Divide and Conquer: How to Setup Your First Droplet Digital PCR Experiment

Written by: Jelena Jankovic

last updated: January 14, 2020

Droplet digital PCR?  It’s easy. Because we’re here to guide you through it. We recently introduced you to the principles of digital PCR technology and how it differs from qPCR. In a nutshell, digital PCR is an end-point PCR technology that divides a single PCR into a large number of partitions, and then perform PCR inside of each partition. Upon completion, the number of partitions with positive reactions is counted.  The final result is expressed as the absolute number of copies in a sample. Partitioning minimizes the competition of DNA templates for the reagents and dilutes any potential inhibitors in the sample, which increases the sensitivity of PCR. The reaction efficiency has little impact on the results.  Furthermore, standard curves are redundant because the method itself calculates the absolute number of DNA copies.

Droplet Digital PCR Creates Partitions with Droplets

There are several different ways of partitioning and one of them is through creating droplets. By using the principles of microfluidics, pushing the aqueous sample through a barrier of oil under pressure traps the sample into nanoliter sized water-in-oil droplets. The droplets present separate partitions in which the PCR is performed. When the reaction is finished, droplets are read in a machine that singulates them as they pass through a detector and measures fluorescence. If the amplitude is high, the droplet is given a positive score. The software then counts the number of positive droplets per sample and calculates the exact number of copies.

Designing a Droplet Digital PCR Experiment

First, you need the instruments

So far, several companies are producing droplet digital PCR machines; the basic experimental design is almost the same for whichever one you opt for. Two machines make up the droplet digital PCR system—one that generates the droplets and one that counts the droplets. You also need a thermal cycler to perform the PCR itself.  You can use any thermal cycler. But if you can choose, use the quantitative one. That way you can track the reaction in real time, and if it doesn’t perform well, you can stop it at that point and avoid going into further labor with counting the droplets.

Next up – the reagents

Apart from the droplet generating oil, all reagents are basically the same as in qPCR; depending on the system, they can be proprietary. Therefore, you can choose either the probe–based assay or a fluorescent dye assay. You also need specific primer pairs for the gene of interest and the reference gene (if you are measuring gene expression). If you want to multiplex, use two probes with two different dyes, because the detector measures fluorescence on two channels. You can also multiplex without a probe—longer amplicons will bind more fluorescent dye and fluoresce more. When using large DNA templates (genomic DNA from fresh tissue/cells), digest the DNA first with endonucleases. Genomic DNA is highly viscous, and this can jeopardize the formation of water-in-oil particles. Genomic DNA from FFPE does not need this pre-processing step as it is already fragmented because of the tissue fixation. When digesting the DNA, make sure the enzymes do not cut within the amplification region. Check the dynamic range (minimum and maximum number of template copies in a sample) of the machine and the dilute the template accordingly. Just for reference, there are approximately 120,000 copies in 400 ng of human DNA, assuming 1 copy/haploid genome. So for estimating the number of copies per sample you can use the formula: m(g) = n(genome size in base pairs) x 1.096×10-21 g/bp (mass of one base pair)

Ok, so What Is the Workflow?

The first step in a droplet digital PCR experiment is to mix the reaction mixture with the oil, load it onto the droplet generator, and wait for the droplets to pop-out under pressure. Some systems may require that you transfer the droplets into a PCR plate. Be very careful with pipetting, because droplets are unstable at this step; use a pipette that has a gentle vacuum. The second step is thermal cycling performed at the conditions specified for the gene of interest. There is one critical parameter to set here. Because droplets are immobile, the rates of normal aqueous thermal diffusion are decreased, meaning that droplets may not heat up simultaneously. To overcome this, you should set the ramp rate to 2.5°C/sec to allow all droplets to reach the correct temperature. Finally, read the droplets in a droplet reader. In the reader, the drops first  pass a quality control test. If they do not meet the size and shape criteria, they are excluded from the measurement. Droplets that contain the target sequence have a fluorescence that passes the threshold, and are assigned the value of 1. The ones without the template have a fluorescence below the threshold, and are assigned the value of 0. The software counts the number of positive and negative droplets, and calculates the exact number of copies. The number of droplets included in the final measurement is not ultimately important, because the result is calculated from the proportion of the negative ones. The actual reading can take several hours. It is good to know that droplets are very stable upon PCR. You can store them for a few days in the fridge if you do not have enough time to read them on the same day.

Data Analysis Employs Poisson Correction

The perk of data analysis in droplet digital PCR is the ease of drawing clear thresholds. There are two reasons for this.  First, the measurement is performed at end point.  Only droplets with the target sequence fluoresce, making the difference in amplitude between the positive and negative droplets very high.  Second, many droplets are measured. This gives a large number of positives and negatives so the results are not biased by a small fraction of droplets that do not reach end point (due to low reaction efficiency). Another advantage is you can vary threshold without affecting calculation of the concentration because of the large difference in the amplitude. We mentioned that the exact number of copies is calculated, not just simply counted. This is because you correct for the possibility of droplets having more than one DNA template. The probability for this to occur rises with the concentration of the DNA in the sample, and the distribution of DNA templates follows Poisson distribution. The data analysis software automatically corrects the result to include the possibility of >1 copy/partition. The higher the concentration, the larger the correction. However, the concentration of the DNA must not be higher than the set limit, since the calculation of copy numbers is based on the fraction of negative droplets. If the concentration is higher than the set limit, there may not be empty droplets and the calculation will not be performed.

But Before You Start, Clearly Define What You Are Looking For

As magnificent as the technology of droplet digital PCR and digital PCR in general sounds, it is not convenient for every PCR experiment. It is superior for “finding a needle in haystack” situations—looking for rare allele mutations, small fold changes in gene expression (like two-fold that qPCR cannot detect), or a small difference in copy numbers (i.e., between 10 and 11). While high throughput experiments and large changes in expression are successfully conducted with conventional or real time PCR, the situations where accuracy and sensitivity is paramount will greatly benefit from the use of the precise digital PCR. Finally, the sensitivity rises with the number of droplets, so when choosing the system, check how many droplets they create and how sensitive your detection needs to be. Or, to extend the maxima from the title—the more you divide, the easier you’ll conquer.

Jelena is a Research Associate with PhD in Biochemistry from the University of Belgrade, working in the field of thyroid oncology with focus on molecular diagnostics.

More 'PCR, qPCR and qRT-PCR' articles