AGBT 2019: A Scalable and Automatable Method for the Extraction of cfDNA

Content Type: Poster
Authors: Lauren P. Saunders; Antonia Hur; Brittany Niccum; Asmita Patel

Introduction

Liquid biopsies represent a promising area of cancer testing as taking blood is less invasive than tumor biopsies. The cell free DNA (cfDNA) present in the blood includes DNA derived from cancer cells and cancer biomarkers can be detected in the extracted cfDNA.

Whole blood also contains genomic DNA, and can be removed by centrifugation, resulting in plasma. cfDNA is present in very small amounts in blood or plasma, and thus larger amounts of plasma are required for many applications. Larger extractions are more challenging to automate, as they require additional pipetting steps.

Here we present a novel cfDNA extraction kit and show its compatibility with extractions from 200 μl – 5 mL. We discuss the optimization of the method and demonstrate automation on a KingFisher Duo. This workflow can also be automated on a Biomek i7 Automated Workstation. We demonstrate that this kit can be used for NGS and produce results comparable to other commercial kits.

Scalable Chemistry with Inputs from 200 μL to 5 mL

Figure 1. DNA isolation from varying plasma amounts. Nuceleosomes were spiked into plasma to ensure enough DNA was present for QC. The Apostle MiniMax kit isolated nearly all the input DNA, while Kit 2 isolated less from larger plasma volumes.

Genomics Poster cfDNA Scalable Chemistry Figure 1Genomics Poster cfDNA Scalable Chemistry Figure 2
Figure 2. Isolation of DNA from 1 – 5 mL of EDTA plasma. Yields were highest with the Apostle MiniMax kit for 3-5 mL, and yields were similar across all kits at 1 mL.

Apostle MiniMax™ Workflow

Genomics Poster cfDNA Workflow Figure 3
Figure 3. Apostle MiniMax Workflow. The Apostle MiniMax kit involves a lysis step, then the addition of magnetic beads to bind the DNA. Once the DNA is bound, it is washed with various wash buffers and finally eluted from the beads.

Removal of PCR inhibitors and gDNA

Genomics Poster cfDNA PCR Inhibitors Figure 4          Figure 4. Comparison of PCR inhibition. The p41 primers and the KAPA hgDNA Quantification and QC kit was used to estimate [DNA]. Undiluted samples were compared to samples diluted 1:4 to measure the effect of PCR inhibitors.

If PCR inhibitors are present, the estimated concentration will be higher in more dilute samples. Similar concentrations estimated from the 1:1 and the 1:4 dilution of KAPA is a sign of low inhibition.

As such, Apostle MiniMax and Kit 1 have low inhibition and significant PCR inhibition is seen in Kit 2.

 

Genomics Poster cfDNA Increasing Inputs Figure 5
Figure 5. Increasing input amounts yield increasing amounts of cfDNA. Bioanalyzer traces show that the increase in DNA yield is due to increasing amounts of a small DNA peak. No contaminating genomic DNA was seen. The high peaks at the beginning and end of the trace are high and low markers.

Protocol Optimization

Genomics Poster cfDNA Protocol Optimization Figure 6

Figure 6. Optimization of Proteinase K. Changes in Proteinase K concentration have significant effects on final [DNA]. Increasing the amount of Proteinase K to 150 μL results in significantly more yield in 5 mLs of plasma collected in EDTA tubes.

Automation on KingFisher Duo

The extraction was automated on the KingFisher Duo instrument after the addition of the binding/ nanoparticle solution. The automated portion of the extraction was 37 min.

Genomics Poster cfDNA KingFisher DuoGenomics Poster cfDNA Automation KingFisher Duo Figure 7

Figure 7. cfDNA was extracted from 1 mL of plasma in concentrations consistent with manual extractions. Little variability was seen between the replicates (standard deviation of 1.0).

Automation on the Biomek i7 Automated Workstation

The Apostle MiniMax™ kit is compatible with automation on Beckman Coulter’s Biomek i7 Automated Workstation** instrument with integrated KingFisher Presto. This configuration is expected to have a run time of 2.5 hr. The Apostle MiniMax™ kit can also be automated on the Biomek i7 without the KingFisher integration, and the run time is expected to be approx. 4.5 hr.

Biomek i7 Automated Liquid Handling Workstation

 

Sample Name Percent Aligned Reads Read Enrichment Uniformity of Coverage (Pct > 0.2* Mean) Target Coverage at 1x Target Coverage at 20x
EDTA Donor 1 Apostle MiniMax™ 99.80% 70.10% 98.40% 100.00% 99.90%
EDTA Donor 2 Apostle MiniMax™ 99.80% 70.80% 97.90% 99.90% 99.80%
EDTA Donor 1 Kit 1 99.70% 67.90% 97.80% 100.00% 99.90%
EDTA Donor 2 Kit 1 99.80% 65.10% 95.70% 99.90% 99.70%
cfDNA Tube 1 Donor 1 Apostle MiniMax™ 99.80% 72.10% 98.60% 100.00% 99.90%
cfDNA Tube 1 Donor 2 Apostle MiniMax™
99.80% 71.90% 98.50% 100.00% 99.90%
cfDNA Tube 1 Donor 1 Kit 1 99.80% 71.10% 97.50% 100.00% 99.80%
cfDNA Tube 1 Donor 2 Kit 1
99.80% 70.20% 95.90% 100.00% 99.90%
cfDNA Tube 2 Donor 1 Apostle MiniMax™ 99.80% 68.10% 97.90% 100.00% 99.90%
cfDNA Tube 2 Donor 2 Apostle MiniMax™
99.80% 69.60% 98.40% 100.00% 99.90%
cfDNA Tube 2 Donor 1 Kit 1 99.80% 68.40% 96.70% 100.00% 99.90%
cfDNA Tube 2 Donor 2 Kit 1 99.70% 67.20% 96.30% 100.00% 99.90%
Table 1. Quality Control of NGS run. Libraries were prepared from 25 ng DNA with the Accel-NGS 2S Hyb DNA Library prep kit for NGS and a target capture library was prepared from that library using the IDT xGen Pan-cancer panel. Libraries were pooled and run on an Illumina NextSeq. Data was analyzed via BWA enrichment. Human genome UCSC hg19 was used as the reference genome. Quality control metrics from all runs are good and comparable between extraction methods.

Detection of Cancer Mutations

Genomics Poster cfDNA Mutations Detected Figure 8

Figure 8. Mutations detected with different extraction methods. NGS of donor 2 were compared (run A EDTA, run B Paxgene and Streck). Mutations detected by both kits are shown with the mutations only detected or detected differently by the different kits.

Sample Name Indels Indel Het/Hom Ratio SNVs SNV Het/Hom Ratio SNV Ts/Tv Ratio
EDTA Donor 1 Apostle MiniMax™

108

4.4 374 1.7 2.3
EDTA Donor 2 Apostle MiniMax™ 127 4.5 453 1.9 2.8
EDTA Donor 1 Kit 1 110 4.8 373 1.7 2.3
EDTA Donor 2 Kit 1 126 4.3 452 1.9 2.7
cfDNA Tube 1 Donor 1 Apostle MiniMax™ 108 3.3 398 1.7 2.6
cfDNA Tube 1 Donor 2 Apostle MiniMax™
99 3.1 403 1.7 2.2
cfDNA Tube 1 Donor 1 Kit 1 108 3.3 401 1.7 2.5
cfDNA Tube 1 Donor 2 Kit 1
99 3.3 402 1.7 2.3
cfDNA Tube 2 Donor 1 Apostle MiniMax™ 115 4.8 410 2.1 2.2
cfDNA Tube 2 Donor 2 Apostle MiniMax™
112 4.3 434 2.5 2.6
cfDNA Tube 2 Donor 1 Kit 1 116 4.8 414 2.1 2.2
cfDNA Tube 2 Donor 2 Kit 1 113 4.4 427 2.5 2.6

Table 2. Mutation Detection with Different Extraction Methods. The detection of indels and SNVs was similar with both extraction methods.

Sample Name Indels Indel Het/Hom Ratio SNVs SNV Het/Hom Ratio SNV Ts/Tv Ratio
cfDNA Tube 1 Donor 2 Apostle MiniMax™ Run A 99 3.1 403 1.7 2.2
cfDNA Tube 1 Donor 2 Apostle MiniMax™Run B
101 3.4 399 1.7 2.3
cfDNA Tube 1 Donor 2 Kit 1 Run A 99 3.3 402 1.7 2.3
cfDNA Tube 1 Donor 2 Kit 1 Run B
104 3.5 402 1.7 2.3
cfDNA Tube 2 Donor 2 Apostle MiniMax™Run A
112 4.3 434 2.5 2.6
cfDNA Tube 2 Donor 2 Apostle MiniMax™Run B
112 4.1 432 2.5 2.7
cfDNA Tube 2 Donor 2 Kit 1 Run A 113 4.4 427 2.5 2.6
cfDNA Tube 2 Donor 2 Kit 1 Run B 114 4.4 431 2.5 2.7

Table 3. Run Variation. The duplicate libraries were sequenced to determine the amount of intra-run variation. As you can see, the variation between runs A and B of the same library have equal or greater variation than the runs observed for the different extraction methods, implying that the two methods sequence equaling well with NGS.

Conclusions

  • DNA can be extracted from 200 μL to 5 mL of plasma
  • The Apostle MiniMax kit removes the PCR inhibitors present in plasma
  • Genomic contamination is not present in the extracted cfDNA
  • Extraction of 1 mL plasma can be automated on a KingFisher instrument with yields similar to manual extraction
  • Similar numbers of mutations were found in cancer plasma with the Apostle MiniMax™ kit and another commercial kit.

The Apostle MiniMax™ kit is a versatile new cfDNA kit that can extract from a wide range of sample amounts and be run either manually or on a variety of automation systems.

 

** Not intended or validated for use in the diagnosis of disease or other conditions.
Beckman Coulter makes no warranties of any kind whatsoever express or implied, with respect to these protocols, including but not limited to warranties of fitness for a particular purpose or merchantability or that the protocol is non-infringing. All warranties are expressly disclaimed. Your use of the method is solely at your own risk, without recourse to Beckman Coulter. Not intended or validated for use in the diagnosis of disease or other conditions. These protocols are for demonstration only, and are not validated by Beckman Coulter. © 2019 Beckman Coulter, Inc. All rights reserved. Beckman Coulter, the stylized logo, and the Beckman Coulter product and service marks mentioned herein are trademarks or registered trademarks of Beckman Coulter, Inc. in the United States and other countries. All other trademarks are the property of their respective owners.

Talk To An Expert