Baby's genome sequenced before birth from mother's blood

Using blood from a mother midway through her pregnancy, researchers have successfully sequenced the complete genome of the woman's unborn child.

Writing in the journalScience Translational Medicine, University of Washington, Seattle,scientist Jay Shendure and his colleagues have published their approach to reading the genetic blueprint of a developing baby by piecing back together DNA fragments washing around in the maternal bloodstream.

The approach, which according to Shendure cost about US$50,000 to complete, involved three separate phases of genetic sequencing.

First, DNA extracted from saliva samples was used to compile the complete genomic fingerprint of the father, while DNA from white cells in the mother's peripheral blood provided the maternal sequence. For the mother, the researchers also painstakingly decoded the DNA sequence from each of her 23 pairs of chromosomes - this is called the haplotype - so that they knew exactly what DNA sequences were sitting alongside one another.

The team then also sequenced the DNA contained in the cell-free plasma from the maternal blood sample; by this stage of pregnancy, over 10% of the DNA it contains is of foetal origin.

This produced a mixture of sequences, some of which were maternal, others foetal. But because the researchers knew the maternal sequence in detail, and what sequences were present in the father, it was possible to pick out and reassemble just the foetal DNA code. And this also meant that 39 new genetic changes (mutations) carried only by the developing baby were also predicted.

To check their results, the researchers matched up their sequences with the a version of the child's genome produced using a blood sample from the umbilical cord collected after the child was born.

The accuracy of the procedure was over 98%, meaning that it could be employed as a minimally invasive means of genetic screening without the miscarriage risks associated with amniocentesis or chorionic villus sampling tests currently employed.

And although the study used samples from 18 weeks' gestation, according to Shendure it should be possible to achieve the same results much earlier in pregnancy just by performing a greater number of sequencing reactions to compensate for lower circulating foetal DNA levels.

But it's not a fait-accomplit. "At the moment it's not practical, because it's a lot of effort to do the maternal haplotyping," explains Shendure. "But we're working on that."