Fig 1.
In ddPCR, a mix of sample and reagents is partitioned into a water-in-oil emulsion consisting of ~20,000 droplets with dilutions of template such that at least some droplets contain no copies of the target of interest. Following end-point PCR amplification, the droplets are screened for fluorescent signal. The outcome is binary, a droplet is either positive (containing target) or negative (no target), based on its fluorescent signal. Quantification of the target concentration is achieved by Poisson correction on the fraction of positive droplets to compensate for the fact that multiple copies of the target may be present in some droplets.
Fig 2.
Design and specificity of the Taqman assays representative for the exon 48–50 deletion transcripts in ddPCR analysis.
(A) Position of Taqman MGB assays that were designed to detect exon 51 skipping in a muscle cell culture from a DMD patient with a deletion of exons 48–50 (Δ48–50). Primers and exon spanning probes were positioned as indicated by the arrows and the red lines. (B) The specificity of both DMD ex51-52.2 and DMD ex47-52 Taqman assays was confirmed by ddPCR analysis using dystrophin cDNA constructs representative for Δ48–50 transcript fragments with or without exon 51. The left panel shows the droplet populations obtained using both assays. The ex51-52.2 assay generated positive droplets when the construct containing exon 51 was used as template, but not when the exon 51 lacking construct was the template. For the ex47-52 assay detection was vice versa, yielding positive droplets using the exon 51 lacking template. The red arrow (in left panel) indicates a cloud of droplets with low fluorescence, arising from partial binding of the Taqman probe used in this assay to exon 47 and/or 52 sequences present but not joined in the template containing exon 51. The right panel shows the concentration of the constructs that was calculated based on the number of positive droplets for both assays. Data is represented as mean with 95% confidence interval (CI) (right panel).
Fig 3.
Temperature gradient ddPCR for both DMD ex51-52.2 and DMD ex47-52 assays.
(A) DMD ex51-52.2 assay and (B) DMD ex47-52 assay. The left panels show how the positive droplet populations change with decreasing annealing temperatures. At higher annealing temperatures the overall fluorescence of the positive droplet population (in blue) is lower, caused by a less efficient PCR amplification, and consequently the separation of the positive and negative droplet (in grey) clouds is less pronounced. In the right panels the corresponding concentrations of the constructs along the temperature gradient are represented. The %CV (coefficient of variation) between the different annealing temperatures is 3–4%, indicating a highly consistent quantification of the construct concentration, even at suboptimal amplification conditions at then higher annealing temperatures. Data is represented as mean with 95% CI (right panel).
Fig 4.
Dilution linearity and reproducibility of ddPCR results.
(A) ddPCR analysis was performed on a serial dilution of dystrophin cDNA templates representative for the Δ48–50, with or without exon 51, using Taqman assays with a probe detecting the exon 51–52 junction (DMD ex51-52.2) or exon 47–52 junction (DMD ex47-52) respectively. Concentrations ranged from >5,000 copies/μl down to <1 copy/μl. Dilution linearity is indicated by the coefficient of determination (R2). (B) ddPCR analysis of 8 technical replicates at different template copy numbers (from 1 to 1,000 copies/μl) using the DMD ex51-52.2 and DMD ex47-52 Taqman assays. Merged data of 8 replicates is indicated by solid markers. Data is represented as mean with 95% CI.
Table 1.
Quantification of ratios of exon 51 lacking/containing constructs using different PCR methods.
Fig 5.
Comparison different PCR methods and impact on quantification.
Graphic representation of the quantification of dystrophin cDNA construct templates lacking or containing exon 51 mixed in different ratios ranging from 0% to 100% using nested PCR (40, 50 and 60 cycles), qPCR (LinRegPCR and Pfaffl methods) and ddPCR. The dashed red line indicates the theoretical template ratios. The solid black ddPCR line is almost indistinguishable from the theoretical line and both lines are superimposed. Data is based on one experiment using 2 replicates for ddPCR and 3 replicates for qPCR.
Fig 6.
Comparison of different PCR methods on patient-derived muscle cell cultures.
Quantification of exon 51 skip levels in patient-derived muscle cell cultures (Δ48–50), transfected with an exon 51 skipping AON (400nM) by (A) ddPCR analysis, (B) qPCR analysis and (C) conventional primary PCR (20/30/40 cycles) and nested PCR (44 cycles). (D) Low background levels of exon 51 skipping were detected by ddPCR and qPCR in the non-treated (NT) samples. N/a indicates that no exon skip percentage could be calculated. Percentages shown are mean (n = 3) with SEM.
Fig 7.
Taqman assay for 4 additional deletions.
The specificity of Taqman assays with a probe detecting the exon junctions with (DMD ex44-52.3/DMD ex48-52/DMD ex49-52.2/DMD ex51-53) and without (DMD ex51-52.2/DMD ex51-53) exon 51 skip for four additional relatively frequent deletions amenable to exon 51 skipping (Δ45–50, Δ49–50, Δ50, and Δ52) was confirmed on representative dystrophin cDNA constructs with or without exon 51 (A). Exon 51 skip quantification in MyoD converted patient fibroblasts (Δ45–50 and Δ52) with and without ex51 AON treatment in triplicate (B). Merged data of 3 replicates is indicated by solid markers. Data is represented as mean with 95% CI (dark grey error bars: poisson CI; light grey error bars: total CI).