Advertisement

PGD and aneuploidy screening for 24 chromosomes by genome-wide SNP analysis: seeing the wood and the trees

  • Alan H. Handyside
    Affiliations
    London Bridge Fertility, Gynaecology and Genetics Centre, 1 St. Thomas Street, London SE1 9RY, UK
    Institute of Comparative and Integrative Biology, Faculty of Biological Sciences, University of Leeds, Leeds LS2 9JT, UK
    Search for articles by this author
Published:September 28, 2011DOI:https://doi.org/10.1016/j.rbmo.2011.09.012

      Abstract

      • Bisignano A.
      • Wells D.
      • Harton G.
      • Munne S.
      PGD and aneuploidy screening for 24 chromosomes: advantages and disadvantages of competing platforms.
      argue that, for preimplantation genetic diagnosis (PGD) of aneuploidy for all 24 chromosomes, microarray-based comparative genomic hybridization (array CGH) is superior to the use of single-nucleotide polymorphism (SNP) genotyping arrays. Published studies indicate that both technologies accurately detect aneuploidy of whole chromosomes or chromosome segments. However, given the extra theoretical resolution and parent-of-origin information provided by SNP-based approaches, these may be particularly suited to certain applications such as PGD of single-gene defects or translocation chromosome imbalance combined with comprehensive detection of aneuploidy. A consensus on how to validate aneuploidy testing and all other clinically relevant information resulting from genome-wide analysis is needed urgently.

      Keywords

      Introduction

      In a commentary in this issue of Reproductive BioMedicine Online,
      • Bisignano A.
      • Wells D.
      • Harton G.
      • Munne S.
      PGD and aneuploidy screening for 24 chromosomes: advantages and disadvantages of competing platforms.
      argue that for preimplantation genetic diagnosis (PGD) of aneuploidy for all 24 chromosomes, microarray-based comparative genomic hybridization (array CGH) is superior to the use of single-nucleotide polymorphism (SNP) genotyping arrays. Specifically, they criticize as unreliable the approach of one particular laboratory, which uses a method referred to as ‘parental support’. This approach combines SNP data from both parents with the use of several complex bioinformatic modelling algorithms to detect whole-chromosome and segmental copy-number changes and to distinguish meiotic from mitotic errors. The problem is that the authors do not have access to these algorithms, which are proprietary, and their arguments include guesswork. The commentary, therefore, is in danger of being perceived simply as part of an intense ongoing debate between commercial laboratories, since the authors are well known to be associated with one of the largest service providers for array CGH. Both technologies have their strengths and weaknesses and there is a risk that unconstructive debate will simply undermine the confidence of clinicians and patients in aneuploidy testing by either methodology.

      Validation of single-cell genetic tests

      There are two fundamental problems with validation of single-cell tests for PGD of aneuploidy. The first is simply that in the process of testing, which invariably involves an amplification step, the DNA within the cell is rapidly degraded and is then not available for reanalysis. The second problem is related to testing human embryos at preimplantation stages of development. Remarkably, although the incidence and origins of aneuploidy have been intensively studied over the last 25 years, it is still not fully understood how errors arise in meiosis and in the first mitotic divisions during cleavage and how these then evolve as the embryo develops.
      Most attempts to validate the genome-wide tests for aneuploidy of all 24 chromosomes published to date focus initially on the use of cell lines with known karyotype abnormalities. (This is despite the fact that the some cell-line collections prohibit the use of these lines for that purpose.) The difficulty with this approach is that the karyotypes of these, mainly transformed, cell lines are not completely stable in culture. Thus, the results tend to demonstrate close to 100% detection of the known abnormality, but are likely to also detect other unrelated aneusomies. Indeed, karyotyping is generally based on identifying consistent abnormalities in a limited number of well-spread metaphases. If the number of metaphases analysed is expanded, however, mosaicism for other aneusomies in a minority of cells is frequently identified, varying in incidence between different lines.
      More meaningful for clinicians and patients is the comparison between the results of testing following biopsy with those from the embryo itself. However, again this comparison has limitations related to how representative of the whole embryo are the aneusomies detected in polar bodies, biopsied single cleavage-stage blastomeres or biopsied blastocyst trophectoderm cells – and how relevant this information is when the embryo has reached later stages of development, implanted and continued to develop. For example, chromosome loss through anaphase lag may result in trisomy rescue, whilst malsegregation during cleavage frequently results in chromosomal mosaicism among blastomeres – not necessarily incompatible with further development.
      One approach to avoid some of these problems is to amplify the whole genome of single cells and use the relatively large quantity of amplified DNA to compare results with different platforms using the same DNA sample. Even this, however, has limitations because the methods used to amplify the whole genome may not be optimal for the different methods being compared, thereby compromising the results.
      There is an urgent need therefore to agree on the best way to validate different technologies and to set up independent external quality-assurance schemes to ensure that all laboratories, large and small, provide a consistently high level of accuracy. One recent initiative is the establishment of a subcommittee of the European Society for Human Reproduction and Embryology (ESHRE) PGD Consortium on the use of microarrays for PGD to provide guidelines on their use and to work towards setting up appropriate external quality-assurance schemes. However, journals also have a role to play in insisting on clear descriptions of the methodologies used (if not on complete disclosure of protected algorithms) and open access to all of the raw data to allow independent assessment, before publishing claims that particular methods have been validated.

      Array comparative genomic hybridization

      Array CGH is now used extensively worldwide for PGD of both whole chromosome aneuploidy and segmental chromosome imbalance in couples carrying balanced reciprocal or Robertsonian translocations (
      • Fiorentino F.
      • Spizzichino L.
      • Bono S.
      • Biricik A.
      • Kokkali G.
      • Rienzi L.
      • Ubaldi F.M.
      • Iammarrone E.
      • Gordon A.
      • Pantos K.
      PGD for reciprocal and Robertsonian translocations using array comparative genomic hybridization.
      ,
      • Fishel S.
      • Gordon A.
      • Lynch C.
      • Dowell K.
      • Ndukwe G.
      • Kelada E.
      • Thornton S.
      • Jenner L.
      • Cater E.
      • Brown A.
      • Garcia-Bernardo J.
      Live birth after polar body array comparative genomic hybridization prediction of embryo ploidy – the future of IVF?.
      ). It has several advantages including a relatively straightforward and rapid protocol for PCR library-based whole-genome amplification, DNA labelling, hybridization and scanning, which can be completed within as little as 9 h if necessary. This approach enables the testing of polar bodies before syngamy, for example, in countries like Germany where testing of embryos is not allowed. Also, equipment costs are relatively low, as high-resolution scanners are not required, so that larger infertility clinics can set up their own facilities in house.
      Moreover, since a large randomized clinical trial (
      • Mastenbroek S.
      • Twisk M.
      • van Echten-Arends J.
      • Sikkema-Raddatz B.
      • Korevaar J.C.
      • Verhoeve H.R.
      • Vogel N.E.
      • Arts E.G.
      • de Vries J.W.
      • Bossuyt P.M.
      • Buys C.H.
      • Heineman M.J.
      • Repping S.
      • van der Veen F.
      In vitro fertilization with preimplantation genetic screening.
      ) and several subsequent trials have demonstrated that the use of fluorescence in-situ hybridization (FISH) for aneuploidy testing for limited numbers of chromosomes on cleavage-stage embryos from women of advanced maternal age results in either the same or reduced live-birth rates, most clinics no longer use this methodology. Instead, many clinics, particularly in Europe, have been turning to array CGH for 24 chromosome analysis using polar body biopsy to detect maternal meiotic errors specifically, which is less invasive and avoids the problem of chromosomal mosaicism at cleavage stages.
      The ESHRE Preimplantation Genetic Screening (PGS) Task Force, established in 2007, also decided to adopt this approach and, as a preliminary step towards a large multicentre randomized clinical trial, set up a pilot study based in two clinics with experience in polar body biopsy (

      Geraedts, J., Montag, M., Magli, M.C., Repping, S., Handyside, A., Staessen, C., Harper, J., Schmutzler, A., Collins, J., Goossens, V., van der Ven, H., Vesela, K., Gianaroli, L., 2011. Polar body array CGH for prediction of the status of the corresponding oocyte: Part I. Clinical results. Hum. Reprod. (Epub ahead of print). doi:10.1093/humrep/der294.

      ). About 20 couples in both clinics requesting aneuploidy testing, mainly for advanced maternal age, were consented for the study in which both the first and second polar bodies were biopsied at the same time following intracytoplasmic sperm injection and were tested separately for aneuploidy by array CGH. If either or both the polar bodies were aneuploid, the corresponding zygote was then analysed by the other centre, blinded to the polar body results, and the results collated independently. Although this pilot study highlighted potential problems, including the possibility of contamination causing false-negative results (

      Magli, M.C., Montag, M., Koster, M., Muzi, L., Geraedts, J., Collins, J., Goossens, V., Handyside, A.H., Harper, J., Repping, S., Schmutzler, A., Vesela, K., Gianaroli, L., 2011. Polar body array CGH for prediction of the status of the corresponding oocyte: Part II. Technical aspects. Hum. Reprod. (Epub ahead of print). doi:10.1093/humrep/der295.

      ), there was a high level of concordance (94%) between the euploid/aneuploid status of the polar bodies and the corresponding zygote. A randomized trial is now planned to start later this year in multiple centres across Europe.
      To validate the use of array CGH for the accurate detection of aneuploidy in single blastomeres,
      • Gutierrez-Mateo C.
      • Colls P.
      • Sanchez-Garcia J.
      • Escudero T.
      • Prates R.
      • Ketterson K.
      • Wells D.
      • Munne S.
      Validation of microarray comparative genomic hybridization for comprehensive chromosome analysis of embryos.
      biopsied a large series of cleavage-stage embryos and then used FISH with appropriate probes to confirm any abnormalities in the other cells of the embryo. Overall, the incidence of aneuploidy was 62% and, following improvements to the protocol, the error rate was 1.9% and the failure or ‘no result’ rate, 2.9%. Since human embryos at this stage are also well known to be mosaic as a result of post-zygotic malsegregation of chromosomes (
      • Delhanty J.D.
      • Harper J.C.
      • Ao A.
      • Handyside A.H.
      • Winston R.M.
      Multicolour FISH detects frequent chromosomal mosaicism and chaotic division in normal preimplantation embryos from fertile patients.
      ,
      • Munne S.
      • Weier H.U.
      • Grifo J.
      • Cohen J.
      Chromosome mosaicism in human embryos.
      ), this analysis depends on certain assumptions about the FISH results among the remaining cells. Nevertheless, these results demonstrate excellent concordance in a study design that mimics clinical embryo biopsy and makes results available within 24 h.

      Single-nucleotide polymorphism arrays

      In contrast, the use of SNP genotyping arrays, which are also used extensively for molecular karyotyping in pre- and post-natal investigations, has so far been limited to a handful of clinics and laboratories, mainly in the USA. The advantage of using these arrays is primarily that it is possible to analyse hundreds of thousands of loci across the genome using a single array, with an average spacing as close as 5 kb, enabling high-resolution analysis. At the same time, the genotype information allows the parental origin of any abnormalities to be identified by reference to the genotypes of the two parents. The disadvantages include the high cost of the equipment, including a high-resolution scanner, and the length and complexity of the protocol, although several groups report that it is possible to reduce the time required to less than 24 h.
      • Bisignano A.
      • Wells D.
      • Harton G.
      • Munne S.
      PGD and aneuploidy screening for 24 chromosomes: advantages and disadvantages of competing platforms.
      focus their criticism on one particular approach, but in fact there are three distinct ways of using SNP data for PGD. The first approach is simply to optimize whole-genome amplification from single cells to minimize amplification bias and allele drop out, and to use the standard algorithms available for molecular karyotyping to detect whole and segmental chromosome imbalance by examining, for example, B-allele frequency at each SNP locus or loss of heterozygosity. This approach is based on measuring the intensity of alleles at thousands of SNP loci across each chromosome. The SNPs analysed on these arrays are biallelic (referred to generically as A or B) and have been carefully selected to have a high heterozygosity ratio. Thus, trisomy or segmental gain can be visualized by two (or more) bands representing the B-allele frequency at heterozygous loci (for example, AB versus ABB) and conversely monosomy or segmental loss, by loss of heterozygosity. This approach has been pioneered by Kearns (
      • Brezina P.R.
      • Benner A.
      • Rechitsky S.
      • Kuliev A.
      • Pomerantseva E.
      • Pauling D.
      • Kearns W.G.
      Single-gene testing combined with single nucleotide polymorphism microarray preimplantation genetic diagnosis for aneuploidy: a novel approach in optimizing pregnancy outcome.
      ) and Treff and their colleagues, and extensively validated for both whole chromosome and segmental imbalance by the latter group (
      • Treff N.R.
      • Su J.
      • Tao X.
      • Levy B.
      • Scott Jr., R.T.
      Accurate single cell 24 chromosome aneuploidy screening using whole genome amplification and single nucleotide polymorphism microarrays.
      ,
      • Treff N.R.
      • Northrop L.E.
      • Kasabwala K.
      • Su J.
      • Levy B.
      • Scott Jr., R.T.
      Single nucleotide polymorphism microarray-based concurrent screening of 24-chromosome aneuploidy and unbalanced translocations in preimplantation human embryos.
      ,
      • Treff N.R.
      • Tao X.
      • Schillings W.J.
      • Bergh P.A.
      • Scott Jr., R.T.
      • Levy B.
      Use of single nucleotide polymorphism microarrays to distinguish between balanced and normal chromosomes in embryos from a translocation carrier.
      ), although they have recently moved on to a real-time PCR-based SNP approach which can be completed within 4 h.
      • Treff N.R.
      • Su J.
      • Tao X.
      • Levy B.
      • Scott Jr., R.T.
      Accurate single cell 24 chromosome aneuploidy screening using whole genome amplification and single nucleotide polymorphism microarrays.
      use a simple algorithm in which each chromosome is assigned a copy number based on the copy number of the majority of SNP on that chromosome. Results with a series of single cells from a number of different cell lines with known karyotype abnormalities demonstrated no false-positive results among 72 cells analysed and a false-negative rate of 4.2% which, after eliminating two results that did not reach minimal levels of consistency, was reduced to 1.4% per cell analysed. A large series of single blastomeres from disaggregated cleavage-stage embryos was then analysed and copy-number calling was demonstrated to be highly consistent and the pattern of aneuploidy revealed as being similar to previous studies. However, no attempt was made to model the real-life clinical situation.
      The second approach is the one highlighted by
      • Bisignano A.
      • Wells D.
      • Harton G.
      • Munne S.
      PGD and aneuploidy screening for 24 chromosomes: advantages and disadvantages of competing platforms.
      , which is exclusively available from a commercial laboratory in the USA and combines SNP data from both parents with the use of several advanced bioinformatic modelling algorithms, to detect whole chromosome and segmental copy-number changes and to distinguish meiotic versus mitotic errors. It is my general understanding that this method leverages information on haplotypes and linkage disequilibrium from, for example, the HapMap project and other sources to model the four parental haplotypes and to identify which are present across each chromosome. In this way, trisomies and monosomies and their parental origins are identified, respectively, by the presence of two haplotypes from one parent and one from the other parent or of only a single parental haplotype.
      • Bisignano A.
      • Wells D.
      • Harton G.
      • Munne S.
      PGD and aneuploidy screening for 24 chromosomes: advantages and disadvantages of competing platforms.
      are right to point out the poor quality of SNP genotype information following whole-genome amplification and the difficulty of predicting haplotypes. For example, irrespective of the method used for whole-genome amplification, at the present time, allele drop-out rates for heterozygous SNPs are typically 40–50%. However, without access to the algorithm it is impossible to assess its capabilities independently. In an initial validation study,
      • Johnson D.S.
      • Gemelos G.
      • Baner J.
      • Ryan A.
      • Cinnioglu C.
      • Banjevic M.
      • Ross R.
      • Alper M.
      • Barrett B.
      • Frederick J.
      • Potter D.
      • Behr B.
      • Rabinowitz M.
      Preclinical validation of a microarray method for full molecular karyotyping of blastomeres in a 24-h protocol.
      report results with a large series of single cells from a trisomy-21 cell line which demonstrated a false-negative rate of 2.1% compared with 1% for karyotyping and no false positives. Unfortunately it is not clear whether the authors used their parental-support algorithms for these cells, which presumably would rely on the availability of cells or DNA from the parents. Also, a series of single blastomeres disaggregated from cleavage-stage embryos were analysed. Again the incidence and pattern of abnormalities is consistent with what is known about embryos at cleavage stages. However, no attempt was made to model the clinical situation or examine consistency within each embryo. Instead, confidence levels for copy-number calling for each chromosome were analysed using statistical methods to examine the consistency of the SNP data as a ‘proxy’ for accuracy. Nevertheless, this analysis predicted very high levels of accuracy using the parental support algorithms.
      • Johnson D.S.
      • Gemelos G.
      • Baner J.
      • Ryan A.
      • Cinnioglu C.
      • Banjevic M.
      • Ross R.
      • Alper M.
      • Barrett B.
      • Frederick J.
      • Potter D.
      • Behr B.
      • Rabinowitz M.
      Preclinical validation of a microarray method for full molecular karyotyping of blastomeres in a 24-h protocol.
      also claim to be able to identify abnormal ploidy, including haploidy and triploidy (but not tetraploidy) and to distinguish meiotic and mitotic errors resulting in aneuploidy or uniparental disomy, none of which is possible with array CGH. Here, this commentary agrees with
      • Bisignano A.
      • Wells D.
      • Harton G.
      • Munne S.
      PGD and aneuploidy screening for 24 chromosomes: advantages and disadvantages of competing platforms.
      that the data presented are less clear cut and highlight the need for additional validation of all clinically relevant information gained by any genome-wide methodology. Although
      • Fiorentino F.
      • Spizzichino L.
      • Bono S.
      • Biricik A.
      • Kokkali G.
      • Rienzi L.
      • Ubaldi F.M.
      • Iammarrone E.
      • Gordon A.
      • Pantos K.
      PGD for reciprocal and Robertsonian translocations using array comparative genomic hybridization.
      claim to be able to detect translocation imbalance to a resolution of at least 2 Mb by array CGH, for example, there are no published studies for either technology systematically examining the resolution of detection of partial gain or loss.
      The third approach is the one published recently, which again uses parental SNP genotypes but in this case uses straightforward Mendelian analysis to generate a karyomap of each chromosome or chromosome segment inherited by the embryo (
      • Handyside A.H.
      • Harton G.L.
      • Mariani B.
      • Thornhill A.R.
      • Affara N.
      • Shaw M.A.
      • Griffin D.K.
      Karyomapping: a universal method for genome wide analysis of genetic disease based on mapping crossovers between parental haplotypes.
      ). To do this, it is necessary, in addition, to establish the phase of each informative heterozygous SNP by genotyping either a close relative or child or, alternatively, simply to use the genotype of one embryo to analyse the others. To establish the phase for each heterozygous SNP, it is necessary to determine which allele (A or B) is present on each of the four parental chromosomes. How is this possible with allele drop-out rates approaching 50% following whole-genome amplification from single cells? The answer is only to use heterozygous calls in the test sample and to rely on the accurate calling of genotypes predicted to be homozygous. Although this approach means that only about 10% of the original SNP analysed are useful, it does identify a set of consecutive markers with a mean spacing of about 100 kb across each chromosome. The markers in this ‘chain’ are still so closely spaced that the chance of a double recombination between any consecutive pair of SNP markers is exceedingly unlikely, a phenomenon known as interference. Thus the power of the microarray to genotype hundreds of thousands of SNP loci is harnessed to overwhelm the errors caused by whole-genome amplification of single or small numbers of cells.
      This method was originally envisaged as a universal linkage-based method for PGD of single-gene defects that avoids the requirement to develop patient- or disease-specific tests. However, single-gene defect analysis combined with limited aneuploidy screening using other molecular markers had indicated that aneuploidy is equally a problem in these patients who are often of advanced maternal age. Another advantage, therefore, is that karyomapping simultaneously detects meiotic trisomies, monosomies and deletions for all 24 chromosomes, which is important for advising the patient about the viability of the embryos. Karyomapping is an entirely genotype-based test, i.e. there is no use for quantitative information, and therefore is blind to any whole chromosome or segmental duplications. However, because the positions of crossovers are accurately identified, meiotic errors resulting in trisomy, in which two chromosomes from one parent are inherited by an embryo, can be detected by identifying both haplotypes from that parent in specific regions of the chromosome. So for example, if an embryo inherits a non-recombinant grandpaternal chromosome from the mother together with a recombinant grandmaternal chromosome, both the mothers’ haplotypes will be present in the region of the centromere while crossing over more distally on the chromosome arms will create successive regions in which only one haplotype is present. If these regions encompass the centromere, as in this example, they can be identified as typical of a meiosis-I error, whereas if these regions are more distal on the arms of the chromosome then they can be identified as typical of meiosis-II errors.
      Arguably this feature of karyomapping to identify meiotic and not mitotic trisomies is an advantage, particularly for blastomere analysis at cleavage stages, since it definitively identifies the fact that two meiotic chromosomes were inherited by the embryo in either the oocyte or the fertilizing spermatozoan. Thus, the patient can be counselled about the incidence of these abnormalities and prospects for success with PGD, whilst at the same time avoiding the possibility of not considering for transfer of an embryo that has a chromosome duplication which may only be present in some cells of a chromosomally mosaic embryo.
      • Bisignano A.
      • Wells D.
      • Harton G.
      • Munne S.
      PGD and aneuploidy screening for 24 chromosomes: advantages and disadvantages of competing platforms.
      argue that some meiotic trisomies may be missed, such as a meiosis-II error of an achiasmate chromosome. This error is very rarely encountered in clinical pregnancies or spontaneous abortions and its incidence in preimplantation-stage embryos remains to be established. What is clear is that the majority of conventional meiotic trisomies are readily detected by karyomapping.
      Validation of karyomapping in a series of clinical cases is ongoing. This study centre has taken the approach of using whole-genome amplification to allow the comparison of results of different approaches on the same DNA sample. For example, in a patient of advanced maternal age, who had cleavage-stage biopsy and FISH analysis for chromosomes 13, 16, 18, 21, 22, X and Y, those embryos identified as aneuploid were lysed, the whole genome amplified and then array CGH and SNP genotyping performed on the same DNA products (Figure 1). In an embryo identified as being monosomy 18, 21 and 22 and trisomy X by FISH (Figure 1a, b), array CGH confirmed all of those aneusomies in the embryo as a whole as well as seven other monosomies (2, 6, 8, 9, 10, 11, 14) and one additional trisomy 17 (Figure 1c). Thus, at the single-cell level, the interpretation would probably have been that the changes were mostly if not all caused by post-zygotic malsegregation resulting in a chaotic pattern in one or more blastomeres, whilst at the whole-embryo level, as revealed by array CGH, it is clear that there has been extensive chromosome loss, presumably at an early stage, and gains of two chromosomes, most probably through errors in meiosis. Conventional, blinded intensity-based analysis of the SNP data confirmed all of the monosomies by detection of loss of heterozygosity across those chromosomes but failed to detect the two trisomies (data not shown). Finally, karyomap analysis (Figure 1d), which identifies the parental origin of copy-number abnormalities, clearly demonstrates that all of the missing chromosomes are of maternal origin, strongly suggesting loss in meiosis or before the first mitotic division following fertilization and confirms the meiotic origin of the trisomy 17 as maternal with a pattern of recombination consistent with an error in the first meiotic division. (Only the autosomes were analysed in this case.) Furthermore, whereas there is a typical pattern of non-recombinant and recombinant paternal chromosomes as would be expected, there are no crossovers in the 11 remaining maternal chromosomes, strongly suggesting a complete failure to form chiasmata in the oocyte. This example is clearly extreme but does demonstrate the importance of analysing copy number for all chromosomes and that the additional information about parental origin of any abnormalities is a powerful tool for understanding the nature of the abnormalities and the clinical implications.
      Figure thumbnail gr1
      Figure 1Comparison of aneuploidy detection by fluorescence in-situ hybridization (FISH) on a single blastomere with reanalysis of the biopsied embryo by array comparative genomic hybridization (CGH) and karyomapping. (a) First hybridization to the interphase nucleus of a single blastomere biopsied from a cleavage-stage embryo on day 3 with specific probes for chromosomes 13 (red), 16 (aqua), 18 (blue), 21 (green) and 22 (gold) (note that the extra aqua signal lower right was considered to be non-specific as it was present with several filter sets). (b) Second hybridization with a probe specific for the X chromosome (aqua) and a different locus-specific probe for chromosome 21 (gold). The FISH data suggest monosomy 18, 21 and 22 and trisomy X. (c) Ratio plot following array CGH of biopsied embryo. These data confirm the aneusomies detected by FISH in the single biopsied blastomere and overall identify additional monosomies for chromosomes 2, 6, 8, 9, 10, 11 and 14 and trisomy 17. (d) Karyomap using SNP-based analysis of the autosomes 1–22 for the biopsied embryo from the same amplified DNA as in (c). The karyomaps for each autosomal pair have been displayed as they would be in a karyotype. For each pair, the paternal chromosome is on the left and the maternal chromosome on the right. The positions of the centromeres are indicated by grey bars. The two paternal chromosome haplotypes are coloured in blue and red and the two maternal haplotypes in yellow and green. Black indicates those autosomes for which neither parental haplotype was detected. Switching within each paternal autosome between parental haplotypes identifies the position of crossovers. Note the absence of maternal switching. This karyomap demonstrates that all of the aneuploidies are maternal in origin, i.e. black bars denoting monosomies for 2, 6, 8, 9, 10, 11, 14, 18, 21 and 22 and purple for trisomy 17 denoting the presence of both maternal haplotypes. (For interpretation of the references to colours in this figure legend, the reader is referred to the web version of this paper.)

      Conclusions

      Array CGH is a relatively low-cost option for the rapid detection of whole-chromosome or chromosome-segment gain and loss for all 24 chromosomes, and PGD testing by either polar body or blastomere biopsy has demonstrated high rates of accuracy in the corresponding embryos. SNP arrays and data analysis are more labour intensive and require optimized protocols and/or specialized algorithms for identifying chromosome copy number following whole-genome amplification from single cells. Several published studies indicate that SNP arrays are also highly accurate for aneuploidy detection. Given their extra resolution and parent-of-origin information, however, SNP-based approaches may be particularly suited to certain applications such as PGD of single-gene defects or translocation chromosome imbalance combined with comprehensive detection of aneuploidy. Also, the ability to distinguish normal from balanced embryos (by identifying the normal and derivative chromosomes at the translocation breakpoints) is important for some couples since they would prefer if possible to transfer embryos with normal chromosomes so that their children do not have the same problems as themselves.
      • Bisignano A.
      • Wells D.
      • Harton G.
      • Munne S.
      PGD and aneuploidy screening for 24 chromosomes: advantages and disadvantages of competing platforms.
      are dismissive of the clinical utility of the additional information provided by SNP analysis. They argue, for example, that reporting uniparental disomy, except for a small set of chromosomes with known imprinted loci-causing disease, is unnecessary. In general, this dismissal is premature and short sighted. Seeing the wood and the trees with SNP array analysis, or a combination of the two complementary technologies, may potentially be very important for understanding why and how abnormalities occur, which in turn will enable patients to be counselled more accurately about recurrence risk and the prospects for a healthy live birth.

      Acknowledgements

      The author would like to thank his colleagues Jon Taylor, Dr Mira Grigorova and Dr Alan Thornhill, London Bridge Fertility, Gynaecology and Genetics Centre, Anthony Brown and Tony Gordon, Bluegnome Ltd, Dr Alem Gabriel and Prof Darren Griffin, University of Kent and Dr Emily Clemente and Prof Nabeel Affara, University of Cambridge, for permission to use unpublished data.

      References

        • Bisignano A.
        • Wells D.
        • Harton G.
        • Munne S.
        PGD and aneuploidy screening for 24 chromosomes: advantages and disadvantages of competing platforms.
        Reprod. Biomed. Online. 2011; 23: 677-685
        • Brezina P.R.
        • Benner A.
        • Rechitsky S.
        • Kuliev A.
        • Pomerantseva E.
        • Pauling D.
        • Kearns W.G.
        Single-gene testing combined with single nucleotide polymorphism microarray preimplantation genetic diagnosis for aneuploidy: a novel approach in optimizing pregnancy outcome.
        Fertil. Steril. 2011; 95: 1786.e5-1786.e8
        • Delhanty J.D.
        • Harper J.C.
        • Ao A.
        • Handyside A.H.
        • Winston R.M.
        Multicolour FISH detects frequent chromosomal mosaicism and chaotic division in normal preimplantation embryos from fertile patients.
        Hum. Genet. 1997; 99: 755-760
        • Fiorentino F.
        • Spizzichino L.
        • Bono S.
        • Biricik A.
        • Kokkali G.
        • Rienzi L.
        • Ubaldi F.M.
        • Iammarrone E.
        • Gordon A.
        • Pantos K.
        PGD for reciprocal and Robertsonian translocations using array comparative genomic hybridization.
        Hum. Reprod. 2011; 26: 1925-1935
        • Fishel S.
        • Gordon A.
        • Lynch C.
        • Dowell K.
        • Ndukwe G.
        • Kelada E.
        • Thornton S.
        • Jenner L.
        • Cater E.
        • Brown A.
        • Garcia-Bernardo J.
        Live birth after polar body array comparative genomic hybridization prediction of embryo ploidy – the future of IVF?.
        Fertil. Steril. 2010; 93: 1006.e7-1006.e10
      1. Geraedts, J., Montag, M., Magli, M.C., Repping, S., Handyside, A., Staessen, C., Harper, J., Schmutzler, A., Collins, J., Goossens, V., van der Ven, H., Vesela, K., Gianaroli, L., 2011. Polar body array CGH for prediction of the status of the corresponding oocyte: Part I. Clinical results. Hum. Reprod. (Epub ahead of print). doi:10.1093/humrep/der294.

        • Gutierrez-Mateo C.
        • Colls P.
        • Sanchez-Garcia J.
        • Escudero T.
        • Prates R.
        • Ketterson K.
        • Wells D.
        • Munne S.
        Validation of microarray comparative genomic hybridization for comprehensive chromosome analysis of embryos.
        Fertil. Steril. 2011; 95: 953-958
        • Handyside A.H.
        • Harton G.L.
        • Mariani B.
        • Thornhill A.R.
        • Affara N.
        • Shaw M.A.
        • Griffin D.K.
        Karyomapping: a universal method for genome wide analysis of genetic disease based on mapping crossovers between parental haplotypes.
        J. Med. Genet. 2010; 47: 651-658
        • Johnson D.S.
        • Gemelos G.
        • Baner J.
        • Ryan A.
        • Cinnioglu C.
        • Banjevic M.
        • Ross R.
        • Alper M.
        • Barrett B.
        • Frederick J.
        • Potter D.
        • Behr B.
        • Rabinowitz M.
        Preclinical validation of a microarray method for full molecular karyotyping of blastomeres in a 24-h protocol.
        Hum. Reprod. 2010; 25: 1066-1075
      2. Magli, M.C., Montag, M., Koster, M., Muzi, L., Geraedts, J., Collins, J., Goossens, V., Handyside, A.H., Harper, J., Repping, S., Schmutzler, A., Vesela, K., Gianaroli, L., 2011. Polar body array CGH for prediction of the status of the corresponding oocyte: Part II. Technical aspects. Hum. Reprod. (Epub ahead of print). doi:10.1093/humrep/der295.

        • Mastenbroek S.
        • Twisk M.
        • van Echten-Arends J.
        • Sikkema-Raddatz B.
        • Korevaar J.C.
        • Verhoeve H.R.
        • Vogel N.E.
        • Arts E.G.
        • de Vries J.W.
        • Bossuyt P.M.
        • Buys C.H.
        • Heineman M.J.
        • Repping S.
        • van der Veen F.
        In vitro fertilization with preimplantation genetic screening.
        N. Engl. J. Med. 2007; 357: 9-17
        • Munne S.
        • Weier H.U.
        • Grifo J.
        • Cohen J.
        Chromosome mosaicism in human embryos.
        Biol. Reprod. 1994; 51: 373-379
        • Treff N.R.
        • Su J.
        • Tao X.
        • Levy B.
        • Scott Jr., R.T.
        Accurate single cell 24 chromosome aneuploidy screening using whole genome amplification and single nucleotide polymorphism microarrays.
        Fertil. Steril. 2010; 94: 2017-2021
        • Treff N.R.
        • Northrop L.E.
        • Kasabwala K.
        • Su J.
        • Levy B.
        • Scott Jr., R.T.
        Single nucleotide polymorphism microarray-based concurrent screening of 24-chromosome aneuploidy and unbalanced translocations in preimplantation human embryos.
        Fertil. Steril. 2011; 95 (1606–12.e1–2)
        • Treff N.R.
        • Tao X.
        • Schillings W.J.
        • Bergh P.A.
        • Scott Jr., R.T.
        • Levy B.
        Use of single nucleotide polymorphism microarrays to distinguish between balanced and normal chromosomes in embryos from a translocation carrier.
        Fertil. Steril. 2011; 96: e58-e65