Nucleic acid techniques

Polymerase chain reaction

The PCR was the first target or sequence amplification method described and revolutionized the molecular diagnostics field. Using PCR, the amount of target nucleic acid present in the sample is increased by synthetic in vitro methods. Of all amplification technologies to date, PCR is the most versatile, most user friendly, and most widely applied in both the research and clinical settings. The commercial availability of thermostable DNA polymerases, reagent kits, and instrumentation has made this method routine in the clinical laboratory with applications that span genetic disease, infectious disease, and oncology testing.

PCR uses successive cycles whereby a DNA polymerase copies target DNA sequences from the input sample template DNA. The amplification products or amplicons of each cycle provide new templates for the next round of amplification, thus increasing the concentration of the target DNA sequence exponentially. PCR requires (1) a thermostable DNA polymerase, (2) deoxynucleotides of each base (collectively referred to as dNTPs ), (3) the target sequence, (4) a pair (forward and reverse) of oligonucleotides primers, that are complementary to opposite strands and that define the target sequence, and (5) reaction buffer containing the appropriate concentration of cofactors such as MgCl ₂ .

Each of the components of the PCR plays a significant role in the success of the amplification reaction. Target sequences can be amplified from isolated genomic DNA or complementary DNA (cDNA) derived from an RNA sample prepared from various sample types. In clinical use, PCR typically amplifies short DNA sequences (70 to 200 bp) making this technology very applicable to nucleic acids isolated from formalin-fixed, FFPE tissues. However, larger targets, even up to 25 kb or more can be amplified by PCR if the target is intact and enough time is provided for polymerase extension.

A DNA polymerase enzyme is required for DNA synthesis during the primer extension step. Taq DNA polymerase, isolated from Thermus aquaticus , is commonly used. Taq polymerase exhibits 5′→3′ polymerase activity, 5′→3′ exonuclease activity, thermostability, and optimum performance at 70 to 80 °C. The half-life of Taq activity at 95 °C is approximately 40 to 60 minutes, and extremely high denaturation temperatures (>97 °C) will significantly reduce its active lifetime. PCR is carried out in a buffer containing divalent cations in the form of Mg ²⁺ . Lower Mg ²⁺ concentrations decreases the rate of dissociation of enzyme from template by stabilizing the enzyme-nucleic acid interaction. While Taq DNA polymerase is ideal for routine PCR, there are several other thermostable DNA polymerases with unique qualities which make them useful for special PCR applications such as amplification of long pieces of DNA or high-fidelity amplification.

The amplified target sequence and specificity of the PCR is determined by the primers which flank the target sequence. Several parameters need to be considered when designing a set of primers for PCR as the choice of primers often dictates the quality and success of the amplification reaction. To select primers, the sequence of the target must be known. Below are some guidelines for primer selection/design:

• 1.

  The percentage of G or C bases in a primer (GC content) is a major factor determining the primer annealing temperature in PCR that is usually set between 50 and 75 °C. Choose primers with a GC content near 50% to favor a greater diversity of sequence that tends to result in better specificity.

• 2.

  Typically, choose primers between 18 and 25 bases that are matched in melting temperature (Tm) to each other. A random primer greater than 17 bases long has a good chance of being unique in the human genome.

• 3.

  Make sure the primers anneal to opposite DNA strands of your target with their 3′-ends directed toward each other. The shorter the distance between the primers, the smaller the amplicon and the easier it is to amplify with high efficiency. By limiting the extension time, shorter products are amplified with greater efficiency than longer products and can result in preferential amplification; short extension periods selectively amplify shorter products because longer products do not have the time to fully extend.

• 4.

  Avoid primers that anneal to themselves or to other primers and particularly avoid complementation at the 3′end of primers.

• 5.

  Choose primers that are specific to the target. Avoid simple sequence repeats and common repeated sequences, such as Alu repeats. If your target has close relatives, design your primers so that they will anneal only to your intended target. Targets that need to be avoided include pseudogenes (for genomic DNA) and related bacterial or viral strains (for microorganisms).

• 6.

  Avoid primers that have sequence complementarity to internal sequences of the intended product, especially at the 3′-ends of the primers.

• 7.

  It is important to avoid sequences similar to your primers that are present in the background DNA likely to be present in your assay. The National Center for Biotechnology Information (NCBI) Basic Local Alignment Search Tool (BLAST) is commonly used to search for similar sequences.

• 8.

  Mismatches are better tolerated at the 5′-end rather than the 3′-end of primers. Mismatches at the 3′-end base are most discriminatory and can be used for allele-specific amplification. For efficient amplification, try to completely match at least the first 6 bases at the 3′-end that are important for polymerase binding. Variation at the 5′-end is better tolerated and can be used to introduce restriction sites as long as Tm differences are considered. Primers can also include 5′-tails that are not homologous to the target for subsequent amplification or detection.

Many primer selection programs are available commercially and others can be freely obtained over the Internet. The most useful is Primer-BLAST at The National Center for Biotechnology Information (NCBI).

The reaction buffer for PCR contains salt, magnesium, and each of four deoxyribonucleotide triphosphates (dNTPs). A common PCR buffer used with Taq polymerase consists of up to 50 mM KCl, 1 to 5 mM MgCl ₂ , 10 to 50 mM Tris-Cl (pH 8.3), and 50 to 200 μM dNTPs. Excessive or limiting concentrations of any of these components in the reaction buffer can result in either a failed amplification or nonspecific amplification. However, optimal component concentrations depend on the polymerase used. For example, N-terminal deletion mutants of Taq lack 5′- to 3′-exonuclease activity and are most active without any KCl, whereas full-length Taq is optimal with 50 mM KCl. The best MgCl ₂ concentration for a given PCR must be determined empirically and also depends on the polymerase. Higher concentrations of dNTPs (or MgCl ₂ ) can lead to errors related to dNTP misincorporation by Taq polymerase and should be avoided. In order to improve the efficiency of the PCR, some additives may be included in the reaction mixture such as gelatin or Bovine Serum Albumin (BSA) at concentrations up to 500 μg/mL that help to protect the polymerase from surface inactivation and concentrate the nucleic acid. Dimethylsulfoxide (DMSO), betaine, glycerol, dimethyl formamide (DMF), formamide, or urea may be included when the target sequence for a PCR is known to be of high GC content as these help to lower the denaturation temperature required for PCR.

Most traditional PCR assays include repetitive cycles of a three-step process: denaturation, annealing, and extension or synthesis. First, the isolated DNA including target is denatured into single strands by heat ( Fig. 64.3 ). Denaturation destabilizes the DNA double helix to generate two separate single strands. Second, annealing or hybridization of the primers to their complementary sequences occurs when the mixture is cooled. The primers which flank the target sequence are provided in great excess (usually over a million times the concentration of the initial target) and are used in each cycle of the PCR. Third, after the primers are annealed, the polymerase synthesizes two new DNA strands by extending each of the two primers at their 3′ end and using the sample DNA as a template. The primers are designed to recognize sequences of the target that are close enough together so that the polymerase can extend each strand far enough to include the priming site of the other primer giving rise to an amplified target sequence or amplicon. One cycle of a PCR includes ramping up and down between temperatures so that each step can be accomplished with optimal results. The second PCR cycle also begins with denaturation, but now twice as many strands (the original genomic DNA and the extension products from the first cycle) are available for primer annealing and subsequent extension. Most PCR assays use a three-step cycling process defined by three different temperatures: (1) a high temperature sufficient to denature the target sequence (90 to 97 °C); (2) a low temperature that allows annealing of the primers to the target (50 to 65 °C); and (3) a third temperature optimal for polymerase extension (65 to 75 °C). A more complete schematic of PCR is shown in Fig. 64.4 , detailing why only amplicons of defined length are generated. The instrument that takes samples through the multiple temperature steps is typically a programmable heat block and is known as a thermocycler.

Repetitive thermocycling results in the exponential accumulation of the short amplicon product consisting of primers and all intervening sequences. If the efficiency of each cycle is perfect, the number of target sequences doubles each cycle (the efficiency is 100% or 1.0). PCR efficiency depends on the primers, the temperature-cycling conditions, and presence of any inhibitors that might limit amplification. Amplicons accumulate exponentially in the beginning cycles of PCR. At some point, however, the efficiency of amplification falls and eventually the amount of product plateaus ( Fig. 64.5 ) as the result of exhaustion of components or competition between primer and product annealing (i.e., the single strands of product are at such high concentrations that they anneal to each other rather than to the primers). In a typical PCR using 0.5 μmol/L of each primer, the maximum DNA concentration achievable is about 100 billion copies/μL.

Using traditional or end point PCR requires a final detection step to identify the amplicon. Gel electrophoresis with ethidium bromide staining is a classical method that separates products by size and may suffice for many qualitative applications. For discrimination of amplicons that differ by a single base or several bases, one of the primers can be fluorescently labeled and the post-PCR amplicons separated by capillary electrophoresis on instruments that are typically used for conventional dideoxy-termination sequencing. Alternatively, some form of hybridization assay can be used to verify or analyze the amplified product. Automated methods are always attractive, and closed-tube methods, where amplified products are never exposed to the open environment are particularly advantageous to avoid contamination of future reactions with the products of a prior reaction. Although there are chemical and enzymatic ways to prevent such “carry-over” contamination, closed systems provide a simpler solution. Adding a fluorescent dye or probe before amplification allows optical monitoring to follow the reaction as it progresses (real-time PCR) or after the reaction is complete (end point melting) without the need to process the sample for a separate analysis step.

Polymerase chain reaction kinetics

Conceptually and programmatically, the three-step multicycle process of PCR occurs at three different temperatures for three different times. Standard thermocyclers that use conical tubes focus on accurate temperature control of the heating block at equilibrium, not on the dynamic control of the sample temperature. As a result, sample temperatures are not well defined during transitions, and long cycle times have become standard to ensure that samples reach optimal temperatures. In addition, reproducibility between instruments and manufacturers is compromised, and PCR may require an hour or more to complete 30 cycles of amplification.

The kinetics of denaturation, annealing, and extension suggest that rapid transitions between temperatures (short heating and cooling times) with minimal or no pauses (temperature plateaus) may be optimal ( Fig. 64.6 ). The use of temperature “spikes” at denaturation and annealing, instead of extended temperature plateaus, allows for rapid cycling with the appropriate instrumentation. The actual time required for PCR depends on the size of the product, but when it is less than 500 base pairs (bp), 30 cycles are easily completed within 15 minutes. Furthermore, rapid amplification improves specificity. Fig. 64.7 shows PCR amplification of the same product amplified at different cycling speeds. With conventional cycling, many nonspecific products are generated (see Fig. 64.7 A). These products disappear as the cycling time is decreased (see Fig. 64.7 B–D). In fact, amplification yield and product specificity are optimal when denaturation and annealing times are minimal. However, cycling too rapidly can compromise PCR efficiency that is critical for quantitative PCR.

Initial denaturation of genomic DNA may be required before PCR cycling, depending on how the DNA sample was prepared. Prior boiling of the sample or an initial denaturation step of a few seconds may be necessary to denature long strands of genomic DNA or those with high GC content. During PCR, however, denaturation of the shorter products occurs very rapidly. Even for long PCR products, denaturation is complete in less than 1 second after the denaturation temperature is reached. Anything greater than a denaturation time of “0” serves only to degrade the polymerase. If longer denaturation times are required, either the sample is not reaching temperature or heat-activated polymerases are being used. Product specificity is optimal when annealing times are less than 1 second. The required extension time for each cycle depends on the length of the PCR product. Extension is not instantaneous, although it is much faster than common practice would suggest. Extension rates of polymerases under in vitro conditions are 50 to 100 bases per second. Indeed, by increasing the primer and polymerase concentrations 10- to 20-fold, a 60-bp fragment of genomic DNA can be amplified with good efficiency, yield, and specificity in less than 15 seconds. Such methodologies are not currently available on commercial instrumentation.

PCR amplicons can be as small as 40 bp to over 40 kilobases (kb). To amplify products longer than 2 kb, mixtures of polymerases that include some 3′-exonuclease activity to edit mismatched nucleotides are usually used. Instead of separate annealing and extension temperatures, both processes can be carried out at the same temperature, resulting in two-temperature, instead of three-temperature, cycling. Taq DNA polymerase has a terminal transferase activity that may add a single unpaired nucleotide to the 3′-end of the amplicons. In the presence of all four dNTPs, dATP is preferentially added. This means that many of the double-stranded products generated by PCR will have a protruding “A” at one or both ends. Although this does not influence most detection protocols, it can complicate analysis of small amplicons or when high size resolution is needed. On the other hand, this feature can be useful for ligation and library construction in MPS.

Polymerase chain reaction optimization

Optimization of a PCR is accomplished by experimentation to assess reagent concentrations and conditions for thermocycling. Annealing temperatures can impact PCR efficiency dramatically and should be determined for each primer set. In addition to temperature-cycling conditions, the specificity of PCR depends on the choice of primers and the Mg ²⁺ concentration. Mg ²⁺ is a polymerase cofactor and also stabilizes the double helix. Titration experiments with MgCl ₂ can impact results of a PCR as low concentrations of Mg ²⁺ favor specificity, while high concentrations favor sensitivity but may result in undesired amplification of alternative products. Typical Mg ²⁺ concentrations in legacy PCR are rather limited (1.5 to 2.5 mmol/L), although greater concentrations may be needed to offset chelation of Mg ²⁺ by dNTPs, ethylenediaminetetraacetic acid (EDTA), or citrate in the sample. Furthermore, higher Mg ²⁺ concentrations are optimal with cycle speeds below 1 second.

PCR sensitivity and specificity can be compromised by the formation of unintended low molecular weight artifacts. This process is initiated before PCR, when the primers, template, and polymerase are all together at temperatures below the annealing temperature of the PCR primers. At low temperatures, if a primer momentarily anneals to another primer or to an undesired region, the polymerase may extend the complex. If the extension product, in turn, is primed and extended, then unintended, double-stranded products can be formed (e.g., primer-dimers) that serve as amplification templates throughout the reaction. Primer-dimers lower the efficiency of the intended amplification and decrease assay sensitivity but can usually be distinguished from the intended target by their small size, and low Tm.

The formation of primer-dimers can be minimized in several ways which limit the activity of polymerase until the temperature is increased, a strategy that is often collectively referred to as hot start . One method of hot start involves the use of antibody (or an aptamer) to bind and inactivate the polymerase at room temperature. The binding agent is released upon heating, allowing polymerase activation. Another method uses wax or paraffin to create a physical barrier between the essential components in the reaction. Finally the polymerase, primers, or dNTPs can be chemically blocked so that extension cannot occur until activation by heat, usually requiring an initial denaturation period of 2 to 20 minutes. While many laboratories rely on commercially available universal master mixes and calculated annealing temperatures for a primer set, these variables should always be a part of assay development and troubleshooting efforts in the lab.

Polymerase chain reaction contamination, inhibition, and controls

Because PCR typically produces about one trillion copies of the target sequence, a small amount of previously amplified product unintentionally introduced into a new sample can easily produce a false-positive result. Thus minute aerosol droplets can contain more than enough target for robust amplification. PCR products can contaminate: (1) reagents, (2) pipettes, (3) glassware, and (4) lab space. Simple laboratory precautions can minimize contamination by using (1) physically separated areas for pre-amplification and post-amplification steps, (2) positive-displacement or filter barrier pipettes to minimize aerosol contamination, and (3) prior preparation and storage of individual or combined (“master mix”) reaction components in small aliquots. Chemical methods can prevent contamination from affecting future reactions. These include substituting “U” for “T” in PCR with subsequent degradation of the incorporated “U” by uracil N-glycosylase or irradiation of DNA product with ultraviolet light. The most effective way of preventing contamination is to contain the product within a closed tube or vessel as in real-time PCR. Even with contamination precautions, a negative control/contamination control/blank (all reactants except target DNA) is one of the most important controls for PCR.

One advantage of PCR is that it does not require highly purified nucleic acid to achieve a successful amplification. In practice, however, clinical samples may contain unpredictable amounts of impurities that can inhibit polymerase activity. Consequently, to ensure reliable amplification in clinical assays, some form of nucleic acid purification is often used (see Chapter 63 ). The diverse nature of PCR inhibitors within clinical specimens requires demonstration that the sample (or preparation of nucleic acid purified from it) will allow amplification. Although such confirmation is automatic with genotyping assays, detection and quantification reactions typically use a control nucleic acid sequence different from the target that is added to the sample. Failure to amplify this positive control indicates that further purification of the sample is required to remove inhibitors. PCR may be inhibited by compounds that change amplicon or primer hybridization stability (salt or organics), chelating a crucial component with anticoagulants (Mg ²⁺ ), or polymerase inhibition by a variety of substances. When a PCR fails, the instinct is to use more of the extracted material. However, more starting material often only increases inhibition, while simply diluting the sample will result in successful amplification.

Theoretical detection limits of the polymerase chain reaction

When PCR is performed under optimal conditions, a single copy of the target can be detected. In practice, however, the statistical probability of distributing at least a single copy from a dilute template solution into the PCR must be considered. The Poisson distribution indicates that if, on average, one target copy is present per tube, 37% of the tubes will have no target, 37% will have one target, and the remainder will have more than one target. If there is an average of two copies per tube, approximately 14% of the tubes will have no template and will be false negatives. About three copies on average are necessary for 95% of the tubes to include at least one copy. Therefore the limit of detection (95% probability) of any single PCR cannot be lower than three copies per reaction. About five copies on average are necessary for 99% of the tubes to include at least one copy. This limitation of low copy analysis holds true for any amplification technique. An exception to this is digital PCR which analyzes thousands of reactions in parallel and can quantify less than 1 copy (on average) per reaction.

Reverse transcriptase-polymerase chain reaction

With the addition of a reverse transcriptase, RNA targets can be converted into complementary DNA (cDNA) and then successfully used as starting material in a PCR. There are several commercially available recombinant derivatives of viral reverse transcriptase enzymes used in standard reverse transcriptase-polymerase chain reaction (RT-PCR) applications including from the Moloney murine leukemia virus (MMLV) and the avian myeloblastosis virus (AMV). Reverse transcription and DNA amplification are most often catalyzed by two different polymerases. In one-step reverse RT-PCR, both enzymes are present in a common buffer, and typically PCR primers are used for both reverse transcription and DNA amplification. Some thermostable enzymes have both DNA polymerase and reverse transcriptase activities so that both steps can be performed in the same tube with the same enzyme. In two-step RT-PCR, the reverse transcription is performed first, usually with random hexamers or a poly-dT oligonucleotide (to prime the poly-A tail of most mRNA). After reverse transcription to cDNA, an aliquot of the cDNA is used in a PCR with specific primers to the target sequence.

Real-time polymerase chain reaction

Since the development of traditional end point PCR, many modifications have been made to the chemistry, instrumentation, and design approaches for different qualitative and quantitative applications. Of these, the development of real-time PCR is the most significant advancement in PCR-based testing. Real-time PCR combines the amplification step with simultaneous detection steps that do not require post-PCR manipulation and thus help to minimize any form of contamination. The presence of amplified products is directly monitored within the reaction vessel using fluorescently labeled primers, probes, or dyes and specialized thermocyclers equipped for fluorescent detection. The amplification process modifies fluorescence by a variety of methods. During real-time PCR, the amount of amplicon present in the reaction vessel is related to the amount of emitted fluorescence and the amount of initial target sequence. In addition, the number of PCR cycles needed to detect the target sequence is inversely proportional to the log of the target concentration in the initial sample. Because of this, real-time PCR reactions can also be quantitative. There are two types of detection chemistries for real-time PCR: (1) those which use intercalating DNA binding dyes such as SYBR green I, and (2) those which use various types of fluorescently labeled probes. Both will be discussed in detail later.

Real-time PCR quickly became the method of choice for most molecular diagnostic laboratories because of its increased sensitivity/specificity and turnaround times. These assays can be either qualitative or quantitative in the assessment of target sequences, as well as distinguish variant from wild type sequences. Another powerful feature of most real-time PCR instruments is the ability to perform melting curve analysis. Melting profiles of probes and amplicons can be obtained immediately after PCR by heating the sample in the same tube. Quantification of specific target sequences for both infectious disease and oncology applications is often performed by real-time PCR. By multiplexing the primers and probes for the target sequence with the primers and probes for a control sequence, accurate assessment of the target amounts can be made using either relative quantification or absolute quantification if external standards of known concentration are used to create a standard curve and determine target copy number. A significant advantage of real-time PCR is the turnaround time with which samples can be analyzed due to the elimination of post-PCR processing steps. Using melting curve analysis instead of gel electrophoresis results in a much faster method for analyzing PCR results.

Additional modifications to the polymerase chain reaction

Multiplex PCR is a powerful technique that enables amplification of two or more products in parallel in a single reaction tube. It employs different primer pairs in the same reaction for simultaneous amplification of multiple targets. It is widely used in applications where detection of different target sequences provides better clinical utility such as in oncology and infectious disease testing. Compared with standard PCR systems using only two primers, multiplex PCR requires extensive optimization of primer design and annealing conditions for optimal amplification of each amplicon while avoiding undesired PCR artifacts.

If PCR is performed with a pair of outer primers and then that amplicon is used as the target and amplified yet again with a set of inner primers, this is referred to as nested PCR . Typically, the first round PCR product is diluted 1000 to 1 million times before the inner (nested) primers are added. The advantage of nested PCR is that both sensitivity and specificity increase. The disadvantage is increased potential for contamination, particularly if the first round of PCR amplicons is handled manually for dilution and transfer.

Conventional PCR uses primers that are present in equal amounts, thereby ensuring that most of the products are double-stranded. Asymmetric PCR uses different concentrations of the two primers to generate more of one strand than of the other. For instance, the use of one primer at 0.5 μmol/L and the other at 0.005 μmol/L produces mostly single-stranded DNA. Yield of the product, however, may be low with this technique. With less extreme ratios (e.g., one primer at 0.5 μmol/L and the other at 0.2 μmol/L), the yield is mostly preserved, with one strand produced in enough excess to make it readily available for probe hybridization. One way to improve the efficiency of asymmetric PCR is to equalize the Tms of the excess and limiting primers. The lower concentration of the limiting primer results in a lower Tm than the excess primer. In linear after the exponential or “LATE-PCR,” the stability of the limiting primer is raised sufficiently (typically by making it longer) to counteract the effect of its lower concentration. As a result, the two primers have a comparable ability to bind the template during the initial cycles of PCR. After exponential amplification, linear amplification provides ample template for downstream hybridization assays.

Allele-specific PCR enables preferential amplification of one genetic allele over another by placing the 3′-end of one primer at the polymorphic site. The variant that is better matched to the primer extends more readily than the other allele. This strategy can be used to distinguish a gene from its pseudogenes and for genotyping of single-nucleotide variants. Allele-specific PCR can also be used to determine haplotypes. PCR can also be modified to enrich variants internal to the primer binding sites. Co-amplification at lower denaturation temperature PCR (COLD-PCR) enriches variants irrespective of their type or position within PCR products smaller than 200 bp. Mismatch-containing sequences have a slightly lower denaturation temperature than fully matched sequences. During PCR cycling ( Fig. 64.8 ) a product hybridization step after denaturation enables mutation-containing product strands to bind to wild-type strands, forming heteroduplexes . The temperature of this step is too high for primers to bind. Next, the temperature of the reaction is raised to a critical denaturation temperature (T _c ) that generates preferential denaturation of the mismatch-containing sequences. The temperature is then lowered to allow primers to bind leading to preferential replication of mutation-containing sequences. By repeating this protocol over several PCR cycles, mutation-containing strands are preferentially amplified over wild-type strands. With identification of the correct Tc to within 0.5°C, variant enrichment of 10- to 20-fold for Tm-decreasing (G:C to A:T) or Tm-retaining (e.g., T:A to A:T) variants and 5- to 10-fold for Tm-increasing (A:T to G:C) variants is typical. The enrichment increases if a second round of COLD-PCR is applied. Since during bisulfite treatment of DNA un-methylated sequences undergo several C>T changes that decrease Tm, COLD PCR may also be used for enriching un-methylated sequences in clinical samples. Among various downstream detection methods, high-resolution melting or sequencing are most often used.

Digital polymerase chain reaction

Conventional PCR averages the amplification results of many individual template molecules. Digital or single-molecule PCR is a technique that uses a dilute solution of template distributed across many reaction compartments or “partitions.” Each partition either has or does not have PCR template molecules to amplify. After PCR, the partitions are scored as either positive (one or more initial templates) or negative (no initial template) for a digital readout. Thousands or even millions of partitions are typically scored. Data are typically displayed on dot plots where each partition is a dot that is plotted on fluorescence versus partition count plots ( Fig. 64.9 ). Assay quality is judged by the separation of positive and negative partition clouds and depends on reaction efficiency and primer and probe concentrations. Poorly optimized reactions can produce “rain” from the positive cloud that can make discrimination from negative clouds difficult. The partitions may be aqueous PCR droplets in oil (droplet digital PCR or ddPCR) or micro-wells formed on chips by microfluidics. As in conventional PCR, fluorescence may be generated either by probes or dyes.

Digital PCR can identify and quantify rare sequence variants, precisely determine copy number changes, establish the concentration of PCR standards, and determine the haplotype of variants that are on the same PCR product. When properly performed, digital PCR does not require the standard curves routinely used in quantitative PCR. Digital PCR is less prone to background DNA competition because competing DNA is divided between positive and negative partitions. For example, if only 0.1% of the partitions are positive for a rare variant, 99.9% of the background DNA is in negative wells. The variant-to-background ratio is increased by a factor of 1000 in the positive wells and better amplification of the variant can be expected. Common PCR inhibitors may not be as apparent in digital PCR because a positive threshold is reached even under conditions of moderate inhibition that would affect bulk quantitative PCR.

Digital PCR results are derived from the average number of initial templates per partition, a value known as λ. Estimates of λ from experimental measurements are determined by Poisson statistics. The precision of the λ estimate increases with the number of partitions. Precision decreases as λ gets too low (only a few partitions are positive), or too high (nearly all partitions are positive) and is optimal when λ equals 1.59. Consequently, for best precision, a prior estimate of concentration is necessary. Coefficients of variation can be estimated by the Poisson distribution or determined more exactly by the binomial expansion. Single molecule amplification and digital analysis has also been reported for other amplification methods including loop-mediated isothermal amplification and recombinase polymerase amplification (RPA).

Transcription-based amplification methods

Transcription-based amplification methods are modeled after the replication of retroviruses. These methods are known by various names, including transcription-mediated amplification (TMA), ^, nucleic acid sequence–based amplification (NASBA), and self-sustained sequence replication (3SR) assays. They amplify their target without temperature cycling (isothermally) and use the collective activities of reverse transcriptase, RNase H, and RNA polymerase. The most widely used is TMA, illustrated in Fig. 64.10 . Two primers, a reverse transcriptase, and an RNA polymerase are used. The primer complementary to the RNA target has a 5′-tail that includes a promoter sequence for RNA polymerase. This primer anneals to the target RNA and is extended by the reverse transcriptase, creating an RNA-DNA duplex. The RNA strand is degraded by the RNAse H activity of the reverse transcriptase, allowing the second primer to anneal. The reverse transcriptase then extends the second primer to create double-stranded DNA (dsDNA) that includes the promoter. RNA polymerase recognizes the promoter and initiates transcription, producing 100 to 1000 copies of RNA for each DNA template. Each strand of RNA then binds and extends the second primer, forming an RNA-DNA hybrid; the RNA in the hybrid is degraded, the promoter primer binds and extends to produce dsDNA that can be transcribed, and the cycle repeats. As in PCR, all reagents are included and amplification is exponential with completion in less than one hour. Unlike PCR, these methods do not require temperature cycling (except for an initial heat denaturation if a DNA template is used). They are particularly advantageous when the target is RNA (e.g., human immunodeficiency virus [HIV] and hepatitis C virus [HCV] in blood bank nucleic acid testing).

Loop-mediated amplification methods

Instead of producing products of a defined length, loop-mediated amplification (LAMP) produces a wide range of different DNA structures with branches and loops. In the basic version, two strand displacement primers and two loop-forming primers recognize six segments in the target. The inner two primers each include a 5′-tail that is complementary to the target sequence. After extension of the inner primers, hairpins or loops can form on each end, one of which will have a free 3′-end that can further extend. This loop formation is similar to self-probing and snapback primers, except that the 3′-end is not blocked. The two outer primers are used for displacement of the inner extension products to produce the starting material for cyclic amplification. The chain reaction includes both extension of the free 3′-ends and additional priming from the inner primers to the exposed single strands in the loops. The amplification results in a mixture of products with ever more loops and branching structures of increasing complexity. In another version of LAMP, allele-specific amplification with five primers and one competitive probe recognizes seven segments in the target. In both versions, a variety of products are formed, and the reactions can be completed in less than 1 hour.

Strand displacement amplification

Similar to LAMP, strand displacement amplification (SDA) requires initial generation of starting material before the chain reaction. DNA is first heat denatured in the presence of four primers: two outer displacement primers and two inner primers with 5’-tails that includes a restriction site. An exonuclease-negative polymerase with good displacement activity is added in the presence of dCTP, dGTP, dUTP, and a modified deoxynucleotide (dATPαS), incorporating both the restriction site and the modified dATPαS into products that are ready to enter into exponential amplification. Exponential amplification occurs at 37 °C by (1) nicking of the restriction enzyme site by the restriction enzyme (double strand cutting is prevented by dATPαS); (2) extension from the nicked site with strand displacement; (3) priming of the displaced strand with the original inner primer that includes the restriction site; and (4) extension of both the primer and displaced strand, forming a new doubled stranded product with the restriction site. Steps 1 through 4 are repeated over and over again for exponential amplification.

Variants of polymerase chain reaction that do not require heat denaturation

Variants of PCR have been developed that replace the need for heat denaturation with enzymatic separation of the double helix. These methods do not require thermal cycling and better reflect the normal DNA replication process, although the end products are the same as in the PCR. For example, the helicase dependent amplification (HDA) uses the unwinding enzyme, helicase, to separate the double helix into single strands. As originally described, additional proteins were needed to stabilize the process that was performed at 37 °C. Using a heat-stable helicase and polymerase allowed amplification at 60 to 65 °C without the need for any other accessory proteins. Another technique, RPA uses a recombinase to scan dsDNA for priming sites, causing strand exchange to anneal the primers and single stranded binding proteins to stabilize the loop structure long enough for strand displacement primer extension. Two opposing primers exponentially replicate a short DNA fragment as in PCR, but the reaction is performed at 37 °C or even at room temperature without temperature cycling.

Rolling circle amplification

If a primer is annealed to a closed circle of DNA in the presence of a processive, displacing polymerase, the complement of the circle will be synthesized over and over again with displacement of the tandem repeats. If two primers are used in opposite orientation, progressively more complex branches will be formed in an exponential reaction. The rolling circle can be formed by ligation of the two ends of a linear probe on template DNA. Ligation may happen directly, after polymerization through a gap, or after annealing of an additional, allele-specific oligonucleotide.

Whole genome and whole transcriptome amplification

Instead of specific amplification of one target to improve sensitivity, methods that amplify all genomic DNA or mRNAs are useful when the target is in short supply. For example, multiple-displacement amplification uses exonuclease-resistant random hexamers and a highly processive polymerase to amplify DNA nonspecifically. Initial DNA denaturation is not necessary, and the reaction proceeds isothermally. Similarly, messenger RNA can be generically amplified with a poly(T) primer modified with an RNA polymerase promoter. After reverse transcription, second-strand DNA synthesis and transcription, antisense RNA is produced. Both whole genome and antisense RNA amplification are also useful as nucleic acid purification methods before amplification or detection.