Detection of CNV with digital PCR (dPCR)

Genomic alterations associated with disease, such as certain copy number variants (CNVs, also known as copy number alterations, or CNAs), can be very challenging to accurately quantify. These variants, in which specific genomic regions are either deleted or repeated—sometimes hundreds of times—can be associated with many diseases, including cancer, cardiovascular disease, and neurodevelopmental disorders (Pös et al., 2021). For many disease-associated CNVs, accurate CNV counts are essential for understanding their potential biological impact (Steele et al., 2022).

There are several traditional methods for CNV analysis, including fluorescent in situ hybridization (FISH), flow cytometry, microarray analysis, and next-generation sequencing (NGS). However, each of these has its drawbacks. For example, FISH and flow cytometry offer less quantifiable results compared to dPCR, meaning that small variants in copy number can easily be missed. NGS analysis (which is costly and complex in terms of workflow and analysis) is often impacted by tandem repeat elements—one of the hallmarks of many CNVs.

Standard quantitative PCR-based methods (qPCR) can also be a powerful tool for detecting and assessing CNVs. However, these methods depend on comparing signal intensity from the target gene to a standard curve, thus increasing the number of required reactions—and qPCR usually only provides detection sensitivity down to about a 2-fold difference. On the other hand, dPCR can provide true quantitative data and molecular counts that enable detection of difference low as 10%—without the need for standard curves.

One example where this elevated sensitivity can be important for understanding disease is the amplification of the HER2 gene in breast cancer. As many as 20% (and perhaps more) of breast cancers (Chen et al., 2022) caused by somatic mutations contain extra copies of this gene.

Whether these CNVs are being detected for research or for diagnostic purposes, one challenge with tumor biopsy samples is that the samples often contain high numbers of normal cells in addition to cancerous cells. This can make it much more difficult to accurately assess the HER2 CNV of the cancerous cells—and highlights the value of the sensitive detection offered by dPCR.

Like other polymerase chain reaction (PCR) workflows, dPCR employs primers flanking target genomic regions to multiply copies of that region for analysis. Following the addition of a DNA polymerase and free nucleotides (dNTPs), the reaction is subjected to a series of several “amplification cycles” during which the target region is copied (amplified); during each round a signal from a light-emitting fluorophore is released (FIGURE 1). In qPCR applications, this signal is detected at each round, and then compared to the signals produced by known concentrations of the same target in a standard curve, providing an estimate of copy number in the original sample.

• First—prior to amplification, the dPCR reaction mixture described above is divided into thousands of tiny microreactors (FIGURE 2, below). This is done either by physically separating the mixture into nanowells on a chip or plate, or by creating oil-based emulsions of tiny “reactor droplets.” This partitioning step isolates rare target molecules from potential background “noise” so they are easier to detect while also diluting potential PCR inhibitors. The result is that some partitions will contain the DNA with the target sequence, and some will not; this separation is what enables the later calculations of absolute input amounts.
  
  
• Second—rather than measuring signal in dPCR after each round of amplification and then comparing the signal to a standard curve (as for qPCR), the fluorophore signals are not detected until after the last amplification cycle (end-point PCR).
  
  
• Third—during signal detection, each dPCR microreaction is counted as either positive (1) or negative (0) for the target molecule. The number of target molecules in the original input is then determined by Poisson analysis.

Thus, dPCR provides sensitive, accurate quantification of target molecules without requiring a set of standards--enabling the detection of rare molecules or even small changes in copy number.

To illustrate how CNV is assessed using dPCR, we’ll use the example of the HER2 gene that we introduced above. In this example (FIGURE 3), a multiplexed reaction was carried out in which two different genes (HER2 and EIF5) were assessed using two differently colored fluorophores (shown as yellow and blue below). EIF5 was used as a baseline; two copies are present per cell, one from each parent. Some of the dPCR partitions contained only HER2 (yellow), some contained only EIF5 (green), and some contained both (shown in green); the actual number of copies present was then evaluated using Poisson calculations. In this example, ~5.8 copies of the HER2 gene were present in this sample.

In summary, dPCR offers a highly sensitive method for accurate quantification of target mutations – such as CNV in both noninvasive liquid biopsy samples and solid tissue samples. It offers a solution that is sensitive, faster, more reliable, and less complex than many other workflows.
 

Learn more about our dPCR system.