In the field of liquid biopsy, one of the challenges is identifying the cellular source of biological material. For example, current methods for analyzing circulating tumor DNA (ctDNA) rely upon genetic differences to distinguish ctDNA from cell-free DNA (cfDNA) released by normal cells. Because ctDNA represents only a small fraction of total circulating DNA, tumor-specific aberrations must be identified to confirm that DNA fragments are tumor-derived. Prenatal screening also relies upon genetic differences between maternal and fetal tissues represented in cfDNA. However, conditions that increase the total amount of cfDNA in the absence of genetic changes, such as heart attack, stroke, chronic kidney disease and autoimmune disorders, cannot be monitored with cfDNA. A recent paper by Snyder and colleagues explores the possibility of using genome-wide nucleosome mapping to identify the cellular source of cfDNA (1).
cfDNA consists of primarily short (<200 base pairs) double-stranded DNA fragments, most likely derived from apoptotic hematopoietic cells. The authors isolated cfDNA from healthy controls using the QIAamp® Circulating Nucleic Acid Kit from QIAGEN and evaluated the size distribution of cfDNA fragments. Their results revealed a peak at 167 base pairs, corresponding to the length of DNA wrapped around a single nucleosome plus a linker histone. They then examined nucleosome positioning throughout the genome by sequencing cfDNA and evaluating the distribution of aligned fragment endpoints. If cfDNA is partially protected from nuclease cleavage by packaging within nucleosomes, then fragment endpoints should cluster around the boundary between nucleosome core particles. In addition, fewer fragment endpoints would be expected within DNA wrapped around the nucleosome core particle. The authors found that their genome-wide map of nucleosome protection in cfDNA correlated with nucleosome positioning maps from other groups. They also found that nucleosomes were strategically positioned relative to transcriptional start sites, intron-exon boundaries and splice sites.
The authors then asked whether nucleosome spacing could be used to determine the source of cfDNA from healthy individuals and in patients with stage IV cancer. Patient samples were compared to a database containing nucleosome mapping data from human cell lines and primary tissues. In healthy controls, nucleosome positioning correlated with cells of lymphoid or myeloid lineage, supporting the hypothesis that most cfDNA in healthy individuals is derived from apoptotic hematopoietic cells. In five patients with stage IV cancer, nucleosome mapping patterns correlated with cancer cell lines, in some cases matching the tissue of origin.
With an expanded reference database, nucleosome footprinting may permit the localization of tumors of unknown origin, as well as monitoring conditions characterized by abnormal cellular proliferation or damage. This technique would expand the realm of liquid biopsy beyond oncology and prenatal testing to include a wider array of pathological states.
- 1. Snyder, M.W. et al. (2016) Cell-free DNA comprises an in vivo nucleosome footprint that informs its tissues-of-origin. Cell 164, 57.