| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Technical Advance |

From the ARUP Institute for Clinical and Experimental Pathology* and the Department of Pathology,
University of Utah School of Medicine, Salt Lake City, Utah
| Abstract |
|---|
|
|
|---|
The fluorescence chemistries used in nucleic acid detection comprise those incorporating non-specific double-stranded DNA (dsDNA) binding dyes, or those using fluorescently-labeled oligonucleotide probes that hybridize specifically to sequences within the target. The non-specific dsDNA binding dyes include ethidium bromide,1 YO-PRO 1,2 and SYBR Green I.3 The binding of dsDNA dyes to double-stranded DNA is accompanied by a dramatic increase in fluorescence, thus the non-specific dsDNA binding dye-based methods are capable of detecting amplification and product accumulation, but are unable to provide unambiguous verification of the identity of the amplified product. The probe-based methods confer additional specificity to the detection of the amplification products3-5 and are capable of distinguishing products differing by only one base.6-8 By comparison, the sequence-specific fluorescent probe-based methods include those using adjacent hybridization probes,3,7 exonuclease (TaqMan),9 hairpin (Molecular beacon),4 and self-probing amplicons (Scorpion).10 In general, the sequence-specific methodologies entail a fluorescence quenching or potentiation interaction between two or more fluorophores,4,11,12 or other more complex interactions involving additional functional groups.13 The synthesis of such dual- or multiple-fluorophore or chemical-group-labeled oligonucleotides for real-time PCR can be technically demanding and expensive. Hence, it is desirable to develop less complex approaches for PCR product detection.
Sequence-specific probe-based designs using only one fluorophore are one method for simplification of fluorescence-based PCR product detection, and represent a significant technical advance over dual-fluorophore based systems. Here, we describe a novel strategy for fluorescence detection and genotyping of PCR products. The method exploits the phenomenon of dequenching of fluorescence of fluorescein-labeled oligonucleotides on hybridization to a complementary DNA target. The reversibility of the phenomenon enables the performance of melting curve analysis, which permits genotyping by Tm. Our novel method yielded 100% concordance when compared with standard mutation detection assays. To our knowledge, this is the first study describing the exploitation of the dequenching phenomenon for PCR detection and genotyping, and establishes the utility of this phenomenon for PCR detection in general.
| Materials and Methods |
|---|
|
|
|---|
Artificial Templates
Oligonuclotides containing sequences from the Factor V gene or the methylenetetrahydrofolate reductase (MTHFR) gene were synthesized by Operon Technologies (Alameda, CA). The Factor V sequence included the Leiden mutation (G1691A) and the MTHFR sequence included the C677T base substitution. The oligonucleotide sequences consisted of either the wild-type or the mutant sequence as listed in Table 1
.
|
For genotypic analysis of the human Factor V, MTHFR and ß-globin loci, DNA was extracted from leukocytes obtained from whole blood samples using the MagNa Pure LC Instrument (Roche Molecular Biochemicals, Indianapolis, IN). All samples were obtained from the archived inventories of ARUP Laboratories (Salt Lake City, UT) with institutional review board approval. For the Factor V locus, we tested 65 wild-types, 23 heterozygotes (G1691A), and 12 homozygotes (G1691A). For the MTHFR locus, we tested 10 wild-types, 5 heterozygotes (C677T), and 5 homozygotes (C677T). For the human ß-globin locus, we tested DNA from 37 wild-types (HbAA), 5 HbSS (homozygous), 3 HbCC (homozygous), and 7 HbSC (compound heterozygous) samples by dequenching PCR. All of the genotypes were confirmed by DNA sequencing and/or a modification of a previously described protocol for Factor V Leiden,6 MTHFR,8 and hemoglobin ß-chain genotyping by high performance liquid chromatography (HPLC).14 The dequenching-based genotyping studies were scored independently from the DNA sequencing, dual-fluorescent probe and HPLC studies, and the concordance between the results of the various approaches was determined after dequenching PCR.
Fluorescence Melting Curve Analysis
Oligonucleotide probes labeled with fluorescein at either the 5' or 3' end were synthesized by Operon Technologies. Probe sequences for the Factor V and MTHFR genes are listed in Table 1
. The fluorescein label was attached to the oligonucleotide probe with an intervening six-carbon spacer. Probes were designed to be complementary to the mutant sequences because this configuration resulted in the greatest Tm shifts from mismatches for both the Factor V and MTHFR genotyping assays.
Melting curve analysis was performed using the LightCycler (Roche Molecular Biochemicals). Each 10 µl reaction contained 0.4 µmol/L template oligonucleotide, 0.2 µmol/L fluorescein-labeled probe, 10 mmol/L Tris-HCl, 50 mmol/L KCl, 3.0 mmol/L MgCl2, and 250 mg/ml BSA. To assess the effect of pH on the fluorescence levels yielded by the dequenching chemistry, melting curves were performed in the presence of buffers ranging from pH 8.3 to 9.2.
The melting protocol consisted of denaturation at 95°C for 10 seconds, a rapid ramp down to 35°C at a rate of 20°C/sec, annealing at 35°C for 30 seconds, and heating to 85°C at 0.3°C/second. Probe melting peak analysis was performed using derivative plots (dF/dT versus T) as previously described.15
PCR Amplification and Genotyping
To determine whether PCR amplification would be compatible with fluorescence-dequenching genotyping, we performed fluorescence-dequenching PCR analysis using Factor V,16
MTHFR and human ß-globin locus specific primers (Table 2)
. The design of the specific probes for each target is illustrated in Figure 3A
(Factor V), Figure 4A
(MTHFR), and Figure 5A
(ß-globin). Oligonucleotides were obtained from Genset Corporation (La Jolla, CA) and Operon Technologies.
|
|
|
|
Software
LightCycler software version 3.3 was used for all analyses. For dequenching melting curve analysis, fluorescence was detected in channel 1 with the gains set at 3.
DNA Sequencing
Automated DNA sequencing of PCR products was performed using dideoxynucleotide termination chemistry and the Applied Biosystems 3100 Genetic Analyzer.
| Results |
|---|
|
|
|---|
The single-fluorophore dequenching reactions are depicted as peaks in the positive scale in the dF/dT versus T graphs (Figure 1)
. For the Factor V Leiden mutation, the homozygous wild-type yielded a melting peak with Tm at 51.4 ± 0.1°C and the homozygous mutant yielded a melting peak with Tm 59.4 ± 0.1°C. For MTHFR, the homozygous wild-type yielded a melting peak with Tm at 64.9°C and the homozygous mutant yielded a melting peak with Tm at 67.0°C (data not shown). These experiments demonstrated the ability of the single-fluorophore dequenching method to discriminate between DNA sequences differing by single nucleotide substitutions.
|
The magnitude of change in fluorescence observed for the dequenching assays was dependent on the pH of the reaction solution. The results for the Factor V wild-type melting peaks in solutions of pH 8.3, 8.6, 8.9, and 9.2 are shown in Figure 2
. The assays were optimally performed at pH 9.2, which was nonetheless compatible with PCR amplification as shown below.
|
Factor V Locus
The single-fluorophore dequenching reactions are depicted as peaks in the positive scale in the dF/dT versus T graphs (Figure 3B)
. The G1691A homozygous samples yielded a probe melting peak with Tm at 58.6 ± 0.9°C and the wild-type samples yielded a melting peak with Tm at 50.3 ± 0.7°C. The heterozygous samples yielded two melting peaks; each corresponding to the respective peaks observed for the mutant and wild-type homozygotes. Using melting profiles obtained from the single probe dequenching experiments, we successfully genotyped 100 of 100 of the cases examined for mutations in wild-type (n = 65), heterozygous mutant (n = 23), and homozygous mutant (n = 12) samples. There was 100% concordance between the results of the single-fluorophore dequenching and a previously described dual-probe based method.6
(Table 3)
.
|
The single-fluorophore dequenching reactions are depicted as peaks in the positive scale in the dF/dT versus T graphs (Figure 4B)
. The C677T homozygous samples yielded a probe melting peak with Tm at 67.8 ± 0.4°C and the wild-type samples yielded a melting peak with Tm at 64.9 ± 0.6°C. The heterozygous samples yielded a broad melting peak spanning the respective Tms observed for the mutant and wild-type homozygotes with a Tm at 66.3 ± 0.8°C. Using melting profiles obtained from the single probe dequenching experiments, we successfully genotyped 20 of 20 of the cases examined for mutations in wild-type (n = 10), heterozygous mutant (n = 5), and homozygous mutant (n = 5) samples. There was 100% concordance between the results of the single-fluorophore dequenching and a previously described dual probe based method8
(Table 3)
.
Human ß-globin Locus
The single-fluorophore dequenching reactions are depicted as peaks in the positive scale in the dF/dT versus T graphs (Figure 5B)
. The HbSS samples yielded a probe melting peak with Tm at 64.2 ± 1.0°C, the wild-type samples yielded a melting peak with Tm at 60.2 ± 2.2°C, and the HbCC samples yielded a melting peak with Tm at 56.5 ± 0.6°C. The HbSC samples yielded two melting peaks; each corresponding to the respective peaks observed for the HbSS and HbCC homozygotes. Using melting profiles obtained from the single-probe dequenching experiments, we successfully genotyped 52 of 52 of the cases examined for mutations in the HbAA (n = 37), HbSS (n = 5), HbCC (n = 3), and HbSC (n = 7). There was 100% concordance between the results of the single-fluorophore dequenching and DNA sequencing or HPLC-based assays (Table 3)
.
| Discussion |
|---|
|
|
|---|
Several fluorescence properties such as intensity, half-life, and emission spectrum are altered as a consequence of hybridization.25,26 For instance, fluorescence quenching occurs during probe hybridization when fluorescein15,27 or BODIPY-FL28 is brought in close approximation to deoxyguanosine nucleotides. The probe can be labeled on either the 3' or the 5' end with similar quenching efficiency. Maximum quenching efficiency is achieved when the probe-target interaction is such that a G is present at the first overhang position on the target strand. Additional neighboring Gs on the target strand increase quenching incrementally, but a G in the first overhang position is most essential.27
The phenomenon of dequenching has recently been described and exploited for the analysis of a number of biological parameters. In this regard, dequenching has been used to measure the dilution of liposome-entrapped fluorophore caused by changes in membrane permeability or membrane fusion.29,30 Dequenching of a self-quenching fluorogenic probe labeled with octadecylrhodamine and specific for a hydrophobic binding pocket of the activator protein that is mutated in G(M2) gangliosidosis, has been used to characterize the oligosaccharide-binding specificity of the activator protein.31 With regard to nucleic acids, dequenching has been used for the analysis of RNA degradation in vitro and in vivo.32 In addition, Lee and colleagues25 have used fluorescence dequenching for kinetic studies of restriction endonucleases. The dequenching-based assay provided an easy and rapid method for acquiring detailed data density necessary for precise kinetic studies. However, dequenching-based strategies have not been used for the detection of single nucleotide polymorphisms or genotyping.
In this study, we show that the phenomenon of hybridization-induced fluorescence dequenching can be used for nucleic acid detection and genotyping. In our experiments, we directly conjugated fluorescein to a guanosine base at either the 5' or the 3' end of an oligonucleotide probe complementary to the target of interest. The close proximity of the fluorescein to the guanine base typically results in quenching of fluorescence.15,27 This deoxyguanosine-mediated quenching effect has previously been considered problematic for the design of FRET-based probes33 and in DNA sequencing.34,35 Nevertheless, hybridization of the fluorescein-labeled probe to the unlabeled complementary strand resulted in dequenching of the fluorophore, and an increase in fluorescence was observed. Although the signal generated by the dequenching approach was weaker than that obtained with the dual probe approach, our studies show that it is robust enough for routine genotyping of clinical samples. In all cases PCR amplification did not exceed 45 cycles, and consequently contamination was not a problem. Interestingly, we noted that the intensity of the fluorescence was influenced by the pH of the reaction solution, with increasing pH favoring increased fluorescence above background up to pH 9.2, at which the analyses were optimally performed and compatible with PCR amplification.
In conclusion, our studies show that the dequenching format of single fluorophore-based reporting systems for nucleic acid detection and PCR monitoring combine the advantages of simplicity, ease of design, and superior specificity to that provided by non-specific DNA binding dyes or intercalators. Further, single-labeled fluorophore based systems obviate the requirement for inclusion of an additional fluorophore without sacrificing specificity of product detection. The fluorescence of single-labeled probes is also reversible and depends only on hybridization of the probe to the target, allowing study of the melting characteristics of the probe from the target, thereby facilitating genotyping by Tm. We anticipate that the single-fluorophore dequenching format will be adapted for a diverse number of applications in molecular research and diagnostics.
| Acknowledgements |
|---|
| Footnotes |
|---|
Supported by grant CA83984 from the National Institutes of Health to K.S.J.E.-J., and by the ARUP Institute for Clinical and Experimental Pathology, Salt Lake City, Utah.
Accepted for publication March 19, 2003.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
E. Lyon and C. T. Wittwer LightCycler Technology in Molecular Diagnostics J. Mol. Diagn., March 1, 2009; 11(2): 93 - 101. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Vinayagamoorthy, K. Mulatz, M. Drebot, and R. Hodkinson Molecular Typing of West Nile Virus, Dengue, and St. Louis Encephalitis Using Multiplex Sequencing J. Mol. Diagn., May 1, 2005; 7(2): 152 - 159. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. P. Vaughn and K. S. J. Elenitoba-Johnson High-Resolution Melting Analysis for Detection of Internal Tandem Duplications J. Mol. Diagn., August 1, 2004; 6(3): 211 - 216. [Abstract] [Full Text] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |