Genomics

Researchers focus on differences between groups to find bad DNA

By David Ewing Duncan|Monday, October 24, 2005

More than 15 million elderly Americans slowly lose their eyesight due to age-related macular degeneration: an accumulation of inflammation-related protein and fat beneath the center of the retina that slowly destroys it. The disorder runs in families, but the gene responsible had eluded scientists. In March three separate teams announced that they had zeroed in on a DNA sequence on chromosome 1 that carries the gene for complement factor H, a protein involved in regulating inflammation. A mutation in this gene may account for about half the cases of macular degeneration in the United States.

This genetic culprit was revealed by a second and little-heralded phase in the Human Genome Project, and it comes five years after a rough draft of the entire human genome was announced. Touted as the key to deciphering the genetic book of life, that initial sequence has proved most useful for finding or confirming genetic mutations that cause rare diseases such as Tay-Sachs disease and Huntington’s. These alterations are relatively easy to identify because they can be traced and isolated in families with a history of the disease. Finding genetic clues to common diseases is much more difficult because many genes—as well as lifestyle and environmental exposures—may be involved. So rather than search the entire genome for genes related to common cancers, heart disease, asthma, and diabetes, scientists have turned to detecting inherited variations in the genomes of different populations and how they may be linked to disease vulnerability.

The human genome contains 6 billion nucleotides, but the differences between one person’s DNA and another’s are slight. Still, the scattered differences that are common to various populations provide a map for hunting down disease. Looking at whether shared variations—single-letter misspellings—in vast chunks of DNA called haplotypes are correlated with certain diseases can help researchers decide which segments of the genome to scour for more specific clues. The project, called the HapMap, is the effort of a six-nation consortium. The group has spent more than two years analyzing DNA from 269 donors with diverse ancestry. They announced their first major milestone—identifying 1 million of these common variations—in February.

To find the macular degeneration gene, researchers compared DNA from people with and without the disease. All three teams found sections of DNA—haplotypes—that differed and ultimately pinpointed a single-letter difference that changed the amino-acid content of complement factor H. HapMap researchers say that refining the map further will speed up such discoveries, and they plan to release a new version this month that will include 4 million single-nucleotide variants.

For now, the HapMap project does not mean new genetic tests for diseases are about to appear, but at least one company is considering whether to develop a test for genetic vulnerability to macular degeneration. Although having the mutation more than doubles a person’s risk of the disorder, which causes loss of central vision, it doesn’t mean that person will automatically develop the disorder. Moreover, because no one is certain about how to minimize risk, a test wouldn’t have much practical value, according to Stephen Daiger, a geneticist at the University of Texas Health Sciences Center.

Yet there is good news for some patients. In July Genentech announced that the drug Lucentis stabilized or even improved vision in a yearlong clinical trial of patients with so-called wet macular degeneration, a form that involves overgrowth of blood vessels. The drug works by inhibiting a protein involved in blood-vessel formation. Lucentis is one of a handful of other drugs that combat this form. No drugs are effective for the other, far more common form of the disease.

Although tests for many rare genetic disorders such as Tay-Sachs disease and Huntington’s disease are available, only a few tests can predict the risk of common diseases. And even where studies have linked a gene to a disease, people carrying the gene or genes will not necessarily develop the disease; they simply face an increased risk. Nor does not having the gene mean the person will remain disease free, because many diseases result from interactions of multiple genes and the environment. Below are examples of genetic tests that are currently available and how predictive each is for individuals. Genetic counseling should accompany these tests.


David Altshuler, director of medical and population genetics at the Broad Institute of Harvard University and MIT and an associate professor of genetics and genetic medicine at Harvard, is hunting for genetic clues to diabetes and prostate cancer.

Can population genetics—and genetic testing of individuals—predict if someone will get a disease?
A: Population genetics is the study of variations in heredity that run in families. We try to find a gene or genes responsible and learn how they cause the variation. It is seldom predictive. Having a gene associated with a disease is an indicator only that you might get that disease. It’s like taking a cholesterol test. It’s a warning that the odds are higher than normal. It can indicate a change in diet or lifestyle or an intervention with a drug.

How accurate is the information?
A: Genetic markers are found all the time that come from comparing people with a disease to those without the disease, but until you figure out what the gene does, it is not useful information. There are a lot of hucksters out there saying that this or that gene will predict a disease—it’s not true for most common diseases. You have genes that definitely cause rare genetic disorders like Huntington’s. And tests for genes for diseases like breast cancer have been validated. But for most common diseases, the data need to be thoroughly understood and the link validated by having many labs replicate the findings. Many times the findings are relevant only to a particular population where the test was done. People should be skeptical of most of these tests.

Does sequencing the genes of animals such as the mouse and the dog help us understand ourselves?
A: If you want to understand human genetics, you need to line up a lot of species to compare genes to find how similar and different they are. This tells us how these genes evolved and gives insight on what the genes do and how mutations might be treated with drugs.

Will we ever have a little card we carry around that has our genome on it?
A: We could already carry cards like that with family histories or diseases and our latest blood tests, but we don’t. I’m less interested in tests than treatments.


Gene Test Report Card

Alzheimer's - Of the three variants of the ApoE gene, ApoE4 carries the highest risk for late-onset Alzheimer’s (other genes are associated with early-onset Alzheimer’s). No cure or effective treatment exists for Alzheimer’s, so ethicists usually advise against genetic testing unless there is a strong family history or a diagnosis is in question.

Breast Cancer - Out of more than 200,000 new cases of breast cancer each year, about 10,000 or more are linked to the BRCA1 or BRCA2 gene. Testing positive for the gene means you have at least a 35 percent chance of developing breast cancer. People with a family history of the disease use the test results to help them decide what steps to take to preserve their health.

Colon Cancer - About 7,000 patients a year have a hereditary colon cancer associated with one or more gene variants. All of these variants are linked to the onset of colon cancer by age 45 in most carriers. Testing may be advisable for people with a strong family history of colon cancer in order to begin screening and management.

Comment on this article
ADVERTISEMENT

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

ADVERTISEMENT
ADVERTISEMENT
Collapse bottom bar
DSCMayCover
+

Log in to your account

X
Email address:
Password:
Remember me
Forgot your password?
No problem. Click here to have it emailed to you.

Not registered yet?

Register now for FREE. It takes only a few seconds to complete. Register now »