The ‘+’ indicates that the previous base is a locked nucleic acid (LNA) base.Droplet digital PCR
All quantified DNA libraries and size standards were prepared for droplet PCR in 25 µL reactions containing 2× ddPCR Master Mix (Bio-Rad), 250 nM TaqMan probe, 900 nM each of the appropriate flanking primers, and ∼10,000 copies of target DNA. Emulsified 1 nL reaction droplets were made by adding 20 µL of each reaction mixture to the sample wells of a droplet generator DG8 cartridge (Bio-Rad) and 70 µL ddPCR Droplet Generation Oil (Bio-Rad) to the oil wells of the cartridge for use in the QX100 Droplet Generator (Bio-Rad). Forty microliters of the generated droplet emulsions were transferred to Twin.tec semi-skirted 96-well PCR plates (Eppendorf, Hamburg, Germany), which were then heat sealed with pierceable foil sheets. To amplify the target DNA, the droplet emulsions were thermally cycled using the following protocol: initial denaturation at 95°C for 10 min, followed by 40 cycles of 94°C for 30 s and 60°C for 1 min. The fluorescence of each thermally cycled droplet was measured using the QX100 Droplet Reader. All measurements were performed in triplicate.Data analysis
The equation of the line fitting the correlation between amplicon size and fluorescence amplitude for the size standards was generated using Microsoft Excel (Redmond, WA) and applied to the measured fluorescence amplitude of each sequencing library to calculate amplicon size. The fastq data files produced by the MiSeq were imported to Sequencher (Gene Codes, Ann Arbor, MI) and aligned to the pET-23a plasmid sequence to generate a sequence alignment/map file (SAM). A perl script was used to count the length of each read pair by retrieving the number corresponding to the “TLEN” field of the SAM file. Only library molecules for which both paired-end reads passed the quality filter were included in the analysis.Results and discussion
The ability of the QuantiSize assay to combine quantification and size determination in a single ddPCR experiment is derived from a correlation that exists between the fluorescence amplitude of droplets and the size of amplicons within them. With standard ddPCR reagent concentrations, DNA amplification is eventually limited by the availability of dNTPs and inhibited by the presence of pyrophosphate (19, 20); thus long DNA templates, which consume more dNTPs and generate more pyrophosphate, will produce fewer products than short templates at the end point of a standard reaction. Because the final number of products generated within a droplet determines its level of fluorescence, the measured fluorescence amplitude of droplets containing short templates will be greater than that of droplets containing long templates. The QuantiSize assay exploits this fact to generate an equation relating fluorescence amplitude to amplicon size by using measurements of known size standards. The equation describing the relationship between fluorescence amplitude and amplicon size can be used to calculate the size of any unknown ddPCR template that shares common primer and probe binding sites with the size standards. Creating size standards that have primer and probe binding sites in common with DNA samples can be accomplished in a number of ways including cloning sample DNA into a vector and appending adapter sequences to both the sample DNA and size standards (21).
We created a set of size standards applicable to Illumina NGS libraries containing inserts ranging from 25 to 1000 base pairs flanked by adapter sequences compatible with the Illumina MiSeq platform. A pair of primers and a fluorescent TaqMan probe were designed to hybridize to the adapter sequences such that the length of each amplicon is 160 bp plus the length of the insert. As primers and probe are specific to the MiSeq adapter sequences, only adapter-ligated molecules that will be amplifiable on the MiSeq flow cell will be quantified.
A ddPCR experiment was performed with the aforementioned size standards in separate wells of a 96-well plate. Droplets containing the target (positive) increased in fluorescence following amplification of the target whereas droplets lacking the target (negative) remained at the background level of fluorescence (Figure 1A). The distribution of droplet amplitudes is consistent across most amplicon lengths, but the 760 and 860 bp amplicons show a broader distribution of amplitudes (Figure 1B). An inverse, linear correlation between amplicon size and mean fluorescence amplitude was observed (R2 = 0.99436) (Figure 1C). The equation describing this correlation allows for the calculation of amplicon size given a measured fluorescence amplitude. The slope of this equation provides a measure of the difference in mean fluorescence amplitude that is expected with a given difference in amplicon size. Maximizing the magnitude of this slope maximizes the resolution of size standards, which is advantageous for the purpose of determining the length of unknown amplicons more accurately. The size standards used for QuantiSize are highly analogous to the standards used in gel and capillary electrophoresis. The size of unknown DNA can be determined by visually comparing the fluorescence amplitude of the size references to that of the unknown DNA or by entering the fluorescence amplitude value into the equation describing the relationship between average fluorescence amplitude and amplicon size for the size standards.