Peptide-nucleic acid-mediated enriched polymerase chain reaction as a key point for non-invasive prenatal diagnosis of β-thalassemia

Abstract

The presence of fetal DNA in maternal plasma can be exploited to develop new procedures for non-invasive prenatal diagnosis. Tests to detect 7 frequent β-globin gene mutations in people of Mediterranean origin were applied to the analysis of maternal plasma in couples where parents carried different mutations. A mutant enrichment amplification protocol was optimized by using peptide nucleic acids (PNAs) to clamp maternal wild-type alleles. By this approach, 41 prenatal diagnoses were performed by microelectronic microchip analysis, with total concordance of results obtained on fetal DNA extracted from chorionic villi. Among these, 27/28 were also confirmed by direct sequencing and 4 by pyrosequencing.

Introduction

Since the identification of fetal DNA in maternal plasma,1 several attempts have been made to develop alternative approaches for non-invasive prenatal diagnosis of genetic diseases. The identification of paternally inherited mutations in maternal plasma is particulary challenging mainly due to the predominant presence of maternal DNA (>95%) which shares the same alleles with the fetus and competes with the fetal DNA during the amplification process. Recently, assays to detect 4 β-globin gene mutations causing thalassemia in Southeast Asians were performed by MALDITOF mass spectrometry analysis.2 Paternally inherited alleles were correctly identified in 20 of the 23 pregnancies. This approach is sufficiently specific to detect fetal mutations directly from maternal plasma, although the instrumentation is too sophisticated and expensive for routine clinical use. Subsequently, a size separation strategy was developed using electrophoresis and excision of the gel fraction containing short-sized DNA for selective enrichment of fetal sequences.3,4 This protocol was used in combination with PCR amplification with a peptide-nucleic-acid (PNA) clamp, a strategy previously proposed by our group to selectively suppress amplification of maternal allele,5,6 and was applied to non-invasive diagnosis of β-thalassemia in 32 pregnancies, with 100% sensitivity and 93.8% specificity in mutation detection.3 Nevertheless, the enrichment of fetal sequences through gel elution is problematic and prone to contamination. As an alternative approach, we evaluated a microelectronic microchip technology that proved to be highly reliable in the identification of mutations in several disease genes7–11 coupled with PNA-clamping of maternal wild-type allele to preferentially amplify fetal mutated sequences. PNAs are DNA analogs in which the ribose-phosphate backbone is replaced by a peptide backbone. PNA/DNA hybrids have much higher thermal stability than corresponding DNA-DNA hybrids, but are more destabilized by single base-pair mismatches. Thanks to their properties, specifically designed PNAs could interfere with amplification of the wild-type allele while still allowing amplification of the mutant allele. Therefore, any PCR-based mutation detection technology can be combined with allelic suppression by sequence specific-PNA clamping.12,13 The strategy was applied to the identification of fetal paternally inherited mutations on DNA extracted from maternal plasma in couples at risk for beta-thalassemia and other hemoglobinopathies undergoing invasive prenatal diagnosis. In addition, we also evaluated the direct sequencing and pyrosequencing14,15 methodologies in a subset of diagnoses.

Design and Methods

Patients

Plasma samples were collected prior to chorionic villus sampling (9–12 weeks of gestation) from 41 women in couples at risk of β-thalassemia where the two parents carried different mutations. Detailed written informed consent, as approved by the local Ethical Review Boards, was obtained for each woman.

Sample preparation

DNA was extracted from 500 μL of maternal plasma16 with the QIAamp DSP Virus kit, Qiagen.

Table 1.Sequences of primers, reporters, stabilizers, oligonucleotides and peptide-nucleic acids.

Polymerase chain reaction conditions

PCR was performed in 50 μL containing 15 μL of plasma DNA, 200 μM dNTPs, 10 mM Tris-HCl, 1.5 mM MgCl2 1.3 U of thermostable DNA polymerase (AmpliTaq Gold; Applied Biosystems), 20 pmoles of each primer and PNA. Cycling conditions were: 1 cycle at 95°C for 10 mins., 47 cycles at 95°C for 45 secs., 60°C for 45 secs., (for PNA annealing),17 45 secs., at the annealing temperature specific for each primer set (for primer annealing and extension), followed by 1 cycle at 72°C for 10 mins. For the Lepore-Boston mutation, primers specific for the rearranged chromosome were use18 (Table 1).

PNA-mediated enriched polymerase chain reaction

PNAs (Applied Biosystems) were designed complementary to the wild-type sequence of 6 beta-globin gene mutations: IVSI-1(G>A) IVSI-6(T>C), IVSI-110(G>A), Cd39(C>T), IVSII-745(C>G) and HbS(A>T) (Table 1). Mutations IVSI-1 and IVSI.6 share the same PNA. For the Lepore-Boston mutation, PNA clamping is not necessary since the rearranged chromosome is not present in maternal DNA. PNAs were designed to have a melting temperature approximately 15°C higher than that of the PCR primer sets.

Microchip analysis

We used the NMW 1000 NanoChip™ Molecular Biology Workstation by Nanogen (San Diego, CA, USA). In this system, the amplified DNA fragment is directed electrophoretically into the chip, and then hybridized with fluorescently labeled DNA probes specific for the mutant and wild type sequences, and the presence or absence of the mutation is detected by the fluorescence signal. Conditions for stabilizer and reporter design, microchip addressing, hybridization, thermal stringency and fluorescence detection have been previously described.7–11

Direct sequencing

Direct sequencing was performed using ABI Prism 3100 Genetic Analyzer (PE Biosystems, Foster City, CA, USA).

Pyrosequencing

Pyrosequencing (Pyrosequencing AB, Uppsala, Sweden) is a real-time sequencing method that relies on the sequential incorporation of nucleotides in a primer-directed polymerase extension reaction.14,15

Conditions were developed for the analysis of IVSI-110, Cd39, IVSI-1 and IVSI-6 mutations on amplicons amplified with the same primers used for the microchip analysis (Table 1). For the pyrosequencing reaction, primer 5'-CACTGACTCTCTCTGCCT-3' was used for IVSI-110, 5'-GGTCTACCCTTGGACC-3' for Cd39 and 5’-TTGGTGGTGAGGCCC-3’ for IVSI-1 and IVSI-6 mutations.

Figure 1.Non-invasive prenatal diagnosis on maternal plasma performed by the microchip system (A and C: left panels), direct sequencing (A and C: right panels) and pyrosequencing (panels B) for the Cd39 (A and B), and IVSI-6 (C) mutations in pregnant women. For the microchip analysis, maternal plasma was analyzed in the absence of PNA and in the presence of increasing amounts of PNA (pmoles in 50 μL final volume of the PCR reaction), in parallel with wild-type (wt) and heterozygous (het) genomic DNA control samples (25 ng). Fluorescence hybridization signals for wild-type (grey) and mutant (black) alleles are shown. In couples analyzed in panels A and B the fetus inherited the paternal mutated allele, in panel C the fetus inherited the paternal wild-type allele.

Results and Discussion

In an attempt to identify efficient and accessible methodologies for non-invasive prenatal diagnosis of genetic diseases we evaluated a microelectronic chip and two sequencing-based techniques. They were all coupled with a PNA-mediated mutant enrichment amplification strategy because these methodologies are not sufficiently sensitive to directly identify fetal mutated alleles in maternal plasma. One of the advantages over other microarrays is that the microelectronic microchip is highly reliable in the detection of sequence variations and can be easily customized by the end-user.10 We also evaluated other highly accurate and user-friendly methodologies. Among accessible technologies for genotyping, nucleic acid sequencing is fast becoming a diagnostic tool in routine use in clinical laboratories. In our study, the sequencing approach proved to be reliable and could be used to confirm results obtained by other methods. Pyrosequencing is also a good approach because it is flexible and easy to use. In the present study, it proved to be highly effective in detecting minority mutated alleles. A total of 41 prenatal diagnoses on maternal plasma were performed by the microchip analysis, with complete concordance with results obtained on fetal DNA extracted from chorionic villi and analyzed with standard procedures. Among these, 27/28 were also confirmed by direct sequencing and 4/4 by pyrosequencing (Table 2 and Figure 1). The results of two independent analyses perfomed by the microchip system did not agree only in one couple tested for the paternal Cd39. The first analysis indicated the presence of a fetal mutated allele in 3 out of 5 PNA-clamped amplified samples. The second analysis showed the absence of a mutated fetal allele in all the 5 PNA-clamped samples. Sequencing confirmed the absence of the mutated allele at all PNA concentrations, and the fetus was diagnosed as having inherited the wild-type paternal allele, in accordance with results on chorionic villi. This discrepant result could be ascribed to a contamination problem since in the same PCR session 25 ng of genomic DNA samples either heterozygous or homozygous for the Cd39 mutation had been amplified as controls. This amount of genomic DNA is much more abundant than plasma DNA and its amplification for 47 cycles may give rise to contamination. To overcome this problem, in all subsequent analyses only plasma positive control samples (father carrying the mutation under investigation) were included in the PCR session. No further contamination problems were observed. The correct identification of the fetal wild-type paternal allele in a couple where a previous fetus had been correctly identified as a carrier of the paternal mutated Lepore-Boston allele further confirms the complete clearance of fetal DNA from maternal plasma after delivery thus preventing diagnostic errors due to persistence of fetal DNA from previous pregnancies.19 A key point of our strategy is the enrichment of minority mutated fetal alleles by PNA-clamping. The major drawback of this approach is the assessment of the optimal PNA concentration which needs to be established for each mutation and may be difficult to be applied for routine analysis and large scale clinical applications despite the potential clinical usefulness. For non-invasive prenatal diagnosis, each maternal plasma sample was tested for the optimal range of PNA concentrations specific for each mutation of interest (over the range shown in Table 1). Due to the variability of total and fetal DNA content in each maternal plasma sample, slight refinements of this scale were sometimes necessary. For each method (microelectronic microchips, direct sequencing and pyrosequencing), independent PCR reactions in the presence of the optimal PNA concentration specific for each mutation of interest (Table 1) were performed on DNA extracted from the same maternal plasma samples. The strategy we propose could provide the basis for developing additional assays to identify paternally inherited β-globin gene polymorphisms to also extend non-invasive diagnosis to couples where both parents carry the same mutation. Our data suggest that a wide range of easy accessible approaches could be used in combination with PNA clamping for non-invasive pre-natal testing of a variety of genetic diseases, with a view to future application in clinical diagnostic settings. Non-invasive approaches could avoid the risk of fetal loss and the costs associated with invasive procedures such as chorionic villus sampling and amniocentesis in the 50% of couples for whom the fetus did not inherit the paternal mutated allele. The use of non-invasive approaches early in pregnancy would still allow the remaining couples time to undergo subsequent invasive procedures.

Table 2.Non-invasive prenatal diagnoses of β-thalassemia on maternal plasma.

Footnotes

• Funding from the European Commission for the Special Non-invasive Advances in Fetal and Neonatal Evaluation (SAFE) Network of Excellence (LSHB-CT-2004-503243), Telethon (Project n. GGP04016) and Regione Piemonte (PF189/2004) are gratefully acknowledged.
• Authorship and Discosures LC contributed to the design, conduction and analysis of the study and wrote the paper; SG contributed to the design, performed the research and collected data; GR, MT, CC, LS, FP and BF contributed to acquisition, analysis and interpretation of data; MF, MS, and AF critically reviewed the manuscript. The authors reported no potential conflicts of interest.

• Received June 26, 2007.
• Revision received October 18, 2007.
• Accepted October 23, 2007.