Clinical Chemistry · 2004 · 14 citations · 2 references
Life SciencesGeneticsMolecular GeneticsGenomicsMedical GeneticsMolecular DiagnosticsDna SequencingMutation AnalysisMutation ScreeningPolymerase Chain ReactionRestriction Enzyme TechnologyVariant DetectionAmplified Fragment Length Polymorphism AnalysisAmplified Fragment Length Polymorphism ValidationGenetic Mutation DetectionPoint MutationsPolymerase Chain Reaction (Pcr) AssaysPrimer DesignRestriction Enzyme DigestionRestriction Fragment Length Polymorphism Analysis
For the detection of point mutations (or variant alleles), PCR followed by restriction fragment length polymorphism (RFLP) detection is widely used and is probably one of the simplest methods for detection of known mutations. This method depends on the fact that the mutation of interest disrupts a recognition sequence for a restriction enzyme. When a restriction site is not present around the mutation, mismatch primers need to be designed that do create a restriction site. Such a mismatch primer is thus usually located directly adjacent to the mutated nucleotide. We validate our PCR-RFLP assays by direct sequencing of the PCR product. In two of our assays, however, we found that none of the three samples found to be heterozygous by PCR-RFLP was confirmed as heterozygous by direct sequencing. Detection of the CYP3A53 polymorphism (A6986G) is based on amplification of a 293-bp DNA fragment digested with Ssp1 (1). For amplification, we used 40 pmol of primers P1 (5′-CATGACTTAGTAGACAGATGAC-3′) and P2 (5′-GGTCCAAACAGGGAAGAAATA-3′) in a final volume of 50 μL. The reverse primer P2 [which contains a mismatch (underlined nucleotide) to create a restriction site for the Ssp1 restriction enzyme] is located directly next to the mutated nucleotide. Fig. 1A shows the PCR-RFLP results for a wild-type (g.6986A), a heterozygous, and a homozygous variant (g.6986G) individual. For validation, the PCR products of six samples were subjected to direct sequencing on an automated ABI 310 capillary sequencer using Big Dye terminator chemistry (Applied Biosystems) with 10 pmol of primer P1 (which was also used in the PCR) and 2 μL of PCR product in a final volume of 20 μL (25-cycle sequencing reaction). The homozygous wild-type and homozygous variant alleles were confirmed by the sequencing results, but the sequencing results for the three heterozygous samples indicated a homozygous variant genotype (Fig. 1B ). PCR-RFLP (A) and direct sequencing results (B and C) for the CYP3A5 allele. (A), PCR-RFLP results for CYP3A5*3. The PCR products were analyzed on a 3% agarose gel with a 50-bp DNA marker. Lane O, uncut PCR product; lane 1, homozygous *3/*3 individual; lane 2, heterozygous *1/*3 individual; lane 3, wild-type *1/*1 individual. (B and C), direct sequencing results for products obtained with primers P1 and P2 (B) and with primers P3 and P4 (C) for a homozygous *3/*3 individual (panel 1), a heterozygous *1/*3 individual (panel 2), and a wild-type *1/*1 individual (panel 3). We performed an alternative CYP345 PCR assay with primers P3 (5′-GGAGAGTGGCATAGGAGATA-3′) and P4 (5′-CCCTTCGATTTGTGAAGACAG-3′) on a homozygous wild-type, a heterozygous, and a homozygous variant DNA sample, as identified by the PCR-RFLP results obtained with primers P1 and P2. These primers anneal to the CYP3A5 gene at other positions farther from the mutation site. We sequenced the PCR products of these samples with primer P3 or P4 in the sequencing reaction, and the sequencing results agreed completely with the PCR-RFLP results (Fig. 1C ). An identical observation was obtained in our PCR-RFLP assay for detection of the CYP3A41B polymorphism [A(−290)G] (2). In this assay, a 334-bp DNA fragment was amplified by use of primers P5 (5′-GGACAGCCATAGAGACAACTGCA-3′) and P6 (5′-CTTTCCTGCCCTGCACAG-3′). In this test, the forward primer P5 [containing a mismatch (underlined nucleotides) to create a restriction site for the Pst1 restriction enzyme] is located directly adjacent to the mutated nucleotide. Direct sequencing with primer P6 confirmed the wild-type [g.(−290)A] and homozygous variant [g.(−290)G] PCR-RFLP results, but the heterozygous PCR-RFLP result indicated the homozygous variant allele. We also performed another PCR test with primers P6 and P7 (5′-AACAGGCGTGGAAACACAAT-3′); P7 anneals to the DNA strand farther from the genetic polymorphism. After direct sequencing of these PCR products with primer P6 or P7, the results totally agreed with the PCR-RFLP results (results not shown). Because the only difference between the two alleles that are amplified in a heterozygous sample is the mutated nucleotide, we hypothesize that the addition of the first nucleotide might stabilize the primer–template–Taq polymerase complex better when a G (or a C) is incorporated (CYP3A53 and CYP3A41B alleles) compared with an A (or a T; CYP3A5 and CYP3A4 wild-type alleles). Subsequently, the G (C) allele is more efficiently amplified, leading to a failure to detect the A (T) allele in heterozygous samples. In densitometric scans of 26 independent PCR-RFLPs, the mean (SD) staining intensity of the 148-bp CYP3A5 fragment from the wild-type allele was 70 (5)% of the staining for the 168-bp CYP3A53 mutant allele-derived fragment. Taking into account that smaller fragments are always stained less intensely, the inhibition of amplification is thus maximally 30%. On the basis of scans of the areas of the sequencing peaks, we conclude that the wild-type peak area is only 10% compared with the mutant peak area. There thus seems to be an additional selection in the sequencing reaction. To determine to what extent this phenomenon was influenced by the use of a mismatch primer in the original PCR, we performed a PCR for CYP3A53 with forward primer P1 and reverse primer P2a (5′-GGTCCAAACAGGGAAGAGATA-3′), in which the mismatch was omitted. We sequenced the PCR products of these samples with primer P1 in the sequencing reaction. Again our genotyping software designated only G at the mutated position, although closer visual inspection revealed the presence of a small A peak (peak area 34% of the mutant peak area compared with 10% of the mutant peak area when mismatched primers were used). However, this signal was not as strong as the signal obtained in the sequencing reaction of the PCR product with the distant primers P3 and P4 (peak area of wild-type nucleotide was 65% of the peak area of the mutant nucleotide). On the basis of these findings, we concluded that the PCR-RFLP tests detected the correct genotype and that direct sequencing of the PCR products obtained with one of the primers located adjacent to a mutated nucleotide may cause unequal amplification of alleles in heterozygous samples. This effect is even stronger when mismatch primers are used. This work therefore describes a potential pitfall in DNA sequencing, indicating that sequencing may not always be the “gold standard”.
2
CYP3A5 Variant Allele Frequencies in Dutch Caucasians
Ron HN van Schaik, Ilse P. van der Heiden, John N. van den Anker et al. · Clinical Chemistry · 2002 · 303 citations · Full text
Medicine, Genetics, Pharmacotherapy +15
Ron H. N. van Schaik, Saskia N. de Wildt, Nienke M van Iperen et al. · Clinical Chemistry · 2000 · 66 citations · Full text