| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
(1 Department of Pharmacology, University of Oxford, Oxford, UK;2 Institute of Medicine, Section of Pharmacology, University of Bergen, Bergen, Norway;
aaddress correspondence to this author at: Department of Pharmacology, University of Oxford, Mansfield Rd., Oxford OX1 3QT, United Kingdom; fax 44-1865-271882, e-mail helga.refsum{at}pharmacology.oxford.ac.uk
Many methods for detection of single-nucleotide polymorphisms (SNPs) are based on fluorescence resonance energy transfer (1). One of these, the Amplifluor SNP genotyping system, combines the use of universal energy transfer-labeled primers (uniprimers) (2) with allele-specific amplification (3). The uniprimer is a modification of the beacon concept (1)(4); it uses the beacon’s hairpin structure, but instead of acting as a probe, the hairpin is part of the primer, and when incorporated into the amplicon, it opens and starts to fluoresce (see Fig. S1 in the Data Supplement that accompanies the online version of this Technical Brief at http://www.clinchem.org/content/vol51/issue9/) (2). The advantage of the technology is that one pair of uniprimers is used for any SNP. Several variants of the uniprimer system for SNP detection are available (5)(6)(7). We modified the method so that it became faster and less expensive and then used it to identify 3 polymorphisms related to homocysteine and B vitamins (8): transcobalamin 2 (TCN2) 776C>G, methylenetetrahydrofolate reductase (MTHFR) 677C>T, and reduced folate carrier 1 (SLC19A1) 80G>A.
For each SNP, 5 primers were used: 2 uniprimers, 2 allele-specific primers, and 1 common primer for the opposite direction. The uniprimers (AmplifluorTM; Chemicon International) were labeled with either fluorescein or sulforhodamine and were provided by Flowgen. The other primers (Sigma) were designed by use of Amplifluor Assay-Architect (Serological Corporations) and Molecular Beacon Design software (Premier Biosoft International). The nonlabeled primers are listed in Table
S1 in the online Data Supplement.
|
Purified DNA was obtained from EDTA-blood samples by use of the Wizard DNA Purification Kit (Promega Corporation). The genotype distribution was investigated in 500 EDTA-blood samples from Norwegian blood donors. For these samples, we used boiled blood supernatant as the DNA source (9). Briefly, 50 µL of whole blood was mixed with 250 µL of Dulbecco phosphate-buffered saline without CaCl2 and MgCl2 (Sigma), boiled for 10 min, and subjected to centrifugation. The supernatant was stored at 4 °C until analysis. All blood samples were collected in Institutional Review Board–approved projects.
PCR was performed in 96-well thin-walled microtiter plates sealed with optical tape (Bio-Rad); 20-µL reactions were used. In the optimized assay, the reaction buffer contained 50 mM KCl, 20 mM Tris-HCl (pH 8.4), 3 mM MgCl2, 0.2 mM of each deoxynucleoside triphosphate, and 0.5 U of iTaq DNA polymerase (from Thermus aquaticus) as supplied in the iQ Supermix (Bio-Rad). For each reaction, we added 375 nM common primer, 25 nM of each allele-specific primer, 25 nM (TCN2 776C>G, SLC19A1 80G>A) or 50 nM (MTHFR 677C>T) of each uniprimer, and 50–100 ng of DNA or 2 µL of boiled blood supernatant.
PCR, fluorescence detection, data collection, and data analyses were carried out in the iCycler iQ real-time PCR instrument (Bio-Rad). After preincubation at 95 °C for 3 min, amplification was carried out for up to 65 cycles (denaturation at 95 °C for 30 s, annealing/extension for 30 s). The optimal annealing temperature, defined as the temperature giving the maximum difference in threshold cycle (Ct) between genotypes and nontemplate control, was found by use of the iCycler’s thermal gradient system. In the optimized method, annealing temperatures were 59 °C for TCN2 776C>G, 70 °C for MTHFR 677C>T, and 68 °C for SLC19A1 80G>A.
For each polymorphism, we had samples with known genotypes. TCN2 genotypes were determined by restriction enzyme digestion and gel electrophoresis (10). MTHFR genotypes were determined with a TaqMan real-time PCR method (11). Samples with known SLC19A1 80G>A genotypes were obtained by sequencing with the BigDye V3.1 Terminator Cycle Sequencing Kit (Applied Biosystems).
We first investigated the uniprimer assay with purified DNA samples, starting with the TCN2 776C>G polymorphism. The conditions were according to recommendations provided by the supplier (12), i.e., standard concentrations of uniprimers and nonlabeled primers, a low concentration of MgCl2, and use of TitaniumTM Taq DNA polymerase (BD Clontech®), and recommended PCR cycling conditions. However, in both gel electrophoresis and real-time PCR, we observed amplification products in nontemplate controls, and there was no differentiation between genotypes. The usual strategies to improve allele specificity and prevent primer-dimerization (12)(13)(14), including optimizing concentrations of various PCR components, use of different primer sets, or addition of a second mismatch in the allele-specific primers, did not help. However, when we diluted the uniprimers, nonspecific amplification was markedly reduced, and genotype discrimination improved (Fig. S2 in the online Data Supplement). The result was surprising: According to theory, an excess of uniprimer over allele-specific primers will ensure that uniprimer is preferentially used (5)(6). However, rather than decrease, the fluorescence yield increased, in particular in heterozygous samples. Subsequently, we have routinely used uniprimer concentrations of 25–50 nM (5- to 10-fold dilution). This modification has made the overall assay cost virtually independent of the labeling system.
For the TCN2 776C>G polymorphism, dilution of uniprimers was sufficient to obtain genotype discrimination. For MTHFR 677C>T and SLC19A1 80G>A, however, both alleles continued to be amplified in homozygous samples. We then investigated another strategy to improve allele specificity, i.e., increasing the annealing temperature (7)(14). This approach markedly improved discrimination for both MTHFR 677C>T and SLC19A1 80G>A (Fig. S3 in the online Data Supplement). The hairpin in the uniprimer unfolds at higher temperatures (2); the recommended annealing temperature is therefore 55–60 °C (12). However, under our conditions, which included a high MgCl2 concentration in the PCR buffer, the assay tolerated 68–70 °C without reduction in yield or an increase in background fluorescence. When we used a lower MgCl2 concentration (1.8 mM) during PCR, optimum discrimination was obtained at 65–67 °C. For MTHFR 677C>T, dilution of uniprimers and use of high temperature (
70 °C) led to excellent discrimination between the genotypes. For SLC19A1 80G>A, the optimal annealing temperature was 68 °C; genotype discrimination was then adequate but not excellent.
Even after we had optimized the PCR conditions, amplification products in nontemplate controls were sometimes observed, but such samples were easily recognized because amplification was either minimal or the curve had a different shape (Fig. S3 in the online Data Supplement).
We have previously used dried blood spots without DNA purification for genotyping (11), but this did not work with the uniprimer system. We then used phosphate-buffered saline–diluted boiled blood supernatant as the DNA source (9). Boiling removes major PCR inhibitors in the blood, such as hemoglobin, lactoferrin, and IgG (15)(16). The Ct increased 5 cycles compared with purified DNA, but each genotype cluster remained well separated. The results for each polymorphism in a typical run are shown in Fig. 1
, and the genotype distributions in a healthy population are listed in Table 1
.
|
We assessed the performance of the modified assay, using boiled blood supernatants. In a standard run, amplification failed or the pattern was unusual in 1–3 of 92 samples. These required repeat analysis, which rarely failed. Repeat analysis of samples with unambiguous results (starting from whole blood) gave 98%–100% concordance, suggesting that the method is precise and robust. For the MTHFR polymorphism, we compared the results obtained with the modified uniprimer system and a TaqMan assay (11). Four of the 92 samples tested (4.3%) gave discrepant results. Repeat analyses with both methods revealed that the uniprimer system had correctly identified 3, whereas the last sample gave a definite genotype with the uniprimer system but failed to amplify in the TaqMan assay.
To assess cost, we compared 3 methods: (a) restriction enzyme digestion and gel electrophoresis, (b) the uniprimer system using conventional concentrations of uniprimers and purified DNA, and (c) our optimized method with diluted uniprimers and boiled blood. We estimated cost based on the following: starting from whole blood, manual sample handling, DNA extraction by use of commercially available reagents, 20-µL PCR reaction, and purchasing of reagents for at least 1000 assays. The total cost included labor, instrumentation, consumables, and reagents. For the gel electrophoresis assay, the total cost/SNP was US $10.50, and genotyping 96 blood samples took 2.5 days. The corresponding values were approximately US $7.60 and 6–7 h for the conventional uniprimer technique and approximately US $3.80 and 5 h for our modified method. Approximately 40% of the cost of the modified assay is related to the real-time instrument and PCR reagents. The real-time instrument has advantages when optimizing new assays, but it is more expensive than a conventional PCR instrument combined with a fluorescence plate reader (5)(6). The latter alternative also allows extensive miniaturization of the assay (7). By combining our modifications with a miniaturized format (7), the cost of genotyping can be reduced by an additional 30%.
In conclusion, we used 2 simple strategies to improve allelic discrimination by the uniprimer system, i.e., reducing the uniprimer concentration 5- to 10-fold and, if necessary, increasing the annealing temperature to 68–70 °C. Moreover, the modified method can use boiled blood supernatant as the DNA source, which markedly reduces labor and the cost of genotyping. Although the method cannot match newer high-technology systems in throughput, it may be suitable in terms of instrument investment and ease of use for the many laboratories that continue to do their own genotyping.
Acknowledgments
We thank The Advanced Research Program of Norway for supporting the study. The funding source had no influence on design, experiments, and analyses of the data or the decision to submit the manuscript for publication. We are grateful to Dr. Joshua Miller (University of California, Davis), who gave us samples with known TCN2 776C>G genotypes.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  H. Refsum, C. Johnston, A. B. Guttormsen, and E. Nexo Holotranscobalamin and Total Transcobalamin in Human Plasma: Determination, Determinants, and Reference Values in Healthy Adults Clin. Chem., January 1, 2006; 52(1): 129 - 137. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
