Understanding Variability in DNA Amplification Reactions

PCR (1) is arguably the most common technique in molecular biology. As such, it has significantly impacted research and development in fields such as biochemistry, medicine, bioengineering and beyond. Success or failure in PCR is influenced by a myriad of factors including primer design, cycling conditions, and the quality and concentration of reaction substrates and solutions. By understanding the interplay of these variables, PCRbased tools and techniques are bolstered, and difficult amplifications become routine.

The range of variables that impact DNA amplification reactions is masked by the relatively high success rate of PCR experiments. It is only after unsuccessful amplification attempts that these variables become evident. Ideally, various components can be altered to achieve success and even to favor a desired outcome (e.g., specificity over yield, sensitivity over specificity, etc.). However, practical issues, such as very high (or low) template GC content, the presence of inhibitors, limitations of primers, source materials or time, can limit the ability either to follow optimized PCR guidelines or to systematically evaluate a sufficient number of variables to ensure success.

The PCR Test Panel:

With the long-standing goal of enabling the research of our customers and our own scientists, NEB continues to devote resources both to basic DNA polymerase research and to the development of new products for DNA amplification. As part of these efforts, we have created a quantitative, microfluidic-based, PCR Test Panel. The goals of the PCR Test Panel are to:

  1. Systematically manipulate the numerous variables present in an amplification reaction.
  2. Understand the contributions of each variable to the desired signal-to-noise outcome.
  3. Provide information to enable researchers to modify a limited set of variables depending on the desired outcome (or inherent limitation) of their particular set of amplifications.
  4. Facilitate the development of enhanced tools (e.g., polymerases, buffers, etc.) that will broaden the definition of "routine PCR" to encompass situations that are currently challenging.

Use of the PCR Test Panel is simplified by a microfluidic, agarose gel-based mimic (LabChip® GX platform, Caliper Life Sciences, Hopkinton, MA) that allows rapid end-point quantitation of amplification reactions in a 96-well plate format. A supporting, internal database has been designed to link the results from the microfluidic analysis (e.g., yield and purity of each expected product) to the detailed contents of each well (e.g., identity and concentration of polymerase, template, buffer, additives, etc.). Cycling conditions, thermocycler ID and other relevant details are also tracked.

Table 1: Examples of Test Panel Variables

By varying large numbers of conditions and quantitating both specificity and yield of a PCR product, the contribution of each reaction component can be systematically evaluated. Some of the variations in template and primer components that are assessed by the PCR Test Panel are described in Table 1. Importantly, this system is easily adapted and can be scaled to assess new components and conditions as they arise.

In practice, the PCR Test Panel involves a typical PCR workflow from experimental planning to data analysis, with crucial steps of database integration. The workflow is comprised of the following steps:

  1. Design PCR experiment and capture relevant details in tracking database.
  2. Run PCR experiments in triplicate (various module/experiment combinations are utilized to make 96-well plates).
  3. Run 96-well plates through quantitative microfluidic analysis.
  4. Export results from microfluidic platform into database to link quantitation with experimental details.

Experiments are performed in triplicate and relevant information about the contents of each well is captured in the database. The data tracking system communicates the expected product sizes for each well/reaction to the microfluidic system. The output of microfluidic analysis is an electropherogram, which for familiarity is also represented as a typical "gel-like" image (Figure 1). The instrument software identifies the expected peaks and quantitates the yield and percent purity of each (described in more detail below). Finally, results from the analysis are imported into the database and connected with well contents and additional experimental information, forming the foundation of an expanding set of data of PCR results. Simple data analysis can be accomplished within the database, and IGOR Pro (Wavemetrics, Inc., Portland, OR) is used for more complex analysis and graphical display.

Figure 1:  Electropherogram and interpreted gel image of primary PCR Test Panel data.

The electropherogram (left) shows overlaid results from two samples (G1, G4) selected from the interpreted gel (right). Expected peaks (EP) are identified by the software from information provided by the database. For product quantitation and size assessment, a DNA ladder is run after every 12 samples and upper and lower markers (UM, LM) are mixed 1:1 with each sample immediately prior to analysis.

Advances in Research and Development as a result of the PCR Test Panel:

The PCR Test Panel has furthered our understanding of many aspects of amplification reactions, one example being the best use for Lambda PCR tests. The functional tests of many commercially available polymerases employ Lambda as a substrate. However, the Lambda amplification reactions that were part of the PCR Test Panel were so consistently robust that they did not serve as good indicators for how a polymerase would perform with real world, complex templates, even when matched for GC content. Interestingly, at typical template concentrations for Lambda, an increase in enzyme concentration resulted in a linear increase in the yield of amplified product (akin to enzyme titer experiments). A similar, linear response was not observed at working concentrations for more complex genomic targets (≤1 kb). Instead, high variation in yield was detected for individual amplicons in response to increased enzyme concentration in the reaction. These results demonstrate that Lambda can be used as an appropriate template for product quality control assays, whereas "real-world" templates are better suited for optimizing product use recommendations.

Recently, the PCR Test Panel was used in the development of the OneTaq line of amplification products. Extensive testing was used not only to guide formulations and eventual usage recommendations, but also to assess performance relative to a broad field of competitors. These comparisons demonstrated that OneTaq DNA polymerases and buffers performed well in "routine" amplifications, but unlike many of their peers, continued to offer robust performance with more difficult amplicons, such as those with high GC or AT content. Figure 2 shows the gel-like image of triplicate reactions with OneTaq Hot Start DNA Polymerase (with and without the High GC Enhancer) and six other hot start polymerases (plus competitor's GC enhancers, where provided) on a single high GC amplicon (68%). This type of output provides familiar visual information of both product yield and purity. However, the quantitative primary output of microfluidic analysis (yield as ng/µl and % product purity as a function of the total signal in each reaction lane) allows a more condensed view of the data, which is more compatible with the scale of the PCR Test Panel.

In Figure 3, the triplicate reactions shown in Figure 2 were condensed into a single panel (boxed area in Figure 3) and are shown as part of a series of reactions with high GC (human genomic amplicons) ranging from 66%-80% GC content. One routine amplicon (55% GC) was also included as a control. Amplifications were all performed according to each manufacturer's specific recommendations. Percent product purity is represented as circles in Panel A. Product yield is expressed as a bar chart in Panel B. Reactions were set up in the absence (solid bars) and presence (striped bars) of GC enhancers, where provided by the manufacturer. Percent purity scores account for contributions from primer-dimer formation and any other non-specific products formed by the reaction. Low purity scores can arise from a prominent secondary product, from a smear of non-specific products, or from significant primer dimer interference (compare boxed area in Figure 3 to Figure 2 which shows the gel-like representation of the same data).

Although it would be convenient to suggest a single solution to all PCR difficulties, the reality is more complex and nuanced. For example, considering only GC content, we have observed that the percent GC content of a template is not always an adequate predictor of reaction difficulty, nor could it alone define the need for a GC-specific buffer or enhancer. One striking example can be seen in Figure 3, where a 73% GC amplicon could only be robustly amplified using OneTaq in GC Buffer with the High GC Enhancer, whereas an 80% GC amplicon did not require the Enhancer. Trends that emerge from studies of this scale help inform product guidelines for a broad range of product applications. For OneTaq DNA Polymerase, these include guidelines for increasing enzyme concentration when amplifying products over 3 kb and the use of a 68°C extension temperature.

Figure 2: Gel view of triplicate reactions on a single 68% GC amplicon for OneTaq Hot Start DNA Polymerase and hot start competitors.

A familiar gel-like output provides visual information about product yield and purity. Data from these triplicate reactions were averaged to create the single panel (boxed area) of the graphs shown in Figure 3. Amplifications were all performed according to each manufacturer's specific recommendations.

Figure 3: Examples of data from the PCR Test Panel: Comparison of OneTaq Hot Start DNA Polymerase to other commercially available hot start polymerases on high GC amplicons.

Reactions containing high GC human genomic DNA templates were set up at room temperature. PCR experiments included 30 cycles. Purity (A) and Yield (B) were calculated via microfluidic analysis from triplicate reactions. OneTaq DNA polymerase was used in the absence (brown solid bar) or presence (brown striped bar) of High GC Enhancer. Competitor polymerases were cycled according to manufacturer's recommendations and included GC enhancers when supplied (striped bars).

Conclusion:

Using a wide variety of conditions and a multitude of primer/template sets has allowed NEB to develop a comprehensive view of its PCR products (and others in the market) that is not possible with a smaller number of amplification reactions. Although it is interesting to study all the variables that affect PCR, we also realize that successful amplification of a desired target is what is important to our customers. The scope and scale of the PCR Test Panel has allowed us to see past the natural variability produced by the rugged PCR landscape, where a single amplicon or reaction condition can be found to prove nearly any point desired. Data at this scale serves as a vantage point to build upon the strengths of various polymerases, buffers and protocols to expand the useful range of PCR conditions and the definition of what is routine. It allows researchers at NEB to develop solutions for challenging amplifications and support future polymerase-based applications.

References

  1. Saiki, et al., (1988) Science 239, 487-491

LabChip® is a registered trademark of Caliper Life Science.
From NEB Expressions Winter 2011 By Nicole M. Nichols, Ph.D., New England Biolab