Category: Automation and High-Throughput Technologies
Formalin-Fixed, Paraffin Embedded (FFPE) tissue derived DNA samples are notoriously difficult to sequence effectively. FFPE tissues are widely used for long term storage due to its ease of archiving, but this comes at a cost to DNA integrity. FFPE tissue preparation can damage DNA in a variety of ways. The highly variable quality and quantity of this sample type leads to uncertainty in NGS library preparation and sequencing. Labor intensive and inefficient protocols have made library preparation challenging to automate. FFPE DNA libraries often lack the molecular complexity needed to achieve high sequencing coverage. Typical sample preparation methodologies cannot rescue suitable amounts of FFPE input material for efficient library preparation and utilize over-amplification to drive final library yields.
AstraZeneca has created a high throughput FFPE DNA pipeline with a proven ability to create more complex sequencing libraries. Replacing the acoustic shearing step with a FFPE specific enzyme fragmentation process allows for the method to be fully automated without the need for special equipment and reduces hands-on time by 80%. We show that enzyme fragmentation is viable for damaged FFPE DNA and that DNA quality determines what size inserts are generated. Increasing the efficiency of the adapter ligation reaction may be the easiest and most effective way to preserve the original sample complexity. The efficacy of adapter ligation has been increased up to 50% by focusing on repairing DNA before ligation and stochastically driving the reaction with tuned adapter concentrations. This pipeline also accepts low-input, poor-quality samples that did not meet input requirements for previous workflows, un-locking many samples that were once deemed non-viable. Throughout all sample processing, variables are adjusted to the batch quality scores. This tailoring effectively translates pre-sequencing QC metrics to sequencing success. This pipeline, incorporating enzymatic fragmentation and optimized ligation reactions, represents a paradigm shift that will change the way FFPE samples are processed and quality controlled.
Methods: FFPE tissue samples are first extracted using a magnetic bead based method then quantified and qualified by a qPCR assay. Samples are then batched based on DNA quality and put through an optimized, automated workflow to prepare whole genome libraries. Samples are then evaluated for library preparation efficiency. Libraries then proceed through exome target enrichment utilizing a hybridization capture technology. Final capture libraries are then sequenced on an IlluminaÒ HiSeq 4000. Integration of all processes and technologies with a Laboratory Information Management System enable sample tracking throughout the entire pipeline.
Barrett Nuttall– Translational Genomics Scientist, AstraZeneca, Waltham, MA
Translational Genomics Scientist
Barrett Nuttall is a molecular biologist working to bring more NGS and automation to oncology. His focus is on simplifying NGS preparation pipelines with new technologies that are faster and more efficient. Forging new NGS Gold Standards with automation!