Faye Boeckman, Larissa Tan, Marni Brisson, Rob Park, and Keith Hamby, Bio-Rad Laboratories, Inc., 2000 Alfred Nobel Dr., Hercules, CA 94547
When researchers need to measure the amount of RNA or DNA in a preparation, they typically employ a traditional biochemical approach. High concentrations of nucleic acid can be estimated by staining agarose or acrylamide gels after electrophoresis. With less material, or for more exact determination, the concentration can be estimated by spectroscopy or fluorometry. Occasionally, the amount of nucleic acid present is so small that none of the traditional measurement methods are appropriate. In these cases, researchers have turned to the polymerase chain reaction (PCR) to make more copies of the desired nucleic acid. After amplification they measure the amount of DNA produced in the reaction and finally calculate the starting amount of nucleic acid based on post-PCR measurements. For a number of technical reasons, accurate calculation of the starting amount of nucleic acid is not possible, and even relative comparisons between 2 amplified samples are generally not valid because of differences in efficiency and specificity of the 2 amplification reactions. In gene expression studies, when the starting amount of nucleic acid is diminishingly small, or for the most demanding sensitivity, such as monitoring the number of viral particles present in a patients blood sample, real-time quantitative PCR is the only reasonable approach. In real-time quantitative PCR, a fluorescent reporter molecule is added to the chemical reaction. This reporter may be specific for the desired amplification product, or it can be nonspecific, used on ly to report the accumulation of double-stranded product. In practice, quantitative PCR is carried out in an instrument that can monitor the reaction mixture and measure the fluorescence of the reporter system during each round of amplification, not just at the end point. During the early rounds of amplification, the change in fluorescence is negligible and beyond the sensitivity of the detector, but at some point, a change will register with the detection system. This point, called the threshold cycle (CT), is proportional to the log of the starting amount of nucleic acid (Heid et al. 1996). When samples of known starting amount are amplified simultaneously under the same conditions as the unknowns, it is possible to construct a standard curve, and from this standard curve determine the starting amount of DNA in each of the unknowns.
In some instances it is desirable to simultaneously quantitate 2 or more genes; for example, in gene expression studies, one of the genes might be a housekeeping gene used as an internal standard. Simultaneous amplification and quantitation can also be more cost efficient and lead to higher throughput. However, optimization of experiments designed to accurately quantitate multiple target genes in one tube is not a trivial undertaking. In this article we discuss the optimization of a multiplex reaction system. We present data showing the simultaneous amplification of 4 targets from genomic DNA in a single tube and also showing the simultaneous amplification of 2 targets when one is present in great excess of the other.
Genomic DNA Multiplex
Prior to performing PCR, human genomic DNA (250 ng/l; Promega) was partially digested with BamHI at 37C for 2 hr, and then heated to 100C for 10 min before immediately cooling on ice. This DNA was then stored at 4C until use. To show that any gene amplified alone had an identical CT when amplified with other genes in multiplex (on the same plate), single-gene reactions with 100 ng human genomic DNA were compared to the 100 ng multiplex reaction. Reaction conditions were identical for single and multiplex reactions as described below. To demonstrate a range of multiplex capabilities, 10-fold serial dilutions, from 500 ng to 50 pg, were assayed for α-tubulin, GAPDH, IL-1 β, and factor VIII DNA. The following 50 l reactions were prepared to demonstrate multiplex capabilities:
25 l 2x Life Technologies custom PLATINUM Super mix (40 mM Tris
pH 8.4, 100 mM KCl, 1.6 mM dNTPs, 6 mM MgCl2, 50 U/ml Platinum Taq polymerase)
2 l 50 mM MgCl2 (Life Technologies)
0.4 l 5 U/l Life Technologies PLATINUM Taq polymerase
0.5 l 100 mM CLONTECH Advantage ultrapure dNTPs
0.2 l 100 M α-tubulin forward primer (5'-CCAAGCTGGAGTTCTCTA-3', Operon)
0.2 l 100 M α-tubulin reverse primer (5'-CAGAGTGCTCCAGG-3', Operon)
0.15 l 100 M α-tubulin probe (5'-FAM-CCCAGGTTTCCACAGCTGTAGTTGACCTGGG- DABCYL-3', Operon)
0.15 l 100 M GAPDH forward primer (5'-CATGTTCCAATATGATTCCAC-3', Operon)
0.15 l 100 M GAPDH reverse primer (5'-CCTGGAAGATGGTGATG-3', Operon)
0.1 l 100 M GAPDH probe (5'-HEX-CAAGGCTGAGAACGGGAAGCTTGTCCAGCCTTG- DABCYL-3', Operon)
0.075 l 100 M IL-1 forward primer (5'-TGCT
0.075 l 100 M IL-1 reverse primer (5'-GTGGTGGTCGGAGATTC-3', Operon)
0.075 l 100 M IL-1 probe (5'-Texas Red-CTCTGCCCTCTGGATGGCGGCAGAG- DABCYL-3', Operon)
0.2 l 100 M factor VIII forward primer (5'-GACAGTGGAAATGTTACC-3', Operon)
0.2 l 100 M factor VIII reverse primer (5'-CATCCCAGCATGTAGATG-3', Operon)
0.15 l 100 M factor VIII probe (5'-Cy5-AGCTGGAATTTGGCGGGTGGAATGTCCAGCT- BH2-3', Integrated DNA Technologies)
10.375 l PCR-grade ddH20
10 l diluted human genomic DNA (Promega)
PCR conditions were 3 min at 95C, followed by 50 cycles of 10 sec at 95C, 60 sec at 55C.
7-Order Concentration Difference
Ten-fold serial dilutions of GAPDH DNA, from 109 to 102 copies per 50 l reaction, were prepared and amplified either alone or with 109 copies of α-tubulin. The following reaction conditions were used:
25 l 2x Life Technologies custom PLATINUM Super mix
2 l 50 mM MgCl2 (Life Technologies)
0.5 l 5 U/l Life Technologies PLATINUM Taq polymerase
0.5 l 100 mM CLONTECH Advantage ultrapure dNTPs
0.15 l 100 M α-tubulin forward primer (5'-GCAAGCTGGCTGAC-3', Operon)
0.15 l 100 M α-tubulin reverse primer (5'-CATAATCAACTGAGAGACG-3', Operon)
0.1 l 100 M α-tubulin probe (5'-HEX-CACCGGTCTTCAGGGCTTCTTGCCGGTG- BH1-3', Biosource International)
0.15 l 100 M GAPDH forward primer (5'-CAACTACATGGTCTA CATGTTC- 3', Operon)
0.15 l 100 M GAPDH reverse primer (5'-CTCGCTCCTGGAAGATG-3', Operon)
0.1 l 100 M GAPDH probe (5'-FAM-CGGCACAGTCAAGGCCGAGAATGGTGCCG- DABCYL-3', Operon)
1.0 l plasmid DNA dilution
20.2 l PCR-grade ddH20
PCR conditions were 3 min at 95C, followed by 50 cycles of 10 sec at 95C, 45 sec at 55C.
Interpretation of Results
Gene expression studies of multiple gene targets require both precise and accurate quantitation of the same samples at the same time. The iCycler iQ system displays powerful multiplexing capabilities that we demonstrate here on genomic and plasmid targets. Furthermore, we demonstrate the need to optimize PCR efficiency when working with varying starting template concentrations.
In order to conduct successful multiplexing you must (1) maximize and equalize the efficiency of each reaction and (2) minimize any cross-reactivity between individual reactions. Maximizing the efficiency of a reaction allows accurate quantitation over a wider range of starting template concentrations and improves the reproducibility of replicate samples. It is equally important that the individual reactions have similar efficiencies because any differences in individual efficiencies will be amplified when the components of the 2 reactions are combined. If you know that the multiplex templates differ in concentration by less than 1,000-fold, then you can accept efficiency differences of about 10%. For example, one reaction may be 96% and the other 87% efficient. If you do not know the template concentration difference, or if it is gre ater than 1,000-fold, then the reaction efficiencies should not differ by more than 5% and the individual reaction efficiencies should be greater than 90%. The most powerful tool for adjusting reaction efficiency is to examine the secondary structures of the PCR product, primers, and probe. It is critical to eliminate all secondary structures that might interfere with DNA:DNA interactions, but the secondary structure must be evaluated at the appropriate temperature. Secondary structures that melt below the annealing temperature are not important. It is often necessary to raise the annealing temperature or relocate one or both primers or the probe in order to avoid secondary structure problems.
Cross-reactivity between primers and targets in the separate reactions can occur due to specific gene interaction effects, such as primer recognition, as well as to competition for substrates. It is straightforward to test for cross-reactivity of the individual reaction components by comparing a set of wells containing one reaction alone to a set containing all reaction mixtures. If there is no crossreactivity, the CT of template A when amplified alone will be identical to the CT of template A in the presence of the other components. A labeled probe is not required to test for cross-reactivity.
Addition of extra polymerase, dNTPs, and MgCl2 improves results for 2 or more individual reactions in multiplex. For a typical 50 l reaction, we use 3.253.75 U Taq polymerase, 450 M each dNTP, and 5 mM MgCl2 (compared to 1.25 U polymerase, 200 M each dNTP, and 3 mM MgCl2 for a single reaction). If cross-reactivity is minimal, but efficiencies of multiplex reactions are lower than th e individual reactions, minor adjustments to the primer concentrations may be necessary.
Genomic DNA Multiplex Discussion
Prior to multiplexing 2 or more PCR targets, it is important to show that amplification of each target is unaltered in a multiplex reaction containing additional PCR targets. We show that amplification of the PCR target is identical whether it is amplified alone or in multiplex. CT values for a single gene target were not significantly different when amplifying the target alone or with 3 other genes under optimized conditions. This was true for plasmid or genomic DNA targets. Figure 1 (left-hand panels) shows the results for each of the 4 human genomic DNA targets, α-tubulin, IL-1 β, GAPDH, and factor VIII. Amplification of the gene alone or in multiplex resulted in identical CT values and identical slopes (which indicate reaction efficiency), showing exponential amplification. This confirms that multiplex conditions do not alter any of the targets amplification under these conditions. The reaction conditions were critical to this observation. Because the PCR efficiency is slightly different for each PCR product in this multiplex under identical reaction conditions (not shown), primer concentrations were adjusted to ensure similar amplification of each target. The right side of the figure shows 4 orders of magnitude amplification for each gene target in the multiplex reaction. The inset shows the standard curve for each reaction. The slopes and corresponding efficiencies of the reactions are 3.12 (~100%), -3.43 (96%), -3.52 (92%) and 3.53 (92%) for factor VIII, GAPDH, IL-1 β, and α-tubulin respectively.
A wide difference between the starting concentrations of 2 gene products targeted for amplification is a common scenario when screening samples with a standard housekeeping gene and a rare message of interest. Here we show optimized conditions for up to a 10 million-fold difference in starting template concentrations between 2 PCR targets. Plasmids containing the human cDNAs for GAPDH and α-tubulin were multiplexed over varying concentrations. The α-tubulin plasmid was held constant at 109 copies per well while varying the concentration of GAPDH from 102 to 109 copies per well. Figure 2 shows the results of multiplexing. There is no significant difference between amplification of GAPDH alone or in multiplex with 109 copies of α-tubulin as shown by the CT values (see table inset in Figure 2). Similar results were obtained by holding GAPDH constant at 109 copies and multiplexing with the same serial dilution of α-tubulin (data not shown). To avoid problems when multiplexing with extreme differences in target gene concentration (e.g., 109 copies α-tubulin with 102 copies of GAPDH), optimization of reaction components is required. We found that the effect of increasing the amount of enzyme and free nucleotides was additive (not shown). The MgCl2 concentration was increased to compensate for the higher concentration of nucleotides, which chelate magnesium ions. In addition, we found that accurate quantitation of small amounts of one target in a reaction with large amounts of a second target depended heavily on the amplification efficiencies of both targets. Large differences required high- (>90%) and equal-efficiency PCR amplifications of both genes. Amplification efficiencies of the GAPDH and α-tubulin plasmids were determined prior to multiplexing and found to be nearly equal and nearly 100% (data not shown).
Gene expression studies of multiple gene targets, as well as other applications, require both precise and accurate quantitation of the same samples at the same time. The optimization of PCR efficiency is essential in performing this accurate quantitation. We have demonstrated that the iCycler iQ system is a powerful tool to achieve this optimization and to perform multiplex real-time PCR with plasmid and genomic targets.
Heid CA et al., Real time quantitative PCR, Genome Res 6, 986994 (1996)
Cy is a trademark of Amersham Pharmacia International PLC. Texas Red is a trademark of Molecular Probes, Inc.
* Practice of the patented polymerase chain reaction (PCR) process requires a license. The iCycler iQ system includes a licensed thermal cycler and may be used with PCR licenses available from Applera Corporation. Its use with authorized reagents also provides a limited PCR license in accordance with the label rights accompanying such reagents. Some applications may require licenses from other parties.
back to top