| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Eragen Biosciences, Inc., 918 Deming Way, Madison, WI 53717-1944.
2 Department of Pathology, University of Michigan, Ann Arbor, MI 48109.
aAuthor for correspondence. E-mail jprudent{at}eragen.com.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: We developed a novel real-time PCR technology that uses universal energy transfer probes constructed from An Expanded Genetic Information System (AEGIS) for both quantification and genotyping analyses.
Results: RNA quantification by reverse transcription-PCR was linear over four orders of magnitude for the simultaneous analysis of ß-actin messenger RNA and 18S ribosomal RNA. A single trial validation study of 176 previously genotyped clinical specimens was performed by endpoint analysis for factor V Leiden and prothrombin 20210A mutation detection. There was concordance for 173 samples between the genotyping results from Invader® tests and the AEGIS universal energy transfer probe system for both factor V Leiden and prothrombin G20210A. Two prothrombin and one factor V sample gave indeterminate results (no calls).
Conclusion: The AEGIS universal probe system allows for rapid development of PCR assays for nucleic acid quantification and genotyping.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Several kinetic PCR chemistries and instruments that support those chemistries have been described, with the most popular methods using fluorescence for detection (1). Specifically, the TaqMan assay was the first homogeneous system described and is likely the most widely used (2)(3). TaqMan generates signal by the exonucleolytic cleavage of a hybridized, amplicon-specific, energy transfer probe. The target-specific cleavage event leads to an increase in signal fluorescence. The number of PCR cycles required for signal generation is dependent on the initial concentration of target. Other systems have followed and include Molecular Beacons (4), Scorpion primers (5), Amplifluor (6)(7), and single-labeled probes (8). To detect the accumulation of just a single amplicon, the addition of a DNA binding dye such as SYBR Green is simple and effective. However, SYBR Green methodology does not allow for the incorporation of an internal reference, an accepted method for normalizing sample-to-sample variation (9).
As with any technology base, every system has benefits and drawbacks. Scientists weigh the plusses and minuses to choose the system that best fits a particular application. In many cases, this decision is based solely on the available instrument. Today there are more than five kinetic PCR instruments that visualize a fluorescent shift during temperature cycling. Despite all this advanced technology, there still seems to be a need for a robust system that (a) does not require target-specific probes, (b) can assay multiple targets in the same reaction, (c) can be easily implemented on preexisting PCR designs, and (d) can be executed on any of today’s standard instruments. With An Expanded Genetic Information System [AEGIS; Ref. (10)], we demonstrate that a system can be constructed to accomplish all four.
Nucleic acids [RNA, DNA, and peptide nucleic acid (PNA)] recognize molecular information using Watson–Crick base pairing rules. Such a simple recognition system is found nowhere else in chemistry and makes DNA ideal for designing analysis platforms. Adding additional base pairs to the two preexisting in nature would seem to be an obvious advancement for such platforms. It is then not surprising that AEGIS base pairs have been exploited in DNA recognition platforms such as branched DNA to create more robust technologies (11). The chemistry to synthesize AEGIS phosphoramidites that perform well in traditional DNA oligonucleotide synthesis chemistries has also been perfected (12). To take this idea one step further, we set out to determine whether AEGIS could be further exploited in other nucleic acid analysis systems that use enzymes, such as PCR. Specifically, we used AEGIS to create a universal probe system for closed-tube PCR monitoring. To demonstrate the utility of the technology, we focused in on two major applications, single base change detection and nucleic acid quantification. For single base change analysis, we developed endpoint tests for the factor V (F5) Leiden mutation and for a common mutation found within the 3'-untranslated region of the prothrombin [factor II (F2)] gene (G20210A). For quantification analysis, we developed a real-time method for the quantification of two RNA species: the 18S ribosomal RNA and the human ß-actin mRNA. All demonstrated tests detect two target sequences simultaneously in a closed-tube format. The technology was performed on a variety of instrument platforms.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
universal energy transfer probes
A variety of universal energy transfer probe oligonucleotides were synthesized as described (Table 1
). Two fluorescent labels were used for the syntheses of these probes: 6-carboxyfluorescein (FAM) and hexachlorofluorescein (HEX). Six probes were used in this study as shown in Table 1
. The 3' ends of energy transfer probe oligonucleotides were capped with biotin to prevent nonspecific polymerase extension from the probes.
|
melting temperature determination for aegis energy transfer probes
Nearest-neighbor numbers for AEGIS base pairs are being collected but are not yet available. Therefore, thermodynamic melting data for AEGIS energy transfer probes were experimentally determined. The AEGIS energy transfer probes listed in Table 1
were mixed with equimolar complementary AEGIS overhang primers at 500 nmol/L in a buffer containing 10 mmol/L bis-Tris-propane (pH 8.9), 2.0 mmol/L magnesium chloride, and 40 mmol/L potassium acetate. DNA duplexes were melted from 25 °C to 90 °C, and the absorbance at 260 nm (A260) was measured using a Beckman DU7500 spectrophotometer equipped with a temperature-controlled cuvette. Duplex melting temperature (Tm) values were determined by first-derivative analyses of plots of A260 vs temperature (°C).
multiple rounds of nuclease cleavage by energy transfer probe turnover
Five synthetic double-flap gap substrates were created by hybridizing a universal AEGIS energy transfer probe with a Tm of 43.4 °C to the complementary 5' AEGIS overhang of hairpin-forming oligonucleotides with either no base or each of the four natural bases mispaired with isoguanosine (iG; Table 1
). Reactions were performed at 48 °C in the presence of 0.1 U/µL Thermus aquaticus DNA polymerase (Promega) in a buffer containing 10 mmol/L bis-Tris-propane (Sigma-Aldrich), pH 8.9, 2.0 mmol/L magnesium chloride, and 40 mmol/L potassium acetate. The probe concentration was 400 nmol/L, whereas the hairpin concentrations were either 5 or 50 nmol/L. Reaction fluorescence was monitored as described for endpoint genotyping before incubation and at ten 2-min intervals, then at 5-min intervals for a total reaction time of 60 min. The pre-read (0 min) relative fluorescence unit (RFU) determination was subtracted from each time point.
endpoint genotyping assays
Primers complementary to the human prothrombin or F2 (OMIM no. 176930) and the F5 (OMIM no. 227400) loci were synthesized as described (Table 1
). Two allele-specific forward primers with AEGIS overhangs complementary to universal energy transfer probes and a standard reverse primer were synthesized as described for each target locus (Table 1
). F5 genotyping assays were performed using 5 zmoles (10-21 moles; 3000 copies) of synthetic oligonucleotides as controls (Table 1
). A heterozygous control consisted of an equimolar ratio of the wild-type and Leiden control oligonucleotides at a total concentration equal to that of the homozygous controls. Both genotyping assays were performed with 5–10 ng (1500–3000 copies) of human genomic DNA samples (Coriell Cell Repositories). The genomic DNA samples were as follows: NA14897 (homozygous F2 and F5 wild type), NA16028 (F2 G20210A heterozygote), NA16000 (F2 G20210A homozygote), and NA14641 (F5 Leiden heterozygote).
PCR was performed with primers and probes at a concentration of 0.2–0.4 µM in a buffer containing 10 mM bis-Tris-propane (pH 8.9), 2.0 mM magnesium chloride, 40 mM potassium acetate, and 50 µM deoxynucleotide triphosphates (A, C, G, T; Promega). The DNA polymerase was AmpliTaq for F5 Leiden and AmpliTaq Gold for prothrombin. Polymerase was used at 0.05 U/µL. Reagents were prepared at twice the working concentration and then added to an equal volume of control oligonucleotide or human genomic DNA sample to a final volume of 10 µL. Reactions were overlaid with 15 µL of mineral oil before cycling. PCR cycling conditions consisted of an initial denaturation step of 5–12 min at 95 °C followed by 36–40 cycles of rapid cycling amplification of 1 s at 95 °C and 1 s at 57 or 58 °C in either a Genius (Techne Inc.) or a 9600 (Applied Biosystems) thermocycler. After cycling, endpoint signal was generated by incubation at 48 °C (F5) or 59 °C (F2) for 30 min. Fluorescence was detected using a Cytofluor 4000 fluorescence plate reader (Applied Biosystems) with photomultiplier tube gain set to 50. For FAM fluorescence, the excitation filter was 485 ± 10 nm and the emission filter was 530 ± 12.5 nm. For HEX fluorescence, the excitation filter was 530 ± 12.5 nm and the emission filter was 580 ± 25 nm. Background RFU of reactions before cycling and nuclease incubation were subtracted from experimental RFU before analysis.
Genotypes were determined based on sufficient assay signal generation and the ratio of FAM RFU to HEX RFU for each reaction before assay validation with clinical samples. The goal was to set robust cutoff values to minimize wrong calls. In control experiments in which the target genotypes were known, the reaction fluorescence was first analyzed to determine whether sufficient signal was generated to make a definitive call. For the prothrombin and F5 systems, any reaction in which the sum of the corrected FAM and HEX signals did not exceed 900 or 500 total RFU, respectively, was scored as "no call" and the ratio was not calculated. Reactions that generated sufficient signal were scored based on the ratio of FAM to HEX signals. For both systems, if the ratio was >3.0, the reaction was scored as homozygous for the allele queried by the FAM-labeled probe. If the ratio was <0.3, the reaction was scored as homozygous for the opposite allele. If the ratio was between 0.5 and 2.0, the reaction was scored as heterozygous. Reactions that did not fall into these ranges were scored as indeterminate or no calls. In reactions in which background subtraction produced negative reaction fluorescence, a value of 1.0 was assigned and used to calculate the ratios.
real-time rna quantification assay
Primers complementary to the human ß-actin (ACTB; OMIM no. 102630) and the mammalian 18S loci were synthesized as described (Table 1
). Standard base oligonucleotides were used to prime first-strand cDNA synthesis. Second-strand cDNA synthesis was primed using oligonucleotides with AEGIS overhangs complementary to energy transfer probes. RNA samples were human cardiac total RNA and rat kidney total RNA (Ambion). A ß-actin control RNA was prepared by PCR amplification of human ß-actin cDNA with a primer containing a T7 RNA polymerase-binding site followed by transcription in vitro with T7 RNA polymerase (Promega). Single-tube RT-PCR reaction conditions were as described above for PCR except that primers were used at concentrations from 0.02 to 0.2 µM, deoxynucleotide triphosphates were used at 100 µM, and reactions contained Moloney murine leukemia virus reverse transcriptase (Promega or Ambion) at 0.5 U/µL.
Standard RT-PCR cycling conditions consisted of a 5-min reverse transcription step at 60 °C and a denaturation/reverse transcription-inactivation step of 5 min at 95 °C followed by 40 cycles of 1 s at 95 °C and 60 s at 60 °C in the iCyclerIQ (Bio-Rad). Reactions were also performed in the SmartCycler (Cepheid) and the ABI Prism 7700 (Applied Biosystems; data not shown) thermocyclers. A 3-min bias optimization step at 4 °C required by the iCycler IQ was inserted before the reverse transcription step when cycling with that instrument. Fluorescence measurements were performed during the 60 °C extension/cleavage step to determine the cycle threshold (Ct) using iCycler IQ software. The Ct is defined as the number of amplification cycles required for the reaction fluorescence to cross a user-defined RFU value. In practice, the threshold was usually set to a value 5–10 times the SD of the background RFU. A calibration curve was generated from a plot of log target copy number vs Ct.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). During extension of the standard reverse PCR primer, the polymerase misincorporates a natural base opposite the iG and then terminates at the iC. The exact ratio of the misincorporated bases has not been determined, but experimental evidence demonstrated that T. aquaticus DNA polymerase can incorporate any one of the four nucleotides opposite of iG in the absence of iCTP (data not shown). Thus, PCR amplification produces a product with a 5' single-stranded overhang and a 3' unpaired single base flap (Fig. 1
). An energy transfer probe complementary to the 5' overhang will hybridize at a temperature near its Tm. For these experiments, the energy transfer probe was synthesized with a 5' fluorophore; a T that mispairs with the iG in the overhang, creating a second unpaired flap; a 5'-penultimate iG required to form a base pair with the iC in the overhang; and a dabcyl quencher located immediately downstream of the iG. The quencher was placed in this position as close as possible to the fluorophore because quenching efficiency for a given dye pair varies inversely with the sixth power of the distance between molecules (13). The final structure contained two flaps and a single-base gap, hence the name double-flap gap (Fig. 1
). The double-flap gap substrate is recognized and cleaved by flap endonucleases, including the 5'–3' nuclease of Taq polymerase. The specific nucleolytic activity separates the fluorophore from the dabcyl quencher, and an increase in fluorescence is observed (Fig. 1
).
|
If the nuclease reaction is performed near the Tm of the probe, multiple rounds of cleavage occur by exchange of cut probes with uncut probes, leading to greater signal generation. This is expected because complementary DNA strands near the Tm exist in a mixed state of hybridized to nonhybridized, with exchange occurring at a rate 
1/s. In experiments where the temperature was held well below the optimum for the enzyme and where the energy transfer probe was in eightfold excess over mock target substrates, cleavage was complete within 50 min, demonstrating turnover (Fig. 2
). There was little or no preference for the substrate 3' base, indicating that no matter what nucleotide the polymerase incorporated opposite iG, the reaction rate stayed virtually constant. An increase in fluorescence was absent when the energy transfer probe was hybridized to a substrate lacking a 3' mismatch. This is consistent with the observation that under our PCR conditions, little or no cleavage occurred when the energy transfer probe was hybridized to unextended AEGIS overhang primers. Fluorescence signal generation appeared to be specific to formation of the double-flap gap.
|
Probe cleavage can be designed to occur during or before PCR by changing the Tm of the energy transfer probe. For endpoint signal generation, we used an energy transfer probe with a Tm below the annealing temperature of the PCR primers. We chose a 
Tm of 10 °C because this is sufficient to keep the vast majority of energy transfer probes from hybridizing to the primers during PCR. After a defined number of cycles, the temperature of the reaction was then lowered to the Tm of the energy transfer probe, and cycling of the energy transfer probes to the overhang yielded a linear increase in signal. The universal energy transfer probe system can be modified so that signal generation occurs in real time. For kinetic PCR, we designed the energy transfer probe to have a Tm near the annealing temperature of the PCR primers. We have observed that a 1-min hold time is sufficient to not only anneal and extend the reverse primer but also to hybridize the energy transfer probe, form the double-flap gap structure, and perform the cleavage reaction.
endpoint genotyping for thrombosis risk: f5 and prothrombin g20210a
The utility of the double-flap gap substrate for single-nucleotide polymorphism (SNP) genotyping was demonstrated by application to two well-known thrombosis-associated mutations within the F5 and prothrombin (F2) genes. A mutation known as F5 Leiden is the most common inherited risk factor for deep vein thrombosis among persons of European descent. The single base substitution within the coding region of the F5 gene from G to A occurs with an approximate allele frequency of 1–5%. F5 Leiden encodes an abnormal F5 protein with an amino acid substitution (Arg506Gln) that is resistant to cleavage by activated protein C (14). The G20210A mutation in the 3'-untranslated region of the F2 or prothrombin gene is also associated with increased venous thrombosis. The F2 G20210A mutation leads to increased concentrations of plasma prothrombin and is found in 1% of persons of European descent (15).
Both the prothrombin and F5 Leiden genotyping assays were designed to query both alleles of a SNP with use of two universal energy transfer probes in a single well. The PCR primers consisted of one reverse primer and two allele-specific primers containing discrete 5' AEGIS overhangs. Two universal energy transfer probes complementary to the allele-specific AEGIS overhangs were synthesized (no post-synthetic modifications were required for the energy transfer probes). In the F5 Leiden system, the energy transfer probe complementary to the overhang on the primer specific for the wild-type G allele was labeled with FAM. The energy transfer probe complementary to the A allele specific primer was labeled with HEX. Oligonucleotides representing both alleles of the F5 gene were synthesized and used as synthetic targets in control reactions. Five zmoles of synthetic target or 5 ng of characterized human genomic DNA (Coriell Laboratories), representing 3000 or 1500 target molecules, respectively, were assayed. In five independent trials comprising 152 observations, no miscalls were observed. Only a single no-call reaction was seen in which signal was not generated in either the FAM or HEX channels.
In the prothrombin system, the energy transfer probe complementary to the primer specific for the wild-type F2 G allele was labeled with HEX, whereas the energy transfer probe complementary to the F2 A-allele-specific primer was labeled with FAM. As controls for F2 genotyping, characterized human genomic DNA samples (Coriell Laboratories) were used. We assayed 10 ng of genomic DNA, representing 3000 target molecules, in nine independent trials comprising 78 total observations. There were no miscalls and only one no call.
validation of f5 leiden and prothrombin 20210a mutation assays in clinical specimens
To validate the AEGIS universal energy transfer probe system for clinical samples, we compared results using clinical human genomic DNA specimens that had been previously genotyped for F5 Leiden and prothrombin G20210A using the Invader® assay (Third Wave Technology). This validation was performed in the Molecular Diagnostics Laboratory at the University of Michigan Health System with approval of the human subjects Institutional Review Board.
The validation study consisted of 176 anonymized samples that had been genotyped as follows using Invader: F5 Leiden A/A (n = 1), A/G (n = 24), and G/G (n = 150); prothrombin 20210 A/G (n = 13) and G/G (n = 161). These samples were genotyped in a single trial using the universal AEGIS energy transfer probe method (Fig. 3
). The operators were blinded to the previously determined genotypes. Genotypes were determined using the criteria described in the methods. With the exception of three no calls (two prothrombin and one F5), complete concordance was observed between the AEGIS universal energy transfer probe system and the Invader System.
|
real-time RNA quantification: ß-actin MRNA and 18S RNA
An important application for real-time PCR is the quantification of RNA. Determination of the quantity of a specific RNA target in any given sample can be achieved by performing reverse transcription to generate cDNA followed by PCR or by running the reactions consecutively in a closed tube (RT-PCR). For this study we applied the RT-PCR methodology to two commonly used control RNA species: human ß-actin mRNA and the 18S ribosomal RNA. The sensitivity and linearity of the assay were determined by varying RNA target concentration in the reactions.
We initially tested our ß-actin and 18S designs singly for dynamic range. We designed PCR primers specific to amplify a region conserved between all mammalian 18S RNA sequences (Fig. 4A
). Regression analysis of log RNA concentration vs Ct indicated a linear correlation over six orders of magnitude of input RNA concentrations with a correlation coefficient (R2) of 0.99 (Fig. 4B
). The buffer-only control also generated fluorescent signal that could be the result of target contamination or template-independent amplification. Another assay specific to the 5' end of ß-actin RNA was designed to be insensitive to genomic DNA or to other actin isoforms. The assay was linear over five orders of magnitude for ß-actin targets with a mean R2 of 0.99 in five independent runs (data not shown). A full-length ß-actin mRNA transcribed in vitro with T7 RNA polymerase was quantified by absorbance at 260 nm and then used as a calibrator for absolute quantification of the amount of actin transcript. A linear regression of log ß-actin standard RNA copy number vs Ct was used to quantify the actin mRNA in unknown samples. In four independent experiments with a total of 22 separate measurements, we found that human cardiac total RNA contains a mean (SD) of 1.3 (0.8) x 108 molecules of ß-actin transcript per microgram of RNA.
|
The performance of the ß-actin and 18S systems in a duplex reaction was also evaluated using human cardiac total RNA. We hypothesized that because the ribosomal target is present at such a high concentration, large amounts of 18S amplicon could accumulate and exhaust a crucial component required for amplification of the ß-actin target. A 10-fold reduction in the 18S primer concentration compared with the ß-actin primers gave quantitative results when both reactions were performed in multiplex. Interestingly, although the Ct for the 18S reaction was significantly increased by the reduction in PCR primer concentration, PCR efficiency was not seriously affected. Apparently either the high copy number of 18S transcript in total RNA or the higher predicted Tm of the 18S primer set permits efficient amplification despite the reduction in primer concentration. When the Cts for both targets were analyzed simultaneously in triplicate, a linear range from 16 pg to 10 ng input sample was observed (Fig. 5
). The assay detection limit was affected by negative-control template-independent amplification (observed in two of three triplicate reactions). The mean (SD) Cts for these two no-target reactions were 35.7 (0.3) and 36.6 (0.1) cycles for the ß-actin and 18S systems, respectively.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The closest alternative technique to the AEGIS system described here may be universal Amplifluor primers (16)(17). Both systems use a universal mode for direct detection of products amplified using tailed primers. However, there is an important distinction between the two that makes the AEGIS method more attractive: the AEGIS detection system uses universal probes, whereas Amplifluor uses energy transfer primers. Universal detection primers require polymerase extension from two nested PCR primers to detect each target sequence. Signal generation with AEGIS is achieved by hybridization and cleavage of a probe that is inert in PCR. The use of universal probes thus reduces the chance of primer-primer interactions. For example, Amplifluor needs five primers for SNP analysis, whereas AEGIS needs only three. With fewer primers in the PCR reaction, the need for optimization of primer design, reaction conditions, and cycling conditions is likely to be reduced.
The AEGIS method is broadly applicable to not only DNA but RNA as well. To make this a general use technology system, we have optimized the chemistry and are currently exploring our options for distribution. Although we demonstrated tests that analyze no more than two target sequences simultaneously, we envision additional chemistries that will broaden the multiplex capability.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  D. R. Christensen, L. J. Hartman, B. M. Loveless, M. S. Frye, M. A. Shipley, D. L. Bridge, M. J. Richards, R. S. Kaplan, J. Garrison, C. D. Baldwin, et al. Detection of Biological Threat Agents by Real-Time PCR: Comparison of Assay Performance on the R.A.P.I.D., the LightCycler, and the Smart Cycler Platforms Clin. Chem., January 1, 2006; 52(1): 141 - 145. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. C. Johnson, D. J. Marshall, G. Harms, C. M. Miller, C. B. Sherrill, E. L. Beaty, S. A. Lederer, E. B. Roesch, G. Madsen, G. L. Hoffman, et al. Multiplexed Genetic Analysis Using an Expanded Genetic Alphabet Clin. Chem., November 1, 2004; 50(11): 2019 - 2027. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Rickert, H. Lehrach, and S. Sperling Multiplexed Real-Time PCR Using Universal Reporters Clin. Chem., September 1, 2004; 50(9): 1680 - 1683. [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Johnson, C. B. Sherrill, D. J. Marshall, M. J. Moser, and J. R. Prudent A third base pair for the polymerase chain reaction: inserting isoC and isoG Nucleic Acids Res., March 29, 2004; 32(6): 1937 - 1941. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Moser and J. R. Prudent Enzymatic repair of an expanded genetic information system Nucleic Acids Res., September 1, 2003; 31(17): 5048 - 5053. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
and 
, A; • and 
, C; 
and 
, G; 
and 
, T; and x, none. Open symbols indicate a hairpin concentration of 5 nM (80-fold excess of probe); closed symbols indicate a hairpin concentration of 50 nM (8-fold excess of probe). RFU monitored before incubation were subtracted from measurements taken at intervals of 5 min for a total reaction time of 60 min. Cleavage of the double-flap gap produces an increase in the reaction fluorescence.
