Date Published: April 11, 2019
Publisher: Public Library of Science
Author(s): Samantha J. McDonough, Aditya Bhagwate, Zhifu Sun, Chen Wang, Michael Zschunke, Joshua A. Gorman, Karla J. Kopp, Julie M. Cunningham, Ruslan Kalendar.
Archival tissues represent a rich resource for clinical genomic studies, particularly when coupled with comprehensive medical records. Use of these in next generation sequencing (NGS) is a priority. Nine formalin-fixed paraffin-embedded (FFPE) DNA extraction methods were evaluated using twelve FFPE samples of varying tissue types. Quality assessment included total yield, percent dsDNA, fragment analysis and multiplex PCR. After assessment, three tissue types from four FFPE DNA methods were selected for NGS downstream evaluation, targeted and whole exome sequencing. In addition, two low input library protocols were evaluated for WES. Analysis revealed average coverage across the target regions for WES was ~20-30X for all four FFPE DNA extraction methods. For the targeted panels, the highest molecular tag coverage was obtained with the Kingfisher FFPE extraction method. The genotype concordance was 99% for the commonly called variant positions between all four extraction methods with the targeted PCR NGS panel and 96% with WES. Assessing quality of extracted DNA aids in selecting the optimal NGS approach, and the choice of both DNA extraction and library preparation approaches can impact the performance of archival tissue in NGS.
Next generation sequencing is rapidly becoming established in the clinic, predominantly in oncology but also as a means of diagnosis in individuals with unresolved medical issues. Archival tissue represents a singularly valuable resource for disease oriented research, particularly when combined with comprehensive medical records such as that of the Mayo Clinic. However, DNA extracted from such samples can vary widely in quality due to age, fixation conditions, DNA-protein crosslinking, and inhibitors, which may impact downstream genomic analyses. Samples are typically obtained in the operating room so how they are handled, as well as time exposed to formalin both contribute to potential DNA damage but are typically outside the control of investigators.
DNA extraction methods for FFPE tissues vary in quality and quantity of resultant DNA, all of which may impact performance in downstream assays. In this report, we evaluated nine methods, including both manual and automated protocols, as the latter are preferred to minimize potential errors in sample handling. QC evaluations were performed to determine the degree of DNA damage and which methods application might procure the best quality data; DNA from four extraction methods (manual and automated) were assayed for low input WES and amplicon based targeted sequencing based on yield, percent dsDNA, DQN and fragment length. Three of these methods, QIAGEN GeneRead (GR-A or GR-M), Promega Maxwell (PM-A) and QIAGEN QIAamp (QA-M or QA-A), were evaluated in Bonnet et al; QA and GR were less fragmented than PM-A, having longer median fragment lengths, in agreement with findings in this report.