Single-nucleotide polymorphisms have been detected by various methods, including allele-specific oligonucleotide hybridization (1), allele-specific amplification (2), and restriction fragment length polymorphism (RFLP) analysis (3). In some cases, sequencing of amplified DNA has been used as a direct method (4)(5). RFLP-based methods appear to be superior in genotyping studies. In fact, recently, Cascorbi and Roots (6) recommended the use of RFLP in preference to other methods for single-nucleotide polymorphism detection in the N-acetyltransferase-2 (NAT2) gene. Some groups have used more than one method in their analyses, and when discrepancies occurred between two methods, they used DNA sequencing as a gold standard (7). In this report, we describe a comparison of RFLP methods with automated DNA sequencing for the detection of three different mutations in the NAT2 gene. Our results indicate that caution is needed in the use of automated sequencing to detect heterozygous mutations accurately at some polymorphic sites in the NAT2 gene. Established PCR-RFLP methods were used to detect NAT2 mutations at the C282T, T341C, and C481T sites. The C-to-T mutations at the 282 and 481 sites were detected with RFLP analysis of a 360-bp DNA fragment with FokI or KpnI digestions, respectively (3). Two different RFLP methods were used for detection of the T341C mutation: a previously described DdeI method (3), and a new complementary NcoI-RFLP. For both methods, a seminested PCR was performed to amplify a 360-bp DNA fragment using a 998-bp DNA fragment as a template. The latter fragment was preamplified as described previously by Hickman and Sim (8), using their Nat-Hu14 and Nat-Hu16 primers. The seminested PCR mixture (50 μL) contained 2 μL of 100-fold diluted 998-bp DNA, 200 μM each dNTP, 1.0 μM primers, 1.25 U of Taq DNA polymerase (Life Technologies), and a buffer consisting of 10 mM Tris-HCl, pH 8.3, 50 mM KCl, and 1.5 mM MgCl2. Amplification was performed for 30 cycles in a thermocycler (Hybaid Express), with the following conditions for each cycle: 30 s at 95 °C, 1 min at 60 °C, and 30 s at 72 °C. For our NcoI method, the Nat-Hu14 primer (8) was used with a new mutagenesis primer, 341RC (5′-C GAC AAT GTA ATT CCT GCC GTC C-3′). In the DdeI method, the same Nat-Hu14 primer was used together with another modified mutagenesis primer, MS341R (5′-GAC AAT GTA ATT CCT GCC CTC A-3′). The underlined bases in the 341RC and MS341R primers represent mismatched bases with template DNA. RFLP analysis was performed using either 5 U of NcoI (New England Biolabs) or 5 U of DdeI (Life Technologies) in a final buffer volume of 20 μL under mineral oil for 16 h at 37 °C. The resulting fragments were resolved on a 10% polyacrylamide gel. Sequence analysis was performed for 20 DNA samples in a commercial laboratory (Waikato, Hamilton, New Zealand) using a gel-purified 998-bp DNA fragment as a template. Both DNA chains were sequenced using the Nat-Hu14 primer (8) in the forward sequencing and a new primer, MS 472R (5′-TTC CTC TCT CTT CTG TCA AGCA-3′), for reverse sequencing. The sequencing reactions were performed with an ABI Prism 377 automated DNA sequencer (PE Applied Biosystem) that utilized the BIG DYE terminator chemistry. Using these automated DNA sequencing and PCR-RFLP methods, we genotyped 20 DNA samples at three polymorphic sites in the NAT2 gene (C282T, T341C, and C481T). Table 1 shows a comparison of these genotyping results. The genotypes determined at the C481T site for all 20 samples were in complete agreement with the sequencing and KpnI-RFLP methods. However, the sequencing and RFLP methods yielded different genotyping results for four samples (Table 1 , samples 8, 9, 10, and 11). These samples were all identified as heterozygous at the C282T and T341C sites by the RFLP methods used, but in the sequencing data, all four samples appeared as homozygous wild type at the T341C site, and one (sample 8) was identified as a homozygous mutant (T/T) at the C282T site. Genotypes at three sites in the NAT2 gene by RFLP and sequencing methods.1 Quantification of some peaks was obtained by cutting and weighing of the peaks (main peak was set as 100%). Sequencing at the 481 polymorphic site was obtained only for one DNA strand because of the location of the sequencing primer used. Genotypes at three sites in the NAT2 gene by RFLP and sequencing methods.1 Quantification of some peaks was obtained by cutting and weighing of the peaks (main peak was set as 100%). Sequencing at the 481 polymorphic site was obtained only for one DNA strand because of the location of the sequencing primer used. RFLP analysis of the T341C mutation in some representative samples using the NcoI method and a complementary DdeI method is shown in Fig. 1 A. A 360-bp DNA fragment was digested with NcoI or DdeI, respectively, and the resulting fragments were separated by electrophoresis on a 10% polyacrylamide gel. A common fragment, 136 bp (in NcoI) or 150 bp (in DdeI), was produced in all samples because of the presence of a nonpolymorphic NcoI or DdeI site within the 360-bp DNA fragment. In addition to these common fragments, two diagnostic fragments were produced in both methods: 221 and 198 bp in the NcoI method, and 211 and 187 bp in the DdeI method. Heterozygous samples contained two diagnostic fragments, whereas homozygous wild-type or mutant samples had only one of the diagnostic fragments. One sample (Fig. 1A , lane 1/15) was prepared as an artificial heterozygote by mixing equal amounts of amplified DNA from sample 1 and sample 15. This artificial mixture was identified correctly as a heterozygote by both RFLP methods used, but not by automated DNA sequencing. RFLP (A) and sequence (B) analysis of the T341C site of the NAT2 gene. (A), 360-bp amplified DNA was digested with NcoI (bottom) and DdeI (top), and the resulting fragments were separated by electrophoresis on a 10% polyacrylamide gel at 100 V for 140 min. The gels were stained with ethidium bromide to visualize the DNA fragments. The lane numbers correspond to the sample numbers shown in Table 1 . Lane 1/15 is an artificial heterozygous mixture, prepared by mixing equal amounts of samples 1 and 15. Lane M contains 100-bp ladder as DNA size markers. (B), sequencing of the NAT2 gene around the T341C site was performed in a commercial laboratory using the ABI Prism 377 DNA sequencer (PE Applied Biosystems) with BIG DYE terminators. Each sample was sequenced for both DNA strands [forward (F) and reverse (R)]. The arrowheads indicate the polymorphic T341C site of the NAT2 gene. Sequencing results for some representative DNA samples, including the artificial heterozygous mixture, are shown in Fig. 1B . The peaks produced at the T341C site (shown with arrowheads) were quantified for both the forward and reverse sequencing reactions (Table 1 ). As expected, several heterozygous DNA samples (Table 1 , samples 4, 7, 12, 14, and 16) all contained two peaks with good yields at the 341 mutation site, whereas homozygous wild-type or mutant samples had only one specific peak (e.g., Table 1 , samples 1 and 15). However, the automated sequencing could not identify the heterozygosity in sample 20, which was an equimolar mixture of samples 1 and 15. A similar problem was observed in the detection of heterozygotes in other four samples at the T341C mutation site, and one sample for the C282T site (Table 1 , samples 8, 9, 10, and 11). We cannot offer a plausible explanation as to why the detection of heterozygotes was missed in a few DNA samples (4 of 20) by automated DNA sequencing. This may be attributable to inherent problems in the fluorescence-based DNA sequencing, especially for heterozygote detection. One problem is possibly related to the type of DNA polymerase used in the procedure. Leren et al. (4) observed that a G/A heterozygote in the low-density lipoprotein receptor (LDLR) gene was identified correctly by sequenase sequencing, but was missed by sequencing with Taq DNA polymerase. The enzyme used in our case (ABI 377; PE Biosystems) was Ampli-Taq DNA polymerase. Another problem may be attributable to the production of uneven peak-height patterns, which are caused by differences in the efficiency of dideoxy termination at different bases and are affected by sequence context as well. Zakeri et al. (5) reported that a heterozygous C peak was much smaller in size than a heterozygous G peak. This is in agreement with our results. In addition to the production of uneven peaks, a high background, produced by impurities in the DNA template, could further complicate the results, especially for heterozygote detection. In our case, the background in the electropherograms was fairly low because all the 998-bp DNA fragments were purified by gel extraction. The quality of sequence data was actually very good for each sample, allowing us to read a sequence of ∼500 nucleotides, which matched completely with the published NAT2 gene sequence (9). However, in a few samples, the signals produced by heterozygosity at the T341C and C282T sites were below the set default value (30%) for heterozygote detection. For example, our quality-control sample (sample 20) could be scored as a borderline heterozygote in the reverse sequencing reaction (27% G and 100% A) but not with the forward primer. If the threshold value of heterozygote detection was set lower, e.g., 20%, then sample 10 could also be scored as a heterozygote, at least in one sequencing direction. When a scoring problem arises for a sample after sequencing in both directions, it should be considered an undefined genotype until an independent method is used for its identification. We therefore used independent PCR-RFLP methods to check the genotypes of all 20 samples at three different polymorphic sites in the NAT2 gene, and discrepancies were found between the sequencing and RFLP methods for only four test samples (Table 1 , samples 8, 9, 10, and 11). In conclusion, there are some inherent problems in automated DNA sequencing, which may lead to inaccurate heterozygote identification in some samples. Until the associated problems are fully resolved, precautions should be taken in the use of automated sequencing for heterozygote detection. If possible, we recommend the use of two complementary PCR-RFLP methods in this type of analysis.
No takes yet. Share an insight, caveat, or question.
Şimşek et al. (2001) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: