IDT_Logo

Towards providing personalized medicine—considerations for reliable NGS data

Research profile: Read how scientists at Geneseeq Technology, Inc. improved their target capture methods to increase accuracy in clinical diagnostics by using optimized blocking oligos and stringent hybridization conditions.

Oct 18, 2017

Revised/updated Mar 29, 2017

Geneseeq is a cancer diagnostics company that provides treatment solutions based on comprehensive, systematic analysis of patients’ genomic mutations. The company mission is to bring precision care to cancer patients through a combination of next generation sequencing (NGS) and an up-to-date cancer knowledge base. As of March 2014, the company had guided treatment decisions for approximately 300 patients.

Background

Geneseeq was founded in 2008 by a group of postdoctoral researchers from University of Toronto, Canada, and Stanford University, CA, USA. Geneseeq headquarters and R&D team are based in the MaRS Discovery District in Toronto. The bioinformatics department is located in Stanford University. In 2013, the company’s Asia office was established in Nanjing, China, and is now Geneseeq’s largest facility, with 10,000 ft2 lab space and 15 scientists.

Choosing target capture technology

The founding members of Geneseeq were IDT customers during their graduate research studies, and they have continued using IDT products since forming the company. The products they have used range from primers for PCR and long DNA fragments to, more recently, NGS xGen® Lockdown® Probes and xGen Blocking Oligos for target capture. Geneseeq scientists use these target capture probes to sequence specific regions of the cancer genome. The resulting data helps clinicians make informed decisions about which therapies are suitable for individual patients, what is known as “personalized medicine”.

The clinical diagnostics field recognizes that there are only several hundred genes in the human genome that can be targeted for cancer treatment. The current trend is to selectively sequence these actionable genes. Methods for isolating these genes from the entire genome have, until recently, been challenging. The scientists at Geneseeq have used both solid phase–based array methods (e.g., Roche NimbleGen SeqCap®), and solution-based probe methods (e.g., Agilent SureSelect® and Illumina TruSeq®). They settled on xGen Lockdown Probes from IDT because these probes offer unique advantages that other kits do not:

  • Scalability—ability to adjust the number of reactions needed;
  • Flexibility—ability to alter the number of enriched genes per reaction; and
  • Capacity for optimization—introduce single additional probes as necessary

Target capture method

Geneseeq was an early adopter of xGen Lockdown Probes and undertook protocol optimization over the first year and a half of use. They tried different hybridization buffers and temperatures, and compared different blocking oligonucleotide sequences. The scientists also tested a variety of sample types: formalin-fixed, paraffin-embedded (FFPE), fresh and frozen tissue, and blood.

Overcoming FFPE samples

Most cancer specimens are formalin-fixed and paraffin-embedded (FFPE), a process that causes varying degrees of DNA damage, depending on the pathology processing protocol and the age of the sample. Typically, FFPE samples are notorious for yielding low quantity and quality nucleic acid, resulting in less effective library construction and low PCR efficiency during library amplification.

Important considerations for working with FFPE tissue are:

Pre-enrichment

  • Method of DNA extraction
  • Method of DNA fragmentation
  • Method of library construction

Enrichment

  • Sample ratios for multiplexing (e.g., ratios for pooling different libraries of variable DNA quality, ratios for libraries when combining FFPE samples with fresh or frozen tissue or with blood samples.)

This is critical because the efficiency of PCR is different when using DNA isolated from FFPE samples than when the DNA comes from blood/tissue samples. To attain the desired coverage for every sample within a multiplex pool, it is important that library input from different sources be adjusted to overcome any difference in PCR efficiency during post-capture amplification.

Post-capture amplification

  • Maximum recovery of the captured library and minimization of PCR cycles for post-capture amplification are required to reduce PCR duplication and bias in downstream NGS.

Geneseeq scientists have developed sophisticated guidelines for working with FFPE samples to ensure uniformity of coverage and consistent quality across samples. In their experience, xylene works best for dissolving paraffin in FFPE tissue sections. They also heat samples at 90°C for 1 hr after proteinase K digestion. This step helps to reverse formaldehyde modification of nucleic acids. Commercial DNA repair kits also help to improve library quality. At Geneseeq, the researchers use PreCR® Repair Mix (NEB) to repair a broad range of DNA damage, including mutations and PCR-blocking modifications.

Choice of hybridization and wash buffer

In initial experiments, Geneseeq scientists used an IDT xGen Target Capture Protocol that included a 48-hour hybridization step. However, they discovered that SSC in the post-capture wash buffer inhibited subsequent on-bead PCR amplification. The researchers then tried an updated IDT protocol which incorporated a 4-hour hybridization at 47°C using the SeqCap® EZ Hybridization and Wash Kit (NimbleGen). They optimized target capture by comparing data obtained from hybridization at different temperatures and found that hybridization at 65°C resulted in greater enrichment for genes of interest (Figure 1).

DO-YR-GeneSeeq Fig 1

Figure 1. Hybridization at higher temperature increases enrichment efficiency. Target capture was performed using xGen Lockdown Probes and SeqCap EZ Hybridization and Wash kit (NimbleGen). Hybridization was performed for 4 hr at 47°C or 65°C, as indicated. Relative levels of 3 target genes after enrichment were measured by real-time qPCR using primers targeting the enriched exon fragments. Relative fold enrichment was calculated by normalizing the relative level of targeted gene fragments to the prehybridization library DNA concentration.

To minimize loss of captured DNA during elution and purification steps, and to shorten the duration of steps before actual sequencing, they also performed on-bead, post-capture PCR amplification. On-bead PCR was not inhibited by the presence of residual wash buffer when the SeqCap wash kit was used. Additionally, similar or greater enrichment of target genes was obtained for on-bead PCR than for PCR performed after elution (Figure 2).

DO-YR-GeneSeeq Fig 2

Figure 2. On-bead post-capture PCR amplification enhances enrichment. Target capture was performed using xGen Lockdown Probes and SeqCap EZ Hybridization and Wash kit (NimbleGen). Post-capture PCR amplification was performed either on-bead or on captured DNA eluted from beads by boiling at 98°C for 10 min. Relative levels of 3 target genes after enrichment were measured by qPCR, using primers targeting the enriched exon fragments. Relative fold enrichment was calculated by normalizing the relative level of targeted gene fragments to the prehybridization library DNA concentration.

 

Choice of blocking oligos

Geneseeq scientists first used IDT xGen Standard Blocking Oligos for Illumina TruSeq and Other Single-Index Adapters. When IDT launched xGen Universal Blocking Oligos, in which a single blocking oligo can effectively block multiple barcoded adapters, Geneseeq performed a thorough comparison of the two IDT blocking oligo products. The researchers found that, for their purposes, the Universal Blocking Oligos were more efficient at reducing background (off-target rate), leading to enhanced enrichment (on-target rate) (Figures 3 and 4).

DO-YR-GeneSeeq Fig 3

Figure 3. xGen Universal Blocking Oligos increase enrichment efficiency. Target capture was performed using xGen Lockdown Probes and different IDT hybridization protocols with xGen Standard or Universal Blocking Oligos, as indicated. Relative levels of 3 target genes were measured by real-time qPCR using primers targeting enriched exon fragments and normalizing to the relative level of total enriched library measured by qPCR using p5 and p7 primers (Illumina). Fold enrichment was calculated by normalizing the relative level of targeted gene fragments to their own relative levels in prehybridization samples (Prehyb). The data show that using Universal Blocking Oligos significantly enhanced enrichment compared to Standard Blocking Oligos. Additionally, combining Universal Blocking Oligos with the xGen Rapid Capture Protocol (NimbleGen buffers; 4 hr hybridization at 65°C) provides up to 5X greater enrichment than the original 48 hr hybridization protocol (IDT buffers) used with Standard Blocking Oligos.

 

Negative controls

For negative controls, Geneseeq scientists use genes, ACTB (β-actin) and GAPDH, not included in their target enrichment panels. The researchers were very impressed with the performance of the xGen Lockdown Probes. They found that immediately after library preparation, they were able to detect the negative controls by qPCR, but after enrichment with xGen Lockdown Probes, when combined with Universal Blocking Oligos and with use of the xGen Rapid Capture Protocol, there was no trace of these amplicons (Figure 4). This demonstrated how effectively the negative controls were depleted by enrichment and indicated that the off-target rate is low.

DO-YR-GeneSeeq Fig 4

Figure 4. xGen Universal Blocking Oligos reduce off-target rates. Target capture was performed using xGen Lockdown Probes with different hybridization protocols and xGen Standard or Universal Blocking Oligos, as indicated. The presence of negative control genes, (A) ACTB and (B) GAPDH, was measured by qPCR using primers targeting the negative control gene fragments and normalizing to the relative level of total enriched library measured by qPCR using p5 and p7 primers (Illumina). Fold enrichment was calculated by normalizing the relative level of control genes after PCR to their relative levels in prehybridization samples (Prehyb). The data show that combining Universal Blocking Oligos with the xGen Rapid Capture Protocol (NimbleGen buffers; 4 hr hybridization at 65°C) provides markedly better specificity compared to the original 48 hr hybridization protocol (IDT buffers) or Standard Blocking Oligos, indicated by the absence of negative control genes.

Performance of xGen Lockdown Probes

IDT xGen Lockdown Probes are individually assessed by mass spectrometry for QC. Geneseeq researchers found that use of these probes, together with the optimized target capture protocol, allowed them to achieve >80% on-target rate with very high uniformity of coverage for different genes. The researchers compared xGen Lockdown Probes to other commercially available target capture reagents and discovered that panels comprising Lockdown Probes combined in a 1:1 molar ratio delivered, on average, ~40% better enrichment efficiency compared to the other reagents.

Sequencing platforms

Initially, the scientists used the SOLiD sequencing system (Applied Biosystems). However, since 2010, they have been using Illumina sequencing platforms, starting with the HiSeq® 2000 instrument, and now the HiSeq® 2500 instrument for sequencing larger genomes, such as the human exome. For clinical assays, the scientists use the MiSeq® system. They have chosen to use the MiSeq system, rather than HiSeq, because the former provides sufficient throughput for their gene panel and produces higher quality sequencing data with longer read lengths in a short period of time. Furthermore, the MiSeq System is the first and only NGS system for in vitro diagnostics approved by the US Food and Drug Administration (FDA). More recently, Geneseeq has been investigating the new NextSeq® 500 system (Illumina) that allows higher throughput and the flexibility to fulfill their growing clinical diagnostics needs.

From R&D to clinical diagnostics

Geneseeq scientists use xGen Lockdown Probes for both research and product development in their labs in Toronto and Stanford. The optimized probe sets and protocols are then used by their clinical diagnostics service lab in Nanjin, to serve the Asian market. They have tested >300 patients in the last year and are following the progress of some of these patients for whom the tests suggested a particular targeted therapy. By accumulating diagnosis and treatment information from large patient numbers, the Geneseeq scientists are generating sufficient data to provide statistical significance to their tests in clinical practice.

Research profile

Xue Wu is a molecular biologist with specialization in signal transduction and cancer biology. She obtained her bachelor’s degree in biological sciences in China and a master’s degree from Dalhousie University, Halifax, Canada, where she investigated molecular mechanisms of mammary gland development and tumorigenesis. Subsequently, she worked at Atlantic Research Centre, Halifax, as a research assistant, studying colorectal cancer. She obtained her PhD from the University of Toronto, studying the molecular pathogenesis of human genetic syndromes caused by germline mutations. Her postdoctoral training was at the Ontario Cancer Institute, Canada, where she studied functional genomics and epigenomics of different cancers, including medulloblastoma, triple negative breast cancer, and lung cancer drug resistance. She has publications in several high impact journals.
Yang Shao is a cancer biologist and bioinformatician. He received his bachelor’s degree in immunology and biochemistry from the University of Toronto, Canada, and has been working on cancer driver gene discovery and diagnostics since 2006. Yang Shao began studying cancer genomics early in his research career, using whole genome landscape profiling methods, initially SNP and gene expression arrays, and since 2008, next generation sequencing methods. Together with other colleagues, he founded Geneseeq in 2008. He has publications in several peer reviewed journals and is a contributing author to chapters in several cancer biology textbooks. His main interest is in accelerating the use of NGS in the clinic to increase societal benefit to patients.
lockup_1
lockup_2