AP IMPACT: Gene sweeps yield discoveries
- AP IMPACT: Gene sweeps yield discoveries
By MALCOLM RITTER, AP Science Writer
March 30, 2008
NEW YORK - Scientists are scanning human DNA with a precision and
scope once unthinkable and rapidly finding genes linked to cancer,
arthritis, diabetes and other diseases.
It's a payoff from a landmark achievement completed five years ago
the identification of all the building blocks in the human DNA.
Follow-up research and leaps in DNA-scanning technology have opened
the door to a flood of new reports about genetic links to disease.
On a single day in February, for example, three separate research
groups reported finding several genetic variants tied to the risk of
getting prostate cancer.
And over the past year or so, scientists have reported similar
results for conditions ranging from heart attack to multiple
sclerosis to gallstones. The list even includes restless legs
syndrome, a twitching condition best known as "jimmy legs" in an
episode of "Seinfeld."
Interviews with scientists at the center of this revolution and a
review of published studies over the past six months by The
Associated Press make clear the rapid adoption of the new technology
and the high expectations for it.
Since 2005, studies with the gene-scanning technique have linked
nearly 100 DNA variants to as many as 40 common diseases and traits,
scientists noted this month in the Journal of the American Medical
"There have been few, if any, similar bursts of discovery in the
history of medical research," two Harvard researchers declared last
summer in the New England Journal of Medicine.
What does all this excitement mean for ordinary people? Not so much
just yet. Simply finding the genes that can raise the risk of an
illness doesn't mean you can prevent the disease. And developing a
treatment for it can take years.
But there have been some payoffs already.
One involves a leading cause of blindness in older people, age-
related macular degeneration. A series of genome-wide scans, the most
recent in 2005, "led to huge breakthroughs in understanding" that
disease, said Stephen Daiger, a Houston scientist.
When scientists implicated a particular gene that's involved in a
system of disease-fighting proteins in the blood, it gave scientists
a "slap-on-the-forehead kind of insight ... into the biology of
what's going on," said Daiger, a vision genetics expert at the
University of Texas Health Sciences Center.
That galvanized research into the disease. And at least one new drug
is being tested in patients now.
What's made this and other hopeful findings possible is the "genome-
wide association study," which lets scientists scan the entire
complement of DNA from thousands of people in unprecedented detail.
While the basic technique is not new, its popularity has exploded
recently because of cost-cutting advances in technology and
discoveries about the genome.
"It lets you go searching for that needle in the haystack," says
Michael Watson, executive director of the American College of Medical
It's a big haystack. DNA is made up of long sequences of building
blocks, sort of like sentences composed from a four-letter alphabet:
A, C, G and T. The human genome contains about 3 billion letters,
about as many as the total number of letters and digits in more than
100 Manhattan phone books.
Scientists have identified the order of the letters in the human
genome, a feat the government declared accomplished in 2003. But of
course, different people have slightly different DNA sequences.
People commonly differ in what letter they have at about 10 million
positions along the full genome. Some folks may have a T where most
people have a C, for example.
And those single-letter variations are key to the genome-wide scans.
Basically, scientists compare DNA from a large number of people, some
sick with a particular disease, and others healthy. They can look at
a half-million or more positions to see what letter appears. If sick
people tend to show a different result than healthy ones say, if
they tend to have a T in some spot more often than healthy people do
it's a red flag.
It suggests that some genetic influence on the risk of that disease
comes from that spot or nearby. So it gives scientists a specific
place to look more closely for a disease-promoting gene.
In practice, genome scans can be big undertakings.
Scientists in Iowa and Denmark are searching blood samples from 7,000
babies and new mothers in the United States and Denmark for genetic
variations that raise the risk for premature birth.
DNA will be extracted, and early this summer, more than half a
million spots on the microscopic strands from each mother and baby
will be assessed for clues to where the genetic variations may lie.
The DNA will be analyzed at the Center for Inherited Disease Research
at Johns Hopkins University in Baltimore. Robots will put a tiny drop
of DNA-bearing solution from each person onto a clear glass slide
roughly the size of a business card, with four drops per slide.
The lab's DNA scanners, blue boxes each about twice as big as a
desktop printer, will reveal what DNA "letter" appears in more than
580,000 spots in the genetic material, said lab director Kimberly
This scan takes about half an hour per sample. Once the results are
available, the scientists will use statistical tests to find the
telltale signs of a possible gene affecting risk of premature birth.
They'll double-check to make sure any such signal shows up in more
than one population.
Even five years ago, such a detailed examination of DNA from so many
people would have been inconceivable.
Genome scans offer some major advantages over previous gene-hunting
techniques. Scientists don't have to start by guessing what genes
might be involved in a disease, or confine themselves to families
where a tendency to an illness is inherited.
And the genome-scan approach reveals genes with only subtle influence
on the risk of getting sick, too slight to be found by earlier
methods. That's just the kind of gene that plays a role in common
illnesses like heart disease.
Even if its impact on risk is small, a newly found gene can be a
bonanza to scientists if it reveals something new about the biology
of a disease. That in turn can give hints for finding new treatments.
For non-scientists, the most direct payoff of finding new disease
genes may be in devising tests to identify people at elevated risk
for a particular disorder.
Most genetic variants found in the genome scans boost a person's risk
by around 50 percent. If the disease risk is fairly low, that's "not
something you'd lose much sleep over," Watson said.
More useful, he said, is the notion of finding variants in maybe a
half-dozen genes that affect the risk for a disease, then testing a
person for all of them at once to come up with a more powerful
Earlier this year, for example, Swedish researchers reported
preliminary evidence that men with four or five particular gene
variants ran more than four times the risk of getting prostate cancer
than men with none of them. When family history was factored in, such
a combined test could identify men who ran a nine-fold higher risk.
An Iceland-based company, deCode Genetics, announced in February that
it is offering a test for eight genetic variants related to prostate
cancer. Altogether, the variants make 10 percent of men run twice the
normal risk of prostate cancer, and 1 percent run three times the
normal risk, the company said.
Dr. Teri Manolio of the National Human Genome Research Institute said
it'll take more work to figure out the value of genetic testing for
prostate cancer. There is no proven treatment to prevent it; the only
advice to a man at higher risk would probably be for more aggressive
screening for the disease.
Then there's the question about what people will do with gene test
results. What if you already know that everybody should watch their
weight, for example, and then a DNA test shows a heightened risk for
diabetes and your doctor tells you to ... watch your weight?
Maybe people would pay more attention to health advice if they knew
they were genetically vulnerable to getting sick otherwise. But maybe
not. It's an open question, Manolio said.
"I think some people will," Watson said. "I think some people just
won't, because they're the kind of people who aren't influenced by
those sorts of things.... I'm not pessimistic or optimistic, but I'm
sure not everybody does the right thing."
On the Net:
Explanation of genome scans: http://www.genome.gov/20019523
More information: http://www.genome.gov/20019523 2
Catalog of published gene scans: http://www.genome.gov/GWAstudies