The polymerase chain reaction (PCR) was discovered and commercialized in the 1980s and revolutionized molecular biology and clinical genetics. The reaction involves the exponential amplification of a target polynucleotide sequence by repeatedly thermocycling a salt-buffered mix of template nucleic acid molecules, oligonucleotide primers, dNTPs, and a thermostable DNA polymerase. Over the years, ongoing development and application of the PCR reaction enabled molecular cloning, engineered transgenic organisms, DNA forensics, clinical diagnostic DNA sequencing, among many additional innovative technologies.
In the traditional PCR reaction, the starting template material is in fact many nucleic acid molecules, usually several nanograms’ worth corresponding to millions of molecules. This starting material is most often a heterogeneous mixture of variant molecules, genetic alleles, or genomes depending on the application. Cancer is a clinically relevant example of this heterogeneity, where a relatively small number of template DNA molecules from cancer cells may harbor a cancer-causing somatic mutation relative to the wild-type (WT) germline DNA.
In the early 1990s, several groups began exploring the possibility of diluting template molecules for the PCR reaction to an extent such that, on average, any single PCR reaction only contained a single template molecule. When many reactions are performed at this level of dilution, the frequency of positive reactions follows the Poisson distribution and allows for calculating the abundance of the target molecule based on dilution factor. Several groups reported using versions of this limiting dilution PCR strategy to study, for example, variation among HIV proviruses1, human genomic haplotyping2, and to quantify the fraction of leukemic cells in patient samples3,4.
In 1999, recognizing the urgent importance for early detection of cancer-causing somatic mutations in clinical samples and the possibility of early cancer diagnosis, Bert Vogelstein and Kenneth Kinzler noted that detecting these mutations depended on isolating them from a large excess of normal wild-type (WT) cells. Reliably detecting these rare somatic mutations could therefore help diagnose primary tumors in patients who are asymptomatic and whose disease was still curable. Vogelstein and Kinzler developed a strategy to selectively amplify these rare mutant template molecules, distinguish them from WT, and quantify the fraction of mutant alleles relative to WT. The dilution PCR strategy allowed one to partition individual template molecules such that resulting amplification was either completely mutant or completely WT.
The Vogelstein and Kinzler strategy, termed digital PCR5, diluted and partitioned the starting DNA template into a 384-well plate, such that any given well contained one-half genome equivalent on average, i.e. half the wells contained one template molecule and half the wells contained zero template molecules. Subsequent amplification by PCR allowed the detection and quantification of the targeted mutant DNA sequences in the partitioned sample.
Vogelstein and Kinzler recognized the variety of potential applications of digital PCR, including detecting somatic single nucleotide variants in cancer, chromosomal translocations, changes in gene expression, allelic discrimination, and allelic imbalance. In 1999 they also noted that the limit of detection was defined by the number of single-template reactions that could be analyzed, paving the way for increased sensitivity of the technology by using high-throughput platforms such as 1,536-well plates, microarrays, and beyond.
PCR in Microfluidic Systems
In the mid-1990s, advances in microfabrication techniques had allowed the production of devices with feature sizes as small a few microns. Micromachining, photolithography, and etching of silicon and glass substrates produced networks of flow channels that could sort reagents and molecules, partition chemical reactions, and act as droplet generators for mixtures of aqueous and oil-based fluids. These microfluidic devices were well-suited for performing biochemical assays, such as single-cell assays, studies of macromolecules e.g. proteins and nucleic acids, and the polymerase chain reaction (PCR).
The partitioning capabilities of microfluidic lab-on-chip devices were particularly useful for PCR, where a microenvironment allows for increased thermocycling speeds and increased reagent concentrations within the reaction mixture1. Several groups developed integrated systems utilizing microfabricated components of glass and silicon to perform restriction enzyme digests, PCR, and electrophoresis within a single chip-based device 2,3. These systems generally relied on thermocapillary pumps to perform PCR in nanoliter-sized droplets within micron-sized channels, and included entry ports, heating elements, and DNA detectors. They did not require any novel hardware, biochemistries or DNA detection methods; rather, they simply miniaturized existing components such as heaters and applied well-known PCR methods in a micron-sized environment on a chip.
By the late 1990s integrated chips utilizing continuous flow of nanoliter-sized plugs through microchannels were becoming common. These systems achieved thermocycling by transporting the reaction plugs through a microcapillary etched into a glass chip that traverses heaters at the respective melting, annealing, and primer extension temperatures appropriate for PCR4,5. Samples were injected and propelled into inlets on the chip by precision syringe pumps.
In late 1999 the first commercial lab-on-a-chip, the HP 2100 Bioanalyzer, was introduced by Hewlett-Packard’s Chemical Analysis Group, which shortly thereafter became Agilent. Based on Caliper Technologies’ LabChip, the 2100 was the first complete instrument that could perform sample handling, separation, fragment sizing, quantitation, and digital data processing for PCR products, restriction enzyme digests, or RNA samples. Following the HP 2100, other lab-on-chip platforms were commercialized by companies including Nanogen Inc., HandyLab Inc., Micronit Microfluidics, among others.
PCR in Droplets
By the early 2000s, water-in-oil emulsions were being studied for potential applications designed to partition biological materials into picoliter-scale droplets6,7. A significant advance was the use of surfactants in the oil-based continuous phase to maintain the integrity of droplets formed after introduction of the aqueous discontinuous phase. Because of surfactants in the oil-filled microchannels, droplets can be pooled in widened channels or reservoirs, touching each other without breaking or fusing6. Extensive fluidic networks of oil-and-surfactant filled channels can be designed to coordinate the intentional mixing of plugs or droplets to mediate biochemical reactions7.
Fluorinated oils combined with fluorinated surfactants were particularly useful for handling biomolecules within aqueous droplets, as these carrier fluids are inert, stable at high temperatures, compatible with the most common lab-on-chip materials and can prevent surface adsorption of charged biomolecules7,8,9. Fluorinated oils and surfactants were commonly used and commercially available, e.g. FluorinertTM by 3M and ZonylTM by Dupont, among other offerings from Fluorochem Ltd., Sigma-Aldrich Inc., and others.
The combination of commercially available lab-on-chip microfluidic networks with advanced picoliter-scaled aqueous droplet generation and stable surfactant-mediated pooling and sorting of droplets provided future opportunities for single-molecule reactions within these droplet microreactors.
Next-Generation Digital PCR
By the mid 2000’s the power of single-molecule digital PCR assays had been combined with microfluidic liquid handling platforms to miniaturize the digital PCR reaction volumes in chambers on a microfabricated chip. It is well known that the sensitivity of digital PCR is constrained by the number of reactions that can be performed, detected, and analyzed. Accordingly, partitioning template molecules into thousands of chambers at nanoliter-scale volumes dramatically enhances sensitivity, along with providing the additional advantages of reagent savings, decreased diffusion distances, and improved statistics. Around that time several groups had used microfluidic digital PCR in chambers for applications as varied as interrogating microbial community diversity1, human genomic copy number variation2,3, and fetal aneuploidy4.
While digital PCR was well-known and widely practiced, in 2006 Fluidigm Corporation became the first company to commercialize the technology in an integrated fluidic circuit. The BioMark system was based on the 12.765 Digital Array, a chip of 12 panels, each panel partitioned into 765 6-nL chambers. After loading PCR reaction mixture through 12 carrier inputs, the chip was thermocycled, fluorescence was detected, and signal was processed and analyzed by the Digital PCR Analysis software.
Digital PCR in Droplets
An alternative approach that further enhanced throughput and sensitivity was to generate monodispersed picoliter-sized microdroplet reactors, offering many more partitions than chamber-based systems. Droplets are generated in a microfluidic system, thermocycled to perform single-molecule digital PCR within the droplets, and amplification is detected and quantified via real-time fluorescence curves5. These droplet-based lab-on-chip systems were also adapted to perform reverse transcription PCR (RT-PCR) to detect single copies of RNA genomes6.
This technology was first commercialized by QuantaLife, Inc. as the ddPCRTM System in 2011. This chip-based ddPCRTM platform can accommodate eight samples per chip, generating 14,000-16,000 droplets per sample.
A further innovation for digital PCR platforms was multiplexing, i.e. targeting multiple alleles within a single PCR mixture by introducing a plurality of primers and fluorescent probes. By 2010, chamber-based and droplet-based systems had the capability of multiplexing4,8,9. Since target molecules are partitioned individually, a given primer pair will only amplify its particular target in a partition that contains that unique allele. Multiplexing may be accomplished by using multiple target-specific colors of fluorescent probes, target-specific concentrations of the fluorescent probes at the beginning of amplification, or a combination of these strategies.11
The Stilla NaicaTM System
Crystal Digital PCRTM (cdPCRTM) is Stilla’s next-generation technology for absolute quantification of nucleic acids. Using cutting-edge microfluidic innovations, this technology uniquely integrates the Digital PCR process in a single consumable.
Crystal Digital PCRTM relies on the use of a single chip to partition samples into 2D droplet arrays and then subjects the Crystal to thermal cycling, both steps taking place in the NaicaTM Geode. The Crystal in then placed in the NaicaTM Prism3 instrument to be read using three-color fluorescence scanning. This novel technology based on three-color multiplexing enables a different approach to data analysis. When taken together with Stilla’s proprietary Crystal MinerTM software for statistically analyzing large arrays of multiplex target data, Stilla’s NaicaTM system workflow is user-friendly, streamlined, intuitive, and cost effective.