Improving Quantitative Power in Digital PCR through Digital High-Resolution Melting

Abstract

Applying digital PCR (dPCR) technology to challenging clinical and industrial detection tasks has become more prevalent because of its capability for absolute quantification and rare target detection. However, practices learned from quantitative PCR (qPCR) that promote assay robustness and wide-ranging utility are not readily applied in dPCR. These include internal amplification controls to account for false-negative reactions and amplicon high-resolution melt (HRM) analysis to distinguish true positives from false positives. Incorporation of internal amplification controls in dPCR is challenging because of the limited fluorescence channels available on most machines, and the application of HRM analysis is hindered by the separation of heating and imaging functions on most dPCR systems. We use a custom digital HRM platform to assess the utility of HRM-based approaches for mitigation of false positives and false negatives in dPCR. We show that detection of an exogenous internal control using dHRM analysis reduces the inclusion of false-negative partitions, changing the calculated DNA concentration up to 52%. The integration of dHRM analysis enables classification of partitions that would otherwise be considered ambiguous "rain," which accounts for up to ∼3% and ∼10% of partitions in intercalating dye and hydrolysis probe dPCR, respectively. We focused on developing an internal control method that would be compatible with broad-based microbial detection in dPCR-dHRM. Our approach can be applied to a number of DNA detection methods including microbial profiling and may advance the utility of dPCR in clinical applications where accurate quantification is imperative.

Keywords: dPCR; high-resolution melt; internal control.

FIG 1

Assessment of cross talk between fluorophores of different detection modalities. (a) qPCR verification of fluorophore detection in the FAM channel. Reaction conditions are indicated according to the color legend. (b) qPCR results are shown for the same three conditions in the Cy5 channel. (c to e) Digital PCR verification for fluorescent cross talk is shown. Individual chips are shown for the intercalating dye only reaction, probe-only reaction, and the combined condition (with both FAM [EvaGreen] and probe loaded). Channels are as indicated.

FIG 2

Comparison of dPCR performance using probe or intercalating dye-based detection. (a) dPCR rain plots for four different theoretical target DNA concentrations using intercalating dye-based detection. Fluorescence intensity is plotted from a triplicate series of chips for each target DNA concentration with both intercalating dye (EvaGreen) and probe in the reaction mixture. Fluorescence thresholds for the FAM channel were calculated using the definetherain algorithm. Blue, positive reactions; gray, negative reactions; red, rain. (b) dPCR rain plots for 4 different concentrations using probe-based detection. Fluorescence was determined from the same chips used for analysis in the experiment described in panel a. Fluorescence thresholds were calculated for the Cy5 channel using the definetherain algorithm. Blue, positive reactions; gray, negative reactions; red, rain. (c) The percent rain for each concentration is shown. A two-way analysis of variance was performed with a Bonferroni posttest, and the most variation was observed between the FAM and Cy5 channels, which were found to be statistically significantly different (P < 0.001). (d) The peak resolution for each concentration was calculated for both intercalating dye detection and probe detection. The peak resolutions between the two detection methods are statistically significantly different when analyzed using a two-way analysis of variance with a Bonferroni posttest (P < 0.001).

FIG 3

Positive reactions determined by each dPCR detection method for an example of each target DNA concentration. Positive melt curves for example chips under the theoretical conditions (number of genomes/reaction mixture) indicated on the left side of the figure are shown. (a) Positive reactions determined by probe detection. (b) Positive reactions determined by dPCR (intercalating dye) detection. (c) Positive reactions determined dPCR (intercalating dye) but determined to be negative by dHRM analysis. (d) Positive reactions determined by concurrent dPCR and dHRM analysis.

FIG 4

Performance comparison between hydrolysis probe, dPCR, and dPCR-dHRM. Truth tables are shown to compare each detection method used from the gold standard method. In this test, we consider the combined dPCR and dHRM detection method to be the absolute truth based on previous literature and our results shown in Fig. 3. True positives (TP), true negatives (TN), false positives (FP), and false negatives (FN) are quantified and compared between detection methods.

FIG 5

Quantification of positives classified by three methods. Four theoretical DNA concentrations were tested for all three detection methods. The measured DNA concentrations are plotted against the theoretical concentrations. Probe detection and dPCR-dHRM showed the most similar coefficients of variance. A one-way analysis of variance and a Bonferroni posttest were conducted to determine statistical significance between methods. The test results showed no statistically significant differences.

FIG 6

dPCR-dHRM of IC and target DNA. (a) The melt progression of a chip loaded with both primer pairs, internal control template, and target DNA template. Corresponding temperatures are indicated at the bottom of each image. At increasing temperatures, fluorescence decreases as the internal control sequence denatures, leaving only target DNA. As the temperature continues to increase beyond 100°C, there is another decrease in fluorescence as the target DNA denatures. (b) Fluorescence loss and corresponding melt curves.

FIG 7

Optimization of IC primer concentration in dPCR-dHRM. (a) Example melt curves of one chip with target DNA and an internal control are shown at increasing IC primer concentrations from left to right. IC template and target DNA template are held at constant concentrations. (b) Quantitative comparison of six IC primer concentrations. Blue bars indicate the loading ratio calculated for each chip and is defined as the ratio of amplified (IC positive) wells to loaded wells. A one-way analysis of variance with a Bonferroni multiple-comparison test was used to determine statistical significance of loading ratio and target DNA concentration between IC primer concentrations. Loading ratios of IC primer conditions at 0.01 and 0.005 μM are statistically significant from the highest IC primer condition (P values of <0.05 and <0.01, respectively). Red bars indicate the calculated concentration of targets. Only the 0.04 μM IC primer condition had a statistically significant concentration difference from the 0 μM IC condition (P < 0.05). Results for all other conditions were nonsignificant.

FIG 8

Change in calculated concentration using three methods. The target DNA concentration is calculated for four different IC primer concentrations using three methods with equation 1. The methods are the following: consideration of all partitions, consideration of loaded partitions only, and consideration of amplified partitions only. Paired t tests were performed between all pairs of methods to determine statistical significance. The t tests showed that the only comparison with statistically significant differences was the consideration of all partitions versus amplified partitions.