The Emergence of Digital PCR

Q: What is digital PCR and how does it compare to traditional PCR?

A: Digital PCR is really a limiting dilution PCR and involves partitioning a sample [of DNA or cDNA] into multiple reactors so that there are individual PCR reactions taking place in parallel. Traditional PCR has copies of a sample, all in one reaction vessel, whereas in digital PCR, each reactor has either a single copy or no copy of a target molecule—and that’s really the fundamental difference. This allows Poisson modeling of the percentage of reactors that show amplification to accurately compute a starting sample titer.

Q: What is the biggest advantage, as well as the obvious limitation, of using digital PCR?

A: The biggest advantage of digital PCR is that it offers an extremely accurate quantitation of the copy number [of the target molecule] in your sample. The obvious limitation of partitioning your sample into different reactors is that you now have to observe and record tens of thousands, if not millions, of reactors instead of just a few cuvettes or samples.

Q: Does digital PCR involve completely different instrumentation and reagents compared to traditional PCR?

A: The reagents, such as enzymes, primers, and probes that you use for standard PCR are the same for digital PCR. However, there are some additional reagents that need to be used. For instance, for systems that use the droplets, you need emulsifying reagents, like oil, and some surfactants.

Q: Is digital PCR more suitable for certain applications than standard PCR?

A: The applications that are most suitable for digital PCR are obviously ones where titer or copy number quantitation is extremely important, or an application where you need to pick out one or a few copy numbers of target from a complex background. One application that comes to mind immediately is in oncology testing, where you are looking for early detection of mutation or metastasis. In digital PCR, the mutated DNA or RNA oligo is partitioned out into different reactors and hence becomes easier to detect as compared to traditional PCR, where it’s part of a bulk, massively complex sample. Hence, the strongest and most obvious use for digital PCR right now is for medical applications. It’s ideal for applications where knowing the titer is very important, such as in monitoring viral load present in very low copy numbers or in HIV testing.

Q: What are you working on in your lab related to digital PCR applications?

A: As a national lab we investigate many parallel threads and look for the right opportunity to put a new capability together. We wanted to develop technologies to detect certain rare targets in complex environmental samples, and our droplet digital PCR technology was developed with that goal in mind. Since then we have moved on to other broader applications. Currently we are looking at enriching microarrays using PCR and targeted laser-induced DNA desorption as one of our experimental platforms.

Q: Do you need extensive training to get started with digital PCR?

A: The next-generation PCR instruments are definitely more user friendly and do not need users with advanced degrees operating them. Vendors are working to provide systems where the workflows are much simpler and easier from the user’s perspective. There is no need for extensive training. A course that provides a good understanding of how digital PCR works and training from the vendor on how to use the instrument is all you should need. The software tools and data analysis packages that come with the instruments are also fairly standard, but users do need some training there to get started.

Q: Is sample preparation still a challenge in digital PCR, like it is in traditional PCR?

A: Sample prep is every bit as important as it used to be, and perhaps more so in digital PCR, since there is no room for error in quantitation. The controls and standards that are used in standard PCR are also equally important in digital PCR when you are initially validating your assay. However, digital PCR has certain advantages if you have inhibitors in your sample, because the inhibitors are partitioned and may not be present in reactors containing targeted oligos. Also, the systems use endpoint detection instead of real-time [detection], so if your amplification efficiency is lower than normal, a digital system may still detect it, while a bulk PCR reaction may not. Hence, sample prep with inhibitors is more robust on a digital PCR system. But sample prep remains important, along with experience in running and calibrating the assay and having primers that are appropriate for the target you are trying to amplify.

Q: Is the cost per sample significantly higher for digital PCR?

A: The cost per sample will certainly be higher with digital PCR because what was run in plastic disposable tubes is now partitioned into tens of thousands or millions of reactors, as droplets or in an array-based architecture. So you have additional substrates and reagents, you have to create emulsions using microfluidics, and you have to be able to read the optical signal after amplification in all those individual reactors. Hence, cost per sample is going to be higher, which is why people should invest in digital PCR only if they need the quantitation accuracy or need to detect a low-copy target in a complex background. If you don’t need those capabilities, you probably don’t need to make the investment unless the costs fall to qPCR levels.

Q: How should people evaluate the different systems that are currently available?

A: There are different types of digital systems out there and they should be evaluated based on your application. If you are looking for top-line performance with the highest statistics possible for applications requiring high titer accuracy, then take a look at the droplet-based systems that provide millions of reactors. There are also array-based systems that offer tens of thousands of reactors, which are sufficient for some medical and environmental modeling applications. Array-based systems tend to be easier and faster to load, so it really depends on the focus of the lab and what assays they are running.

Q: Where are the big gaps in this field? What needs to be improved upon?

A: By and large, the gaps are being worked on in terms of making systems user friendly and the workflow smoother. However, no matter how easy it gets, there is still more engagement and time needed making emulsions, performing additional steps, and taking more readings. Digital PCR is more expensive in terms of cost and labor. Another gap that is being filled is in assay development. More assays are being modified to the digital PCR format, and you will see people making that investment if the assay is for an application that requires accurate quantitation. There are a lot of good choices for digital PCR on the market today. It’s a proven and user-friendly technology, and if they have a need for it, people should not be afraid to take a look.

Dr. Reginald Beer is the medical diagnostics initiative leader at Lawrence Livermore National Laboratory, where he develops lab-on-chip technologies for molecular diagnostics applications. He demonstrated the first real-time digital PCR in monodisperse droplets in 2007, and has developed additional sorting, trapping, and amplification tools to aid in droplet digital PCR. His research in digital PCR, on-chip miniaturization, and selective microarray dehybridization has led to several new technologies offering improved diagnostics capabilities.