Full COLD-PCR Protocol for Noninvasive Prenatal Diagnosis of Genetic Diseases

To the Editor: After the discovery of fetal DNA in maternal plasma, investigators reported different strategies for the noninvasive prenatal diagnosis of genetic diseases (1). Despite the advances in improving the analytical sensitivity of methods, distinguishing between fetal and maternal sequences remains very challenging, and the field of noninvasive prenatal diagnosis of genetic diseases has yet to attain a routine application in clinical diagnostics. An innovative strategy that has been reported is based on COLD-PCR (coamplification at a lower denaturation temperature PCR),1 which exploits melting temperature (Tm) differences between variant or mismatched sequences and wild-type sequences. This approach uses a critical denaturation temperature (Tc) lower than the Tm to selectively amplify minority mutated alleles (2). We have developed assays for the identification of fetal paternally inherited mutations in maternal plasma. These assays use full COLD-PCR for the detection of IVSI.110 (G>A) and Cd39 (C>T) HBB (globin, beta) gene mutations that cause β-thalassemia. Full COLD-PCR is based on the generation of heteroduplexes between mutant and wild-type sequences. These heteroduplexes melt at lower temperatures than the wild-type homoduplexes. They are selectively denatured at the Tc and then subsequently amplified. Given that both the fetal and maternal DNA content in maternal plasma has been demonstrated to vary from pregnancy to pregnancy (3), we used a range of Tc values and several replicates of the same sample to develop a protocol that was specific for the identification of each mutation. To simplify the Tc evaluation, we amplified plasma DNA from wild-type controls with a conventional PCR protocol and several replicates of the same sample and varied the denaturation temperature. The lowest denaturation temperature that still allowed amplification of wild-type samples (81.9 °C for IVSI.110, 83.8 °C for Cd39) was chosen as the upper limit of a Tc interval, which spanned 1 °C. We scaled the denaturation temperatures for 8 replicates of the maternal plasma sample down by increments of 0.1 °C–0.2 °C, beginning with the upper Tc value identified. The use of a range of potential Tc values with several replicates instead of a single Tc ensured that at least 1 replicate exhibited differential denaturation of mutant and wild-type alleles, thereby allowing robust PCR amplification. The study was approved by the local ethics review boards. After obtaining written informed consent from couples in which the parents carried different mutations, we collected 6 mL of maternal blood before chorionic villus sampling. DNA was extracted from 500 μL of plasma (4) and eluted in a 60-μL volume. COLD-PCR was performed in a 25-μL final volume containing 5 μL of eluted DNA, 200 μmol/L of each deoxynucleoside triphosphate, 10 mmol/L Tris–HCl (pH 8.3), 50 mmol/L KCl, 1.5 mmol/L MgCl2, 1.5 U of FastStart Taq DNA polymerase (Roche Applied Science), and 10 pmol of each primer. Cycling conditions were as follows: 95 °C for 4 min; 25 cycles of 95 °C for 15 s, 54 °C for 30 s, and 72 °C for 1 min; 72 °C for 10 min; 95 °C for 4 min; 35 cycles of 95 °C for 15 s, 70 °C for 8 min (to generate heteroduplexes), specific Tc temperature for 3 s, and 72 °C for 1 min; 72 °C for 7 min. For IVSI.110, the Tc interval was 81.9 °C–80.9 °C; for Cd39, the Tc interval was 83.8 °C–82.8 °C. The amplification primers for IVSI.110 were 5′-TAAGGAGACCAATAGAAACT-3′ (forward) and 5′-GTAGACCACCAGCAGC-3′ (reverse); the amplicon size was 119 bp. The primers for Cd39 were 5′-GTCTATTTTCCCACCCTT-3′ (forward) and 5′-AGCACTTTCTTGCCATGA-3′ (reverse); the amplicon size was 134 bp. After COLD-PCR, samples were sequenced directly with the 3730 DNA Analyzer (Applied Biosystems). We performed 35 diagnoses—for 21 cases in which the father carried the Cd39 mutation (the fetus inherited the paternal mutation in 10 of these cases) and for 14 cases in which the father carried the IVSI.110 mutation (the fetus inherited the paternal mutation in 12 of these cases). In all cases, the fetal paternal mutated allele was not detectable by conventional PCR, whereas the mutation was evident after COLD-PCR. Fig. 1 shows examples of samples lacking the paternally inherited mutated allele, as well as the lowest and highest enrichments of fetal paternally inherited mutated alleles for both mutations. The results obtained with COLD-PCR were in complete concordance with those obtained for fetal DNA extracted from chorionic villi (Fig. 1). Indicated are lowest (1) and highest (2) enrichment of the fetal mutated allele after COLD-PCR and sequencing; results for fetal DNA extracted from chorionic villi (3); fetal mutated allele undetectable with conventional PCR and sequencing (4); and no visible peak of mutated allele when the fetus inherited the paternal wild-type allele (5). Du et al. described the use of fetal circulating DNA and fast COLD-PCR to detect congenital paternal abnormalities in a single individual (5). Fast COLD-PCR cannot detect all mutations, however. To our knowledge, the present report is the first to indicate the feasibility of using full COLD-PCR for prenatal diagnosis in a comprehensive study. We have provided evidence that COLD-PCR enables straightforward and reliable identification of inherited mutated alleles without the need for sophisticated and costly equipment. The method might be extended to noninvasive prenatal diagnosis of genetic diseases, and it has the potential to be easily transferable to clinical diagnostic laboratories. coamplification at lower denaturation temperature PCR melting temperature critical denaturation temperature.