Whole Exome Sequencing of Cell-Free DNA from Plasma

Qiang Fu, Ph.D., J.P. Jerome, Ph.D., Kamran Shazand, Ph.D., Jamie Wibbenmeyer, Ph.D.

Rubicon Genomics, Inc., 4743 Venture Drive, Ann Arbor, MI 48108,USA

Whole genome and exome sequencing (WGS and WES) of circulating cell-free DNA (cfDNA) from plasma has recently been used in landmark proof-of-concept studies as a technique to assay circulating tumor DNA (ctDNA) and circulating fetal DNA. Generally, only a few nanograms of fragmented (approximately 170 bp) cfDNA are recovered per milliliter of plasma. Because of this, a sensitive library preparation method that performs well with relatively short fragments, such as the ThruPLEX® library preparation technology,must be used. Although ThruPLEXhas been validated for cfDNAsequencing applicationsin the laboratory of current customers, we wanted to assess the quality of whole exome sequencingof ThruPLEX libraries made from low input of plasma cfDNA using multiple capture platforms.

Plasma-derived cfDNA from healthy individuals and sheared cell-line control DNA were used to generate ThruPLEX libraries from 0.5, 2, and 10 ng inputs. Whole exome capture was performed using the three commercially available SureSelect® platforms from Agilent: XT, QXT and XT2. ThruPLEX libraries made from plasma-isolated DNA showed adiscrete low molecular weight size distribution representative of circulating cfDNA after both library preparation and library capture based on BioAnalyzer trace profiles. Sequencing performed on Illumina MiSeq® and NextSeq™ 500 indicated that ThruPLEX libraries may be captured and sequenced from low input amounts of DNA, but that cfDNA-derived libraries are captured with slightly less efficiency than sheared control DNA-derived libraries. In addition, the metrics of capture using the highly streamlined SureSelect QXT platform were found to be greater than XT and XT2, but with an increased AT-dropout in sequence data, likely indicating some GC-based bias during the shortened hybridization. Overall, approximately 7 gigabase of sequence data (approximately 100M 75 base reads) was needed to obtain20-fold coverage of at least 80% of the exome from 10 ng of cfDNAstarting material, which is approximately 20% more sequence than was needed for the same input of sheared DNA (Figure).

In conclusion, ThruPLEX library preparation integrated with Agilent’s SureSelect capture reagentsallows researchers to access the exome of cell-free DNA. This will enable the development of potential non-invasive diagnostic assays and biomarker discovery.

Figure