October 1, 2014 (Vol. 34, No. 17)
Richard A. A. Stein M.D., Ph.D.
Various approaches have been developed to identify and quantitate copy-number variation. A new, high-resolution approach that already shows promise is digital PCR.
As illustrated by many research efforts, digital PCR presents multiple advantages over other strategies; at the same time, digital PCR faces its own very specific challenges.
With copy-number variants becoming increasingly important in research, biotechnology, and medicine, significant attention has focused on using digital PCR to accurately characterize these structural variants in a variety of organisms. Integrating digital PCR with other techniques promises to provide a particularly powerful approach for characterizing copy-number variation, with benefits that are extending into multiple areas of science and clinical medicine, from research to diagnostic and therapeutic applications.
“Our work focuses on developing a new generation of HDx™ reference standards, based on precisely defined cell lines,” says Jonathan Frampton, Ph.D., global product manager at Horizon Diagnostics (a business unit of Horizon Discovery Group). Cell lines, which are almost ubiquitously used in biomedical laboratories, are often characterized by significant genomic heterogeneities, making analyses and comparisons challenging.
Independently validated reference standards provide information about the limits of specific assays and allow their reproducibility to be monitored and maintained. This represents a critical aspect of quality control, which becomes even more important considering the powerful drive that exists for standardization across laboratories. “Talking to a number of diagnostic assay developers, we learned that they often use cell lines from commercial sources, and these cell lines are not always well characterized,” confides Dr. Frampton.
Investigators at Horizon Diagnostics are using their gene-editing platform, GENESIS™, to develop cell-line reference standards that carry rare genetic changes. “The strength of our approach is that we use digital PCR as the core technique to test our reference materials, and this is followed by using alternative technologies such as next-generation sequencing or qPCR to confirm that the digital PCR was correct and to provide great confidence to investigators who take these reagents to routinely validate their respective platforms,” explains Dr. Frampton.
The gene-editing capabilities of GENESIS include SNPs and point mutations, deletions, translocations, amplifications, and barcoding. The end point of this process is a clonal heterozygous mutant cell line that is identical with the initial wild-type clone, except for the mutation that is introduced at the endogenous gene loci.
By using this approach, investigators at Horizon Diagnostics generated over 550 cell lines harboring mutations engineered in various genes, including B-Raf, EGFR, N-Ras, and K-Ras. “We are looking to provide these extensively characterized reference standards to support work that is expanding the applications of digital PCR into copy-number variation assays,” notes Dr. Frampton.
One of the challenges in next-generation sequencing, particularly when discriminating higher copy numbers, is the need to generate standard curves in the dynamic range that is required by the particular experiment or set of genes. “Digital PCR provides absolute quantitation, and it allows investigators to distinguish copy numbers in various ranges,” states Dr. Frampton.
Clinically, particularly for certain genes, the number of chromosomal copies is an essential determinant of disease severity, and accurately quantitating copy numbers is therefore critical for diagnostic and therapeutic purposes. “We have performed digital PCR down to 20 copies of a mutant allele in a wild-type background of 50,000 copies, and we can definitely go lower, with additional assay optimization,” asserts Dr. Frampton.
Copy-number variation can also be visualized by fluorescence in situ hybridization (FISH) and other approaches. FISH is used routinely in cytogenetics settings, but its inability to offer quantitative information is one of its major shortcomings. “FISH can only reveal that a chromosome is modified, but it is not informative about the extent of the modification, and this makes digital PCR or even next-generation sequencing more attractive,” remarks Dr. Frampton.
Despite the very high resolution that digital PCR provides at the genomic level, some genetic variants are not expressed, and detecting protein expression by approaches such as immunohistochemistry might be clinically more relevant. “This is why the field, in the next few years, may study and analyze these approaches alongside each other,” concludes Dr. Frampton.
Copy Number of Oncogenes
An important technical aspect surrounding PCR is that, in addition to the number of molecules that can be detected, an even more important parameter is the number of molecules of a certain type that one can visualize in the background of all the other molecules. “This is where digital PCR has no match, because it can detect one mutant molecule in a pool of thousands or millions of other similar molecules if the sample is diluted,” says Jo Vandesompele, Ph.D., CSO at Biogazelle and professor of functional genomics and applied bioinformatics at Ghent University.
Efforts at Biogazelle and in Dr. Vandesompele’s laboratory are focusing on accurately detecting increased copy number of oncogenes in the DNA obtained from surgically removed malignant tumor samples or from circulating nucleic acids in the serum or plasma of cancer patients. The majority of the DNA that is circulating in the plasma originates from normal cells, but in cancer patients, a certain amount of circulating DNA also originates from the tumor. “It is in that small fraction where one can detect an increased copy number of amplified oncogenes,” notes Dr. Vandesompele.
Because differences between normal and pathologically elevated copy numbers can be very small, especially in circulation, it is essential to ensure that experiments provide a sufficient number of reactions and replicates. “One of the disadvantages that we encountered when using digital PCR is the problem of integrating multiple replicates from independent experiments,” concedes Dr. Vandesompele. “A way to combine these replicates and provide confidence about the results has not been worked out yet.”
As part of a collaborative endeavor with researchers in mathematics and statistics at Ghent University, Dr. Vandesompele and colleagues developed a mathematical framework that helps predict the reaction, replicate, and other numbers that must be met to ensure that copy-number differences can be determined with confidence.
“This platform helps integrate multiple replicates and applies a statistical test to ensure that the results reflect stringent confidence and significance measures,” asserts Dr. Vandesompele. By measuring tumor loads at diagnosis, and then following them longitudinally, this strategy provides the opportunity to predict patients who stop responding to therapy or present relapses.
Over the past few years, several candidate viral vaccines were designed by the deletion of essential genes from the full genomes of viruses. This approach typically results in replication-defective viruses, which propagate only if complement cell lines can be established that rescue the replication defects present in the viral strains. Such cell lines are engineered to carry genes that code for proteins needed for viral replication.
“Genetic stability of the introduced transgenes is an important concern,” says Ali Azizi, Ph.D., a scientist in the analytical research department at Sanofi Pasteur. “Instability may lead to inconsistent expression of complementing gene products.”
Conventional methods of assessing transgene stability already pose challenges. Additional problems, however, may arise with certain transgenes, particularly if they are to be used with cell lines for which full genomic sequences have not yet been published.
Dr. Azizi and his colleagues examined the accuracy and the specificity of a digital PCR-based approach to determine the transgene copy numbers and, subsequently, the stability in a complement Vero cell-line expressing HSV genes. Their data revealed a similar number of HSV-1 gene copies both in the master-cell bank and also in the extended cell bank of the transgenic Vero cell line, confirming the stability of these transgenes.
Genetic Linkage Analysis
“We are focusing on a group of receptors involved in modulating viral infections and setting up immune defense responses,” says James A. Traherne, senior research associate at the Cambridge Institute for Medical Research. These receptors, encoded by the highly polymorphic killer-cell immunoglobulin-like receptor (KIR) gene complex on the long arm of the human chromosome 19, modulate the development and the diversity of the Natural Killer (NK) cells, in part as a result of copy-number variations.
One of the prevailing theories is that genomic variation, including copy-number variation, in these and other polymorphic immune gene clusters, shape the progression of several medical conditions. “These genes have expanded and diversified in humans, and this is one of the reasons why there is a strong interest in their evolutionary history,” explains Dr. Traherne.
The KIR gene locus is one of the most rapidly evolving regions of the human genome, and over 100 different haplotypes have been described to date. The head-to-tail organization of the KIR genes as tandem repeats facilitates recombination events, but it also makes sequencing of this chromosomal region challenging.
Standard sequence-specific PCR and subsequently real-time quantitative PCR (qPCR) provided copy-number information and facilitated initial disease association studies. In a proof-of-principle analysis, Dr. Traherne and colleagues developed a next-generation digital PCR assay that allows KIR haplotypes and copy-number variation to be examined by genetic linkage analysis.
In this approach, a limiting number of DNA molecules are stochastically placed into a large number of droplet PCR nanoreactors. Each chamber of the reactor contains either one or no copies of the target PCR molecule. Positive or negative droplets for the PCR targets are counted using a flow cytometric device, allowing the number of targets to be determined without a calibration curve.
When two PCR targets are not physically linked on the chromosome, the two DNA confinement processes are stochastically independent and the droplets may contain one, two, or no DNA targets. In the presence of genetic linkage, the two processes are not independent and double-positive droplets are seen more frequently than in the absence of linkage.
“Digital PCR offers an excellent platform to type gene copy numbers and provide genetic linkage information,” comments Dr. Traherne. “When optimized, this method also allows us to discriminate high copy numbers much better than qPCR.”
“We’ve used digital PCR to explore copy-number changes in cancer cells,” says Rory L. Cochran, Ph.D., medical student at the University of California, San Diego School of Medicine. “It is similar to qPCR but a lot more precise, since we are able to count individual molecules directly as opposed to getting the results indirectly.”
In a case study on a patient with a familial gastrointestinal stromal tumor syndrome, Dr. Cochran and colleagues used droplet digital PCR. The investigators revealed that KIT, one of two genes causally linked to the condition, demonstrated amplification of the mutant allele in the patient’s tumor cells, as compared to surrounding, nontransformed cells.
In another recent study, Dr. Cochran and colleagues analyzed loss-of-heterozygosity (LOH) in two breast cancer patients harboring a rare missense BRCA2 variant of unknown clinical significance (VUS). While conventional PCR pointed toward a change in the allelic abundance in archival formalin-fixed, paraffin-embedded samples, digital PCR did not find evidence for LOH, revealing the power of this approach to quickly and accurately examine allelic ratios from archival tissue samples.
In contrast to next-generation sequencing, digital PCR is easy to perform and does not require extensive computational expertise. “One challenge will be to make digital PCR even more user-friendly,” comments Dr. Cochran, who suggests that digital PCR platforms should work with multiple computer operating systems. “The current platforms are operating system specific.”