| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Brief Communications |
Departments of 1 Obstetrics/Gynaecology and 2 Clinical Chemistry, VU University Medical Center, Amsterdam, the Netherlands;
aAddress correspondence to this author at: Department of Clinical Chemistry, VU University Medical Center, De Boelelaan 1117, 1081 HV Amsterdam, the Netherlands. Fax 31-20-444-3895; e-mail cbm.oudejans{at}vumc.nl.
Abstract
Background: Measuring the allelic ratios of placental transcripts in maternal plasma permits noninvasive prenatal detection of chromosomal aneuploidy. Current methods, however, require highly specialized equipment (MALDI-TOF), limiting the widespread implementation of this powerful RNA single-nucleotide polymorphism (SNP) strategy in routine diagnostic settings. We adapted and applied the Transgenomic WAVE System and quencher extension (QEXT) for this purpose.
Methods: The expressed SNP (rs2187247) in exon 2 of the placentally expressed C21orf105 gene (chromosome 21 open reading frame 105) on chromosome 21 was tested in a trisomy 21 model system in which we obtained RNA selectively released from syncytiotrophoblasts of normal and trisomy 21 placentas during first trimester.
Results: In identifying heterozygous samples, we observed an exact correspondence between sequencing results and results obtained with the WAVE System. With respect to the analysis time required, the WAVE System was superior. In addition, the real-time QEXT assay (as optimized and validated with calibration standards consisting of 262-bp C21orf105 cDNA amplicons) accurately measured allele ratios after we optimized fragment purification, concentrations of input DNA and quencher label, and calculations of reporter signals. Finally, the optimized and validated QEXT assay correctly distinguished normal placentas from trisomy 21 placentas in tests of the following clinically relevant combinations: diploid homozygous (CC), diploid heterozygous (AC), triploid homozygous (AAA), and triploid heterozygous (AAC or ACC).
Conclusion: The QEXT method, which is directly adaptable to current real-time PCR equipment, along with rapid identification of informative samples with the WAVE System, may facilitate routine implementation of the RNA-SNP assay for noninvasive aneuploidy diagnostics.
Methods of noninvasive prenatal screening for numeric chromosomal abnormalities that are based on indirect markers typically detect the associated epiphenomena and therefore lack diagnostic power (1)(2). Direct assessments of fetal chromosome dosage with maternal plasma should use molecular markers transcribed or derived from the chromosome of interest, e.g., placental RNA transcribed from chromosome 21 to detect Down syndrome (1)(2). Lo and coworkers introduced a powerful, robust modification of this direct approach: the RNA–single-nucleotide polymorphism (SNP) allelic-ratio strategy (2). The relative quantification and comparison of allelic-expression ratios for a placentally expressed gene encoded by chromosome 21 enables the clear detection of differences in expression by 2 or 3 copies of the chromosome. The design of such comparisons dramatically increases the predictive power of this particular assay and other direct molecular assays. The use of a single marker (rs8130833) expressed by the PLAC4 (placenta-specific 4) gene enabled correct, noninvasive detection of fetal trisomy 21 in 90% of the cases (sensitivity) and excluded 96.5% of control samples (specificity) (2). The RNA-SNP ratio strategy is currently limited to a population subset, the nature of which is dictated by the heterozygosity state of the SNP used and the ethnicity of the studied population (e.g., white, Asian, African). We expect that population coverage can be increased by including other SNPs within PLAC4 or adding other chromosome 21–encoded transcripts that are expressed in the placenta and detectable in maternal plasma [e.g., COL6A2 (collagen, type VI, alpha 2), COL6A1 (collagen, type VI, alpha 1), C21orf105 (chromosome 21 open reading frame 105)] (2)(3). Technically, the RNA-SNP assay developed and used by Lo and coworkers is based on the differential extension of the polymorphic site to generate small but very specific allele-dependent differences in size. This approach, however, requires highly specialized equipment [matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) analyses], limiting its widespread implementation in routine diagnostic settings (2)(4).
We tested and validated 2 methodologic adaptations: the quencher extension (QEXT) reaction and the WAVE System (Transgenomic). The QEXT reaction is a novel single-step, real-time method to quantify SNPs and is directly adaptable to current real-time PCR equipment (5)(6). In the QEXT assay, a probe with a 5' reporter dye (FAM) is extended by a single base with a dideoxynucleotide (ddNTP) containing a quencher dye (TAMRA). Enzyme-mediated extension by Thermo Sequenase DNA polymerase (USB Corporation) takes place only if the target SNP allele is present. The extension is recorded in real time by the quenching (i.e., reducing the fluorescence) of the reporter dye. The relative amount of a specific SNP allele is determined by measuring the nucleotide-incorporation rate in a thermocycling reaction. Because TAMRA can also serve as a fluorescence donor, depending on the 5' fluorescent reporter, measurement by monitoring increases in fluorescence is also possible, as are multiplex reactions (5)(6).
The WAVE System allows rapid identification of informative (i.e., heterozygous) samples by partially denaturing HPLC of preformed homo- and heteroduplexes. At the temperature optimal for the formation and detection of heteroduplexes, these complexes, which are formed only if an SNP is present, will elute off the cartridge before the homoduplexes and will be visible as 2 additional peaks. If no SNP is present, all of the homoduplex DNA fragments elute as a single peak on the chromatogram. In routine clinical samples, the WAVE method detects DNA variants with a better detection limit, at a lower cost, and in a shorter time than direct sequencing (7).
To verify if these methods could be implemented in the RNA-SNP ratio assay for placentally expressed genes on chromosome 21 (1)(4), we used an SNP (rs2187247) located in exon 2 of C21orf105 as a model system. For this investigation, we isolated total RNA from normal placentas (n = 2) and trisomy 21 placentas (n = 3) obtained from pregnancies terminated at a gestational age of about 13 weeks (after obtaining informed consent). These placentas were denuded in vitro by controlled digestion with trypsin (8)(9). The nature of this treatment, which selectively releases the cellular contents, including RNA, from syncytiotrophoblasts (8)(9), mimics the specificity seen in vivo for the cellular origin of placental molecules released into maternal plasma (10). We confirmed the karyotypes of all placentas by G-banding analysis of metaphase spreads of cultured villi samples.
We started by testing whether the WAVE System could identify informative heterozygous samples with a specificity equal or superior to that of conventional DNA sequencing. We generated cDNA fragments (262 bp, intron-spanning rs2187247 located at position 221) from all samples by reverse transcription–PCR, identified the variant alleles of rs2187247 by cycle sequencing, and compared the results with data obtained with the WAVE System (see the methods in the Data Supplement that accompanies the online version of this Technical Brief at http://www.clinchem.org/content/vol54/issue2). Two of the 3 trisomy 21 placentas and 1 of the 2 controls were heterozygous (see Fig. 1 in the online Data Supplement). The frequency of rs2187247 heterozygotes in our population, as determined by sequencing of genomic DNA of white control individuals (n = 104), was 0.5. The temperature (62.7 °C), predicted by the Navigator program (Transgenomic) corresponded exactly with the actual optimal temperature for distinguishing heteroduplexes and homoduplexes by partially denaturing HPLC with the WAVE System. In addition, we found an exact correspondence between the sequencing results and results obtained with the WAVE System. Finally, the WAVE System was superior with respect to analysis time: a single analysis requires only 2.5 min.
We subsequently generated calibration standards to optimize, calibrate, and validate the QEXT assay (see the methods in the online Data Supplement). We cloned cDNA amplicons confirmed to contain either the A or C allele of rs2187247 into pTOPO vectors (Invitrogen), reamplified with the PCR, and mixed the purified allele fragments in different A-to-C ratios (7:1, 6:1, 5:1, 4:1, 3:1, 2:1, 1:1, 1:2, 1:3, 1:4, 1:5, 1:6, 1:7). The total quantities tested ranged between 10 and 100 ng. Negative controls consisted of homozygous DNA containing the A or C allele and QEXT reactions performed with TAMRA-ddGTP or -ddUTP, respectively. The QEXT method of Rudi et al. (5)(6) accurately measured allelic ratios (see Fig. 2 in the online Data Supplement) when the following modifications were implemented: (a) cDNA PCR-fragment purification with exonuclease I and shrimp alkaline phosphatase (ExoSAP-IT; USB Corporation) was extended to 60 min to ensure the complete removal of unincorporated nucleotides and primers. (b) After Qiagen purification (1), amplicons were eluted with Milli-Q–purified water (Millipore) to remove contaminants that interfere with downstream analysis (i.e., WAVE, QEXT). (c) A 50-ng DNA input was an optimal and practical compromise between specificity and cost-effectiveness. (d) Correction for the observed difference in incorporation rate between TAMRA-labeled ddUTP and ddGTP required a 10-fold higher concentration of the labeled ddUTP than the labeled ddGTP to generate identical thermocycling efficiencies. (e) The FAM reporter signal corrected for assay imprecision with the ROX reference dye (normalized reporter) showed the best correspondence with the predicted outcome; inclusion of or correction for the quencher molecule, TAMRA, is unnecessary and in fact contraindicated. (f) After 40 cycles, we found that the labeling reaction, which is theoretically linear because of the absence of amplification, leveled off after about 15 cycles; therefore, calculating the slope ratios for the linear part of the trajectory allowed specific and reproducible measurement of allele ratios within the following required A:C windows: 1:1 (AC), 1:2 (ACC), and 2:1 (AAC).
Finally, to test the proof of principle that the QEXT assay is a specific method for measuring allele-expression ratios in clinical samples, we tested the optimized and validated QEXT assay in the trisomy 21 model system by evaluating all of the clinically important (i.e., informative) combinations: diploid heterozygous (AC) and triploid heterozygous (AAC or ACC) samples. We also analyzed diploid homozygous (CC) and triploid homozygous (AAA) samples as controls. We tested each combination in a double-double manner; that is, we measured each allele ratio identically in duplicate on 2 different occasions. We obtained first-order regression curves by polynomial regression by using reporter (FAM) signals normalized for assay imprecision with the ROX reference dye and for calculations we used the initial slopes (15 cycles), which corresponded to the linear portions of the reaction trajectories. The measurements of the duplicates indicated that the reproducibility of the curves was good; the correlation coefficients for the A and C reactions were between 0.97 and 1.09. In addition, the profiles of the in vitro (calibrator) and in vivo (placenta model system) reactions were nearly identical for the tested combinations (diploid, triploid, homozygous, heterozygous; Fig. 1
). Finally, the allelic ratios permitted clear differentiation between diploidy (range, 0.7–1.3) and triploidy (0.7 < R > 1.3) in the trisomy 21 model system (see Fig. 1
in the online Data Supplement).
|
We conclude that the WAVE and QEXT methods we have described and validated in our trisomy 21 model system with C21orf105 as a representative direct molecular marker may facilitate routine implementation of the RNA-SNP assay for noninvasive aneuploidy diagnostics (2). The WAVE System permits rapid identification of informative heterozygous samples. Our QEXT method is directly adaptable to current real-time PCR equipment. Besides this practical advantage, the design permits real-time multipoint measurements, multiplex reactions, or the monitoring of increases in fluorescence. Preliminary observations indicate that with minor modifications (use of the Agilent BioAnalyzer; see Fig. 3A in the online Data Supplement) both the WAVE and QEXT methods (see Fig. 3, B and C, in the online Data Supplement) can be applied to clinical plasma samples (placental RNA from maternal plasma) and adapted easily to related placentally expressed genes on chromosome 21.
Acknowledgments
Grant/funding Support: Part of this work is supported by the SAFE network (project no. LSHB-CT-2004-503243). O.T.B. is supported by the Dutch Society for Scientific Research (ZonMW/NWO) (grant no. VIDI 917.56.349).
Financial Disclosures: None declared.
Acknowledgment: The continuous support by the Department of Obstetrics and Gynaecology is greatly appreciated.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
