dPCR
The first time I heard about digital PCR was in 2014 and I was still a postdoc researcher at UC Berkeley. I remember the vivacious speaker, a product manager from Bio-Rad announced: "the next decade is the era of digital PCR. If you are still using conventional PCR, you are out." For long I couldn't wrap my brain around the concept of PCR being digitalized. By then I have been a molecular biologist working on various of PCR reactions for years. And digital PCR? Why? How? And what is it after all? I secretly boycott the idea, as it was bold and novel, and I felt being mocked in face on my ignorance. Now I regret it.
In the following two years at Berkeley, I witnessed an explosion of biological discoveries made available by the new genomic tools: Next Generation Sequencing (NGS), DNA MicroArray, single cell manipulation and sequencing, RNA-seq. The throughput of the assays has increased by hundreds and thousands of folds, so much so that bioinformatics, a discipline on the methodology and development of software tools to understand biological data, is typically employed to make sense of the results instead of human researchers.
Among the buzzwords, digital PCR is not prominent but distinct. It is like the peacock in the zoo: it doesn't represent any species, climate, or regions but every zoo should have one.
Polymerase Chain Reaction (PCR)
PCR is probably one of the most widely used molecular biology tools to amplify specific DNA sequences. The 1983 Nobel Prize in Chemistry was awarded to American biochemist Kary B. Mullis "for his invention of the polymerase chain reaction (PCR) method".
Essentially PCR is an enzyme catalyzed DNA synthesis from deoxyribonucleotides (dATP, dCTP, dGTP, dTTP, or together as dNTPs), the building blocks of DNA. The special enzyme is a DNA polymerase, isolated from the thermophilic bacteria Thermus aquaticus for its heat-resistant activity. Because significant amounts of a sample of DNA are necessary for some molecular and genetic analyses, studies of isolated pieces of DNA are nearly impossible without PCR amplification. Once amplified, these DNA products can be used in many different laboratory procedures. For example, most mapping techniques in the Human Genome Project (HGP) relied on PCR.
In the reaction, double stranded DNA will first be physically separated by heating (1, DNA denaturation or melting). Then the reaction is cooled down to allow primers, which are carefully designed small DNA fragments complementary to the flanking regions of target gene, to anneal to each of the single strand of template DNA (2. annealing). Afterwards, the DNA polymerase carries out the extension of the complementary DNA strand (3. elongation) and completes one PCR cycle. Such cycling process of PCR is fully automated nowadays, and the typical process consists of 25-35 cycles. With the target fragment doubled in each cycle, the amplification can yield millions to billions copies of DNA products in just a few hours.
Quantitative PCR (qPCR)
Digital PCR utilizes the PCR reaction but goes above and beyond. The goal is detection and quantification of target DNA or RNA in complex biological backgrounds. And its modern counterpart is real-time PCR or quantitative PCR (qPCR).
Real-time PCR or qPCR, which measures the accumulation of PCR product as it occurs during amplification, can deliver a great deal of information about a sample in a short amount of time.
How does the target DNA get quantified in real time?
Fluorescent label, mathematical modeling, and standards. In the PCR master mix, a product specific fluorescent reporter is added and the product formation could be traced in real time. As the cycling reaction goes on, an amplification curve (plot below) unfolds as fluorescent signal vs. cycle number.
Because the product doubles itself in each PCR cycle, the initial phase of the curve is exponential after the fluorescent signal exceeds the detection limit. The chain reaction will gradually deplete its ingredients, e.g. primers, dNTP, and slow down. And fluorescence signal draws into the saturated plateau phase. The mathematical modeling should be focused in the linear region of the exponential growth. By setting the threshold line in the plot, a Ct value (the threshold cycle) is generated from the intersection of each reaction curve. The Ct value is inversely proportional to the concentration of the template DNA in the sample. In other words, the high of template DNA concentration, the earlier the PCR fluorescence signal exceeds the threshold. By comparing the Ct value of unknown target with standards or endogenous control, the quantity of the template DNA could be relatively estimated.
Current real-time PCR technology has applications in a huge range of scientific areas including molecular biology, microbiology, genetics, diagnostics, clinical laboratories, forensic science, environmental science, food science, hereditary studies, paternity testing, and many other areas of the life sciences.
Although qPCR can be used for a wide variety of applications, the relative quantification often renders the sample preparation burdensome (standard curve preparation and plotting) and error-prone. qPCR also suffers from a discrimination power of at least 2 fold difference. Alternative approach is needed for applications requiring precision, high sensitivity and reproducibility. And digital PCR offers the solution.
Digital PCR (dPCR)
Digital PCR employs the same master mixture, enzyme, fluorescent reporter as required in qPCR. The chemistry of the reaction is essentially identical in digital PCR. The only difference is the reaction vessel. Starting for similar reaction volume, dPCR partitions the reaction into thousands of individual and paralleled reactions. Absolute quantification is achieved by counting the positive/negative reactions in each partition and modeled on Poisson distribution at the end of reaction.
The first paper on dPCR was published by Dr Alec Morley and Pamela Sykes in 1992. The purpose was to quantify PCR targets and measure the absolute lowest number of leukemic cells in a patient with leukemia. By monitoring the residual disease in leukemia patients, physicians can treat the patients at the earliest possible moment of disease detection.
Further evolutions of the technology allowed for more widespread distribution of this technique, with small partitions created by emulsion droplets and microfluidics.
Examples of these applications are:
Rare allele detection
Copy number variation
Gene expression for <2-fold differences
Quantification of NGS libraries
Detecting low-abundance RNA
Pathogen detection
Viral load detection