Molecular Human Reproduction, Vol. 9, No. 7, 411-420,
July 2003
© 2003 European Society of Human Reproduction and Embryology
Article |
Detailed investigation of factors influencing amplification efficiency and allele drop-out in single cell PCR: implications for preimplantation genetic diagnosis
Submitted on February 4, 2003; accepted on March 19, 2003
1 Department of Obstetrics and Gynaecology, University College London, 86–96 Chenies Mews, London WC1E 6HX, UK, 2 Department of Obstetrics and Gynaecology, Faculty of Medicine, Chiang Mai University, 110 Intawaroros Road, Sriphum, Mueang, Chiang Mai, 50200, Thailand and 3 The Institute for Reproductive Medicine and Science, Saint Barnabas Medical Center, 101 Old Short Hills Road, West Orange, NJ 07052, USA
4 To whom correspondence should be addressed. e-mail: dagan.wells{at}embryos.net
 |
ABSTRACT |
|---|
 Top ABSTRACT IntroductionMaterials and methodsResultsDiscussionREFERENCES |
|---|
Preimplantation genetic diagnosis (PGD) of single gene disorders relies on PCR-based tests performed on single cells (polar bodies or blastomeres). Despite the use of increasingly robust protocols, allele drop-out (ADO; the failure to amplify one of the two alleles in a heterozygous cell) remains a significant problem for diagnosis using single cell PCR. In extreme cases ADO can affect >40% of amplifications and has already caused several PGD misdiagnoses. We suggest that an improved understanding of the origins of ADO will allow development of more reliable PCR assays. In this study we carefully varied reaction conditions in >3000 single cell amplifications, allowing factors influencing ADO rates to be identified. ADO was found to be affected by amplicon size, amount of DNA degradation, freezing and thawing, the PCR programme, and the number of cells simultaneously amplified. Factors found to have little or no affect on ADO were local DNA sequence, denaturing temperature (94 or 96°C) and cell type. Consideration of the causal factors identified during this study should permit the design of PGD protocols that experience little ADO, thus improving the accuracy of PGD for single gene disorders.
Key words: allele drop-out/preferential amplification/preimplantation genetic diagnosis/single cell PCR/prenatal diagnosis
|
Introduction |
|---|
|
TopABSTRACTIntroductionMaterials and methodsResultsDiscussionREFERENCES |
|---|
Preimplantation genetic diagnosis (PGD) can be used as an alternative to prenatal diagnosis of inherited disease. The technique involves the biopsy and testing of single cells (blastomeres) derived from embryos generated using assisted reproductive techniques (Wells and Delhanty, 2001). The cell is usually biopsied on day 3 post-fertilization, at which time embryos typically consist of 6–10 cells. For diagnosis of single gene disorders the cell is then subjected to PCR, the causative gene or linked polymorphisms are amplified and the PCR product is subjected to mutation analysis (Wells and Sherlock, 1998). The result gained from the biopsied cell is used to infer the genotype of the embryo. Only unaffected embryos are transferred to the mother and consequently any resulting pregnancy should be unaffected.
Analytical techniques used for PGD need to be rapid, as most clinics favour embryo transfer on days 3 or 4, giving <36 h for diagnosis. The speed and accuracy of most of the reported PGD protocols are impressive, but several significant problems still face single cell PCR. The major difficulties are contamination, amplification failure (AF) and allele drop-out (ADO). Of these, contamination is perhaps the least enigmatic and consequently the easiest to address. With extreme care, appropriate facilities and much practice contamination of the sample with extraneous DNA can be almost eliminated. The use of DNA fingerprinting, to identify the rare occasions when contamination still occurs, further reduces the risk of misdiagnosis caused by contaminants (Piyamongkol et al., 2001a).
PCR amplification of single cells may also experience total failure of amplification. Amplification failure generally affects 5–10% of single cell subjected to PCR, but exceeds >20% in some reports. It is difficult to empirically determine the reason for amplification failure although there are several likely causes: the isolated cell might have been lost during transfer to the PCR tube, the cell could have been anucleate or in the process of degeneration, or the DNA might not have been accessible to the PCR reagents due to a failure of cell lysis.
Amplification failure and contamination are also encountered during PCR of larger DNA samples, albeit much less frequently. However, a problem that is unique to PCR of minute quantities of DNA is ADO (Ray et al., 1994; Findlay et al., 1995; Rechitsky et al., 1996), which is the failure of PCR to amplify one of the two alleles present in a cell. Only a single allele is amplified and detected after PCR, giving a heterozygous cell the appearance of homozygosity. ADO can affect either of the alleles of a given locus and strikes at random. For PGD of a recessive disease, where both mutations are located within the same fragment, the occurrence of ADO would not normally lead to the transfer of an affected embryo, but could reduce the number of heterozygous (unaffected carrier) embryos detected. If significant numbers of embryos are mistakenly excluded from transfer in this way, then the pregnancy rate may be reduced. For a dominant disease, failure to amplify the mutant allele is more serious, potentially leading to the transfer of affected embryos.
Of all the potential problems facing PGD, ADO has caused the most problems and has resulted in several misdiagnoses (Grifo et al., 1994; Verlinsky, 1996). In most cases the diseases being diagnosed were recessive (mostly cystic fibrosis), but the embryos were compound heterozygotes and the diagnostic protocols used only allowed the detection of one of the two mutations. If this ‘diagnostic’ mutation were subject to ADO then the embryos would appear normal regardless of their actual genotype. Compound heterozygosity can also lead to misdiagnosis in cases where the two mutations are both detectable, but cannot be encompassed by a single set of primers. In such cases PGD strategies sometimes employ multiplex amplification of two different fragments, each containing one of the mutations. If ADO affects the mutant allele in either of these fragments an affected embryo will be diagnosed as an unaffected carrier. On rare occasions when both fragments are subject to ADO an affected embryo may even appear to be homozygous normal.
There are several theories as to the origin of ADO. Some have speculated that it is caused by DNA degradation; leading to PCR-refractory breaks in both DNA strands. Another possibility is that ADO results from inaccessibility of the DNA template due to imperfect PCR conditions or incomplete cell lysis. Methods for improving amplification efficiency and minimizing ADO have been proposed and include use of highly sensitive fluorescent PCR (F-PCR) (Findlay et al., 1995), increasing the PCR denaturation temperature (Ray and Handyside, 1996), and the use of different cell lysis buffers (El-Hashemite and Delhanty, 1997; Thornhill et al., 2001). However, none of these measures consistently eliminates ADO. The most experienced PGD laboratories are generally able to reduce ADO rates to 5–15% (i.e. one of the two alleles affected in 5–15% of amplifications). However, ADO rates higher than this are not uncommon and in rare cases they may affect more than 40% of amplifications, severely compromising diagnosis.
Here we report a comprehensive analysis of factors influencing amplification efficiency, ADO and preferential amplification in single cells and suggest improved strategies for single cell PCR based on our results.
|
Materials and methods |
|---|
|
TopABSTRACTIntroductionMaterials and methodsResultsDiscussionREFERENCES |
|---|
Single cell PCR was performed using single buccal cells from subjects heterozygous for one or more of seven loci. Amplification efficiency, ADO and the amount of preferential amplification were assessed for each locus and compared to various physical or experimental parameters in over 3000 single cell PCR amplifications. Blastomeres analysed during this study were obtained from embryos undergoing PGD or from embryos that were surplus to IVF and donated for research by patients attending the Assisted Conception Unit of University College Hospital, London.
Unless stated otherwise, all cells used during this study were identical in terms of their treatment and the age of the sample. To control for any affects of freezing and thawing all cells were lysed and then frozen a single time, even if they were intended for immediate use. The thermal cycler and the reagents used for PCR were also kept constant for all experiments.
Single cell isolation and DNA extraction from single cells
Buccal cells, isolated by micromanipulation, were transferred into 5 µl droplets of phosphate-buffered saline (PBS; Gibco BRL, UK), 4% bovine serum albumin (Sigma, UK) on a Petri dish in a laminar flow cabinet. Cells were washed in a minimum of four fresh PBS droplets, while visualizing under a dissecting microscope, and were then transferred to 200 µl thin-wall microcentrifuge tubes, preloaded with 1 µl of 17 µmol/l SDS (Sigma) and 2 µl of 125 µg/ml proteinase K (PK) (Roche Diagnostics Ltd, UK) (El-Hashemite and Delhanty, 1997). A drop of lightweight mineral oil (Sigma) was added as an extra barrier against contamination and to prevent evaporation. Samples were incubated at 37°C for 1 h, proteinase K was then inactivated by heating at 99°C for 15 min. 2 µl of the last washing drop was also taken as a blank (negative control) for each blastomere assessed. For buccal cells, one wash drop blank was taken for every 10 cells tested. After lysis the extracted single cell DNA was stored at –80°C, unless otherwise stated.
Fluorescent PCR on single cell DNA
Extracted DNA from single cells was amplified by PCR using one, two or occasionally three sets of primers (see Table I). The loci amplified were: the repetitive region of the myotonic dystrophy gene, DMPK, on chromosome 19; APOC2, a microsatellite linked to DMPK; D21S1414 and D21S11, microsatellites on chromosome 21; ‘bthalW1’ and ‘outer ß-thalassaemia’, two partially overlapping ß-globin fragments; HUMTH01, a microsatellite polymorphism linked to ß-globin on chromosome 11 (details in Table I). The PCR mixture consisted of 0.2 µmol/l of each primer, 200 µmol/l deoxynucleotide triphosphates (dNTP) (Pharmacia Biotech, UK), 1xGeneAmp Buffer (10xGeneAmp Buffer contains 100 mmol/l Tris–HCl, pH 8.3, 500 mmol/l KCl, 15 mmol/l MgCl2; Applied Biosystems, UK) and 1.5 IU AmpliTaq Gold (Applied Biosystems) and was made up to a total volume of 25 µl with distilled deionized water. The amplifications were performed with the conditions: denaturation at 94°C for 45 s (96°C for the first 10 cycles), annealing at 60°C for 45 s and extension at 72°C for 1 min for 40 cycles. These were preceded by primary denaturation at 94°C for 12 min to activate the AmpliTaq GoldTM enzyme and followed by final extension at 72°C for 10 min. Other thermal programs (denaturation/annealing/extension) of 15 s/15 s/25 s and 30 s/30 s/45 s and other modifications as outlined in Table II were also carried out.
|
|
GeneScanTM analysis on ABI PrismTM310
The multiplex amplified products from single cells were tagged with fluorochromes by using labelled primers. This allowed analysis to be performed on an automated fluorescent sequencer, ABI PrismTM310 (Applied Biosystems). DMPK and ß-globin fragments were labelled with the green fluorescent dye (TET®; Applied Biosystems), and APOC2 and HUMTH01 fragments with the blue fluorescent dye (6-FAM®; Applied Biosystems). A mixture of 1 µl fluorescent PCR products, 12 µl of deionized formamide and 0.5 µl of size standard [Genescan®-500 (TAMRA); Applied Biosystems] was prepared and denatured at 95°C for 5 min. The denatured samples were subjected to capillary electrophoresis using Performance Optimised Polymer 4 (POP-4TM; Applied Biosystems; 5 s injection time, 15 000 V, 60°C, 24 min). The data were analysed by GeneScanTM analysis software (Applied Biosystems) (Figure 1). The peak areas of the alleles of each sample were compared and defined into different amplification characteristics, as described below.
|
Nested PCR and single strand conformational polymorphism (SSCP) analysis for ß-thalassaemia, IVSI-110 mutation
After multiplex amplification using outer ß-thalassaemia primers and HUMTH01 primers, 2 µl of the reaction was added to a second PCR, utilizing nested ß-thalassaemia primers (El-Hashemite et al., 1997). The PCR products from this second (inner) amplification were analysed by SSCP, while the HUMTH01 PCR products, present in the outer reaction, were analysed by electrophoresis on an ABI PrismTM310. For SSCP 2 µl of formamide was added to 1.5 µl of PCR product and the mixture denatured at 95°C for 10 min. The denatured samples were kept on ice until loading onto a 12.5% polyacrylamide gel (GeneGel® Excel; Pharmacia Biotech, UK). The SSCP analysis was carried out by gel electrophoresis on the GenePhorTM Electrophoresis Unit (Pharmacia Biotech), with the following conditions: 5°C, 600 V, 15 W, 25 mA, 1 h 30 min DNA bands were visualized by silver staining (Harvey et al., 1995). The band patterns of the ß-globin alleles of each sample were compared and defined into different amplification characteristics, as described below.
Definition of amplification characteristics
Symmetrical amplification: the two alleles are considered equally amplified. The quantity of each allele differs by <10%, as determined by analysis of the area of each fluorescent peak after electrophoresis.
Preferential amplification: the quantity of each allele differs by >10%. One allele is over-amplified with respect to the other, but both are above the threshold of detection.
Accurate amplification: both alleles are amplified to a detectable level, but not necessarily to an equal extent (i.e. cases of symmetrical amplification + cases of preferential amplification).
Amplification failure: neither allele can be detected; the locus has failed to amplify all together.
Allele drop-out (ADO): only one allele is detectable; the other has failed to amplify.
ADO rate: calculated by dividing the number of reactions displaying ADO by the number of reactions in which one or more alleles were detected (i.e. cases of total amplification failure are excluded from this calculation).
Amplification efficiency: defined as the proportion of tests in which one or more alleles is successfully amplified. This includes cases of ADO and preferential amplification, as well as cases of symmetrical amplification.
|
Results |
|---|
|
TopABSTRACTIntroductionMaterials and methodsResultsDiscussionREFERENCES |
|---|
Freezing and thawing
The influence of freezing and thawing on amplification failure and ADO was assessed by comparing two groups of 50 buccal cells for amplification of the DMPK and D21S1414 loci. Both sets of cells were isolated from the same sample at the same time, but one group was lysed and then frozen prior to PCR, while the other group was lysed and then used immediately. Amplification rates for fresh cells averaged 94% for the two loci and ADO was seen in 19% of amplifications. Frozen cells were successfully amplified in 85% of reactions and displayed 30% ADO (Table II, Study A). The difference between the two sets of cells was statistically significant (
2, P = 0.0091).
Amount of DNA necessary to avoid ADO
Experiments using precise numbers of isolated cells were employed to determine how much DNA (or how many cells) is necessary to entirely eliminate ADO. The ADO rates and amplification efficiencies were calculated for the DMPK and D21S1414 loci in 40 single cell amplifications, 40 two-cell amplifications and 40 three-cell amplifications. The combined amplification efficiencies and ADO rates for the two loci are given in Table II, Study B. The reduction in ADO produced by the use of two cells verses one cell is statistically significant (2, P < 0.0001). The use of three cells rather than two improved results further, with ADO rates approaching 0% and amplification efficiencies of 
100%. However, a statistical reduction in ADO for three cells versus two could not be demonstrated with the number of cells used here.
Cell type/quality and the affect of DNA degradation
The amplification of various loci (DMPK, APOC2, D21S1414, ß-globin) from 132 blastomeres was compared with identical reactions performed on 450 buccal cells. These experiments revealed no significant differences between the amplification of the two cell types. Average ADO rates for buccal cells and blastomeres were 9.5 and 10.4% respectively, while average amplification efficiencies were 96.2 and 94.7% (2, P = 0.706). More important than cell type was the condition of cells. Fifty buccal cells were aged at room temperature for 7 days prior to isolation and PCR amplification. After ageing, the HUMTH01 locus gave an ADO rate of 14.9% and an amplification efficiency of 94%. Fresh cells gave lower ADO rates of 7.4% and gave an amplification efficiency of 96.4%.
For the amplification of blastomeres the presence of a visible nucleus was also a critical factor in amplification efficiency. In clinical PGD cases, using nucleated blastomeres, we have obtained successful amplification on 94.4% of occasions (100/106 loci amplified). This compares to a 0% (0/8) amplification efficiency for blastomeres that had no visible nucleus.
Examination of preferential amplification
Two polymorphic loci (APOC2 and DMPK) were each amplified in 100 single cells using a duplex PCR. The size of the amplified alleles varied according to the different loci and patients tested. The differences between the size of alleles at a single locus ranged from 4 to 45 bp. The amplification of polymorphisms, in which alleles vary in length, demonstrated that both longer and shorter alleles could be subject to preferential amplification. However, in 66% of amplifications the shorter allele was relatively over-amplified, compared with only 19% of amplifications in which the longer allele was over-amplified. Despite the differences in preferential amplification, ADO rates for long or short alleles of the same locus did not differ significantly (ADO rates for the short and long alleles were 3.9 and 5% respectively).
Multiplex PCR
In the course of this study four distinct multiplex PCR protocols were investigated. Comparison of single cell amplification using individual pairs of primers with multiplex PCR using combinations of primer pairs revealed that in the vast majority of cases the addition of an extra set of primers did not significantly affect ADO rates or amplification efficiency. To demonstrate this empirically we amplified 50 single cells using D21S1414 primers and 50 more cells using a combination of D21S1414 and DMPK primers. Amplification efficiency and ADO rates for D21S1414 were 70 and 32% when amplified alone and 74 and 34% when amplified in a duplex reaction. Similar results were seen with amplification of the DMPK locus. This gave amplification efficiency and ADO rates of 94 and 12.8% respectively when amplified alone; 95 and 9.5% when amplified simultaneously with the APOC2 locus and 98 and 4.1% when amplified with D21S1414. The pooled data for these combinations of loci is given in Table II, study C.
The only combination of loci that encountered problems during multiplex PCR contained primers for the ß-globin (‘outer ß-thalassaemia’) and HUMTH01 loci. In this case no HUMTH01 product was amplified. The problem could not be rectified despite the use of numerous different thermal cycling conditions in >300 single cell PCR. Ultimately efficient amplification of both loci was achieved, but only after a modified Taq polymerase (AmpliTaq GoldTM) was employed. Use of this polymerase provided 96.7% amplification efficiency with 9.0% ADO for the HUMTH01 locus (150 single buccal cells tested), not significantly different from rates obtained using HUMTH01 primers alone. Similar results (amplification efficiency of 95.8 and 8.7% ADO rate) were obtained using 24 blastomeres.
Deviation from expected amplification characteristics
The amplification characteristics of 950 single cell duplex PCR were analysed to determine whether ADO and amplification failure affect loci amplified in the same reaction in an independent fashion, or whether amplification failure/ADO at one locus increases the likelihood that amplification failure/ADO will be observed at the other (Table III). On the basis of the amplification failure rates observed for individual loci (57/950 for locus 1 and 130/950 for locus 2) it was predicted that failure of both loci to amplify should occur in 0.8% of amplifications. However, coincident amplification failure was actually observed in 3.7% of PCR, a highly statistically significant increase above the expected.
|
Not only was there an increased coincidence of amplification failures, but ADO also affected both loci more often than expected. Fifty amplifications displayed simultaneous ADO at both loci, rather than the expected 38 [expected coincidence of ADO = (153/950x234/950)x950]. However, even the large number of cells tested here were insufficient to reveal whether this difference was significant, or just a chance variation.
PCR cycling conditions
The affect of different denaturing temperatures and PCR cycle lengths on ADO and amplification efficiency was investigated by adjusting the PCR programme. The amplification of DMPK and D21S1414 loci from 100 cells, using denaturing temperatures of 94°C throughout the programme, was compared with a programme that employed a 96°C denaturation for the first 10 PCR cycles and 94°C for the remaining cycles. The amplification efficiencies for DMPK and D21S1414 achieved using the 94°C programme averaged 88%, while for the 96°C programme they averaged 85%. ADO rates averaged 30% for both the 94 and 96°C programmes (Table II, Study D). The differences are not statistically significant for either locus (2, P = 0.591 and 0.308 for the DMPK and D21S1414 loci, respectively) and indicate that an increase in denaturing temperature from 94 to 96°C has little if any effect on ADO.
Two hundred single buccal cells from subjects heterozygous for the DMPK triplet repeat and the polymorphic APOC2 locus were duplex amplified using one of three different thermal cycle programmes: denaturation/annealing/extension steps lasting 15 s/15 s/25 s, 30 s/30 s/45 s or 45 s/45 s/1 min (Table II, Study E). It was observed that the rates of amplification failure and ADO for both loci were gradually reduced with increasing PCR program length. The longest PCR program displayed the lowest ADO rates, averaging 5% for the two loci, compared with 16% for the shortest program (2, P < 0.0001).
To ascertain which element of the cycle benefits most from an extended duration, different components were independently tested. An increase in extension time from 1 min to 2 min (for the first 10 cycles only) was tested on 100 cells. There was no significant difference in the amplification efficiency and ADO rates between both groups for the DMPK or D21S1414 loci (Table II, Study F).
The effect of altering the length of the denaturing phase was also tested (Table II, Study G). One hundred cells were subjected to PCR for the D21S1414 and DMPK loci, half using denaturing times of 45 s and half using 2 min (for the first 10 cycles only). For cells treated with the extended denaturing time the averaged ADO rates and amplification efficiencies were 22 and 91% respectively. This compares with rates of 30 and 85% for cells denatured for just 45 s. Both loci show an apparent trend towards improved amplification with increased denaturing time; however, this is not statistically significant for the number of cells tested in this study (2, P = 0.1159).
DNA sequence composition and primers
To determine whether DNA sequence influences the occurrence of ADO, the base composition, melting temperature (Tm) and tendency to form secondary structures was analysed for seven amplicons and compared to their ADO rates (Table IV). GC contents ranged from 33–63%, but no correlations related to ADO were identified. As well as the amplified fragments, the sites of primer annealing and flanking regions (200 bp upstream and 200 bp downstream) were also tested. No association between ADO rate and any of these characteristics was detected despite a wide range of Tm and GC contents. Loci containing repetitive sequences displayed no significant difference in ADO and amplification efficiency rates when compared with loci containing only unique sequence.
|
To further investigate the affect of differences in DNA sequence, the amplification characteristics of D21S1414 and D21S11 were compared. These loci share 63% of their sequence, have the same GC composition and share a primer-annealing site. The only difference is that D21S1414 is amplified using a reverse primer downstream of that used for amplification of D21S11. Consequently the fragments generated are an average of 135 bp longer than their D21S11 counterparts. If differences in DNA sequence were significant determinants of ADO, one would expect these loci to behave similarly. One hundred cells were isolated from the same sample, at the same time, and half were amplified with each set of primers. ADO and amplification efficiency rates were 32 and 70% for D21S1414 and 20 and 90% for D21S11, and thus differed significantly (
2, P = 0.004; Table II, Study H).
Fragment length
To explore the influence of amplified fragment size on ADO and amplification efficiency seven loci were amplified from 650 single buccal cells. A significant increase in ADO and a small increase in amplification failure with fragment size were observed (Figure 2 and Figure 3), suggesting that ADO, amplification failure and fragment length are intimately linked. During these studies particular attention was focused on D21S1414 and D21S11, two loci that share extensive regions of sequence, but differ in length.
|
|
|
Discussion |
|---|
|
TopABSTRACTIntroductionMaterials and methodsResultsDiscussionREFERENCES |
|---|
Anyone involved in PGD, or any other areas that utilize PCR amplification of single cells, should be acutely aware of the problem of ADO—the random failure of one of the two alleles in a heterozygous cell to amplify. Although the existence of ADO has been acknowledged for several years, progress in combating it has been slow, and no way to entirely eliminate the problem has yet been discovered (Findlay et al., 1995). There has been little prospect that this situation will improve, largely because the basis of the phenomenon is so poorly understood. Review of the many single cell PCR methods in the literature provides little assistance in formulating theories as to its origin, because while two published methods might appear to be analogous they will certainly contain numerous subtle, but critical, differences that prevent direct comparison. In this study we have employed carefully controlled tests allowing evaluation of many factors likely to affect the incidence of ADO. We have analysed large numbers of loci (generally 100–200 for each set of conditions); however, subtle alterations of ADO rates (changes of <5%) can only be statistically verified when several hundred loci are tested. Consequently it is possible that some factors might have been overlooked by this study, although their influence on ADO is predicted to be minor.
PCR conditions
It has been suggested that sub-optimal reaction conditions during PCR might increase ADO rates. Specifically, inadequate denaturation of DNA samples has been linked to the occurrence of ADO in a previous study (Ray and Handyside, 1996). For this reason many PGD units now employ increased denaturing temperatures for the first 5–10 cycles of amplification, a modification said to significantly reduce ADO rates. However, in this study we found no difference in ADO rates when comparing denaturing temperatures of 94 and 96°C (Table II, Study D), suggesting that the temperatures typically used in routine PCR are above the threshold necessary for efficient denaturation of most DNA templates. The improvement in ADO rates originally reported might have been observed because a wider range of temperatures was tested, 90, 93 and 96°C. The use of a different type of polymerase in this study might also have reduced the significance of temperature differences. Although we were unable to confirm that increased denaturing temperature reduces ADO, our data suggest that longer denaturing times might have a beneficial effect on ADO rates, a possibility that warrants further investigation.
To explore the possibility that the melting temperature of the target affects ADO, and to determine whether base sequence or secondary structure have any influence, we analysed primer sites, flanking sequences and amplicons for melting temperature, sequence composition and propensity to form secondary structures. No association between any of these factors and ADO rate was detected (Table IV). Our experiments using whole genome amplification (WGA) performed on single cells have also indicated that the site of primer annealing and the local secondary structure are unlikely to influence most cases of ADO. Whole genome amplification methods such as primer extension preamplification initiate replication of a given locus from different sites each cycle, and are not dependent on the annealing of a primer to one specific location. Theoretically this should overcome many problems specific to the sequence around and within the priming sites. However, existing data suggests that the use of WGA does not reduce the incidence of ADO (Hahn et al., 1998; D.Wells and J.K.Sherlock; unpublished observations).
Cell lysis
Several studies have shown that the choice of cell lysis protocol can have a considerable impact on ADO rates. This presumably reflects variation in the ability of different methods to ensure that the DNA template is made accessible to the PCR reagents. Most centres involved in PGD use either proteinase K or alkaline buffers to achieve cell lysis. Both methods appear to reduce levels of ADO when compared to lysis in distilled water alone (Sermon et al., 1995; El-Hashemite and Delhanty, 1997). Unfortunately the optimal methodology remains unclear, as data in the literature are contradictory. One study claimed that proteinase K digestion could accomplish a complete elimination of ADO (El-Hashemite and Delhanty, 1997). A later study also suggested a beneficial effect of proteinase K digestion (compared with alkaline lysis the use of proteinase K more than halved the ADO rate at the two loci tested), but was unable to replicate a 0% ADO rate (Abou-Sleiman, personal communication). A third study found no appreciable difference between ADO rates using the two methodologies (Verlinsky and Kuliev, 1998), while a fourth found that alkaline lysis was superior to proteinase K, granting a marked reduction in ADO for two out of three loci tested (Thornhill et al., 2001). Laboratories that have attempted comparison between lysis using proteinase K and alkaline buffers seem to achieve the best results with the technique that they already use routinely. This might suggest that neither technique is particularly robust, both performing poorly in the hands of those unfamiliar with their use.
Multiplex PCR
Many single cell diagnostic strategies now employ multiplex PCR, a method that allows amplification of several distinct DNA fragments in a single reaction. The amplification of multiple fragments permits redundant diagnoses to be performed. It is unlikely that ADO will affect all of the fragments amplified in a multiplex reaction and consequently the probability of misdiagnosis as a result of ADO is much reduced (Ao et al., 1998; Kuliev et al., 1999; Piyamongkol et al., 2001a;b). The development of more sophisticated PGD protocols has produced more robust methodologies for single cell testing. However, some groups have been concerned that the increased complexity of the reaction might adversely affect amplification of individual loci and thus be counterproductive.
Our analysis of four different primer combinations during this study, and our wider experience of single cell PCR, suggest that incompatible mixtures of primers will occasionally be encountered. However, such problems are easily identified during preliminary testing of new protocols and may sometimes be overcome by the use of a modified Taq polymerase enzyme (e.g. AmpliTaq Gold). Importantly, the addition of an extra set of primers was found to have little if any affect on ADO rates (Table II, Study C). Our data confirm that duplex PCR can be used with confidence at the single cell level, provided preliminary testing is undertaken. However, it should be noted that increasing the number of primer sets added to a multiplex PCR (e.g. triplex, quadruplex, etc) will inevitably increase the probability of incompatibilities being encountered.
Cell number
Most PGD centres currently employ biopsy of 1–2 cells from the 8–10 cell embryo on day 3 post-fertilization. However, recent improvements in blastocyst culture have improved the feasibility of sampling a larger quantity of cells on day 4 or 5. Our data confirm that diagnostic accuracy would benefit significantly from the biopsy of a larger number of cells. However, the greatest reduction in ADO rates is achieved when the number of cells amplified is increased from one to two. Sampling of three cells provides a slight improvement over that seen with two cells. Our data suggest that for most loci it would only be necessary to take 3 cells to achieve 100% amplification efficiency and ADO rates that approach zero (Table II, Study B).
Cell type
The variety of cells amplified is also said to influence ADO rates (Rechitsky et al., 1998), a significant factor since preliminary testing of new PGD protocols usually focuses on buccal cells or lymphocytes rather than blastomeres. Although we observed some variation related to cell type (e.g. buccal cells performed slightly better than blastomeres), we found that these discrepancies were minor in comparison to differences caused by the way the cells had been handled and treated. Anucleate blastomeres and cells from fragmented embryos were seen to display a high frequency of total amplification failure as previously reported (Ray et al., 1998; Kanavakis et al., 1999), but these blastomeres account for a small minority (8%) of the total analysed. Theoretically blastomeres could give a higher than expected ADO rate due to the possibility that the cell sampled is haploid or monosomic. Cytogenetic analyses predict 7–15% of blastomeres to be haploid (Harper et al., 1995; Kuo et al., 1998). However, the extremely high ADO rates in blastomeres relative to other cell types, reported by others (Rechitsky et al., 1998), were not observed in this study, suggesting that the differences these authors observed could have been due to factors other than cell type. Our results from blastomeres may be superior because the cells tested were always fresh. The use of fresh cells should give results analogous to the performance of blastomeres in a PGD case.
Freezing and thawing
We believe that some of the variation in ADO rates, which we observed in earlier work, could have been a consequence of differences in the number of times cells had been frozen and thawed. This is important, as the patient cells used to evaluate new PGD protocols are often stored in the freezer prior to use. ADO rates are then inferred from work done on these cells, but this might not be representative of the ADO rates that will be observed when fresh blastomeres are amplified. There are two ways in which freezing could affect ADO and these highlight two of the principal factors suspected of causing this phenomenon. Firstly, the accessibility of the DNA could be affected, as the freezing process might help to rupture cell membranes thereby improving access to the DNA and possibly reducing ADO. Alternatively, freezing and thawing might damage the DNA and consequently prevent amplification. This latter possibility is supported by our data, which clearly indicate that freezing is detrimental.
Preferential amplification
Preferential amplification of one of the two alleles in a heterozygous cell is a common phenomenon. In cases of marked preferential amplification the results may resemble ADO, unless sensitive detection methods such as fluorescent PCR (F-PCR) are used to reveal the under-amplified allele (Findlay et al., 1995). Preferential amplification affects alleles at random, although there is a tendency towards preferential amplification of shorter DNA fragments in PCR where the alleles differ in size (Walsh et al., 1992). This is purely an effect of PCR reaction kinetics.
For qualitative analysis, as in PGD, most cases of preferential amplification are of little importance. However, it interferes with attempts to employ quantitative fluorescent PCR (QF-PCR) at the single cell level. Quantitative fluorescent PCR has been successfully applied to the detection of aneuploidy in larger DNA samples (e.g. prenatal samples), but relies on alleles being amplified proportionately (Mansfield, 1993). Our data and those of others suggests that <75% of single cells give accurate, quantitative amplification of alleles (Sherlock et al., 1998). For this reason single cell trisomy screening can only be accomplished using QF-PCR if multiple polymorphic loci are examined, increasing the likelihood that a locus with no preferential amplification, or a locus that has three distinct alleles, will be detected (Katz et al., 2002).
Association of ADO and amplification failure
It is clear from our data that the phenomena of amplification failure and ADO are complex and are influenced by multiple factors. The two are, however, clearly related; loci displaying high ADO also tend to have high amplification failure rates. It is logical that whatever causes ADO will sometimes cause amplification failure (i.e. simultaneous ADO of both alleles). The incidence of amplification failure caused in this way is predicted to equal (ADO rate)2. Review of our data from multiple loci indicates that ADO can explain some cases of total amplification failure, but not all. Indeed, depending on the locus examined, amplification failure affects 2–7% more reactions than would be anticipated on the basis of ADO alone. Factors unrelated to ADO, responsible for this 2–7% discrepancy, probably include cell loss during transfer to the PCR tube, accidental sampling of an anucleate cell, and failure of cell lysis, all of which are likely to prevent any amplification from taking place. These factors also explain why duplex reactions suffer simultaneous failure of both loci to amplify in 3% more reactions than predicted from the amplification efficiencies of each locus (Table III). Together, these data suggest that there is a baseline amplification failure rate of 2–7% (due to cell loss, anucleate cells and failed lysis), with additional amplification failure attributable to simultaneous ADO of both alleles. For most purposes the baseline amplification failure rate will be the same for all loci. However, there may be some variation between laboratories (due to differences in micromanipulation and cell lysis techniques) and between samples (due to different amounts of DNA degradation).
The relationship between ADO and total amplification failure was emphasized by our analysis of almost 1000 single cell duplex PCR amplification. When comparing data from the two simultaneously amplified loci, more cells than expected were found to display a combination of ADO at the first locus and amplification failure at the second, hinting at a shared basis for these phenomena. Cells with ADO at one locus also had a slightly higher incidence than expected of ADO at the second locus. Differences such as these might indicate that a small proportion of cells (1%) are predisposed to unusually high levels of ADO, perhaps due to degraded DNA in dead or apoptotic cells. The existence of such cells will have a small impact on the accuracy of PGD protocols that utilize more than one diagnostic locus, making redundant diagnoses slightly less effective than anticipated.
DNA degradation
DNA degradation can explain both ADO and some cases of amplification failure. A double-stranded break in the DNA is refractory to PCR, as is the coincidence of two single-stranded breaks (one on the antisense strand and one on the sense strand) if they both fall between the primers used for DNA amplification. Depending on whether one allele or both is affected by such DNA damage, ADO or total amplification failure will result. Theoretically DNA degradation could also cause some cases of preferential amplification. A single-strand break, present from the beginning of the first PCR cycle, is predicted to cause a 2:1 preferential amplification in favour of the undamaged allele.
The affect of DNA degradation may be particularly relevant to the use of buccal cells during the development of new PGD protocols. Such cells, collected by mouthwash, are shed towards the end of their life or after death and may contain DNA degraded by apoptotic processes, digested by bacterial action, or damaged by the cells’ own enzymes following breakdown of internal membranes. There have been occasions when buccal cell samples have been shipped to our laboratory and have spent several days at room temperature before delivery, potentially leading to further degradation. The effect of poor sample quality was illustrated by our comparison of buccal cells aged for 7 days with fresh buccal cells. The 7 day old cells were found to have increased levels of amplification failure and ADO rates were increased by 7.5%. A relationship between cell quality ADO and amplification failure has been noted previously (Ray et al., 1998).
Fragment size
If DNA degradation is a significant cause of ADO and amplification failure then it seems likely that these problems should affect long amplified fragments more often than shorter fragments. Breaks in the DNA strand are expected to occur at random positions, consequently the longer the DNA fragment the greater the chance of a break occurring between the two PCR primers. There is some anecdotal evidence suggesting that large amplicons do indeed have an increased amplification failure rate (Sermon et al., 1996; Vrettou et al., 1999). Our analysis of seven different loci has now provided confirmation that long fragments have higher ADO rates as well as more amplification failure when compared with small amplicons. This is shown in Figures 2 and 3 and was further illustrated by our comparison of the amplification of D21S1414 and D21S11 (Table II, Study H), overlapping loci that share most of their DNA sequence and one of their priming sites, but differ in size. Extrapolation from our data on ADO, amplification failure and fragment length suggests that amplification of fragments <40 bp in length could cause ADO rates to approach zero, and decrease amplification failure to a baseline (2–7%) level, caused by factors unrelated to ADO.
On the basis of our data, the most accurate strategy for PGD is to take two fresh, nucleated blastomeres, use an efficient lysis protocol, and then amplify small fragments containing the mutation sites. We have designed single cell PCR tests for cystic fibrosis and ß-thalassaemia that utilize minisequencing, a mutation detection method that permits screening of PCR fragments as small in size as 45 bp (Bermúdez et al., 2003). Because of the small size of the amplified fragment it should be possible to complete amplification and electrophoretic analysis extremely rapidly, as well as reducing ADO and amplification failure rates. In the event of persistent ADO this test could be performed as a multiplex PCR, allowing simultaneous amplification of a linked marker to provide a back-up diagnosis.
|
Acknowledgements |
|---|
Dr Wirawit Piyamongkol was funded by the Royal Thai Government and Faculty of Medicine, Chiang Mai University, Thailand. Dr Dagan Wells was initially supported by an MRC research fellowship and later by the Institute for Reproductive Medicine and Science of Saint Barnabas Medical Center.
|
REFERENCES |
|---|
|
TopABSTRACTIntroductionMaterials and methodsResultsDiscussionREFERENCES |
|---|
Ao, A., Wells, D., Handyside, A.H., Winston, R.M. and Delhanty, J.D. (1998) Preimplantation genetic diagnosis of inherited cancer: familial adenomatous polyposis coli. J. Assist. Reprod. Genet., 15, 140–144.[CrossRef][Web of Science][Medline]
Bermúdez, M.G., Piyamongkol, W., Tomaz, S., Dudman, E., Sherlock, J.K. and Wells, D. (2003) Single cell sequencing and mini-sequencing for preimplantation genetic diagnosis. Prenat. Diagn. (in press).
Brook, J.D., McCurrach, M.E., Harley, H.G., Buckler, A.J., Church, D., Aburatani, H., Hunter, K., Stanton, V.P., Thirion, J.P., Hudson, T. et al. (1992) Molecular basis of myotonic dystrophy: expansion of a trinucleotide (CTG) repeat at the 3' end of a transcript encoding a protein kinase family member. Cell, 68, 799–808.[CrossRef][Web of Science][Medline]
El-Hashemite, N. and Delhanty, J.D. (1997) A technique for eliminating allele specific amplification failure during DNA amplification of heterozygous cells for preimplantation diagnosis. Mol. Hum. Reprod., 3, 975–978.
El-Hashemite, N., Wells, D. and Delhanty, J.D. (1997) Single cell detection of beta-thalassaemia mutations using silver stained SSCP analysis: an application for preimplantation diagnosis. Mol. Hum. Reprod., 3, 693–698.
Findlay, I., Ray, P., Quirke, P., Rutherford, A. and Lilford, R. (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, 1609–1618.
Grifo, J.A., Tang, Y.X., Munné, S., Alikani, M., Cohen, J. and Rosenwaks, Z. (1994) Healthy deliveries from biopsied human embryos. Hum. Reprod., 9, 912–916.
Hahn, S., Garvin, A.M., Di Naro, E. and Holzgreve, W. (1998) Allele drop-out can occur in alleles differing by a single nucleotide and is not alleviated by preamplification or minor template increments. Genet. Test., 2, 351–355.[Web of Science][Medline]
Harper, J.C., Coonen, E., Handyside, A.H., Winston, R.M., Hopman, A.H. and Delhanty, J.D. (1995) Mosaicism of autosomes and sex chromosomes in morphologically normal, monospermic preimplantation human embryos. Prenat. Diagn., 15, 41–49.[Web of Science][Medline]
Harvey, C.B., Pratt, W.S., Islam, I., Whitehouse, D.B. and Swallow, D.M. (1995) DNA polymorphisms in the lactase gene. Linkage disequilibrium across the 70-kb region. Eur. J. Hum. Genet., 3, 27–41.[Web of Science][Medline]
Kanavakis, E., Vrettou, C., Palmer, G., Tzetis, M., Mastrominas, M. and Traeger-Synodinos, J. (1999) Preimplantation genetic diagnosis in 10 couples at risk for transmitting beta-thalassaemia major: clinical experience including the initiation of six singleton pregnancies. Prenat. Diagn., 19, 1217–1222.[CrossRef][Web of Science][Medline]
Katz, M.G., Mansfield, J., Gras, L., Trounson, A.O. and Cram, D.S. (2002) Diagnosis of trisomy 21 in preimplantation embryos by single-cell DNA fingerprinting. Reprod. Biomed. Online, 4, 43–50.[Medline]
Kuliev, A., Rechitsky, S., Verlinsky, O., Ivakhnenko, V., Evsikov, S., Wolf, G., Angastiniotis, M., Georghiou, D., Kukharenko, V., Strom, C. and Verlinsky, Y. (1998) Preimplantation diagnosis of thalassemias. J. Assist. Reprod. Genet., 15, 219–225.[CrossRef][Web of Science][Medline]
Kuliev, A., Rechitsky, S., Verlinsky, O., Ivakhnenko, V., Cieslak, J., Evsikov, S., Wolf, G., Angastiniotis, M., Kalakoutis, G., Strom, C. and Verlinsky, Y. (1999) Birth of healthy children after preimplantation diagnosis of thalassemias. J. Assist. Reprod. Genet., 16, 207–211.[CrossRef][Web of Science][Medline]
Kuo, H.C., Ogilvie, C.M. and Handyside, A.H. (1998) Chromosomal mosaicism in cleavage-stage human embryos and the accuracy of single-cell genetic analysis. J. Assist. Reprod. Genet., 15, 276–280.[CrossRef][Web of Science][Medline]
Mansfield, E.S. (1993) Diagnosis of Down syndrome and other aneuploidies using quantitative polymerase chain reaction and small tandem repeat polymorphisms. Hum. Mol. Genet., 2, 43–50.
Piyamongkol, W., Harper, J.C., Sherlock, J.K., Doshi, A., Serhal, P.F., Delhanty, J.D. and Wells, D. (2001a) A successful strategy for preimplantation genetic diagnosis of myotonic dystrophy using multiplex fluorescent PCR. Prenat. Diagn., 21, 223–232.[CrossRef][Web of Science][Medline]
Piyamongkol, W., Harper, J.C., Delhanty, J.D.A. and Wells, D. (2001b) Preimplantation genetic diagnostic protocols for alpha- and beta-thalassaemias using multiplex fluorescent PCR. Prenat. Diagn., 21, 753–759.[CrossRef][Web of Science][Medline]
Ray, P.F. and 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, 213–218.
Ray, P.F., Winston, R.M.L. and Handyside, A.H. (1994) Single cell analysis for diagnosis of cystic fibrosis and Lesch Nyhan syndrome in human embryos before implantation. Miami Bio/Technology European Symposium, Advances in Gene Technology: Molecular Biology and Human Genetic Disease, 5, 46.
Ray, P.F., Ao, A., Taylor, D.M., Winston, R.M. and Handyside, A.H. (1998) Assessment of the reliability of single blastomere analysis for preimplantation diagnosis of the delta F508 deletion causing cystic fibrosis in clinical practice. Prenat. Diagn., 18, 1402–1412.[CrossRef][Web of Science][Medline]
Rechitsky, S., Freidine, M., Verlinsky, Y. and Strom, C.M. (1996) Allele dropout in sequential PCR and FISH analysis of single cells (cell recycling). J. Assist. Reprod. Genet. 13, 115–124.[CrossRef][Web of Science][Medline]
Rechitsky, S., Strom, C., Verlinsky, O., Amet, T., Ivakhnenko, V., Kukharenko, V., Kuliev, A. and Verlinsky, Y. (1998) Allele dropout in polar bodies and blastomeres. J. Assist. Reprod. Genet., 15, 253–257.[CrossRef][Web of Science][Medline]
Sermon, K., Lissens, W., Joris, H., Van Steirteghem, A. and Liebaers, I. (1996) Adaptation of the primer extension preamplification (PEP) reaction for preimplantation diagnosis: single blastomere analysis using short PEP protocols. Mol. Hum. Reprod., 2, 209–212.
Sermon, K., De Vos, A., Van de Velde, H., Seneca, S., Lissens, W., Joris, H., Vandervorst, M., Van Steirteghem, A. and Liebaers, I. (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, 791–796.
Sermon, K., Lissens, W., Nagy, Z.P., Van Steirteghem, A. and Liebaers, I. (1995) Simultaneous amplification of the two most frequent mutations of infantile Tay–Sachs disease in single blastomeres. Hum. Reprod., 10, 2214–2217.
Sharma, V. and Litt, M. (1992) Tetranucleotide repeat polymorphism at the D21S11 locus. Hum. Mol. Genet., 1, 67.
Sherlock, J., Cirigliano, V., Petrou, M., Tutschek, B. and Adinolfi, M. (1998) Assessment of diagnostic quantitative fluorescent multiplex polymerase chain reaction assays performed on single cells. Ann. Hum. Genet., 62, 9–23.[CrossRef][Web of Science][Medline]
Thornhill, A.R., McGrath, J.A., Eady, R.A., Braude, P.R. and Handyside, A.H. (2001) A comparison of different lysis buffers to assess allele dropout from single cells for preimplantation genetic diagnosis. Prenat. Diagn., 21, 490–497.[CrossRef][Web of Science][Medline]
Verlinsky, Y. (1996) Preimplantation genetic diagnosis. J. Assist. Reprod. Genet., 13, 87–89.[CrossRef][Web of Science][Medline]
Verlinsky, Y. and Kuliev, A. (1998) Progress in preimplantation genetics. J. Assist. Reprod. Genet., 15, 9–11.[CrossRef][Web of Science][Medline]
Vrettou, C., Palmer, G., Kanavakis, E., Tzetis, M., Antoniadi, T., Mastrominas, M. and Traeger-Synodinos, J. (1999) A widely applicable strategy for single cell genotyping of beta-thalassaemia mutations using DGGE analysis: application to preimplantation genetic diagnosis. Prenat. Diagn., 19, 1209–1216.[CrossRef][Web of Science][Medline]
Walsh, P.S., Erlich, H.A. and Higuchi, R. (1992) Preferential PCR amplification of alleles: mechanisms and solutions. PCR Methods Appl., 1, 241–250.[Medline]
Weber, J.L. and May, P.E. (1989) Abundant class of human DNA polymorphisms which can be typed using the polymerase chain reaction. Am. J. Hum. Genet. 44, 388–396.[Web of Science][Medline]
Wells, D. and Delhanty, J.D. (2001) Preimplantation genetic diagnosis: applications for molecular medicine. Trends Mol. Med., 7, 23–30.[CrossRef][Web of Science][Medline]
Wells, D. and Sherlock, J.K. (1998) Strategies for preimplantation genetic diagnosis of single gene disorders by DNA amplification. Prenat. Diagn., 18, 1389–1401.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Ling, G. Zhuang, B. Tazon-Vega, C. Zhang, B. Cao, Z. Rosenwaks, and K. Xu Evaluation of genome coverage and fidelity of multiple displacement amplification from single cells by SNP array Mol. Hum. Reprod., November 1, 2009; 15(11): 739 - 747. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Lin, M. Takata, H. Murata, Y. Goto, K. Kido, S. Ferrone, and T. Saida Polyclonality of BRAF Mutations in Acquired Melanocytic Nevi J Natl Cancer Inst, October 21, 2009; 101(20): 1423 - 1427. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Curtis, B. S. Stewart, and S. A. Karl Sexing Pinnipeds with ZFX and ZFY Loci J. Hered., June 4, 2007; (2007) esm023v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Gardner, E. Oitmaa, A. Messner, L. Hoefsloot, A. Metspalu, and I. Schrijver Simultaneous Multigene Mutation Detection in Patients With Sensorineural Hearing Loss Through a Novel Diagnostic Microarray: A New Approach for Newborn Screening Follow-up Pediatrics, September 1, 2006; 118(3): 985 - 994. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Altarescu, B. Brooks, Y. Kaplan, T. Eldar-Geva, E.J. Margalioth, E. Levy-Lahad, and P. Renbaum Single-sperm analysis for haplotype construction of de-novo paternal mutations: application to PGD for neurofibromatosis type 1 Hum. Reprod., August 1, 2006; 21(8): 2047 - 2051. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Rund and E. Rachmilewitz {beta}-Thalassemia N. Engl. J. Med., September 15, 2005; 353(11): 1135 - 1146. [Full Text] [PDF]
|
|
|
|
|
|
A.R. Thornhill, C.E. deDie-Smulders, J.P. Geraedts, J.C. Harper, G.L. Harton, S.A. Lavery, C. Moutou, M.D. Robinson, A.G. Schmutzler, P.N. Scriven, et al. ESHRE PGD Consortium 'Best practice guidelines for clinical preimplantation genetic diagnosis (PGD) and preimplantation genetic screening (PGS)' Hum. Reprod., January 1, 2005; 20(1): 35 - 48. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. De Rycke, I. Georgiou, K. Sermon, W. Lissens, P. Henderix, H. Joris, P. Platteau, A. Van Steirteghem, and I. Liebaers PGD for autosomal dominant polycystic kidney disease type 1 Mol. Hum. Reprod., January 1, 2005; 11(1): 65 - 71. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. E. Pierce, J. E. Rice, J.A. Sanchez, and L. J. Wangh Detection of cystic fibrosis alleles from single cells using molecular beacons and a novel method of asymmetric real-time PCR Mol. Hum. Reprod., December 1, 2003; 9(12): 815 - 820. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||