Stephen Quake knows firsthand the wrenching decision parents face when a doctor recommends an amniocentesis or chorionic villus sampling test to check for chromosomal abnormalities in a fetus. His wife underwent one of each test with her two pregnancies. The invasive tests use a needle to collect fetal cells through the mother's abdomen. Both tests carry a low, but real, risk of miscarriage (about 1 in 400). They are most often used to find out if a fetus has abnormal copy numbers of certain chromosomes, such as trisomy 21, which leads to Down syndrome, and trisomies 13 and 18, which have very low survival rates after birth.

So when Quake read an article about researchers trying to capture information from fetal DNA circulating in the mother's blood, he immediately saw a possibility for making the invasive tests obsolete.

Quake, an HHMI investigator at Stanford University, had recently published the first single-molecule DNA sequencing method. It occurred to him that he could harness the power of modern sequencing techniques to solve what is essentially a problem of counting molecules.

“Sequencing is going through an amazing revolution right now and there is opportunity for great creativity in using it for things other than straightforward genome sequencing,” he says. Quake's group devised a method to amplify fetal DNA from a mother's blood sample and sequence the millions of tiny DNA fragments. Next, they matched up each fragment to its corresponding chromosome—as if each fragment gets to vote for a particular chromosome.

“Then we look for ‘voter fraud,’” he says. “In other words, is any sequence overrepresented, indicating that there are three copies of that chromosome rather than two?” In a 2008 pilot study, the method correctly identified the number of fetal chromosomes in 18 pregnancies—there were 12 cases of trisomy—and it worked as early as the 10th week of pregnancy. This work was published in October 2008 in Proceedings of the National Academy of Sciences.

Quake says the simple blood test could be available to doctors in the next couple of years. He thinks the test could work as early as the 7th or 8th week of pregnancy—the time when most women have their first prenatal doctor's visit. Such a test would carry no risk to the fetus and would allow families to make decisions about a trisomy pregnancy much earlier.

“What we went through as a family definitely sensitized me to the problem,” says Quake, the father of a 4-year-old boy and 7-year-old girl. “I still get e-mail every day from people who want to get the noninvasive test. It affects a lot of people.”
—K.P.

HHMI INVESTIGATOR