December 8, 2010 – Original source: New Scientist
Image: SMC Images/Getty
For the first time, a fetus has had its entire genome mapped from a sample of its mother’s blood. This technical tour de force could open the door to new methods of prenatal genetic diagnosis.
In 1997, researchers led by Dennis Lo of the Chinese University of Hong Kong showed that “floating” fetal DNA can be detected in maternal blood plasma – it passes across the placenta from fetal cells that have broken down.
Lo’s discovery sparked a lot of interest, because it raised the possibility of diagnosing genetic problems in a fetus without the need for invasive procedures such as chorionic villus sampling (CVS) or amniocentesis to extract fetal cells, both of which carry a small risk of inducing a miscarriage.
But it’s hard to distinguish fetal sequences from the larger quantity of a woman’s own DNA. This has so far largely limited practical applications of the technique to unambiguous situations in which particular fetal genes are not carried by the mother. For instance, fetal sex can be determined by detecting sequences from the male Y chromosome. It’s also possible to identify fetuses at risk of rhesus disease, where the mother’s immune system attacks a protein on her fetus’s red blood cells, by looking for the gene for this rhesus protein in the blood of women who are rhesus negative.
Lo has previously worked on methods to detect fetuses with Down’s syndrome from floating fetal DNA. Now, through a combination of brute-force DNA sequencing and sophisticated bioinformatics, his team has shown that it should be possible to detect any genetic disease from a sample of a pregnant woman’s blood.
Match and contrast
Lo recruited a couple who were at risk of having a child with beta-thalassaemia, an inherited form of anaemia. By comparing the father’s genome and fetal DNA extracted by CVS with billions of fragments of DNA from the woman’s blood, Lo was able to construct maps of the entire fetal and maternal genomes. This revealed that the fetus was a carrier of beta-thalassaemia, but was not itself afflicted by the condition.
Of course, the whole point of sampling maternal blood is to avoid performing CVS or amniocentesis. But Lo says that this was just a proof of principle – in practice it should be possible to distinguish fetal and maternal sequences by comparing the fragments obtained from the woman’s blood sample with DNA sequenced from her relatives.
Showing that the entire fetal genome is present in a pregnant woman’s blood is an important development, says Diana Bianchi, a specialist in prenatal genetic diagnosis at Tufts University in Boston. “This paper is beautiful,” she says.
However, at present the analysis is too cumbersome and expensive for clinical use. “At this moment, it would probably cost $200,000 per case,” says Lo. “Cutting costs will be very important.”
While sampling the entire fetal genome for genetic defects may remain prohibitively expensive for some while, Lo hopes within a year to develop a test focused on about five important genetic conditions, with the sequencing costing around $2000.
Still, Bianchi believes that the bioinformatics involved in reliably distinguishing fetal from maternal DNA sequences from a blood sample may prove impractical for many clinical labs. She also points out that the latest estimates put the risk of miscarriage associated with amniocentesis as low as 0.06 per cent. “At some point, someone’s going to need to do an elegant cost-benefit analysis,” she says.
Journal reference: Science Translational Medicine, DOI: 10.1126/scitranslmed.3001720