Researchers from the University of California, San Diego have developed a sequencing method using Pacific Biosciences' RS machine to identify breakpoints for structural variations, which they think may eventually have applications in cancer, including early diagnosis and monitoring patients' response to treatment.
The team published a proof of concept of the method using cancer cell lines in Genome Research earlier this month.
"Our goal is to have a method that basically tries to look at structural variations and accommodate for the fact that in every cancer patient you can have different breakpoints that cause the same structural variant," lead author Anand Patel, a graduate student in Vineet Bafna's UCSD lab, told In Sequence.
Now, the team is working to automate and increase the throughput of the method, as well as to test it in primary tumor samples. Patel said that eventually the group would like to commercialize it, but he did not have a timeline.
The basic idea behind the method — dubbed AmBre, for amplification of breakpoints — is that structural variants are pervasive across many different tumor types, but the specific breakpoints of even common structural variants differ from person to person and could be useful as patient-specific biomarkers, explained Patel.
The approach involves using PCR to amplify the breakpoint-containing DNA then sequencing on the PacBio RS, which Patel said is an ideal platform due to the long reads that enable it to span entire structural variants.
While the method targets known structural variants, the specific breakpoints are unknown, said Patel. Thus, the researchers had to develop bioinformatics capabilities for the primer design process in order to capture all possible breakpoints within the target region while minimizing off-target amplification. The resulting amplicons are then sequenced on the RS and the reads correspond with an amplicon to generate a consensus amplicon sequence.
In the Genome Research study, the team analyzed four cancer cell lines for known deletions in the CDKN2A region. The AmBre design suggested 16 primer designs that were 6 kilobases apart to cover the 100 kb region. The 16 primers were used across each of the cell lines and each cell line produced a unique amplicon. Those were then sequenced on a single SMRT cell of the RS.
Sequencing resulted in deep coverage of all the breakpoints. Cell line A549, an adenocarcinoma, had the lowest coverage with 400 fragments, while cell line CEM, an acute lymphoblastic leukemia, had the highest coverage with 18,000 fragments. However, the authors attributed the difference in coverage to amplicon length. The cell lines with shorter amplicons had higher coverage. On previous versions of the PacBio RS, shorter amplicons were easier to load onto the SMRT cell than longer amplicons, but that problem has since been alleviated with PacBio's MagBead station for amplicon loading.
In three out of the five cell lines, the results from the AmBre method were concordant with published results. In the other two cell lines — the breast cancer cell line MCF7 and the glioblastoma line T98G — the structural variant breakpoints had not been characterized despite previous efforts including whole-genome sequencing of the MCF7 cell line. However, the authors were able to confirm the structural variants and corresponding breakpoints in the two cell lines with Sanger sequencing. "The ease of the discovery in our experiment attests to the value of a targeted approach to [structural variant] detection," they wrote.
The method may be especially amenable for extremely heterogeneous cancer samples, Patel said, because of its targeted nature and also because it homes in on structural variants rather than single point mutations.
When looking for single point mutations from heterogeneous samples, researchers must look for "only a one-nucleotide change … compared to a high background of normal DNA," Patel said. Structural variants are more unique to tumor DNA, so with appropriate primer design, the AmBre method enriches for tumor DNA, making it easier to find the breakpoint sequence responsible for the structural variant.
Aside from the CDKN2A deletions, the team also demonstrated the method on a more complex structural variant, a gene fusion between RUNX1 and RUNX1T1, the result of an interchromosomal translocation between chromosomes 8 and 21.
The researchers used AmBre to design 10 reverse primers in the RUNX1 region and 18 forward primers in the RUNX1T1 region that were spaced about 3 kb apart from each other to capture breakpoints on chromosome 8. To capture chromosome 21 breakpoints, they designed 10 forward and 19 reverse primers in the RUNX1 and RUNX1T1 regions, respectively.
Next, they tested the primer design on the acute myeloid leukemia cell line Kasumi-1. The primers closest to the breakpoints produced a 3.8-kb amplicon and 2.7-kb amplicon from chromosomes 8 and 21, respectively. "Both reactions resulted in a strong signal and virtually no background noise, despite there being close to 30 primers in each reaction," the authors wrote.
Finally, the team tested the method's ability to assess breakpoints from heterogeneous mixtures, successfully amplifying a 2.2-kb CDKN2A deletion from the A549 cell line and a 3.6-kb CDKN2A from the MCF7 line in mixture ratios as low as one cancer cell in 1,000.
Moving forward, Patel said that the team plans to incorporate microfluidics in order to carry out hundreds of PCR reactions at the same time and to begin testing the method on primary tumors as well as on different types of structural variants.