*A.J.S. and S.A. contributed equally to this work
The polymerase chain reaction (PCR) is an enzymatic technique for the in vitro synthesis of targeted nucleic acid regions (1), and it is an integral method in most aspects of molecular biology and genetics. PCR relies on thermally stable DNA polymerase to synthesize DNA by adding nucleotides complementary to those on a single-stranded template. By repeated cycling through 3 key temperatures (usually 95°C, ~60°C, and 72°C), exponential amplification of a target template is obtained (2). Many molecular applications, including forensics and clinical diagnostics, are highly reliant on PCR, requiring consistent, accurate, and repeatable assays (3,4). In such instances, the generation of misleading or false results may have profound implications and, therefore, a high level of analytical accuracy is required.
The development of hot-start DNA polymerases was a significant PCR innovation. Hot-star t polymerases remain inactive until the reaction is heated to a temperature at which oligonucleotide primers can no longer anneal to the DNA template (5). This prevents non-specific amplification and primer-dimer formation, and allows reaction mixes to be prepared at room temperature without the negative impact of spurious polymerase activity (6,–10).
During establishment of an assay for measuring telomere length, we observed that several commercial hot-star t polymerases mediated formation of primer-dimers (11). Here, we expand these observations and describe a simple method for detecting polymerase activity prior to the initial heat denaturation step. In this assay, 2 different primer pairs (Table 1) are used, and each primer pair can form a 3-bp overlap capable of annealing at room temperature. If a DNA polymerase extends either pair of primers prior to thermal activation, an ~79-bp primer-dimer will be formed which acts as the template for amplification in subsequent cycles (Figure 1). However, an effective hot-start enzyme will not produce a primer-dimer, as the enzyme activation and subsequent PCR cycling temperatures are higher than the annealing temperature of the 3-bp overlap. No genomic DNA template is required in this simple assay format.
PCR was carried out in a Mastercycler pro thermal cycler (Eppendorf, Stevenage, UK). A total reaction volume of 30 ml containing 1 × buffer, MgCl2, dNTPs, hot-start polymerase, and primers (IDT, Singapore) was used, at the concentrations recommended by the manufacturers. A robust amplicon was used as a positive control for polymerase activity (Table 1) in a reaction containing ~30 ng of human genomic DNA (Figure 1C). The reactions were prepared at room temperature, and hot-start polymerase activity was assayed using touchdown PCR (12). The initial denaturation time and extension temperature were adjusted according to the enzyme suppliers’ specifications and are specified in Table 2. Cycling conditions consisted of denaturation at 95°C for 15 s, annealing for 15 s, and extension for 45 s. The initial annealing temperature was 65°C, and this was decreased by 1°C per cycle for 10 cycles, followed by 25 cycles at 55°C. A final extension was performed for 5 min.
In total, 17 hot-start DNA polymerase enzymes (Table 2) were assayed for activity prior to thermal denaturation using 2 primer pairs (Primer Sets 1 and 2), in addition to a positive control primer pair to confirm the overall polymerase activity in each preparation. The presence of a primer-dimer at ~79 bp indicates activity of the hot-start enzyme prior to thermal activation. The absence of a primer-dimer indicates that the enzyme was only active after thermal activation, as would be expected for a hot-start enzyme. Analysis using Primer Set 1 revealed a primer-dimer for 7 of the polymerases tested, and analysis using Primer Set 2 revealed a primer-dimer for 12 of the polymerases tested (Figure 1). Polymerases that produced a primer-dimer with Primer Set 1 also produced primer-dimers using Primer Set 2, suggesting that Primer Set 2 was more sensitive. The difference in detection sensitivity between primer pairs likely reflects the relative stability of the 3-bp overlap or the Tm of each primer, which may have relevance to amplification efficiency during the subsequent PCR cycles. These results were consistent over multiple assays (data not shown). We have not disclosed the performance of individual enzymes as our findings are based on testing of individual batches of the hot-start DNA polymerases, and we do not know if there is batch-to-batch variation for each manufacturer’s hot-start DNA polymerase.Table 2.
To ensure that the observed amplicon was indeed a primer-dimer, two experiments were performed. First, a manual hot-start PCR was carried out where Taq polymerase was added to aliquots from a PCR master mix at either 95°C or room temperature, prior to thermal cycling. The primer-dimer was only apparent when the polymerase was added at room temperature, which was indicative of polymerase activity prior to thermal activation (Figure 2, Lanes 1 and 2). Second, the origin of this band was verified by restriction digestion. The design of Primer Set 1 was such that formation of a true primer-dimer would generate a recognition site for EcoRI (Table 1). Digestion of the PCR product with EcoRI did indeed lead to cleavage (Figure 2, Lane 3), confirming formation of a true primer-dimer.
We found that many commercial enzyme preparations marketed as “hot-start” exhibit polymerase activity prior to thermal activation. PCR applications, par ticularly in forensic or molecular diagnostic settings, rely on accurate and consistent assays. The failure of many hot-start enzymes to perform as expected could, therefore, have quite profound implications. Although enzyme suppliers provide evidence of tests for polymerase activity in documents accompanying their products, there is no evidence that they test for polymerase activity prior to thermal activation. We provide here a simple and effective assay that can be used to screen hot-start polymerases for this undesirable property.Author contributions
A.S. carried out much of the experimental work and draf ted the manuscript. S.A. made initial observations, contributed to experimental work and editing of the manuscript. M.K. contributed to design of the study and editing of the manuscript.
This work was funded by the Carney Centre for Pharmacogenomics and University of Otago.
The authors declare no competing interests.
Address correspondence to Martin A. Kennedy, Department of Pathology & Carney Centre for Pharmacogenomics, University of Otago, Christchurch, PO Box 4345, Christchurch, New Zealand. E-mail: