- Email: [email protected]
Validating a rapid, real-time, PCR-based direct mutation detection assay for preimplantation genetic diagnosis
Gene 548 (2014) 299–305
Contents lists available at ScienceDirect
Gene journal homepage: www.elsevier.com/locate/gene
Methods Paper
Validating a rapid, real-time, PCR-based direct mutation detection assay for preimplantation genetic diagnosis Hsin-Fu Chen a,b, Shun-Ping Chang c,d, Sheng-Hai Wu d, Wen-Hsiang Lin c, Yi-Chung Lee e, Yen-Hsuan Ni f, Chi-An Chen a, Gwo-Chin Ma c,g, Norman A. Ginsberg h, En-Min You c, Feng-Po Tsai i, Ming Chen a,c,d,f,j,⁎ a
Department of Obstetrics and Gynecology, College of Medicine, and Hospital, National Taiwan University, Taipei, Taiwan Graduate Institute of Medical Genomics and Proteomics, College of Medicine, National Taiwan University, Taipei, Taiwan Department of Genomic Medicine, Changhua Christian Hospital, Changhua, Taiwan d Department of Life Science, National Chung-Hsing University, Taichung, Taiwan e Department of Neurology, Taipei Veterans General Hospital, Taipei, Taiwan f Department of Medical Genetics, National Taiwan University Hospital, Taipei, Taiwan g Institute of Biochemistry and Biotechnology, Chung Shan Medical University, Taichung, Taiwan h Department of Obstetrics and Gynecology, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA i Poyuan Women Clinic, Changhua, Taiwan j Department of Life Science, Tunghai University, Taichung, Taiwan b c
a r t i c l e
i n f o
Article history: Received 6 January 2014 Received in revised form 8 July 2014 Accepted 12 July 2014 Available online 14 July 2014 Keywords: ARMS-qPCR Cleavage-stage embryo biopsy Fresh embryo transfer (FET) Pre-implantation genetic diagnosis (PGD) Trophectoderm biopsy
a b s t r a c t Although co-amplification of polymorphic microsatellite markers is the current gold standard for preimplantation genetic diagnosis (PGD) of single-gene disorders (SGD), this approach can be hampered by the lack of availability of informative markers. We recently (2011) devised a novel in-house assay for PGD of aromatic L-amino acid decarboxylase deficiency, based on an amplification refractory mutation system and quantitative PCR (ARMS-qPCR). The objective of the present study was to verify ARMS-qPCR in a cohort of 20 PGD cycles with a diverse group of SGDs (15 couples at risk for 10 SGDs). Day-3 cleavage-stage embryos were subjected to biopsy and genotyping, followed by fresh embryo transfer (FET). The diagnostic rate was 82.9%; unaffected live births were achieved in 9 of 20 FET cycles (45%), with only one false negative (among 54 transferred embryos). Overall, the ARMS-qPCR had frequent allele-dropout (ADO), rendering it inappropriate as the sole diagnostic method (despite a favorable live-birth rate). Regardless, it has the potential to complement the current gold-standard methodology, especially when trophectoderm biopsy becomes a preferred option and genotyping needs to be timely enough to enable FET. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Pre-implantation genetic diagnosis (PGD) is a genetic analysis performed on cells biopsied from oocytes or developing embryos to detect specific genetic abnormalities known to exist in one or both parents (Brezina et al., 2012; Sermon et al., 2004). As an early form of prenatal genetic diagnosis (PND), PGD is an option for parents at risk of
Abbreviations: PGD, preimplantation genetic diagnosis; SGD, single-gene disorders; ARMS-qPCR, amplification refractory mutation system and quantitative PCR; FET, fresh embryo transfer; ADO, allele-dropout; PND, prenatal genetic diagnosis; Ct, threshold cycle; wt, wild-type; MU, mutant allele; SMA, spinal muscular atrophy; AIP, acute intermittent porphyria; XLMTM, X-linked myotubular myopathy; CMT2E, Charcot–Marie– Tooth type 2E; AADC, aromatic L-amino acid decarboxylase deficiency; ADPKD, autosomal dominant polycystic kidney disease; ARPKD, autosomal recessive polycystic kidney disease; DFNB4, autosomal recessive non-syndromic hearing loss; CNMX, X-linked centronuclear myopathy; DMD, Duchenne muscular dystrophy. ⁎ Corresponding author at: No. 176, 3F, Chunghua Rd., Changhua 500, Taiwan. E-mail addresses: [email protected], [email protected] (M. Chen).
http://dx.doi.org/10.1016/j.gene.2014.07.039 0378-1119/© 2014 Elsevier B.V. All rights reserved.
transmitting a genetic abnormality, without facing termination of an affected pregnancy (Adiga et al., 2010; SenGupta and Delhanty, 2012). In the last two decades, PGD has been done by one of the following: polar body biopsy (days 0/1), cleavage-stage biopsy (day 3 embryo; blastomere biopsy), and blastocyst biopsy (days 5/6; trophectoderm biopsy; Harper and Sengupta, 2012; Xu and Montag, 2012). Although both the polar body and blastocyst biopsies are increasing in popularity, until recently, cleavage-stage biopsy was the prevalent approach in PGD (Harper et al., 2010; Harton et al., 2011; Scott et al., 2013a,b). Technologies have been developed for PGD of various genetic indications, including most monogenic defects (namely autosomal dominant, autosomal recessive or X-linked recessive disorders). In the last decade, PGD involving amplification of DNA has been broadly applied to numerous mutation-detection strategies; these approaches usually included PCR (amplification step) and a post-PCR assay (detection step; Adiga et al., 2010; Harper and Sengupta, 2012; Harper et al., 2010; Spits and Sermon, 2009). During these two steps, technical difficulties can arise, including amplification failure, contamination, and allele dropout
300
H.-F. Chen et al. / Gene 548 (2014) 299–305
(ADO), which result in misdiagnosis (Adiga et al., 2010; Harper and Sengupta, 2012). Establishing appropriate guidelines for PGD is important to optimize the diagnostic service and patient care (Harton et al., 2011). The current gold-standard genotyping strategy for PGD has been the co-amplification of polymorphic microsatellite markers, linked or unlinked with the targeted mutation, in a fluorescent and multiplex PCR. Although it is indirect in nature and errors may still arise due to recombination, perhaps the error rate can be reduced by combining more markers (Chang et al., 2013; Chen et al., 2011; De Rycke et al., 2005; Harton et al., 2011; Korzebor et al., 2013; Laurie et al., 2010; Pickering et al., 1994; Zachaki et al., 2011). Conversely, direct mutation detection occasionally warrants the use of Sanger sequencing (e.g., if mutations are single- or oligo-base pair changes, namely point mutations or small insertion/deletions); however, this takes longer, and in such PGD cycles, fresh embryo transfer (FET) is not feasible. In this study, we evaluated an effective PCR-based genotyping assay, namely duplex-nested amplification refractory mutation system quantitative polymerase chain reaction (ARMS-qPCR), to directly detect mutations for which the mutant alleles are more easily and conventionally detected by Sanger sequencing of the amplified amplicon which contains the mutation (Moutou et al., 2003, 2007; Newton et al., 1989). The ARMS-qPCR consists of ARMS, followed by qPCR. The ARMS enables DNA amplification (when the target allele is contained within the sample), making it specifically suitable for PGD, whereas qPCR is designed to reduce the incidence of ARMS-induced ADO. In our previous publication, biopsied blastomeres from five Taiwanese couples with aromatic L-amino acid decarboxylase deficiency (AADC), an autosomal recessive disease, were correctly classified as affected (homozygous mutant), carrier (heterozygous for mutant and wild-type alleles), or normal (homozygous wild-type) by ARMS-qPCR within one working day (Kuo et al., 2011). Therefore, we intended to evaluate the performance of this novel ARMS-qPCR assay by including more PGD cycles in a diverse group of single-gene disorders. Furthermore, gold-standard methods were retrospectively used to verify the accuracy of PGD in embryos designated as affected (and not transferred) and in biopsies derived from established pregnancies. 2. Materials and methods 2.1. Patients and case preparation In the present study, 15 Taiwanese couples were referred to our central lab for PGD of various single-gene disorders (Table 1). None of these couples had a history of infertility or subfertility. All PGD cases in this study were performed in accordance with the standards for transport PGD centers in the USA (Gutiérrez-Mateo et al., 2009). Before
enrollment, we clearly stated that the genotyping strategy was novel and that our in-house designed genotyping assay would be the sole method of PGD (consequently, the risk of misdiagnosis was uncertain). However, couples were offered retrospective confirmatory prenatal diagnoses. Couples were enrolled only after they had provided informed consent. Parental genetic analyses were performed. All mutations (paternal or maternal) of the particular single-gene disorders included in this study are listed in Table 1. 2.2. Pre-PGD validations of ARMS-qPCR For detection of genetic diseases in this study, preclinical validation of the ARMS-qPCR methodology was performed as described (Kuo et al., 2011). Briefly, the efficacy of ARMS-qPCR was evaluated in 20 lymphocytes isolated from a normal and an affected individual (for an autosomal dominant disease due to a specific mutant allele), and from a normal, carrier, and affected individuals for an autosomal or X-linked recessive disease. Initially, DNA was extracted from peripheral lymphocytes of all parents and examined. The ARMS-qPCR procedure was used to detect the parental origin of mutations, thereby serving as a quality-control measure to ensure feasibility when subsequently used for PGD in specific couples at risk for specific loci of causative genes for various SGDs. 2.3. ARMS-qPCR PGD protocol Single cells biopsied from cleavage-stage embryos (one or two blastomeres from each embryo) on day-3 (eight-cell stage) with good morphology were received on the same day (that the biopsy was done) and were subjected to ARMS-qPCR. Embryonic DNA was extracted from these blastomeres using the Genomic DNA Mini Kit (Geneaid, New Taipei City, Taiwan). Primer sets for ARMS-qPCR for PGD were designed and evaluated, as we reported (Kuo et al., 2011). The ARMS-qPCR procedure required two rounds of PCR: in the first-round, two primer sets were used for amplification by a duplex-nested PCR (Fig. 1a), whereas in the second-round, two sequence-specific forward primers for ARMSqPCR were modified with a mismatch at the penultimate nucleotide position of the mutation site to increase specificity of the PCR reaction (Fig. 1b). Finally, wild-type and mutant alleles were distinguished by assessing the threshold cycle (Ct) value through the qPCR (Fig. 1c). Theoretically, contamination or ADO could be detected during the PCR process described above (Pickering et al., 1994). Diagnostic reports were usually completed within one day after receipt of the samples. Only blastomeres classified as “unaffected” were selected for transfer on the same day that the reports were available. The qPCRs were performed on a Roche LC 480 system (Basel, Switzerland).
Table 1 Summary of our PGD series: disease entities and inheritance patterns. Disorder
Gene
Genetic patterna
Couples (n = 15)
Cycles (n = 20)
Mutations
Charcot–Marie–Tooth type 2E (CMT2E) Acute intermittent porphyria (AIP) Autosomal dominant polycystic kidney disease (ADPKD) Aromatic L-amino acid decarboxylase deficiency (AADC) Autosomal recessive polycystic kidney disease (ARPKD) β-Thalassemia
NEFL HMBS PKD1 DDC PKHD1 HBB
AD AD AD AR AR AR
1 2 1 1b 1 4
1 3 2 1 2 5
Autosomal recessive non-syndromic hearing loss (DFNB4) Spinal muscular atrophy (SMA)
SLC26A4 SMN1/SMN2
AR AR
1 2
1 3
X-linked myotubular myopathy (XLMTM) Duchenne muscular dystrophy (DMD)
MTM1 DMD
XL XL
1 1
1 1
c.23CNG (p.P8R) c.848GNA (p.W283X) c.12220_12221delCT IVS6+4ANT c.8425GNA (p.G2809R), c.5517delT c.125_128delTCTT, c.52ANT, c.135dupC, c.−78ANG, IVS2+654 (CNT) c.1229CNT, c.2168ANG 1:2 carrier 1:1 carrier c.1160CNA (p.S387Y) c.10019GNA (p.C3340Y)
a b
AD, autosomal dominant disease; AR, autosomal recessive disease; XR, X-linked recessive disease. Kuo SJ, Ma GC, Chang SP et al. Taiwan J Obstet Gynecol 2011; 50: 468–73.
H.-F. Chen et al. / Gene 548 (2014) 299–305
301
Fig. 1. Schematic diagram of a duplex-nested amplification refractory mutation system quantitative polymerase chain reaction (ARMS-qPCR) for PGD of the c.23CNG point mutation in the NEFL gene. (a) Duplex-nested PCR was used to amplify the region from NEFL promoter to exon 1, including the position of the c.23CNG mutation (*). Primers were designed using the reverse-strand sequence. OF and OR, outer primer set; IF and IR, inner primer set. (b) ARMS-qPCR for separating mutant and wild-type alleles. WTF and MUF, two primers specific to amplify wild-type (WT) and mutant (MU) alleles, respectively. (c) A representative ARMS-qPCR for the following genotypes: WT/MU (a wild-type and a mutant allele), WT/WT (two wild-type alleles). M, size marker. Wt (the lane in gel), PCR with WTF primer. Mu (the lane in gel), PCR with MUF primer. The qPCR amplification plots are shown at the bottom. The WTF and MUF primer for qPCR were distinguished with different and duplicate labels (red and green lines indicate PCR with WTF primer, whereas blue and yellow lines indicate PCR with MUF primer).
2.4. Accuracy of PGD Determination or confirmation of the genetic status of the couples was performed by PCR amplification of the region encompassing the
targeted gene, followed by direct Sanger sequencing of each PCR amplicon. If pregnancy progressed to the gestational age appropriate for confirmatory prenatal diagnosis, fetal DNA was extracted from amniocytes or villi and fetal genetic status was verified by prenatal
302
H.-F. Chen et al. / Gene 548 (2014) 299–305
genotyping, including direct sequencing, coupled with microsatellite linkage analysis. In addition, non-transferred embryos classified as “affected” were genotyped to confirm the accuracy of PGD. 2.5. Outcome measurement Diagnostic rate (the percentage of samples for which a diagnosis was made), clinical pregnancy rate, and successful live birth delivery rate (excluding spontaneous abortions) were calculated. 3. Results
reduction was performed and the remaining unaffected twins were born uneventfully. Therefore, the false-negative rate in our ARMSqPCR group was 1 in 54 (b 2%). For the remaining 38 embryos classified as “affected”, follow-up genotyping of the discarded embryos indicated that there were no false positives in our series. Since it was not feasible to confirm genotypes of transferred embryos that did not achieve clinical pregnancy, it was impossible to estimate the actual misdiagnosis rate in our series. For the one misdiagnosis (false-negative), three informative microsatellite markers (D8S1104, D8S1477, and D8S492) were used to determine that the pregnancy was affected. Therefore, this error could have presumably been avoided if a gold-standard molecular assay had been used (Fig. 2).
3.1. Patients and cycles 3.4. Clinical outcomes In total, 15 couples were enrolled in our study and underwent a total of 20 PGD cycles (Table 1) from 2011 to 2012. The number of PGD cycles per couple ranged from 1 to 2; the majority (10) underwent only one. Maternal age ranged from 28 to 38 years. 3.2. Single gene disorders tested In this study, PGD was performed for 10 single-gene disorders, including five autosomal recessive disorders (50%), three autosomal dominant disorders (30%), and two X-linked disorders (20%; Table 1). The most frequent disorder tested was β-thalassemia (four couples, five cycles), followed by spinal muscular atrophy (SMA; two couples, three cycles) and acute intermittent porphyria (AIP; two couples, three cycles; Table 1). For some of the inherited single-gene disorders (AIP and X-linked myotubular myopathy, XLMTM) included in this study, PGD has apparently not been reported. 3.3. Efficiency of PCR amplification and accuracy of diagnosis In all 10 monogenic disorders, PGD was only performed after preclinical validation when both the successful amplification rate and the correct diagnosis rate were N 90%. A total of 111 embryos were biopsied for PGD (Table 2), with a diagnosis made for 92 (82.9%; Table 2). For specific single-gene disorders, the diagnostic rates ranged from 71.4 to 100% (Table 2). Among the 92 embryos with a diagnosis, 38 were classified as “affected” and 54 were classified as “unaffected” (Table 2). In addition, 19 of the total 111 embryos were classified as “no diagnosis”, at least some of which may have originated from PCR amplification failures. Regarding the misdiagnosis rate, one fetus in a triplet pregnancy in a couple with CMT2E was confirmed by prenatal diagnosis to have been erroneously classified (false-negative; refer to Fig. 2). Selective fetal
A total of 54 embryos classified as “unaffected”, derived from 20 PGD cycles, were transferred (range, 1–4 embryos per cycle). Following transfer of the above 54 “unaffected” embryos, 10 cycles had positive β-hCG concentrations (50% clinical pregnancy rate per couple). After excluding one biochemical pregnancy, nine pregnancies were established in 20 cycles with no subsequent miscarriages or ectopic pregnancies (Table 2). Ultimately, 13 healthy infants were born, including five singletons and four dizygotic twins (twin pregnancy rate was 44.4%). The live birth rate per cycle (i.e. take home baby rate per cycle) in this series was 45%. 3.5. Disease-specific outcomes of our PGD series 3.5.1. β-Thalassemia (the most common disease in our series) The largest group of PGD in our study was performed for βthalassemia with ARMS-qPCR containing customized primers, designed for targeted mutations of four couples. The paternal/maternal heterozygote mutations of these couples were c.135dupC/c.−78ANG, c125_128delTCTT/c.52ANT, and c125_128delTCTT/c.−78ANG. A total of 16 unaffected embryos were transferred back to the four patients, resulting in five healthy children. 3.5.2. Single-gene disorders with the first PGD attempt in the literature The PGD tests of single-gene disorders including acute intermittent porphyria (AIP) and X-linked myotubular myopathy (XLMTM) were apparently the first attempts at PGD to have been reported for these conditions. Among these two disorders, only one couple (with a heterozygous maternal mutation; c.848GNA) (p.W283X) of AIP gave birth (one healthy girl).
Table 2 Overall clinical outcome of PGD data. Genetic patterna
Disorder
Embryos diagnosed
Unaffected embryos
Affected embryos
No diagnosis
Diagnostic rate (%)
Pregnancies/no. of children born
No. cycles
Live-birth rate per cycle (%)
AD
Charcot–Marie–Tooth type 2E (CMT2E) Acute intermittent porphyria (AIP) Autosomal dominant polycystic kidney disease (ADPKD) Subtotal Aromatic L-amino acid decarboxylase deficiency (AADC) Autosomal recessive polycystic kidney disease (ARPKD) β-Thalassemia Autosomal recessive non-syndromic hearing loss (DFNB4) Spinal muscular atrophy (SMA) Subtotal X-linked myotubular myopathy (XLMTM) Duchenne muscular dystrophy (DMD) Subtotal
7 19 14 40 5 6 23 6 17 57 8 6 14 111
4 6 5 15 2 1 16 3 10 32 2 4 6 53
1b 9 6 16 3 5 2 3 3 16 5 2 7 39
2 4 3 9 0 0 5 0 4 9 1 0 1 19
71.4 78.9 78.6 77.5 100 100 78.3 100 76.5 84.2 87.5 100 92.9 82.9
1/2 1/1 1/1 3/4 1/1 0/0 3/5 0/0 2/3 6/9 0/0 0/0 0/0 9/13
1 3 2 6 1 2 5 1 3 12 1 1 2 20
100 33.3 50 50 100 0 60 0 66.7 50 0 0 0 45
AR
XR
Total a b
AD, autosomal dominant disease; AR, autosomal recessive disease; XR, X-linked recessive disease. One affected embryo, misdiagnosed as unaffected one by ARMS-qPCR, was confirmed and detected during subsequent prenatal genotyping (direct sequencing).
H.-F. Chen et al. / Gene 548 (2014) 299–305
303
Fig. 2. Prenatal genotyping for Charcot–Marie–Tooth type 2E (CMT2E) performed by linkage analysis. (a) Pedigree of a couple carrying CMT2E mutations and three fetuses. Three informative microsatellite markers are ordered from telomere (top) to centromere (bottom). The numbers in microsatellite markers represent the size of PCR products (bp). Numbers in bold are the predicted alleles linked to the mutation. The underlined number 219 of Fetus 1 is the allele of D8S492 with recombination. (b) Capillary electrophoresis of fluorescent PCR products obtained after multiplex PCR of three informative markers linked to the NEFL gene. On top of the electropherogram, the marker name is located above the corresponding alleles (peak). Numbers next to each peak represent the size of the allele (bp). The marker profiles from top to bottom are listed in the order of father, mother, normal embryo 1, affected embryo 2, and normal embryo 3, respectively.
3.5.3. The remaining single-gene disorders in our PGD series The PGD tests of single-gene disorders including AADC, Charcot– Marie–Tooth type 2E (CMT2E), autosomal dominant polycystic kidney disease (ADPKD), autosomal recessive polycystic kidney disease (ARPKD), autosomal recessive non-syndromic hearing loss (DFNB4), SMA, and XLMTM were reported. Among these disorders, five couples with CMT2E, AADC, ADPKD, and SMA, respectively, each had a healthy birth. The first couple with the heterozygous paternal mutation of CMT2E (c.23CNG (p.P8R)) gave birth to a healthy twin pair of girls. The second couple, both with the same heterozygous mutations of AADC (IVS6+4ANT), gave birth to a live, unaffected boy, as we recently reported (Kuo et al., 2011). The third couple with paternal heterozygous mutation (c.12220_12221delCT) of ADPKD had a healthy girl. Finally, the last two couples who were SMA carriers (SMN1: SMN2 to be 1:2 or 1:1) had live births of a singleton and twins, respectively.
4. Discussion 4.1. Diagnostic rate and ADO Reviewing previous reports in the literature, the diagnostic rates and ADO for PGD per locus were mostly N 80% and b 20%, respectively (Chang et al., 2013; Drury et al., 2001; Fiorentino et al., 2006; Gutiérrez-Mateo et al., 2009; Moutou et al., 2007; Ray and Handyside, 1996; Sermon et al., 1998; Vrettou et al., 2004; Wells and Sherlock, 1998). Recently, following the Guidelines for Best Practice PGD published by the European Society for Human Reproduction and Embryology (ESHRE) PGD Consortium, the recommended amplification efficiency (diagnostic rate) and ADO rate of pre-clinical validation should be ≥90% and b10% for each marker, respectively (Harton et al., 2011). In this study, the overall diagnostic rate using ARMS-qPCR was
304
H.-F. Chen et al. / Gene 548 (2014) 299–305
82.9%, which was below the recommended standard in the abovementioned guidelines. However, diagnostic rates for specific disorders varied from 71.4% (CMT2E) to 100% (AADC, ARPKD, DMD). The variation among various monogenic disorders may have been due to variable amplification efficiency of ARMS-qPCR. Unfortunately, the sensitivity and specificity of the primer sequences were sometimes unsatisfactory, perhaps due to adjacent nucleotide sequences (detailed sequences are presented in Supplementary Table S1). We inferred that embryos for which we failed to get a rapid diagnosis (17.1% of all embryos tested) were mostly due to negative PCR signals of ARMS (i.e. ADO; data not shown). Although we subsequently confirmed all clinical pregnancies, we were unable to confirm the actual status of embryos that failed to progress to a clinical pregnancy. Consequently, the definitive ADO rate remains unknown. Given the nearby single nucleotide polymorphism (SNP) loci are co-amplified in this ARMS-qPCR platform, it is possible to more correctly estimate ADO rates. Therefore, this deserves future investigation.
4.2. Autosomal-dominant (AD) diseases Recommended standards for diagnostic rate and ADO for AD diseases are ≥90% and b10% per locus, respectively (Harton et al., 2011). For the three AD diseases in the present study, the overall diagnostic rate 77.5% was below the suggested standard (Table 2); therefore, for confirmatory prenatal diagnosis, we verified the genotypes not only by direct sequencing, but also microsatellite linkage analysis. For example, special attention was given to the disease with the poorest performance (CMT2E), in which the father had a heterozygous mutation P8R (c.23CNG) which was diagnosed by ARMS-qPCR with a diagnostic rate of 71.4% (Table 2), below the minimum recommended level (Harton et al., 2011). However, after a successful fresh embryo transfer, a triplet pregnancy was established. Unfortunately, based on prenatal genotyping by direct sequencing, one fetus had the paternal allele of mutation P8R and was terminated immediately, resulting in a 14.3% (1 in 7) misdiagnosis rate for this locus, which is unacceptable. One of the likely explanations for this single misdiagnosis was ADO during the amplification step of ARMS-qPCR (Fig. 2). In that regard, we subsequently determined that Fetus 1 contained the mutant allele (219 bp on D8S492 locus; Fig. 2a).
4.3. Autosomal-recessive (AR) diseases The recommended standard for diagnostic rate for AR diseases is ≥90%. However, the ADO rate for these diseases could be slightly higher than that of AD diseases, since it needs both mutant alleles to have an affected child (Harton et al., 2011). In the present study, although the overall diagnostic rate of 84.2% (Table 2) for five AR diseases was below the recommended level, we achieved a 100% diagnostic rate for three of them (i.e., AADC, ARPKD, and DFNB4; Table 2). Regarding the live-birth rate (per FET cycle) in this series, live-birth rates for AR and AD diseases were identical (50%; Table 2).
4.4. X-linked recessive (XR) diseases The overall diagnostic rate for the two XR diseases (Duchenne muscular dystrophy and XLMTM) in this study was 92.9% (Table 2). Although this met the recommended standard for PGD, no live birth was achieved in the two FET cycles (Table 2; Harton et al., 2011). Among the two XR diseases, XLMTM (MIM number 310400), also known as X-linked centronuclear myopathy (CNMX), is an X-linked congenital myopathy characterized by slowly progressive muscular weakness and wasting (Bitoun et al., 2005).
4.5. Comparison of ARMS-qPCR and microsatellite multiple linkage analysis One reason that microsatellite multiple linkage analysis is the current gold standard is this strategy can provide flexibility, namely transfer of either fresh or frozen embryos, since this methodology is PCRbased and results can be available within one working day (facilitating FET after trophectoderm biopsy of day-5 blastocysts). In contrast, another proposed protocol involving trophectoderm biopsy, followed by whole genome amplification and genotyping with minisequencing, reduced or eliminated the option of FET (Chang et al., 2013). It is noteworthy that our genotyping platform ARMS-qPCR can be done as quickly as the current gold standard. A recent randomized study that blastocyst biopsy yielded better pregnancy rates than day-3 cleavage stage embryo biopsy provided additional support for our ARMS-qPCR as a method of direct mutation screening, especially when FET is desired (Scott et al., 2013a,b). An example to demonstrate the comparisons between our novel assay and the gold-standard assay is as follows: in PGD of Duchenne muscular dystrophy (DMD), an XR disease, six embryos were tested by ARMS-qPCR (100% diagnostic rate [Table 2], which exceeded the gold standard of 90%; Harton et al., 2011). To compare the two methodologies, additional microsatellite linkage markers were selected and co-amplified. We designed and pre-tested 4 microsatellite markers flanking the DMD gene. However, only two markers were informative and they were used in this study. Although the availability of informative microsatellite markers may be limited, in general, at least 2 to 4 microsatellite markers closely linked to the target gene are preferably selected and applied (Chang et al., 2013; De Rycke et al., 2005; Harton et al., 2011; Korzebor et al., 2013). Although its diagnostic rate is expected to be higher, multiplex microsatellite linkage analysis is more timeconsuming than ARMS-qPCR. In addition, a PGD cycle with microsatellite multiple-marker analysis using fluorescence-based and multiplex PCR usually costs at least US$1000 (based on US$250 for each fluorescent-labeled primer) compared to only US$100 for ARMS-qPCR. Therefore, ARMS-qPCR is more economic than the gold standard genotyping strategy (microsatellite linkage analysis) and may be useful due to its lower cost and when informative microsatellite markers are not readily available. 4.6. Future perspectives Additional studies are needed to more fully explore the performance of ARMS-qPCR. For example, to circumvent ADO this method could be tested with the use of other types of polymorphic markers (e.g. SNP), which can be co-amplified in the same platform. In addition, since ADO is presumably dependent on the number of cells examined in PGD (Findlay et al., 1995), perhaps the performance of ARMS-qPCR would be improved if more cells were available. As an example, a trophectoderm biopsy could be done in lieu of the day-3 cleavagestage embryo biopsy used in the present study. This could also improve the live-birth rate, as it was recently reported that cleavage-stage embryo biopsy significantly impaired human embryonic implantation potential, whereas a blastocyst biopsy did not (Scott et al., 2013a). Otherwise, whole genome amplification (WGA) for the day-3 biopsied cell may also improve the act of ARMS-qPCR given that single cell WGA provides improvements of ADO over PCR-based methods (Zong et al., 2012). Based on the performance of the ARMS-qPCR method in the present study (in particular the frequent ADO), we concluded that it should not be used as the sole genotyping platform for PGD of AD diseases, but it might be useful as an adjunct testing procedure. For example, detecting mutations with both indirect (i.e. linkage markers) and direct methods (e.g. ARMS-qPCR or mini-sequencing) could improve the diagnostic rate and reduce the incidence of misdiagnosis. Subsequent to completion of the current study, we are currently using ARMS-qPCR coupled with multiple informative linked microsatellite markers. In addition,
H.-F. Chen et al. / Gene 548 (2014) 299–305
another pilot study to evaluate the performance of day-5 trophectoderm biopsy, followed by rapid genotyping (by ARMS-qPCR coupled with multiple informative linked microsatellite markers), and fresh embryo transfer, is planned for the near future. 5. Conclusions The present report evaluated the performance of our novel ARMSqPCR assay and determined the feasibility of this genotyping strategy in the PGD of a diverse group of single-gene disorders with varying types of inheritance. Genotyping results were sometimes available within one working day and the overall PGD diagnostic rate was 82.9% (range, 71.4 to 100%). It was most accurate for XR disorders (100%), followed by AR (84.2%), whereas its performance in AD disorders was less optimal (77.5%). Since ARMS-qPCR is susceptible to allele dropout (ADO), it is unlikely to replace the current gold-standard assay. However, for AR and XR diseases it may be used to complement the current gold standard in PGD. Conflicts of interest The ARMS-qPCR methodology has been filed and patent granted in Taiwan, R.O.C. (a member of the World Trade Organization). The authors claim no conflicts of interest. Author's contribution HFC, SHW, YHN, CAC, and MC conceived and designed the study; SPC, WHL, GCM, and MC conducted the experiments; and HFC, SPC, WHL, YCL, GCM, NAG, and MC wrote the paper. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.gene.2014.07.039. Acknowledgments We gratefully acknowledge critical comments from Dr. Pao-Lin Kuo (National Cheng-Kung University, Tainan, Taiwan) and Dr. John Kastelic (University of Calgary, Canada). We appreciate the collaboration of Dr. Tsung-Hsien Lee (Chung-Shan Medical University, Taichung, Taiwan), Dr. Horng-Der Tsai, Dr. Cheng-Hsuan Wu, Dr. Yu-Ching Chen, and Dr. Hsin-Hung Wu (Changhua Christian Hospital, Changhua, Taiwan) in recruiting patients. This study was kindly supported by two grants (100-CCH-ICO-04 and 101-CCH-IRP-43) from the Changhua Christian Hospital, and the grant from the National Science Council (NSC 1012314-B-371-004-MY3) to M. Chen. References Adiga, S.K., Kalthur, G., Kumar, P., et al., 2010. Preimplantation diagnosis of genetic diseases. J. Postgrad. Med. 56 (4), 317–320. Bitoun, M., Maugenre, S., Jeannet, P.Y., et al., 2005. Mutations in dynamin 2 cause dominant centronuclear myopathy. Nat. Genet. 37 (11), 1207–1209. Brezina, P.R., Brezina, D.S., Kearns, W.G., 2012. Preimplantation genetic testing. BMJ 345, e5908. Chang, L.J., Huang, C.C., Tsai, Y.Y., et al., 2013. Blastocyst biopsy and vitrification are effective for preimplantation genetic diagnosis of monogenic diseases. Hum. Reprod. 28 (5), 1435–1444. Chen, Y.L., Hung, C.C., Lin, S.Y., et al., 2011. Successful application of the strategy of blastocyst biopsy, vitrification, whole genome amplification, and thawed embryo transfer for preimplantation genetic diagnosis of neurofibromatosis type 1. Taiwan. J. Obstet. Gynecol. 50 (1), 74–78.
305
De Rycke, M., Georgiou, I., Sermon, K., et al., 2005. PGD for autosomal dominant polycystic kidney disease type 1. Mol. Hum. Reprod. 11 (1), 65–71. Drury, K.C., Liu, M.C., Lilleberg, S., et al., 2001. Results on single cell PCR for Huntington's gene and WAVE product analysis for preimplantation genetic diagnosis. Mol. Cell. Endocrinol. 183, S1–S4. Findlay, I., Ray, P., Quirke, P., et al., 1995. Allelic drop-out and preferential amplification in single cells and human blastomeres: implications for preimplantation diagnosis of sex and cystic fibrosis. Hum. Reprod. 10 (6), 1609–1618. Fiorentino, F., Biricik, A., Nuccitelli, A., et al., 2006. Strategies and clinical outcome of 250 cycles of preimplantation genetic diagnosis for single gene disorders. Hum. Reprod. 21 (3), 670–684. Gutiérrez-Mateo, C., Sánchez-García, J.F., Fischer, J., et al., 2009. Preimplantation genetic diagnosis of single-gene disorders: experience with more than 200 cycles conducted by a reference laboratory in the United States. Fertil. Steril. 92 (5), 1544–1556. Harper, J.C., Sengupta, S.B., 2012. Preimplantation genetic diagnosis: state of the art 2011. Hum. Genet. 131 (2), 175–186. Harper, J.C., Coonen, E., De Rycke, M., et al., 2010. ESHRE PGD Consortium data collection X: cycles from January to December 2007 with pregnancy follow-up to October 2008. Hum. Reprod. 25 (11), 2685–2707. Harton, G.L., Magli, M.C., Lundin, K., et al., 2011. ESHRE PGD consortium/embryology special interest group-best practice guidelines for polar body and embryo biopsy for preimplantation genetic diagnosis/screening (PGD/PGS). Hum. Reprod. 26 (1), 41–46. Korzebor, A., Derakhshandeh-Peykar, P., Meshkani, M., et al., 2013. Heterozygosity assessment of five STR loci located at 5q13 region for preimplantation genetic diagnosis of spinal muscular atrophy. Mol. Biol. Rep. 40 (1), 67–72. Kuo, S.J., Ma, G.C., Chang, S.P., et al., 2011. Preimplantation and prenatal genetic diagnosis of aromatic L-amino acid decarboxylase deficiency with an amplification refractory mutation system-quantitative polymerase chain reaction. Taiwan. J. Obstet. Gynecol. 50 (4), 468–473. Laurie, A.D., Hill, A.M., Harraway, J.R., et al., 2010. Preimplantation genetic diagnosis for hemophilia A using indirect linkage analysis and direct genotyping approaches. J. Thromb. Haemost. 8 (4), 783–789. Moutou, C., Gardes, N., Viville, S., 2003. Duplex PCR for preimplantation genetic diagnosis (PGD) of spinal muscular atrophy. Prenat. Diagn. 23 (8), 685–689. Moutou, C., Gardes, N., Nicod, J.C., et al., 2007. Strategies and outcomes of PGD of familial adenomatous polyposis. Mol. Hum. Reprod. 13 (2), 95–101. Newton, C.R., Graham, A., Heptinstall, L.E., et al., 1989. Analysis of any point mutation in DNA. The amplification refractory mutation system (ARMS). Nucleic Acids Res. 17 (7), 2503–2516. Pickering, S.J., McConnell, J.M., Johnson, M.H., et al., 1994. Use of a polymorphic dinucleotide repeat sequence to detect non-blastomeric contamination of the polymerase chain reaction in biopsy samples for preimplantation diagnosis. Hum. Reprod. 9 (8), 1539–1545. Ray, P.F., Handyside, A.H., 1996. Increasing the denaturation temperature during the first cycles of amplification reduces allele dropout from single cells for preimplantation genetic diagnosis. Mol. Hum. Reprod. 2 (3), 213–218. Scott Jr., R.T., Upham, K.M., Forman, E.J., et al., 2013a. Cleavage-stage biopsy significantly impairs human embryonic implantation potential while blastocyst biopsy does not: a randomized and paired clinical trial. Fertil. Steril. 100 (3), 624–630. Scott Jr., R.T., Upham, K.M., Forman, E.J., et al., 2013b. Blastocyst biopsy with comprehensive chromosome screening and fresh embryo transfer significantly increases in vitro fertilization implantation and delivery rates: a randomized controlled trial. Fertil. Steril. 100 (3), 697–703. SenGupta, S.B., Delhanty, J.D., 2012. Preimplantation genetic diagnosis: recent triumphs and remaining challenges. Expert. Rev. Mol. Diagn. 12 (6), 585–592. Sermon, K., De Vos, A., Van de Velde, H., et al., 1998. Fluorescent PCR and automated fragment analysis for the clinical application of preimplantation genetic diagnosis of myotonic dystrophy (Steinert's disease). Mol. Hum. Reprod. 4 (8), 791–796. Sermon, K., Van Steirteghem, A., Liebaers, I., 2004. Preimplantation genetic diagnosis. Lancet 363, 1633–1641. Spits, C., Sermon, K., 2009. PGD for monogenic disorders: aspects of molecular biology. Prenat. Diagn. 29, 50–56. Vrettou, C., Traeger-Synodinos, J., Tzetis, M., et al., 2004. Real-time PCR for single-cell genotyping in sickle cell and thalassemia syndromes as a rapid, accurate, reliable, and widely applicable protocol for preimplantation genetic diagnosis. Hum. Mutat. 23 (5), 513–521. Wells, D., Sherlock, J.K., 1998. Strategies for preimplantation genetic diagnosis of single gene disorders by DNA amplification. Prenat. Diagn. 18 (13), 1389–1401. Xu, K., Montag, M., 2012. New perspectives on embryo biopsy: not how, but when and why? Semin. Reprod. Med. 30 (4), 259–266. Zachaki, S., Vrettou, C., Destouni, A., et al., 2011. Novel and known microsatellite markers within the β-globin cluster to support robust preimplantation genetic diagnosis of β-thalassemia and sickle cell syndromes. Hemoglobin 35 (1), 56–66. Zong, C., Lu, S., Chapman, A.R., et al., 2012. Genome-wide detection of single-nucleotide and copy-number variations of a single human cell. Science 338 (6114), 1622–1626.
Contents lists available at ScienceDirect
Gene journal homepage: www.elsevier.com/locate/gene
Methods Paper
Validating a rapid, real-time, PCR-based direct mutation detection assay for preimplantation genetic diagnosis Hsin-Fu Chen a,b, Shun-Ping Chang c,d, Sheng-Hai Wu d, Wen-Hsiang Lin c, Yi-Chung Lee e, Yen-Hsuan Ni f, Chi-An Chen a, Gwo-Chin Ma c,g, Norman A. Ginsberg h, En-Min You c, Feng-Po Tsai i, Ming Chen a,c,d,f,j,⁎ a
Department of Obstetrics and Gynecology, College of Medicine, and Hospital, National Taiwan University, Taipei, Taiwan Graduate Institute of Medical Genomics and Proteomics, College of Medicine, National Taiwan University, Taipei, Taiwan Department of Genomic Medicine, Changhua Christian Hospital, Changhua, Taiwan d Department of Life Science, National Chung-Hsing University, Taichung, Taiwan e Department of Neurology, Taipei Veterans General Hospital, Taipei, Taiwan f Department of Medical Genetics, National Taiwan University Hospital, Taipei, Taiwan g Institute of Biochemistry and Biotechnology, Chung Shan Medical University, Taichung, Taiwan h Department of Obstetrics and Gynecology, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA i Poyuan Women Clinic, Changhua, Taiwan j Department of Life Science, Tunghai University, Taichung, Taiwan b c
a r t i c l e
i n f o
Article history: Received 6 January 2014 Received in revised form 8 July 2014 Accepted 12 July 2014 Available online 14 July 2014 Keywords: ARMS-qPCR Cleavage-stage embryo biopsy Fresh embryo transfer (FET) Pre-implantation genetic diagnosis (PGD) Trophectoderm biopsy
a b s t r a c t Although co-amplification of polymorphic microsatellite markers is the current gold standard for preimplantation genetic diagnosis (PGD) of single-gene disorders (SGD), this approach can be hampered by the lack of availability of informative markers. We recently (2011) devised a novel in-house assay for PGD of aromatic L-amino acid decarboxylase deficiency, based on an amplification refractory mutation system and quantitative PCR (ARMS-qPCR). The objective of the present study was to verify ARMS-qPCR in a cohort of 20 PGD cycles with a diverse group of SGDs (15 couples at risk for 10 SGDs). Day-3 cleavage-stage embryos were subjected to biopsy and genotyping, followed by fresh embryo transfer (FET). The diagnostic rate was 82.9%; unaffected live births were achieved in 9 of 20 FET cycles (45%), with only one false negative (among 54 transferred embryos). Overall, the ARMS-qPCR had frequent allele-dropout (ADO), rendering it inappropriate as the sole diagnostic method (despite a favorable live-birth rate). Regardless, it has the potential to complement the current gold-standard methodology, especially when trophectoderm biopsy becomes a preferred option and genotyping needs to be timely enough to enable FET. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Pre-implantation genetic diagnosis (PGD) is a genetic analysis performed on cells biopsied from oocytes or developing embryos to detect specific genetic abnormalities known to exist in one or both parents (Brezina et al., 2012; Sermon et al., 2004). As an early form of prenatal genetic diagnosis (PND), PGD is an option for parents at risk of
Abbreviations: PGD, preimplantation genetic diagnosis; SGD, single-gene disorders; ARMS-qPCR, amplification refractory mutation system and quantitative PCR; FET, fresh embryo transfer; ADO, allele-dropout; PND, prenatal genetic diagnosis; Ct, threshold cycle; wt, wild-type; MU, mutant allele; SMA, spinal muscular atrophy; AIP, acute intermittent porphyria; XLMTM, X-linked myotubular myopathy; CMT2E, Charcot–Marie– Tooth type 2E; AADC, aromatic L-amino acid decarboxylase deficiency; ADPKD, autosomal dominant polycystic kidney disease; ARPKD, autosomal recessive polycystic kidney disease; DFNB4, autosomal recessive non-syndromic hearing loss; CNMX, X-linked centronuclear myopathy; DMD, Duchenne muscular dystrophy. ⁎ Corresponding author at: No. 176, 3F, Chunghua Rd., Changhua 500, Taiwan. E-mail addresses: [email protected], [email protected] (M. Chen).
http://dx.doi.org/10.1016/j.gene.2014.07.039 0378-1119/© 2014 Elsevier B.V. All rights reserved.
transmitting a genetic abnormality, without facing termination of an affected pregnancy (Adiga et al., 2010; SenGupta and Delhanty, 2012). In the last two decades, PGD has been done by one of the following: polar body biopsy (days 0/1), cleavage-stage biopsy (day 3 embryo; blastomere biopsy), and blastocyst biopsy (days 5/6; trophectoderm biopsy; Harper and Sengupta, 2012; Xu and Montag, 2012). Although both the polar body and blastocyst biopsies are increasing in popularity, until recently, cleavage-stage biopsy was the prevalent approach in PGD (Harper et al., 2010; Harton et al., 2011; Scott et al., 2013a,b). Technologies have been developed for PGD of various genetic indications, including most monogenic defects (namely autosomal dominant, autosomal recessive or X-linked recessive disorders). In the last decade, PGD involving amplification of DNA has been broadly applied to numerous mutation-detection strategies; these approaches usually included PCR (amplification step) and a post-PCR assay (detection step; Adiga et al., 2010; Harper and Sengupta, 2012; Harper et al., 2010; Spits and Sermon, 2009). During these two steps, technical difficulties can arise, including amplification failure, contamination, and allele dropout
300
H.-F. Chen et al. / Gene 548 (2014) 299–305
(ADO), which result in misdiagnosis (Adiga et al., 2010; Harper and Sengupta, 2012). Establishing appropriate guidelines for PGD is important to optimize the diagnostic service and patient care (Harton et al., 2011). The current gold-standard genotyping strategy for PGD has been the co-amplification of polymorphic microsatellite markers, linked or unlinked with the targeted mutation, in a fluorescent and multiplex PCR. Although it is indirect in nature and errors may still arise due to recombination, perhaps the error rate can be reduced by combining more markers (Chang et al., 2013; Chen et al., 2011; De Rycke et al., 2005; Harton et al., 2011; Korzebor et al., 2013; Laurie et al., 2010; Pickering et al., 1994; Zachaki et al., 2011). Conversely, direct mutation detection occasionally warrants the use of Sanger sequencing (e.g., if mutations are single- or oligo-base pair changes, namely point mutations or small insertion/deletions); however, this takes longer, and in such PGD cycles, fresh embryo transfer (FET) is not feasible. In this study, we evaluated an effective PCR-based genotyping assay, namely duplex-nested amplification refractory mutation system quantitative polymerase chain reaction (ARMS-qPCR), to directly detect mutations for which the mutant alleles are more easily and conventionally detected by Sanger sequencing of the amplified amplicon which contains the mutation (Moutou et al., 2003, 2007; Newton et al., 1989). The ARMS-qPCR consists of ARMS, followed by qPCR. The ARMS enables DNA amplification (when the target allele is contained within the sample), making it specifically suitable for PGD, whereas qPCR is designed to reduce the incidence of ARMS-induced ADO. In our previous publication, biopsied blastomeres from five Taiwanese couples with aromatic L-amino acid decarboxylase deficiency (AADC), an autosomal recessive disease, were correctly classified as affected (homozygous mutant), carrier (heterozygous for mutant and wild-type alleles), or normal (homozygous wild-type) by ARMS-qPCR within one working day (Kuo et al., 2011). Therefore, we intended to evaluate the performance of this novel ARMS-qPCR assay by including more PGD cycles in a diverse group of single-gene disorders. Furthermore, gold-standard methods were retrospectively used to verify the accuracy of PGD in embryos designated as affected (and not transferred) and in biopsies derived from established pregnancies. 2. Materials and methods 2.1. Patients and case preparation In the present study, 15 Taiwanese couples were referred to our central lab for PGD of various single-gene disorders (Table 1). None of these couples had a history of infertility or subfertility. All PGD cases in this study were performed in accordance with the standards for transport PGD centers in the USA (Gutiérrez-Mateo et al., 2009). Before
enrollment, we clearly stated that the genotyping strategy was novel and that our in-house designed genotyping assay would be the sole method of PGD (consequently, the risk of misdiagnosis was uncertain). However, couples were offered retrospective confirmatory prenatal diagnoses. Couples were enrolled only after they had provided informed consent. Parental genetic analyses were performed. All mutations (paternal or maternal) of the particular single-gene disorders included in this study are listed in Table 1. 2.2. Pre-PGD validations of ARMS-qPCR For detection of genetic diseases in this study, preclinical validation of the ARMS-qPCR methodology was performed as described (Kuo et al., 2011). Briefly, the efficacy of ARMS-qPCR was evaluated in 20 lymphocytes isolated from a normal and an affected individual (for an autosomal dominant disease due to a specific mutant allele), and from a normal, carrier, and affected individuals for an autosomal or X-linked recessive disease. Initially, DNA was extracted from peripheral lymphocytes of all parents and examined. The ARMS-qPCR procedure was used to detect the parental origin of mutations, thereby serving as a quality-control measure to ensure feasibility when subsequently used for PGD in specific couples at risk for specific loci of causative genes for various SGDs. 2.3. ARMS-qPCR PGD protocol Single cells biopsied from cleavage-stage embryos (one or two blastomeres from each embryo) on day-3 (eight-cell stage) with good morphology were received on the same day (that the biopsy was done) and were subjected to ARMS-qPCR. Embryonic DNA was extracted from these blastomeres using the Genomic DNA Mini Kit (Geneaid, New Taipei City, Taiwan). Primer sets for ARMS-qPCR for PGD were designed and evaluated, as we reported (Kuo et al., 2011). The ARMS-qPCR procedure required two rounds of PCR: in the first-round, two primer sets were used for amplification by a duplex-nested PCR (Fig. 1a), whereas in the second-round, two sequence-specific forward primers for ARMSqPCR were modified with a mismatch at the penultimate nucleotide position of the mutation site to increase specificity of the PCR reaction (Fig. 1b). Finally, wild-type and mutant alleles were distinguished by assessing the threshold cycle (Ct) value through the qPCR (Fig. 1c). Theoretically, contamination or ADO could be detected during the PCR process described above (Pickering et al., 1994). Diagnostic reports were usually completed within one day after receipt of the samples. Only blastomeres classified as “unaffected” were selected for transfer on the same day that the reports were available. The qPCRs were performed on a Roche LC 480 system (Basel, Switzerland).
Table 1 Summary of our PGD series: disease entities and inheritance patterns. Disorder
Gene
Genetic patterna
Couples (n = 15)
Cycles (n = 20)
Mutations
Charcot–Marie–Tooth type 2E (CMT2E) Acute intermittent porphyria (AIP) Autosomal dominant polycystic kidney disease (ADPKD) Aromatic L-amino acid decarboxylase deficiency (AADC) Autosomal recessive polycystic kidney disease (ARPKD) β-Thalassemia
NEFL HMBS PKD1 DDC PKHD1 HBB
AD AD AD AR AR AR
1 2 1 1b 1 4
1 3 2 1 2 5
Autosomal recessive non-syndromic hearing loss (DFNB4) Spinal muscular atrophy (SMA)
SLC26A4 SMN1/SMN2
AR AR
1 2
1 3
X-linked myotubular myopathy (XLMTM) Duchenne muscular dystrophy (DMD)
MTM1 DMD
XL XL
1 1
1 1
c.23CNG (p.P8R) c.848GNA (p.W283X) c.12220_12221delCT IVS6+4ANT c.8425GNA (p.G2809R), c.5517delT c.125_128delTCTT, c.52ANT, c.135dupC, c.−78ANG, IVS2+654 (CNT) c.1229CNT, c.2168ANG 1:2 carrier 1:1 carrier c.1160CNA (p.S387Y) c.10019GNA (p.C3340Y)
a b
AD, autosomal dominant disease; AR, autosomal recessive disease; XR, X-linked recessive disease. Kuo SJ, Ma GC, Chang SP et al. Taiwan J Obstet Gynecol 2011; 50: 468–73.
H.-F. Chen et al. / Gene 548 (2014) 299–305
301
Fig. 1. Schematic diagram of a duplex-nested amplification refractory mutation system quantitative polymerase chain reaction (ARMS-qPCR) for PGD of the c.23CNG point mutation in the NEFL gene. (a) Duplex-nested PCR was used to amplify the region from NEFL promoter to exon 1, including the position of the c.23CNG mutation (*). Primers were designed using the reverse-strand sequence. OF and OR, outer primer set; IF and IR, inner primer set. (b) ARMS-qPCR for separating mutant and wild-type alleles. WTF and MUF, two primers specific to amplify wild-type (WT) and mutant (MU) alleles, respectively. (c) A representative ARMS-qPCR for the following genotypes: WT/MU (a wild-type and a mutant allele), WT/WT (two wild-type alleles). M, size marker. Wt (the lane in gel), PCR with WTF primer. Mu (the lane in gel), PCR with MUF primer. The qPCR amplification plots are shown at the bottom. The WTF and MUF primer for qPCR were distinguished with different and duplicate labels (red and green lines indicate PCR with WTF primer, whereas blue and yellow lines indicate PCR with MUF primer).
2.4. Accuracy of PGD Determination or confirmation of the genetic status of the couples was performed by PCR amplification of the region encompassing the
targeted gene, followed by direct Sanger sequencing of each PCR amplicon. If pregnancy progressed to the gestational age appropriate for confirmatory prenatal diagnosis, fetal DNA was extracted from amniocytes or villi and fetal genetic status was verified by prenatal
302
H.-F. Chen et al. / Gene 548 (2014) 299–305
genotyping, including direct sequencing, coupled with microsatellite linkage analysis. In addition, non-transferred embryos classified as “affected” were genotyped to confirm the accuracy of PGD. 2.5. Outcome measurement Diagnostic rate (the percentage of samples for which a diagnosis was made), clinical pregnancy rate, and successful live birth delivery rate (excluding spontaneous abortions) were calculated. 3. Results
reduction was performed and the remaining unaffected twins were born uneventfully. Therefore, the false-negative rate in our ARMSqPCR group was 1 in 54 (b 2%). For the remaining 38 embryos classified as “affected”, follow-up genotyping of the discarded embryos indicated that there were no false positives in our series. Since it was not feasible to confirm genotypes of transferred embryos that did not achieve clinical pregnancy, it was impossible to estimate the actual misdiagnosis rate in our series. For the one misdiagnosis (false-negative), three informative microsatellite markers (D8S1104, D8S1477, and D8S492) were used to determine that the pregnancy was affected. Therefore, this error could have presumably been avoided if a gold-standard molecular assay had been used (Fig. 2).
3.1. Patients and cycles 3.4. Clinical outcomes In total, 15 couples were enrolled in our study and underwent a total of 20 PGD cycles (Table 1) from 2011 to 2012. The number of PGD cycles per couple ranged from 1 to 2; the majority (10) underwent only one. Maternal age ranged from 28 to 38 years. 3.2. Single gene disorders tested In this study, PGD was performed for 10 single-gene disorders, including five autosomal recessive disorders (50%), three autosomal dominant disorders (30%), and two X-linked disorders (20%; Table 1). The most frequent disorder tested was β-thalassemia (four couples, five cycles), followed by spinal muscular atrophy (SMA; two couples, three cycles) and acute intermittent porphyria (AIP; two couples, three cycles; Table 1). For some of the inherited single-gene disorders (AIP and X-linked myotubular myopathy, XLMTM) included in this study, PGD has apparently not been reported. 3.3. Efficiency of PCR amplification and accuracy of diagnosis In all 10 monogenic disorders, PGD was only performed after preclinical validation when both the successful amplification rate and the correct diagnosis rate were N 90%. A total of 111 embryos were biopsied for PGD (Table 2), with a diagnosis made for 92 (82.9%; Table 2). For specific single-gene disorders, the diagnostic rates ranged from 71.4 to 100% (Table 2). Among the 92 embryos with a diagnosis, 38 were classified as “affected” and 54 were classified as “unaffected” (Table 2). In addition, 19 of the total 111 embryos were classified as “no diagnosis”, at least some of which may have originated from PCR amplification failures. Regarding the misdiagnosis rate, one fetus in a triplet pregnancy in a couple with CMT2E was confirmed by prenatal diagnosis to have been erroneously classified (false-negative; refer to Fig. 2). Selective fetal
A total of 54 embryos classified as “unaffected”, derived from 20 PGD cycles, were transferred (range, 1–4 embryos per cycle). Following transfer of the above 54 “unaffected” embryos, 10 cycles had positive β-hCG concentrations (50% clinical pregnancy rate per couple). After excluding one biochemical pregnancy, nine pregnancies were established in 20 cycles with no subsequent miscarriages or ectopic pregnancies (Table 2). Ultimately, 13 healthy infants were born, including five singletons and four dizygotic twins (twin pregnancy rate was 44.4%). The live birth rate per cycle (i.e. take home baby rate per cycle) in this series was 45%. 3.5. Disease-specific outcomes of our PGD series 3.5.1. β-Thalassemia (the most common disease in our series) The largest group of PGD in our study was performed for βthalassemia with ARMS-qPCR containing customized primers, designed for targeted mutations of four couples. The paternal/maternal heterozygote mutations of these couples were c.135dupC/c.−78ANG, c125_128delTCTT/c.52ANT, and c125_128delTCTT/c.−78ANG. A total of 16 unaffected embryos were transferred back to the four patients, resulting in five healthy children. 3.5.2. Single-gene disorders with the first PGD attempt in the literature The PGD tests of single-gene disorders including acute intermittent porphyria (AIP) and X-linked myotubular myopathy (XLMTM) were apparently the first attempts at PGD to have been reported for these conditions. Among these two disorders, only one couple (with a heterozygous maternal mutation; c.848GNA) (p.W283X) of AIP gave birth (one healthy girl).
Table 2 Overall clinical outcome of PGD data. Genetic patterna
Disorder
Embryos diagnosed
Unaffected embryos
Affected embryos
No diagnosis
Diagnostic rate (%)
Pregnancies/no. of children born
No. cycles
Live-birth rate per cycle (%)
AD
Charcot–Marie–Tooth type 2E (CMT2E) Acute intermittent porphyria (AIP) Autosomal dominant polycystic kidney disease (ADPKD) Subtotal Aromatic L-amino acid decarboxylase deficiency (AADC) Autosomal recessive polycystic kidney disease (ARPKD) β-Thalassemia Autosomal recessive non-syndromic hearing loss (DFNB4) Spinal muscular atrophy (SMA) Subtotal X-linked myotubular myopathy (XLMTM) Duchenne muscular dystrophy (DMD) Subtotal
7 19 14 40 5 6 23 6 17 57 8 6 14 111
4 6 5 15 2 1 16 3 10 32 2 4 6 53
1b 9 6 16 3 5 2 3 3 16 5 2 7 39
2 4 3 9 0 0 5 0 4 9 1 0 1 19
71.4 78.9 78.6 77.5 100 100 78.3 100 76.5 84.2 87.5 100 92.9 82.9
1/2 1/1 1/1 3/4 1/1 0/0 3/5 0/0 2/3 6/9 0/0 0/0 0/0 9/13
1 3 2 6 1 2 5 1 3 12 1 1 2 20
100 33.3 50 50 100 0 60 0 66.7 50 0 0 0 45
AR
XR
Total a b
AD, autosomal dominant disease; AR, autosomal recessive disease; XR, X-linked recessive disease. One affected embryo, misdiagnosed as unaffected one by ARMS-qPCR, was confirmed and detected during subsequent prenatal genotyping (direct sequencing).
H.-F. Chen et al. / Gene 548 (2014) 299–305
303
Fig. 2. Prenatal genotyping for Charcot–Marie–Tooth type 2E (CMT2E) performed by linkage analysis. (a) Pedigree of a couple carrying CMT2E mutations and three fetuses. Three informative microsatellite markers are ordered from telomere (top) to centromere (bottom). The numbers in microsatellite markers represent the size of PCR products (bp). Numbers in bold are the predicted alleles linked to the mutation. The underlined number 219 of Fetus 1 is the allele of D8S492 with recombination. (b) Capillary electrophoresis of fluorescent PCR products obtained after multiplex PCR of three informative markers linked to the NEFL gene. On top of the electropherogram, the marker name is located above the corresponding alleles (peak). Numbers next to each peak represent the size of the allele (bp). The marker profiles from top to bottom are listed in the order of father, mother, normal embryo 1, affected embryo 2, and normal embryo 3, respectively.
3.5.3. The remaining single-gene disorders in our PGD series The PGD tests of single-gene disorders including AADC, Charcot– Marie–Tooth type 2E (CMT2E), autosomal dominant polycystic kidney disease (ADPKD), autosomal recessive polycystic kidney disease (ARPKD), autosomal recessive non-syndromic hearing loss (DFNB4), SMA, and XLMTM were reported. Among these disorders, five couples with CMT2E, AADC, ADPKD, and SMA, respectively, each had a healthy birth. The first couple with the heterozygous paternal mutation of CMT2E (c.23CNG (p.P8R)) gave birth to a healthy twin pair of girls. The second couple, both with the same heterozygous mutations of AADC (IVS6+4ANT), gave birth to a live, unaffected boy, as we recently reported (Kuo et al., 2011). The third couple with paternal heterozygous mutation (c.12220_12221delCT) of ADPKD had a healthy girl. Finally, the last two couples who were SMA carriers (SMN1: SMN2 to be 1:2 or 1:1) had live births of a singleton and twins, respectively.
4. Discussion 4.1. Diagnostic rate and ADO Reviewing previous reports in the literature, the diagnostic rates and ADO for PGD per locus were mostly N 80% and b 20%, respectively (Chang et al., 2013; Drury et al., 2001; Fiorentino et al., 2006; Gutiérrez-Mateo et al., 2009; Moutou et al., 2007; Ray and Handyside, 1996; Sermon et al., 1998; Vrettou et al., 2004; Wells and Sherlock, 1998). Recently, following the Guidelines for Best Practice PGD published by the European Society for Human Reproduction and Embryology (ESHRE) PGD Consortium, the recommended amplification efficiency (diagnostic rate) and ADO rate of pre-clinical validation should be ≥90% and b10% for each marker, respectively (Harton et al., 2011). In this study, the overall diagnostic rate using ARMS-qPCR was
304
H.-F. Chen et al. / Gene 548 (2014) 299–305
82.9%, which was below the recommended standard in the abovementioned guidelines. However, diagnostic rates for specific disorders varied from 71.4% (CMT2E) to 100% (AADC, ARPKD, DMD). The variation among various monogenic disorders may have been due to variable amplification efficiency of ARMS-qPCR. Unfortunately, the sensitivity and specificity of the primer sequences were sometimes unsatisfactory, perhaps due to adjacent nucleotide sequences (detailed sequences are presented in Supplementary Table S1). We inferred that embryos for which we failed to get a rapid diagnosis (17.1% of all embryos tested) were mostly due to negative PCR signals of ARMS (i.e. ADO; data not shown). Although we subsequently confirmed all clinical pregnancies, we were unable to confirm the actual status of embryos that failed to progress to a clinical pregnancy. Consequently, the definitive ADO rate remains unknown. Given the nearby single nucleotide polymorphism (SNP) loci are co-amplified in this ARMS-qPCR platform, it is possible to more correctly estimate ADO rates. Therefore, this deserves future investigation.
4.2. Autosomal-dominant (AD) diseases Recommended standards for diagnostic rate and ADO for AD diseases are ≥90% and b10% per locus, respectively (Harton et al., 2011). For the three AD diseases in the present study, the overall diagnostic rate 77.5% was below the suggested standard (Table 2); therefore, for confirmatory prenatal diagnosis, we verified the genotypes not only by direct sequencing, but also microsatellite linkage analysis. For example, special attention was given to the disease with the poorest performance (CMT2E), in which the father had a heterozygous mutation P8R (c.23CNG) which was diagnosed by ARMS-qPCR with a diagnostic rate of 71.4% (Table 2), below the minimum recommended level (Harton et al., 2011). However, after a successful fresh embryo transfer, a triplet pregnancy was established. Unfortunately, based on prenatal genotyping by direct sequencing, one fetus had the paternal allele of mutation P8R and was terminated immediately, resulting in a 14.3% (1 in 7) misdiagnosis rate for this locus, which is unacceptable. One of the likely explanations for this single misdiagnosis was ADO during the amplification step of ARMS-qPCR (Fig. 2). In that regard, we subsequently determined that Fetus 1 contained the mutant allele (219 bp on D8S492 locus; Fig. 2a).
4.3. Autosomal-recessive (AR) diseases The recommended standard for diagnostic rate for AR diseases is ≥90%. However, the ADO rate for these diseases could be slightly higher than that of AD diseases, since it needs both mutant alleles to have an affected child (Harton et al., 2011). In the present study, although the overall diagnostic rate of 84.2% (Table 2) for five AR diseases was below the recommended level, we achieved a 100% diagnostic rate for three of them (i.e., AADC, ARPKD, and DFNB4; Table 2). Regarding the live-birth rate (per FET cycle) in this series, live-birth rates for AR and AD diseases were identical (50%; Table 2).
4.4. X-linked recessive (XR) diseases The overall diagnostic rate for the two XR diseases (Duchenne muscular dystrophy and XLMTM) in this study was 92.9% (Table 2). Although this met the recommended standard for PGD, no live birth was achieved in the two FET cycles (Table 2; Harton et al., 2011). Among the two XR diseases, XLMTM (MIM number 310400), also known as X-linked centronuclear myopathy (CNMX), is an X-linked congenital myopathy characterized by slowly progressive muscular weakness and wasting (Bitoun et al., 2005).
4.5. Comparison of ARMS-qPCR and microsatellite multiple linkage analysis One reason that microsatellite multiple linkage analysis is the current gold standard is this strategy can provide flexibility, namely transfer of either fresh or frozen embryos, since this methodology is PCRbased and results can be available within one working day (facilitating FET after trophectoderm biopsy of day-5 blastocysts). In contrast, another proposed protocol involving trophectoderm biopsy, followed by whole genome amplification and genotyping with minisequencing, reduced or eliminated the option of FET (Chang et al., 2013). It is noteworthy that our genotyping platform ARMS-qPCR can be done as quickly as the current gold standard. A recent randomized study that blastocyst biopsy yielded better pregnancy rates than day-3 cleavage stage embryo biopsy provided additional support for our ARMS-qPCR as a method of direct mutation screening, especially when FET is desired (Scott et al., 2013a,b). An example to demonstrate the comparisons between our novel assay and the gold-standard assay is as follows: in PGD of Duchenne muscular dystrophy (DMD), an XR disease, six embryos were tested by ARMS-qPCR (100% diagnostic rate [Table 2], which exceeded the gold standard of 90%; Harton et al., 2011). To compare the two methodologies, additional microsatellite linkage markers were selected and co-amplified. We designed and pre-tested 4 microsatellite markers flanking the DMD gene. However, only two markers were informative and they were used in this study. Although the availability of informative microsatellite markers may be limited, in general, at least 2 to 4 microsatellite markers closely linked to the target gene are preferably selected and applied (Chang et al., 2013; De Rycke et al., 2005; Harton et al., 2011; Korzebor et al., 2013). Although its diagnostic rate is expected to be higher, multiplex microsatellite linkage analysis is more timeconsuming than ARMS-qPCR. In addition, a PGD cycle with microsatellite multiple-marker analysis using fluorescence-based and multiplex PCR usually costs at least US$1000 (based on US$250 for each fluorescent-labeled primer) compared to only US$100 for ARMS-qPCR. Therefore, ARMS-qPCR is more economic than the gold standard genotyping strategy (microsatellite linkage analysis) and may be useful due to its lower cost and when informative microsatellite markers are not readily available. 4.6. Future perspectives Additional studies are needed to more fully explore the performance of ARMS-qPCR. For example, to circumvent ADO this method could be tested with the use of other types of polymorphic markers (e.g. SNP), which can be co-amplified in the same platform. In addition, since ADO is presumably dependent on the number of cells examined in PGD (Findlay et al., 1995), perhaps the performance of ARMS-qPCR would be improved if more cells were available. As an example, a trophectoderm biopsy could be done in lieu of the day-3 cleavagestage embryo biopsy used in the present study. This could also improve the live-birth rate, as it was recently reported that cleavage-stage embryo biopsy significantly impaired human embryonic implantation potential, whereas a blastocyst biopsy did not (Scott et al., 2013a). Otherwise, whole genome amplification (WGA) for the day-3 biopsied cell may also improve the act of ARMS-qPCR given that single cell WGA provides improvements of ADO over PCR-based methods (Zong et al., 2012). Based on the performance of the ARMS-qPCR method in the present study (in particular the frequent ADO), we concluded that it should not be used as the sole genotyping platform for PGD of AD diseases, but it might be useful as an adjunct testing procedure. For example, detecting mutations with both indirect (i.e. linkage markers) and direct methods (e.g. ARMS-qPCR or mini-sequencing) could improve the diagnostic rate and reduce the incidence of misdiagnosis. Subsequent to completion of the current study, we are currently using ARMS-qPCR coupled with multiple informative linked microsatellite markers. In addition,
H.-F. Chen et al. / Gene 548 (2014) 299–305
another pilot study to evaluate the performance of day-5 trophectoderm biopsy, followed by rapid genotyping (by ARMS-qPCR coupled with multiple informative linked microsatellite markers), and fresh embryo transfer, is planned for the near future. 5. Conclusions The present report evaluated the performance of our novel ARMSqPCR assay and determined the feasibility of this genotyping strategy in the PGD of a diverse group of single-gene disorders with varying types of inheritance. Genotyping results were sometimes available within one working day and the overall PGD diagnostic rate was 82.9% (range, 71.4 to 100%). It was most accurate for XR disorders (100%), followed by AR (84.2%), whereas its performance in AD disorders was less optimal (77.5%). Since ARMS-qPCR is susceptible to allele dropout (ADO), it is unlikely to replace the current gold-standard assay. However, for AR and XR diseases it may be used to complement the current gold standard in PGD. Conflicts of interest The ARMS-qPCR methodology has been filed and patent granted in Taiwan, R.O.C. (a member of the World Trade Organization). The authors claim no conflicts of interest. Author's contribution HFC, SHW, YHN, CAC, and MC conceived and designed the study; SPC, WHL, GCM, and MC conducted the experiments; and HFC, SPC, WHL, YCL, GCM, NAG, and MC wrote the paper. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.gene.2014.07.039. Acknowledgments We gratefully acknowledge critical comments from Dr. Pao-Lin Kuo (National Cheng-Kung University, Tainan, Taiwan) and Dr. John Kastelic (University of Calgary, Canada). We appreciate the collaboration of Dr. Tsung-Hsien Lee (Chung-Shan Medical University, Taichung, Taiwan), Dr. Horng-Der Tsai, Dr. Cheng-Hsuan Wu, Dr. Yu-Ching Chen, and Dr. Hsin-Hung Wu (Changhua Christian Hospital, Changhua, Taiwan) in recruiting patients. This study was kindly supported by two grants (100-CCH-ICO-04 and 101-CCH-IRP-43) from the Changhua Christian Hospital, and the grant from the National Science Council (NSC 1012314-B-371-004-MY3) to M. Chen. References Adiga, S.K., Kalthur, G., Kumar, P., et al., 2010. Preimplantation diagnosis of genetic diseases. J. Postgrad. Med. 56 (4), 317–320. Bitoun, M., Maugenre, S., Jeannet, P.Y., et al., 2005. Mutations in dynamin 2 cause dominant centronuclear myopathy. Nat. Genet. 37 (11), 1207–1209. Brezina, P.R., Brezina, D.S., Kearns, W.G., 2012. Preimplantation genetic testing. BMJ 345, e5908. Chang, L.J., Huang, C.C., Tsai, Y.Y., et al., 2013. Blastocyst biopsy and vitrification are effective for preimplantation genetic diagnosis of monogenic diseases. Hum. Reprod. 28 (5), 1435–1444. Chen, Y.L., Hung, C.C., Lin, S.Y., et al., 2011. Successful application of the strategy of blastocyst biopsy, vitrification, whole genome amplification, and thawed embryo transfer for preimplantation genetic diagnosis of neurofibromatosis type 1. Taiwan. J. Obstet. Gynecol. 50 (1), 74–78.
305
De Rycke, M., Georgiou, I., Sermon, K., et al., 2005. PGD for autosomal dominant polycystic kidney disease type 1. Mol. Hum. Reprod. 11 (1), 65–71. Drury, K.C., Liu, M.C., Lilleberg, S., et al., 2001. Results on single cell PCR for Huntington's gene and WAVE product analysis for preimplantation genetic diagnosis. Mol. Cell. Endocrinol. 183, S1–S4. Findlay, I., Ray, P., Quirke, P., et al., 1995. Allelic drop-out and preferential amplification in single cells and human blastomeres: implications for preimplantation diagnosis of sex and cystic fibrosis. Hum. Reprod. 10 (6), 1609–1618. Fiorentino, F., Biricik, A., Nuccitelli, A., et al., 2006. Strategies and clinical outcome of 250 cycles of preimplantation genetic diagnosis for single gene disorders. Hum. Reprod. 21 (3), 670–684. Gutiérrez-Mateo, C., Sánchez-García, J.F., Fischer, J., et al., 2009. Preimplantation genetic diagnosis of single-gene disorders: experience with more than 200 cycles conducted by a reference laboratory in the United States. Fertil. Steril. 92 (5), 1544–1556. Harper, J.C., Sengupta, S.B., 2012. Preimplantation genetic diagnosis: state of the art 2011. Hum. Genet. 131 (2), 175–186. Harper, J.C., Coonen, E., De Rycke, M., et al., 2010. ESHRE PGD Consortium data collection X: cycles from January to December 2007 with pregnancy follow-up to October 2008. Hum. Reprod. 25 (11), 2685–2707. Harton, G.L., Magli, M.C., Lundin, K., et al., 2011. ESHRE PGD consortium/embryology special interest group-best practice guidelines for polar body and embryo biopsy for preimplantation genetic diagnosis/screening (PGD/PGS). Hum. Reprod. 26 (1), 41–46. Korzebor, A., Derakhshandeh-Peykar, P., Meshkani, M., et al., 2013. Heterozygosity assessment of five STR loci located at 5q13 region for preimplantation genetic diagnosis of spinal muscular atrophy. Mol. Biol. Rep. 40 (1), 67–72. Kuo, S.J., Ma, G.C., Chang, S.P., et al., 2011. Preimplantation and prenatal genetic diagnosis of aromatic L-amino acid decarboxylase deficiency with an amplification refractory mutation system-quantitative polymerase chain reaction. Taiwan. J. Obstet. Gynecol. 50 (4), 468–473. Laurie, A.D., Hill, A.M., Harraway, J.R., et al., 2010. Preimplantation genetic diagnosis for hemophilia A using indirect linkage analysis and direct genotyping approaches. J. Thromb. Haemost. 8 (4), 783–789. Moutou, C., Gardes, N., Viville, S., 2003. Duplex PCR for preimplantation genetic diagnosis (PGD) of spinal muscular atrophy. Prenat. Diagn. 23 (8), 685–689. Moutou, C., Gardes, N., Nicod, J.C., et al., 2007. Strategies and outcomes of PGD of familial adenomatous polyposis. Mol. Hum. Reprod. 13 (2), 95–101. Newton, C.R., Graham, A., Heptinstall, L.E., et al., 1989. Analysis of any point mutation in DNA. The amplification refractory mutation system (ARMS). Nucleic Acids Res. 17 (7), 2503–2516. Pickering, S.J., McConnell, J.M., Johnson, M.H., et al., 1994. Use of a polymorphic dinucleotide repeat sequence to detect non-blastomeric contamination of the polymerase chain reaction in biopsy samples for preimplantation diagnosis. Hum. Reprod. 9 (8), 1539–1545. Ray, P.F., Handyside, A.H., 1996. Increasing the denaturation temperature during the first cycles of amplification reduces allele dropout from single cells for preimplantation genetic diagnosis. Mol. Hum. Reprod. 2 (3), 213–218. Scott Jr., R.T., Upham, K.M., Forman, E.J., et al., 2013a. Cleavage-stage biopsy significantly impairs human embryonic implantation potential while blastocyst biopsy does not: a randomized and paired clinical trial. Fertil. Steril. 100 (3), 624–630. Scott Jr., R.T., Upham, K.M., Forman, E.J., et al., 2013b. Blastocyst biopsy with comprehensive chromosome screening and fresh embryo transfer significantly increases in vitro fertilization implantation and delivery rates: a randomized controlled trial. Fertil. Steril. 100 (3), 697–703. SenGupta, S.B., Delhanty, J.D., 2012. Preimplantation genetic diagnosis: recent triumphs and remaining challenges. Expert. Rev. Mol. Diagn. 12 (6), 585–592. Sermon, K., De Vos, A., Van de Velde, H., et al., 1998. Fluorescent PCR and automated fragment analysis for the clinical application of preimplantation genetic diagnosis of myotonic dystrophy (Steinert's disease). Mol. Hum. Reprod. 4 (8), 791–796. Sermon, K., Van Steirteghem, A., Liebaers, I., 2004. Preimplantation genetic diagnosis. Lancet 363, 1633–1641. Spits, C., Sermon, K., 2009. PGD for monogenic disorders: aspects of molecular biology. Prenat. Diagn. 29, 50–56. Vrettou, C., Traeger-Synodinos, J., Tzetis, M., et al., 2004. Real-time PCR for single-cell genotyping in sickle cell and thalassemia syndromes as a rapid, accurate, reliable, and widely applicable protocol for preimplantation genetic diagnosis. Hum. Mutat. 23 (5), 513–521. Wells, D., Sherlock, J.K., 1998. Strategies for preimplantation genetic diagnosis of single gene disorders by DNA amplification. Prenat. Diagn. 18 (13), 1389–1401. Xu, K., Montag, M., 2012. New perspectives on embryo biopsy: not how, but when and why? Semin. Reprod. Med. 30 (4), 259–266. Zachaki, S., Vrettou, C., Destouni, A., et al., 2011. Novel and known microsatellite markers within the β-globin cluster to support robust preimplantation genetic diagnosis of β-thalassemia and sickle cell syndromes. Hemoglobin 35 (1), 56–66. Zong, C., Lu, S., Chapman, A.R., et al., 2012. Genome-wide detection of single-nucleotide and copy-number variations of a single human cell. Science 338 (6114), 1622–1626.