Josh P. Roberts has an M.A. in the history and philosophy of science, and he also went through the Ph.D. program in molecular, cellular, developmental biology, and genetics at the University of Minnesota, with dissertation research in ocular immunology.
Digital PCR (dPCR) is largely seen as the next step in the evolution of PCR—it enables reliable, absolute quantification of rare alleles in minute quantities of sample, fast. With such a pedigree, why bother with anything else?
True, quantitative real-time PCR is, well, quantitative. And that well-established technique does offer many benefits over the upstart dPCR. For instance, “real-time [quantification] is less expensive and has the potential for much higher throughput,” notes Jim Huggett, science leader for nucleic acid metrology at LGC, the UK’s designated national measurement institute for chemical and bioanalytical measurement. “It’s pretty much perfectly fine for the vast majority of things you’d need to quantify using a PCR method.”
Yet, there are applications for which dPCR might well represent the only practicable solution—for example, searching for rare mutations in a sea of wild-type sequences, monitoring subtle changes in viral load or tracking copy-number increases. Here we review the different dPCR platforms available and their relative benefits and pitfalls, as well as tips for getting meaningful, reproducible data.
Defining dPCR
Digital PCR begins with essentially the same master mix as qPCR, plus primers and template DNA, explains George Karlin-Neumann, director of scientific affairs at Bio-Rad Laboratories’ Digital Biology Center. “You’re just subdividing it into a lot of very tiny, equal-sized partitions.” Each partition thus becomes a unique reaction that either will or will not fluoresce following thermocycling, depending on whether it contains a target molecule. Hence the term “digital.”
Search Digital PCR Systems
Find, compare and review dPCR systems
from different suppliers Search
Literally counting the number of positive vs. negative partitions—and then massaging that number with Poisson statistics—yields the target concentration in the original solution, considered “absolute” because (unlike qPCR) it does not need to be calibrated to a standard (Ct) curve.
Of the systems currently on the market, Bio-Rad Laboratories’ and RainDance Technologies’ instruments partition the reactions into oil droplets, an approach termed droplet digital PCR (ddPCR) or emulsion dPCR, while Fluidigm and Life Technologies partition in holes or wells on a microchip surface (chip-dPCR).
dPCR applications
Although both qPCR and dPCR are quantitative, there are several broad applications for which dPCR outshines its predecessor. One is detecting minority targets—the so-called “specificity issue,” says Huggett. The problem with qPCR-based approaches is that the probe or primer can sometimes pick up the wild-type sequence as well as the desired target and thus cannot typically detect sequences below about 5% abundance. dPCR, on the other hand, “simply reduces the background signal from the wild type by diluting out all your DNA, so the ratio of wild type to mutant becomes much lower, and you have more chance of finding it.”
As a result, dPCR can detect exceptionally rare targets, and researchers have applied the technique for fetal screening, detection of cancer mutations, organ transplant rejection monitoring, viral tracking and other applications requiring the ability to quantify DNA with high sensitivity in the presence of more abundant, related sequences.
dPCR’s sensitivity also enables customers to detect small differences in copy number (seven from eight copies, say, or even 10 from 11)—changes that cannot be detected using qPCR, explains Iain Russell, senior product manager at Life Technologies.
But as a metrologist, Huggett is perhaps most excited about the fact that with dPCR, “it’s possible that you and I can measure the same sample—me in London and you in Minneapolis—and we can get a very similar result without actually worrying about a calibration curve,” he says. “There is a lot of work in real time showing that you can get wildly different results just because of the calibration strategy you’re using.”
A framework for reporting data
Huggett and his colleagues recently developed and published a set of best practices, called the Digital MIQE Guidelines, which establish the “Minimum Information for Publication of Quantitative Digital PCR Experiments” [1]. Emphasizing the use of adequate controls, including biological replicates, the dMIQE guidelines provide a framework for reporting experimental details and data so they may be reproduced by other laboratories, and so conclusions drawn follow from the experiment undertaken and reported.
Though largely similar to an earlier set of guidelines laid down for qPCR, the two are not identical. For example, it’s not necessary to report information of quantification cycle or amplification patterns in dPCR, but it is important to note such variables as partition size and number. It’s also important to estimate the concentration of the sample ahead of time, because variance and precision varies with concentration far more significantly in dPCR than in qPCR.
The original “MIQE guidelines were published about 10 years too late,” says Huggett. “We felt it would be better to get in there early to encourage people to do the reporting [and] to find the transparency.”
Balancing the pros & cons
Unlike many more mature technologies, there are still some substantial differences among dPCR platforms. For example, with 10 million partitions per sample, RainDance has by far the largest dynamic range and sensitivity. “There is a lot of interest in the area of circulating tumor cells and cell-free DNA, where there’s just not as much material to play with,” says Bill Keating, director of business development for RainDance Technologies. “You’re dealing with rare events, and you want to find them, if they’re there.”
But, notes Huggett, RainDance’s system also takes the longest to run; in the standard configuration you can only do two runs of up to eight samples per day on the RainDrop™ Digital PCR System. It’s a “specialist instrument … like a super camera that they have on satellites.”
Fluidigm’s BioMark HD™ has been on the market longest and is a “very good instrument, very easy to use,” says Huggett. It is real-time capable, which is useful for troubleshooting and obviates the need for downstream processing. But it has a “massive dead volume” and, at only 770 partitions per sample, has the lowest sensitivity of any instrument.
The Bio-Rad QX200™ Droplet Digital™ PCR System and Life Technologies’ QuantStudio™ 3D Digital PCR System each boast 20,000 partitions per sample. The former, with flow-cytometry-like imaging of droplets, has the potential to isolate a droplet for further analysis (a feature that has been demonstrated in the published literature but is not currently offered on the commercial platform), while QuantStudio 3D enables users to re-analyze a chip if something goes wrong with the initial read.
Comparing cost is another matter. One industry report lists four different cost metrics—the initial instrument cost plus three different usage costs: cost per run, cost per sample and cost per 10,000 reactions [2]. RainDrop, for example, lists for $125,000 and costs $240 per run, but it costs only three cents per 10,000 reactions, which is far lower than Bio-Rad’s $150 per 10,000 reactions. But the Bio-Rad QX200 can run up to 96 samples at a per-sample cost of only $3. Meanwhile, a complete QuantStudio 3D system, including enough consumables to run 384 samples, lists for $44,000—about half the initial price of a Bio-Rad system.
If you need absolute quantification, reproducibility and high-sensitivity measurements from minimal sample, dPCR could well be the way to go. Of course, you can’t have everything in a single instrument, so you’ll need to balance speed, throughput, sensitivity, real-time capabilities, supported chemistries, initial and ongoing costs, as well as a host of other factors, to decide which one is right for you.
Your best bet is to take a test-drive. Only by seeing how the instrument performs with your samples can you be sure how it will serve your lab and its needs.
References
[1] Huggett, JF, et al., “The digital MIQE guidelines: Minimum Information for Publication of Quantitative Digital PCR Experiments,” Clin Chem, 59:892-902, 2013. [PubMed ID: 23570709]
[2] Insight Pharma Reports, Digital PCR Technology Report: An Insight to Vendors, Costs & the End User Community, May 2013.
Image: Life Technologies' QuantStudio 3D Digital PCR chip.