NEW YORK (GenomeWeb) − Australian scientists have optimized protocols to capture and sequence methylated free circulating DNA from human plasma, an approach they are now exploring for cancer diagnostics.
The researchers, led by Goli Samimi, head of the ovarian cancer research group at the Garvan Institute of Medical Research in Darlinghurst, Sydney, published their method in BMC Genomics earlier this month. According to Samimi, this is the first study to sequence methylated free circulating DNA in total, rather than specific target sequences.
Free circulating DNA, or fcDNA, and RNA in blood have many potential clinical applications as biomarkers in areas such as reproductive medicine, diagnostics, and transplant medicine because they can be collected by a simple blood draw rather than an invasive method, like a tissue biopsy.
But because fcDNA is only present in minute quantities – typically, between 1 and 27 nanograms per milliliter – its genomewide analysis presents many technical challenges, Samimi told In Sequence, and no standards currently exist for collecting, processing, and sequencing fcDNA.
"We wished to develop optimized methodology for the capture [and] processing of high-quality methylated free circulating DNA, library preparation, and downstream genomewide next-generation sequencing," he said.
According to the paper, it is difficult to capture methylated fcDNA by affinity methods because its concentration is so low and other DNA binds non-specifically. In addition, less than 10 percent of the input DNA is usually recovered, so large volumes of blood are needed to obtain enough material for library construction and next-gen sequencing.
For their study, the researchers analyzed fcDNA from healthy donors. They first compared how much total fcDNA they could obtain from plasma and how much from serum, where clotting has occurred. While the yield of fcDNA from serum was about 10 times higher than from plasma, much of it likely came from leukocytes that had lysed during the clotting process.
Next, they investigated how storage of blood samples prior to processing affects fcDNA concentrations and found that after several hours of storage, fcDNA levels increased in both sample types due to cell lysis, but less so in plasma. To prevent such contamination with genomic DNA, they recommend using plasma rather than serum for fcDNA studies.
Following their initial studies, they captured and sequenced methylated fcDNA from the blood of five volunteers. They first isolated fcDNA from 35 milliliters of plasma using a modified protocol for Qiagen's QIAamp Circulating Nucleic Acids kit. They then captured methylated fcDNA from about 50 nanograms of input fcDNA, using a methyl-binding capture protocol that they had modified to minimize the amount of non-specific DNA binding. Using this protocol, they recovered up to 15 percent of the DNA.
Using a modified ChIP-seq protocol, they then prepared Illumina sequencing libraries from the methylation-enriched fcDNA, using 4 nanograms of input DNA, and sequenced them on a HiSeq 2000. In each sample, more than 50 percent of the total mappable reads were unique, showing that they had obtained "a good level of library complexity."
They also validated that the DNA they sequenced came from methylated fcDNA by performing bisulfite sequencing on selected regions from one sample and found that the results were concordant with those from MBD capture and sequencing.
"These results provide further support that whole-genome analysis of even small amounts of fcDNA can provide high-quality, validated genomic data that strengthen the potential of fcDNA utility in clinical applications," the authors concluded.
"The results presented here are impressive," said Jay Shendure, an associate professor of genome sciences at the University of Washington, who did not participate in the research, pointing out the challenges associated with working with such small amounts of DNA. Two years ago, Shendure's team published a study in which they sequenced the genome of a fetus from free circulating DNA in the mother's blood. "The next challenge will of course be to look in depth at whether methylation patterns in cfDNA can be associated with phenomena of interest, [for example] cancer," he added.
That is what Samimi and his team are currently pursuing. They are using their approach to identify an fcDNA methylation signature that can differentiate between plasma from women with ovarian cancer and plasma from controls. "In the long-term, once our cancer-specific methylation signature is validated in independent cohorts, we hope to apply this method to develop a test for early detection of ovarian cancer in women who are at high risk," Samimi explained.
DNA methylation often happens during cancer development, and promoters in particular become hypermethylated as cancer progresses, he said. Because DNA methylation assays are highly sensitive and specific, they are "an ideal biomarker for early cancer detection, as well as tracking disease progression [and] recurrence."