| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
Genetic and Molecular Epidemiology Unit, Department of Preventive Medicine, School of Medicine, University of Valencia, Valencia, Spain;
aaddress correspondence to this author at: Department of Preventive Medicine, School of Medicine, Avda. Blasco Ibañez, 15, 46010 Valencia, Spain; fax 34-963864166, e-mail francesc.frances{at}uv.es
The recently identified apolipoprotein A-V gene (APOA5) has been shown to play an important role in hypertriglyceridemia (1). Genetic variation in APOA5 has been consistently associated with plasma triglyceride concentrations in several studies (2)(3)(4). Moreover, some studies have demonstrated additional associations with lipoprotein subclasses, remnant-like particles, and cardiovascular disease risk (4)(5)(6). Several single-nucleotide polymorphisms (SNPs) in the human APOA5 gene have been detected with differing frequencies depending on the population analyzed (7)(8), and Klos et al.(7) have also suggested context-dependent associations in different populations. Overall, 5 common SNPs, –1131T>C, –3A>G, 56C>G, IVS3 + 476G>A, and 1259T>C, have been widely reported. Apart from the 56C>G SNP, the other SNPs are in strong linkage dysequilibrium, producing 3 haplotypes (11111, 22122, and 11211) representing 
98% of the population in Caucasians (5)(9); therefore, 2 independent APOA5 SNPs (56C>G and –1131T>C) can be analyzed in association studies as indicators of the corresponding haplotypes. The former consists of a nonsynonymous substitution, changing codon 19 from serine to tryptophan (S19W), whereas the latter is a T-to-C substitution 1131 nucleotides upstream of the initiation codon. To date, in most published reports, these SNPs have been genotyped by PCR with restriction fragment length polymorphism (RFLP) analysis (3).
A system for high-throughput genotyping using fluorescence melting curve analysis, the LightTyperTM (Roche), has recently become commercially available. This instrument (10) offers higher throughput than the original LightCyclerTM (11). The LightTyper is designed explicitly for melting curve analysis to perform rapid, straightforward, reliable allelic discrimination. The system, which uses 384-well plates, provides postamplification genotyping within 10–15 min and performs genotyping automatically. A variety of probe chemistries are compatible for genotyping, including single-labeled probes and fluorescence resonance energy transfer probes (10). SimpleProbesTM are designed to specifically hybridize to a target sequence that contains the SNP of interest (12). Once hybridized, the SimpleProbe emits a larger fluorescent signal than when it is not hybridized to its target sequence. SimpleProbes are more cost-effective than fluorescence resonance energy transfer probes and represent a major advance in decreasing the cost and complexity of SNP analysis (12). We report here the validation of an assay based on the LightTyper system and SimpleProbes to rapidly screen for the 56C>G and –1131T>C SNPs in the APOA5 gene.
In the validation study we genotyped 825 randomly selected individuals (age range, 18–75 years) from the Spanish Mediterranean population. All participants gave informed consent, and the study protocol was approved by the Ethics Committee of the School of Medicine of the University of Valencia.
DNA, isolated from blood, was first genotyped for the 56C>G SNP in the APOA5 gene by melting curve analysis with the LightTyper; the results were then compared with those obtained with the classic RFLP method. For the melting curve analysis, a 136-bp fragment containing the 56C>G SNP was amplified with primers 5'-AGAGCCCAGGCCCTGATTA-3' and 5'-CATCTTCTGCTGATGGATCTGCT-3' (TIB MOLBIOL) together with the SimpleProbe Flq-TCTCCACAGCGTTTTCGGCC-p (TIB MOLBIOL), where Flq represents a fluorescence quencher. PCR was carried out in 384-well plates with a total volume of 10 µL per well in a Thermocycler (Mastercycler-ep380®; Eppendorf). The reaction mixture used in the PCR consisted of 40 ng of genomic DNA, 0.2 µL of each primer (10 µM), 1 µL of the SimpleProbe (1.6 µM), 5 U of Fast Start Taq Polymerase (Roche Diagnostics GmbH), 1.2 µL of 25 mM MgCl2, and 200 µM each deoxynucleotide triphosphate (Roche Diagnostics). After an initial denaturation at 94 °C, 34 PCR cycles were performed with 30 s of denaturation at 94 °C, 45 s of annealing at 55 °C, and 45 s of extension at 72 °C; final extension was at 72 °C for 10 min. A final melting cycle was performed on the LightTyper by heating to 85 °C and cooling to 40 °C at a ramping rate of 0.2 °C/s, with fluorescence data collected continuously. The assay exploits the thermal properties of DNA, i.e., the melting temperature (Tm). We designed the probe to match the wild-type DNA (56C allele), so that the wild-type DNA is 100% complementary to the fluorescent probe, making this complex more stable and thus giving it a higher Tm (65.5 °C). The presence of the mutation gives a lower Tm (56 °C). This difference in Tm allows genotypes to be assigned. The fluorescence signal (F) is plotted in real time against the temperature (T) to generate melting curves for each sample. From these curves, melting peaks (Fig. 1
) are generated by plotting the negative derivative of F with respect to T against T (–dF/dT against T) (13). The melting peak for homozygous CC samples occurs at a higher temperature than the melting peak for homozygous GG samples. In CG heterozygotes, both temperature peaks can be detected.
|
The genotyping results for the 825 individuals from the Mediterranean population are shown in Table 1
. To validate these results, we subsequently determined the 56C>G polymorphism in these 825 individuals by RFLP analysis. PCR was carried out in a total volume of 25 µL in the same thermocycler under standard conditions. The following forward and reverse primers were used for amplification (3): 5'-GGCTCTTCTTTCAGGTGGGTCTCCG-3' and 5'-GCCTTTCCGTGCCTGGGTGGT-3' (TIB MOLBIOL). This amplification was designed to force a G>A (T in the reverse primer, shown underlined), which introduced a TaqI restriction site in the rare allele (3). After restriction enzyme digestion with TaqI (Promega) at 65 °C for 2 h, the common C allele gave fragments of 134 and 23 bp, whereas the G allele gave a single 157-bp product. The fragments were separated by electrophoresis on a 4% Metaphor® agarose gel. The PCR conditions consisted of initial denaturing at 96 °C for 5 min, followed by 30 cycles at 96 °C for 30 s, 63 °C for 30 s, and 72 °C for 45 s, with final extension at 72 °C for 10 min. The genotype results obtained are shown in Table 1
.
|
Although the genotype distributions of the 825 samples successfully tested by both the LightTyper and RFLP analysis were identical and did not differ from the Hardy–Weinberg expectations (
2 =1.97; P = 0.161), we found 2 discrepant samples: 1 typed as GG by the LightTyper and GC by the RFLP method, and 1 typed as GC by the LightTyper and GG by the RFLP method. This represents a concordance rate of 99.75%. These 2 discordant samples were sequenced directly by standard methods (14) on an ABI Prism Automated DNA sequencer. Direct sequence analysis confirmed the melting curve analysis results for 1 of the 2 discrepant samples, and in the other sample direct sequencing confirmed the RFLP result. After individually checking the discrepant results, we found that the error in the RFLP method was the result of a partial digestion, which led to a spurious heterozygote result, whereas the error in the LightTyper was the result of background noise for the probe, which generated a "pseudopeak" corresponding to the mutant Tm. How-ever, the small amplitude of the peak and the deviation by 2 °C from the correct Tm suggested its spurious condition in the second revision of results. Regarding the RFLP method, although standard positive digestion controls were included, when hundred of samples are tested, inhibitory factors present in some samples can decrease the enzyme activity, leading to incorrect results. One possible solution is to generate a internal positive control for each sample, as proposed Danneberg et al. (15). Overall, our results indicate that the LightTyper system can successfully detect SNPs with an accuracy similar to that of the RFLP method. The cost in reagents is similar for both methods; therefore, because the LightTyper is quicker, it appears to be a better choice than the standard RFLP method.
Having validated the melting curve analysis obtained with LightTyper for high-throughput genotyping under real biomedical laboratory conditions, using the 56C>G polymorphism as an example, we also developed an assay to detect the –1131T>C polymorphism. We used the LightTyper and SimpleProbe approach with the primers 5'-CACATCCCTCTTTATGAAACAAT-3' and 5'-GTAGACGGAGTGGGTGTGTCA-3' (TIB MOLBIOL) to amplify a 229-bp fragment. PCR was carried out in 384-well plates in a total volume of 10 µL containing 40 ng of genomic DNA, 0.2 µL of each primer (10 µM), 1.6 µM SimpleProbe (Flq-AGGAACTGGAGCGAAAGTAAGATTT-p; TIB MOLBIOL), 5 U of Fast Start Taq Polymerase (Roche Diagnostics), 1.75 µL of 25 mM MgCl2, and 200 µM each deoxynucleotide triphosphate. The PCR conditions were as follows: initial denaturation at 96 °C for 5 min, followed by 40 cycles of 96 °C for 45 s, 53.5 °C for 45 s, and 72 °C for 45 s, with final extension at 72 °C for 10 min. The melting curve analysis was performed as indicated above. The Tms were 63.5 °C for the wild-type allele (T) and 58 °C for the variant allele (C); accordingly, we found 736 individuals who were homozygous TT, 86 who were heterozygous CT, and 3 who were homozygous CC.
Carriers of the less common allele were grouped, and a crude association analysis was carried out (results not shown). The 56C>G polymorphism was statistically associated (P = 0.04) with higher mean triglyceride concentrations, but the association for the –1131T>C SNP did not reach statistical significance (P = 0.21). Nevertheless, a more detailed analysis including control for covariates and an increased sample size is needed in this population. In this regard, it is interesting to note that our allele frequencies for the variant alleles (particularly for the –1131T>C polymorphism) were slightly lower than those reported in other populations (7)(8).
Finally, to take a full advantage of this new technique, we also tested several primer concentrations in asymmetric PCR assays and found that forward/reverse primer ratios of 0.26 for 56C>G and 0.20 for –1131T>C improved PCR efficiency, decreasing the failure rate near to 0%.
Acknowledgments
We gratefully acknowledge F. Gimenez-Fernández for laboratory assistance. F.F. is the recipient of a fellowship from the University of Valencia (V Segles). This work was supported by Grants PI02/1096 and PI042239 from the FIS, Instituto de Salud Carlos III, Spain.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
