NEW YORK (GenomeWeb News) – In a proof-of-principle paper appearing online today in Science Translational Medicine, researchers from China and the US reported that it's possible to do whole-genome analyses of fetal DNA circulating in maternal blood.
The research team, which included Sequenom Chief Scientific Officer Charles Cantor, sequenced DNA from a pregnant woman's blood sample, obtaining sequence information from both the woman and her fetus. They then combined this sequence data with genotype information for both parents, using a relative haplotype dosage method to map the fetal genome.
"[W]e have used paired-end massively parallel sequencing to study the genomic sequence and size distribution of fetal DNA in maternal plasma," corresponding author Dennis Lo, a chemical pathology researcher at the Chinese University of Hong Kong, and co-authors wrote. "We further constructed a genome-wide genetic map of a fetus from the maternal plasma DNA sequences and from information about the paternal genotype and maternal haplotype."
Some 10 percent of the DNA found in maternal blood represents fetal DNA, raising the possibility that this DNA might be used for non-invasive prenatal testing rather than relying on DNA samples obtained through invasive methods such as amniocentesis or chorionic villus sampling, which pose a slight miscarriage risk.
To date, much of the research on prenatal testing with circulating fetal DNA has focused on detecting aneuploidies such as Down syndrome or trisomy 21 — an application that Lo's team and others have published on in the past.
For the new study, the team got a more comprehensive view of a fetal genome in maternal DNA, Lo told GenomeWeb Daily News, calling the work a "quantum jump" forward.
"Trisomy 21 is actually relatively simple to detect compared to what we do here," he explained, since it involves the duplication of an entire chromosome. "If you do a sequencing-based test, any sequence — any fragment — that comes from chromosome 21 will contribute towards the eventual diagnosis."
In contrast, Lo said, the current study focuses on SNPs, which are much more difficult to identify, and requires an ability to distinguish between maternal DNA and nearly identical bits of fetal DNA inherited from the mother.
To tackle such problems, the researchers used samples from a couple who came to a Hong Kong clinic for prenatal beta-thalassemia testing.
Beta-thalassemia is an autosomal recessive blood condition caused by mutations in the hemoglobin beta subunit gene HBB that is particularly common in Hong Kong, Lo noted.
The mother-to-be carried a single nucleotide mutation in HBB, while the father carried a four base deletion in the gene.
Researchers obtained the blood sample when the woman was about three months pregnant and before CNV samples were taken and then used paired-end sequencing with the Illumina Genome Analyzer IIx to sequence maternal and fetal DNA in the blood sample to an average of 65 times coverage.
They also got blood samples from the father and used the Affymetrix Genome-Wide Human SNP 6.0 array to genotype each parent's sample, as well as the sample collected during CVS, at roughly 900,000 SNPs.
By bringing together genotype and sequence data, the team was able to distinguish between maternal and fetal sequences, in part because the fetal genome contains alleles present in the father but not the mother.
For example, the researchers detected nearly 94 percent of the SNPs at which the mother and father were homozygous for different alleles and for which the fetus was hemizygous.
Overall, they explained, their findings indicated that "the entire fetal genome is represented in maternal plasma, and is present in a constant relative proportion to maternal DNA in maternal plasma."
"It's reassuring to actually see that the whole thing is there," Lo said, noting that their data also allowed them to look at the size distribution of fetal DNA compared to DNA present in maternal blood overall.
The researchers then used genotype and relative haplotype dosage, or RHDO, data to create a map of the fetal genome, determining the parent of origin as they worked their way across the genome.
"[T]o deduce a genome-wide fetal genetic map, one has to analyze the sequencing data in the context of parental genetic maps and to assemble these data in a series of fetal inheritance blocks," the authors noted.
Rather than looking at one SNP at a time, Lo explained, the RHDO strategy allowed the team to assess one haplotype at a time.
When they specifically looked at the HBB gene, the team found that the fetus was a carrier for beta-thalassemia but was not prone to get the disease since the mutated version of the gene inherited from the father was offset by a wild type version from the mother.
If and when personal sequencing becomes more widespread and genome sequence data is available for parents-to-be, Lo noted, it will be possible to create much more refined fetal genome maps. Moreover, he and his co-authors explained, a similar strategy may be useful for analyzing circulating DNA originating from cancers or even transplanted tissue.
Given their findings so far, those involved in the study believe it will be possible to use targeted sequencing of fetal DNA to simultaneously test for mutations behind several diseases — a strategy that's expected to significantly decrease the cost of testing. The method used in the current study comes with a price tag of roughly $200,000 per case.
Lo said the team has filed patent applications related to the technology used in the study and are pursuing related commercialization opportunities.