Author(s): Hallie Gaitsch, NINDS/NIH (NIH Oxford-Cambridge Scholars Program); Jing-Ping Lin, PhD, NINDS/NIH; Daniel Reich, MD, PhD, NINDS/NIH //

ABSTRACT: Cell-free DNA (cfDNA) is present in human plasma during both health and disease states. These DNA fragments are derived from intra- and extracellular nuclease cleavage of DNA in cells undergoing apoptosis or necrosis. Sequence differences in circulating tumor DNA, including known tumorigenic mutations, have been of particular interest in the field of oncology for their potential use as prognostic indicators and early-stage biomarkers for screening. However, the nongenetic characteristics of circulating cfDNA can also provide useful information that is applicable to diseases which do not involve acquired genetic mutations, including many neurologic disorders. Epigenetic signatures such as DNA methylation patterns can be used to identify the tissue- and even cell-type of origin for cfDNA samples. Given that ~28 million unique CpG methylation sites exist within the human genome, the use of methylation array technology and deconvolution algorithms make it possible to identify differentially methylated regions corresponding to various cell types and levels of gene expression. This also opens the possibility of using the cfDNA of cell-types which are minority populations in their tissue, including oligodendrocytes and oligodendrocyte progenitor cells (OPCs) which are central players in multiple sclerosis (MS) pathology, as targets for liquid biopsy development. The low concentration of cfDNA in human blood and high potential for genomic DNA contamination necessitate careful handling during nucleic acid extraction from fluid samples. The objective of this study was to develop an optimized and scalable workflow for isolating cfDNA from human samples for downstream use in a methylation array. We tested various methods of blood collection, plasma extraction, cfDNA isolation, and cfDNA quantification/purity analysis using whole blood samples from healthy volunteers and plasma samples from MS patients. We found that blood collection via Streck tubes, EDTA tubes, and heparinized syringes all performed equally well in terms of cfDNA yield and purity when blood was processed within an hour of collection. If there is a need to store extracted plasma for longer periods of time, we would recommend Streck tubes, which contain a preservative to prevent the release of genomic DNA (gDNA). However, if this storage time is not necessary, heparinized syringes and EDTA tubes are ubiquitous and easy-to-use alternatives. Following plasma extraction from whole blood, use of a second centrifugation step is recommended after thawing frozen plasma samples to dispose of cryogels prior to enzymatic digestion. When isolating cfDNA, two tested commercial kits performed equally well, however, the generic bio-on-magnetic-bead method yielded less consistent results with more high molecular weight gDNA contamination. Post-isolation, we used the SPRISelect (Beckman Coulter) method, which was successfully tailored to select for the target cfDNA size range and to concentrate samples. Quantification of sample concentrations was carried out using Qubit (Thermo Fisher) followed by analysis of sample purity using the 2100 Bioanalyzer Instrument (Agilent). Following our optimized protocol, we see consistent, clear cfDNA peaks at ~160-170 bp with occasional dimer/trimer peaks in Bioanalyzer electropherograms. This protocol will allow for more accurate and informative downstream analysis of cfDNA samples derived from human plasma.

Source of Funding: NIH Oxford-Cambridge Scholars Program