per cent. Most genetic variants do not work like that,
and to find them scientists have turned to a new method called whole genome
association.
Such studies use unrelated subjects and need to be much larger. The Wellcome
Trust Case Control Consortium, which found the FTO gene and publishes more
results next week, compares 2,000 people with particular diseases with 3,000
unaffected controls. As the subjects are not related, scientists must also
look at hundreds of thousands of DNA markers to find any effects.
This approach can find common variants, which raise risks by as little as 10
to 20 per cent. The drawback is the vast amount of data that must be
processed – it has become possible only with the advent of microarrays or
“gene chips” that can scan hundreds of thousands of genetic markers at once.
It is this that has enabled the spectacular recent advances. “It’s been known
in principle that whole genome association studies should be more powerful,
but only in the last two years has it become affordable to do them with the
large sample sizes we need,” said Peter Donnelly, of the University of
Oxford, who chairs the Case Control Consortium.
“Since the genome was published, progress has been slower than people had
hoped, but the pace has zoomed now. Each week and month, exciting new
findings for important diseases are being published. That pace will
continue. What we knew just a year ago is enormously different from what we
know now.”
This enhanced knowledge is changing the way scientists think about genes and
disease. It is now recognised that there are very few conditions that are
caused by single genes with big effects – the influence usually comes from
dozens of genes, each of which has only a minor impact on its own. “We are
mostly finding that for any given disease there are zero or at best one or
two genes with large effects,” said Mark McCarthy of the University of
Oxford, who led the team that found FTO’s effect on obesity.
“Then there is a sprinkling of genes, perhaps five to ten, with modest effects
of 10-20 per cent, and there may be many hundreds with even smaller effects.”
For many conditions, indeed, these smaller effects may be all there are.
Robert Plomin of the Institute of Psychiatry in London, who specialises in
behavioural disorders such as autism, doubts whether whole genome
association will even find many moderate effects in his field.
This means that the genome will rarely provide medicine with “magic bullets”
that tackle disease by correcting faulty genes.
First, most of the variants that contribute to a high risk of breast cancer or
diabetes are common, and are thus likely to have other important functions.
We would use genetic engineering or gene therapy to alter them at our peril.
Then there is the problem of numbers. So many different genes are playing a
minor role that it is implausible to think we might correct them all.
The real value of this knowledge will thus be in understanding the basic
biology of common diseases that affect millions, so scientists can design
better drugs to target the processes that our genes set in train. It should
also allow the development of tests for gene combinations that create a
significant risk.
That would then allow patients and their doctors to manipulate a factor that
almost invariably interacts with genes to affect disease, and which is much
easier to control: the environment. Women with a high breast cancer risk
could be referred for more frequent screening. Those in danger of heart
disease might change their diet and exercise patterns. In some cases, drugs
will be appropriate, to change the biochemical environment in which genes
operate in the body.
“What we are going to get out of this in the end is not genetic engineering,
but environmental engineering,” Dr Plomin said. “A lot is going to be about
changing behaviour, about education. It’s not always going to be: you’ve got
this genetic problem and here’s a pill. Most genetic effects are going to be
too small for that.”
Giving up their secrets
Breast cancer
— New data to be published in the journal Nature this week identified
four new genes and one region of the genome that raise the risk of the
disease
— One in six women carries the riskiest variant of the FGFR gene, which
raises the lifetime risk by 40%-60%
Obesity
— Details of the FTO gene were published in Science in April. This gene
comes in two varieties: 16% of people have two copies of the high-risk
variant, 50% have one high-risk and one low-risk, and 34% have two low-risk
variants
— Those with two high-risk copies have a 70% greater risk of obesity
than those with two low-risk copies. They weigh an average of 3kg more
— Those with one high-risk copy have a 30% greater risk. They weigh an
average of 1.2kg more
Diabetes
— Details of three common genes that raise the risk of type 2 diabetes
by up to 20% were published in Science last week. The discovery raises the
total number of genes implicated in the disease to nine
— The Wellcome Trust Case Control Consortium is conducting further
research on type1 diabetes. More results on both forms of the disease are
expected to be published in Nature next week
Heart disease
A region was discovered on chromosome 9 last month that raises the risk of an
early heart attack by between 40%-60%
— The most damaging variant is carried by between a fifth and a quarter
of white Caucasians
Hypertension
— Under study by case control consortium; data to be published next week
Bipolar disorder
— Under study by case control consortium; data to be published next week
Rheumatoid arthritis, Crohn’s disease
— Under study by case control consortium; data to be published next week
Multiple sclerosis, auto-immune thyroid disease, ankylosing spondylitis
— Subsidiary projects of case control consortium; data expected soon
Sources: Nature, Science, Wellcome Trust
Has the promised report on Gene Therapy and Bi-Polar Disorder been published yet and was it good news?.
Mary Keane, London, GB
Having worked as a molecular biologist on the development of novel viral vectors for cancer gene therapy I found myself becoming increasingly skeptical over the years about many of the underlying assumptions upon which this technique was founded. You pointed out in your article that very few conditions are caused by single genes and the idea therefore of a disease therapy based on using vectors to replace faulty genes starts to appear rather simplistic. Yet I wonder if a much deeper problem is that we’re still not clear what exactly constitutes a gene – is it simply a defined linear sequence of nucleotides being transcribed into an RNA chain specifying a polypeptide? What about non-coding regulatory regions, RNA modification and processing, enzymatic repair of DNA, the chemical modification of chromatin proteins bound to the DNA and regulation by RNAi?
Dr K, Leeds,
Regarding Dr. Plomin's statement that: “What we are going to get out of this in the end is not genetic engineering, but environmental engineering. A lot is going to be about changing behaviour, about education." With all respect to the good doctor, it is much more likely that cures will come from gene manipulation than diet, exercise, or other environmental fixes. The article points out that people with 2 copies of the worst FTO variant have a 70% chance of developing obesity. What about the other 30%? Why aren't they obese? Are they all excercising 5 days per week and eating fiber at each meal? It is much more likely that these fortunate folks possess yet-undiscovered gene s that provide a protective effect powerful enough to cancel the impact of 2 bad copies of FTO. It is in the discovery of these protective genes that the greatest hope for effective treatments will depend. My advice: Focus research on the 30% and the genes that protect them.
R Sheinberg, Torrance, USA / CA