NEW YORK (GenomeWeb) – A number of groups have recently reported the development of targeted next-generation sequencing panels using Pacific Biosciences' single-molecule sequencing technology, due to the ability of the technology to sequence through repetitive regions, areas with high sequence homology, and other complex regions that are difficult to analyze with short-read sequencing technology.
Reporting recently in BMC Genomics and BMC Cancer, researchers from George Washington University and Uppsala University described panels created for ovarian hyperstimulation syndrome and for detecting mutations in BCR-ABL1 gene fusion transcripts in patients with chronic myeloid leukemia, respectively.
In addition, PacBio recently developed its own targeted sequencing workflow that uses the Roche NimbleGen SeqCap enrichment technology with DNA fragments up to 6 kb.
Funda Suer, lead author of the GWU study, told GenomeWeb that her group chose PacBio sequencing technology because it was designing a panel that "targeted a lot of hormone receptors, which are known to have high amounts of sequence homology and GC-rich regions" that are difficult to analyze with short-read sequencing technologies.
Ovarian hyperstimulation syndrome (OHSS) is a condition that some women experience as a result of fertility medicines known as controlled ovarian hyperstimulation (COH). In some cases, COH can overstimulate the ovaries and cause them to swell and leak fluid.
The condition can be "very drastic," Suer said, and sometimes fatal.
Genetics is a known contributing factor that determines which women will develop OHSS and which will not in response to COH, however there has not been a comprehensive analysis of all the potential genes suspected to be involved, and currently there is only one genetic test available that analyzes one gene.
Suer said that the goal of this study was to design a panel that would query all candidate genes, including the 5' and 3' flanking sequences to search for biomarkers that could be used to develop a diagnostic.
The team used RainDance Technologies target enrichment to create 1,951 amplicons averaging 1 kb in length and covering 44 genes totaling 3.18 mb of sequence. They next tested it on 20 DNA samples from patients with clinically defined OHSS.
The researchers included genes implicated in COH response, those associated with OHSS, and genes involved in regulating gonadotropin action or ovarian angiogenesis.
Sequencing was done in just 45 minutes. Looking initially at just the raw reads, they achieved average lengths of 1,178 bases and covered 100 percent of 1,816 out of 1,951 amplicons. Filtering for circular consensus reads increased mean read length to 3,200 bases, and the mean mapped CCS read accuracy was 97 percent.
Importantly, the authors reported, they were able to sequence through both GC-rich and non-GC rich regions. The VEGFA gene, which has about 72 percent GC content, was covered by 42 amplicons and had "uniform coverage."
Next, they sought to validate base calling. Looking at 19 variants, they were able to validate 18 with Sanger sequencing, including a novel missense variant in the hormone receptor gene LHCGR that was found in two severe cases of OHSS.
"These data show excellent promise for follow-up studies with a larger number of OHSS cases," the authors reported.
Suer said the next step is to expand the work to include larger numbers of patients and to compare those that develop OHSS to those that do not.
In the BMC Cancer study, researchers at Uppsala University in Sweden made use of PacBio's Iso-Seq technology to analyze BCR-ABL1 fusions in CML patients receiving tyrosine kinase inhibitor therapy, which targets those fusions. Specifically, the team was looking for mutations within that fusion transcript, which can lead to resistance to the therapy.
The team chose to use PacBio technology for its assay because of its long read lengths, which could sequence through the entire BCR-ABL1 transcript. Sequencing with shorter read lengths requires that smaller fragments of the fusion transcript be sequenced in multiple amplification rounds, which "limits the analysis to a portion of the transcript [and] is likely to introduce a bias in the resulting mutation frequencies," the authors wrote.
The group used long-range PCR amplification to amplify the BCR-ABL1 transcript and created SMRTbell libraries using the 2 kb template preparation protocol. The researchers used an older version of the PacBio chemistry, C2/P4, for sequencing.
They tested the protocol on 22 samples from six patients at various time points throughout their treatment course. All patients had been diagnosed with CML at Uppsala University Hospital, had all received imatinib as first line therapy, and had showed little or no response to tyrosine kinase inhibitor treatment.
Sequencing on a single SMRT cell generated a uniform mapped read coverage of about 10,000x across the entire length of the BCR-ABL1 amplicon, and an average of 32,000 CCS reads per sample, the authors reported.
The researchers identified 13 mutations, all of which had previously been implicated in resistance to one or more tyrosine kinase inhibitors. The assay confirmed all mutations that had previously been detected with Sanger sequencing and identified an additional five mutations that were below 5 percent frequency — below the limit of detection for Sanger. They also detected splice isoforms from full-length CCS reads that spanned the entire transcript.
To evaluate sensitivity and specificity the researchers did serial dilutions from one patient with two ABL1 mutations and found that the mutations could be detected down to 1 percent and 0.5 percent frequency with no false positives, which they said could be explained by the "random distribution of sequencing errors inherent to the PacBio technology, which results in highly accurate base calls from molecules sequenced at high coverage."
Overall, the researchers said that the assay "allows for a simple, efficient, and streamlined workflow conducive to clinical routine." The workflow has a two-day turnaround time, including one day for library preparation and two to three hours of sequencing time per sample.
The authors noted that although their sample size is small, one potential benefit of the assay may be its ability to detect resistance mutations earlier than the conventional Sanger test. For instance, they were able to identify a mutation in one patient four months earlier than with the Sanger assay, indicating that "an NGS screen could be informative when performed at earlier time points, possibly in patients with no or limited responses to TKI therapy."
Although further validation work must be done, the authors wrote that the assay could "potentially be introduced into clinical practice to guide therapeutic decisions for TKI resistant patients."