New frontiers in PCR: going digital or not?

Tanuja Koppal, PhD discusses regular versus digital PCR with Drs. Emir Hodzic, Keith R. Jerome, and Ruth Hall Sedlak.

Q: What do you use PCR for in veterinary sciences?

Hodzic: Our real-time PCR (polymerase chain reaction) facility at UC Davis is one of the oldest PCR core facilities and has been around for nearly 15 years. The facility has two parts, one for research and the other for diagnostics.

The research part includes services such as consultation, help with experimental design, and performing PCR analysis and data analysis, and these are done for people across the campus and beyond. Then there is the veterinary diagnostics part, where we use PCR to diagnose diseases in small and large animals. Here we receive samples from clinicians on campus, as well as from people outside the state and overseas. We have designed nearly 100 PCR assays as molecular diagnostics for a variety of pathogens, including bacteria, viruses, fungi, and parasites that infect small and large animals. We have respiratory, neural, and gastrointestinal panels for detection of pathogens in equine, feline, and canine animals, which is more convenient than doing assays for single pathogens.

Q: What is the most challenging aspect of the PCR work that you do?

Hodzic: The most challenging aspect is designing the assay that is specific for a particular pathogen. Pathogens like viruses undergo several mutations, and finding all the different strains of the species in one assay is difficult. At the same time, all our assays have high efficiency, often between 95 and 100 percent. Finding the right controls for an assay can be challenging, and so is data interpretation. PCR is a very sensitive technique, but sometimes the results need to be validated using some other technique.

Q: How do you address the issues surrounding contamination and reliability of data?

Hodzic: We have very high standards for quality control and quality assurance. Each and every step, from sample receiving to sample processing, from oligonucleotide extractions to performing PCR and finally the data analysis, everything is well controlled. That takes time and effort, and it’s certainly not cheap. We have everything set up in such a way that there is no room for contamination. A part of the lab is set up for receiving samples, while the sample extraction and PCR are done in a different lab. All our reagents, primers, probes, and master mixes are stored in a clean room, and people moving between labs have to follow certain procedures to make sure there is no contamination. We have standard operating procedures (SOPs) for every single step and for the different sample types that we analyze. We regularly check our equipment and the surfaces that we work on to make sure there is no contamination. That’s the only way we can stand behind our data.

Q: How important is training when it comes to PCR work?

Hodzic: Training is very important, and our personnel have to go through several months of training. There are many quantitative PCR (qPCR) seminars and courses offered on campus. Real-time qPCR setup is very easy, but then each and every step that follows is extremely important. When we meet with our clients, we always spend time discussing how to design the PCR experiment, what sample types need to be used, and how to store and collect them. The sample handling, the oligonucleotide extraction, the quality control—everything needs to be carefully assessed. All our personnel are trained in how to follow these PCR steps and how to troubleshoot if there is a problem. However, it’s always good to exchange information and brainstorm with colleagues no matter how trained and knowledgeable one might be.

Q: Have there been a lot of changes in PCR in recent years? Have you considered using digital PCR?

Hodzic: The instrumentation for PCR has not changed very much in the past decade. There are new technologies like digital PCR that have been introduced, but we don’t see [digital PCR] being used in our lab anytime soon. Digital PCR does not have the capacity to analyze multiple pathogens from the same sample or to look at 30 to 40 samples at the same time. It’s labor-intensive and is more suitable for research use, not for diagnostics. Although I do not see real-time PCR changing much in the next few years, the probe systems that are being used will evolve, and PCR will certainly be used more for diagnostic applications.

Q: How long have you been using digital PCR in your lab?

Sedlak: We have been using the droplet digital PCR instrument for a few years now. We started out using it in our clinical laboratory to see if it would be advantageous for clinical care and later started using it for our research, particularly for viral load quantitation and in gene editing, looking at mutations introduced by meganucleases or CRISPR/Cas systems. What we have found is that digital PCR works better for certain applications where you need precise measurements.

Q: What applications warrant the use of digital PCR?

Jerome: Digital PCR is facilitating our work in developing a cure for chronic viral infections such as HIV, hepatitis B, and herpes. We are developing targeted endonucleases, of which one class is the CRISPR/Cas protein system. These nucleases are able to find the viral DNA in a background of cellular DNA and induce specific mutations that are lethal to the virus. It’s an extremely powerful and promising technology. For this work, we need two critical measurements. First, we need to find out precisely how much virus is present in our experimental models, how much of this virus are we able to affect, and how much virus is then destroyed. We need a very precise technology to measure that. qPCR is really best at detecting log scale differences and cannot differentiate between a few percentage point changes. Digital PCR, on the other hand, offers exquisite precision to evaluate viral loads within a couple of percentage point variations. Second, we can use digital PCR to monitor specific mutations that we induce in the virus to target sites that we want to knock out. We can carry out allele-specific PCR or allelic discrimination PCR using qPCR, but with the digital technology we can find a very small percentage of mutations in the background of wild-type viruses. As we get more efficient, we can get very precise measurements of what percentage of the virus has been mutated. This particular application really leverages the tremendous precision that digital PCR offers compared with qPCR.

Q: What are the limitations of using digital PCR?

Jerome: The limitations of digital PCR are fairly minor. In some ways, it’s more forgiving than qPCR for different sample types. It’s more resistant to sample inhibition, which can be a problem in qPCR. If you have a well-designed qPCR assay, then very little additional validation needs to be done to transition it to digital PCR. Hence, the barrier to implementing digital PCR for a lab that already does qPCR well is quite low. The main drawback in some of these first-generation digital PCR systems is that the overall throughput is low compared with clinical-grade qPCR systems that use 96-well plates. Digital PCR takes more hands-on time, but for research applications, that’s not a problem, because we are not running thousands of samples.

Sedlak: The hands-on time required depends on how your lab is set up to do qPCR. Our clinical lab is highly efficient in doing qPCR since we do high-volume clinical testing. Hence, for our lab, the digital PCR instrument that is set up to do 96-well plates takes about three times longer than qPCR. The drawback with digital PCR is that it is not a real-time measurement like qPCR. It’s an end point measurement, and that adds another couple of hours to the overall experiment. The up-front setup of the reactions takes about 45 minutes more than with qPCR. However, there is robotic instrumentation available as an option for droplet generation that reduces the setup time, and our research lab uses it when generating the reaction plates. Similar to qPCR, the results with a digital instrument are only going to be as good as the assay you have developed. However, in digital PCR, it becomes obvious if your assay is not working properly, and that’s a good thing because you can then optimize the assay further. You can also clearly see efficiency problems in digital PCR that are often masked in qPCR.

Q: How do traditional and digital PCR compare in terms of costs?

Sedlak: The cost of consumables and reagents is about $3 to $5 per well for a digital assay, which is not that much more, depending on your throughput, than qPCR. The cost is not prohibitive, and the kind of information you get is something you cannot obtain with qPCR.

Jerome: Digital PCR has aspects that are clearly superior to qPCR, but for many applications those advantages don’t come into play. Most of our routine clinical viral loads, for instance, are still done using qPCR, which offers high throughput and simplicity. The numbers that you get will be more precise with digital PCR, but in a small to medium-sized study, we were not able to demonstrate superiority in terms of patient care or management. So one needs to carefully choose applications where the advantages of digital PCR can be translated into better research or patient care.

Q: How do you choose which type of digital PCR system to use?

Sedlak: For us, the droplet digital instrument offered better throughput and less consumable costs when compared with the chip-based instruments. The chip-based systems have fewer partitions but more costly consumables per partition. However, the chip-based systems can sometimes perform real-time PCR experiments and give a digital readout at the end. We have no need for real-time results in our digital PCR experiments, but if that’s something you need, then it’s better to use a chip-based format that provides the option.

Jerome: The other option that the solid, chip-based media approaches offer, which is appealing to the research field, is that one can recapture the positive wells in the reaction and later sequence the nucleic acids in each well. The droplet system does not allow us to capture positive droplets and sequence them individually.

Q: What are the gaps in digital PCR that you would like to see filled?

Sedlak: Improvements in throughput and hands-on time and reducing the variability introduced by the various hands-on steps are certainly areas that can be improved in digital PCR.

Jerome: My hope is that the manufacturers increase the number of channels that can be analyzed for the droplet digital PCR. Ruth has spent a lot of time developing clever multiplexing approaches using differential labeling, but currently we have only two colors available. Hence, higher-order multiplexing of analytes is very difficult in digital PCR at this time. Multiplexed testing for multiple viruses, using many different channels and colors, will be helpful.

Q: Any advice for people who are looking to invest in digital PCR?

Jerome: I would advise people to think hard about their applications and see if the advantages of digital PCR are valuable for that application. If precision is important, then digital PCR is fantastic. If you need to find and quantitate rare events in the background of wild-type events, then digital is great. However, if you just want to measure how much analyte is present, then qPCR offers better ease of use and throughput. Going digital adds complexity, and one needs to think hard [about whether] that’s really necessary. Most of the point-of-care tests done today are only qualitative, and moving those to qPCR would in itself be a big step up. Adding precision to the measurements with digital PCR is something that can always be considered later.

Emir Hodzic, DVM, PhD, graduated from the veterinary school at the University of Sarajevo, Bosnia and Herzegovina, where he obtained a master’s degree and a PhD. After a postdoctoral fellowship at Yale, he moved to the Center for Comparative Medicine at the University of California, Davis. Since 2005, he has been director of the Real-Time PCR Research and Diagnostics Core Facility.

Dr. Keith R. Jerome is the head of the virology division in the Department of Laboratory Medicine at the University of Washington and a member of the combined program in infectious disease sciences/virology at the Fred Hutchinson Cancer Research Center. He received his MD and PhD degrees from Duke University. He completed his postgraduate training in laboratory medicine and virology at the University of Washington.

Dr. Ruth Hall Sedlak is a research scientist in the molecular virology laboratory in the Department of Laboratory Medicine at the University of Washington. She earned her PhD in microbiology and nanotechnology at the University of Washington in 2011. After completing her postdoctoral training with Dr. Keith Jerome on herpes virus diagnostics and applications of digital PCR, she continued in the lab as a diagnostic development scientist.