ORIGINAL ARTICLE: REPRODUCTIVE BIOLOGY

Simultaneous assessment ofaneuploidy, polymorphisms, andmitochondrial DNA content in humanpolar bodies and embryos with theuse of a novel microarray platform

Michalis Konstantinidis, Ph.D.,a Samer Alfarawati, Ph.D.,a Douglas Hurd, Ph.D.,b Marta Paolucci, Ph.D.,b

John Shovelton, B.Sc.,b Elpida Fragouli, Ph.D.,a,c and Dagan Wells, Ph.D.a,c

a Reprogenetics UK; b Oxford Gene Technology; and c Nuffield Department of Obstetrics and Gynaecology, University ofOxford, Oxford, United Kingdom

Objective: To develop a microarray platform that allows simultaneous assessment of aneuploidy and quantification of mitochondrialDNA (mtDNA) in human polar bodies and embryos.Design: Optimization and validation applied to cell lines and clinical samples (polar bodies, blastomeres, and trophectoderm biopsies).Setting: University research laboratory and a preimplantation genetic diagnosis (PGD) reference laboratory.Patient(s): Samples from 65 couples who underwent PGD for aneuploidy and/or a single-gene disorder.Intervention(s): None.Main Outcome Measure(s): 1) Comparison of aneuploidy screening results obtained with the use of the new microarray with thosederived from two well established cytogenetic techniques. 2) mtDNA quantification. 3) Analysis of single-nucleotide polymorphisms.Result(s): The fully optimized microarray was estimated to have an accuracy ofR97% for the detection of individual aneuploidies andto detect 99% of chromosomally abnormal embryos. The microarray was shown to accurately determine relative quantities of mtDNA.Information provided from polymorphic loci was sufficient to allow confirmation that an embryo was derived from specific parents.Conclusion(s): It is hoped that methods such as those reported here, which provide information on several aspects of oocyte/embryogenetics, could lead to improved strategies for identifying viable embryos, thereby increasing the likelihood of successful implan-tation. Additionally, the provision of genotyping information has the potential to reveal DNA contaminants and confirm parental

Use your smartphone

origin of embryos. (Fertil Steril� 2014;-:-–-. �2014 by American Society for Reproduc-tive Medicine.)Key Words: Preimplantation genetic diagnosis, array comparative genomic hybridization,mitochondrial DNA, aneuploidy screening, single-nucleotide polymorphisms

Discuss: You can discuss this article with its authors and with other ASRM members at http://fertstertforum.com/konstantinidism-microarray-aneuploidy-polymorphisms-mitochondrial-dna/

to scan this QR codeand connect to thediscussion forum forthis article now.*

* Download a free QR code scanner by searching for “QRscanner” in your smartphone’s app store or app marketplace.

I t has been suggested that embryosproduced with the use of in vitrofertilization (IVF) should undergo ge-

netic analysis, with only those predicted

Received February 27, 2014; revised July 13, 2014; acM.K. has nothing to disclose. S.A. has nothing to dis

nology, Oxford, United Kingdom. M.P. has noGene Technology, Oxford, United Kingdom. Egrant from Merck Serono.

M.K. and S.A. should be considered similar in authorReprint requests: DaganWells, Ph.D., University of O

Business Park North, OxfordOX4 2HW, United K

Fertility and Sterility® Vol. -, No. -, - 2014 0015-Copyright ©2014 American Society for Reproductivehttp://dx.doi.org/10.1016/j.fertnstert.2014.07.1233

VOL. - NO. - / - 2014

to be diploid considered for transfer tothe uterus (1). In theory, ensuring thatthe embryo transferred is chromosomallynormal should improve the chances of

cepted July 14, 2014.close. D.H. is an employee of Oxford Gene Tech-thing to disclose. J.S. is an employee of Oxford.F. has nothing to disclose. D.W. has received a

order.xford, Institute of Reproductive Sciences, Oxfordingdom (E-mail: dagan.wells@obs-gyn.ox.ac.uk).

0282/$36.00Medicine, Published by Elsevier Inc.

implantation, decrease the miscarriagerate, and reduce the risk of a child bornwith problems associated with aneu-ploidy such as Down syndrome (1–3).Recent randomized clinical trials withthe use of comprehensive methods ofchromosome analysis have yieldedencouraging data (4–7). These studiessuggest that high-accuracy techniques,capable of assessing the copy number ofevery chromosome, are associated withsignificantly improved implantationrates and reduced risk of miscarriage.

1

ORIGINAL ARTICLE: REPRODUCTIVE BIOLOGY

Despite the promising results reported with the use ofmodern chromosome screening technologies, it remains thecase that even the transfer of a morphologically perfect, chro-mosomally normal embryo can not guarantee that a viablepregnancy will be established. Studies published in the past12 months indicate that implantation failure rates for euploidblastocysts range from 25% to 50% (4–8). It is clear thataspects of preimplantation embryo biology that are poorlydefined or invisible with the use of current methods ofembryo evaluation must also play an important role indetermining the potential of an embryo to produce a child.

One aspect of embryo biology of potential relevance toviability is mitochondrial DNA (mtDNA) content. Mitochon-dria are responsible for the production of most of the energyneeded by the cell for biosynthetic, metabolic, and physiologicprocesses. A number of clinical and animal studies have high-lighted the essential role ofmitochondria in oogenesis and pre-implantation development (9, 10). Decreased amounts ofmtDNA and diminished bioenergetic capacity have beenlinked to fertilization failure, impaired oocyte quality, andabnormal organization and function of the meiotic spindle(11–14). Furthermore, bioenergetic capacity has beencorrelated with embryo developmental potential and IVFoutcome (15, 16).

In response to the need for improved methods of embryoselection, this study sought to create a new microarray plat-form capable of extending genetic evaluation beyond chromo-some analysis. This microarray includes probes for aneuploidydetection but also uses mitochondria-specific probes for therelative quantification of mtDNA. In addition, the option ofincluding probes for the analysis of single-nucleotide poly-morphisms (SNPs) was investigated, an approach that poten-tially allows DNA fingerprinting and may provide otherinformation of diagnostic relevance.

MATERIALS AND METHODSCell Culture

Single cells derived from a variety of aneuploid fibroblast celllines were used during the optimization and validation of themicroarray. The cell lines had the following karyotypes:47,XY,þ13; 47,XY,þ15; 47,XY,þ18; 47,XY,þ21; 45,X; and48,XXY,þ18.

Cell lines were cultured under standard conditions. Singlecells were isolated in 1 mL phosphate-buffered saline solution(PBS with 0.1% polyvinyl alcohol) and transferred to sterile0.2-mL microcentrifuge tubes.

Selection ofWhole-genomeAmplificationMethodand Development of a Customized Microarray

Development of the customized microarray involved thetesting of different whole-genome amplification (WGA)methods, selection of the appropriate oligonucleotide probesand array format, and identification of the best method forlabeling the amplified products. The different WGA methodsassessed were Picoplex (Rubicon Genomics), Genomeplex(Sigma-Aldrich), and multiple displacement amplification(MDA). Picoplex and Genomeplex were performed on isolated

2

single cells according to the manufacturers' instructions. ForMDA, the Repli-g Midi kit (Qiagen) was used following themanufacturer's protocol with the use of a modified alkalinelysis method. This MDA protocol involved incubation of thesamples in a thermocycler at 30�C for 2 hours followed byenzyme inactivation at 65�C for 5 minutes.

Ultimately, the optimized microarray contained 14,334unique probes for aneuploidy detection and 261 uniqueprobes for mtDNA analysis. In contrast to the most widelyused microarray for the purpose of preimplantation aneu-ploidy screening, which uses probes composed of bacterialartificial chromosomes, the new microarray used syntheticoligonucleotides probes of �60 nucleotides in length.

The final protocol involvedWGA of the samples and con-trol male DNA with the use of Picoplex. Control DNA wasprepared by amplifying 1 mL male genomic DNA (1 ng/mL).Picoplex-amplified DNA samples and control samples (8 mL)were labeled with the use of the Cytosure labeling kit, whichincorporates Cy-dCTP using exonuclease-free Klenowenzyme (Oxford Gene Technology [OGT]). After labeling,the DNA was purified with the use of either Cytosure purifica-tion columns (OGT) or Cytosure purification plates (OGT). Thesample and reference DNA were pooled together, dried, andthen resuspended in hybridization mix (Agilent). The com-bined sample was applied on the microarray slide and incu-bated overnight in a rotating incubator at 65�C and 0.054 g.Post-hybridization washes were carried out with the use ofAgilent Oligo microarray comparative genomic hybridization(aCGH) Wash Buffer 1 for 5 minutes at room temperature andAgilent Oligo aCGH Wash Buffer 2 for 1 minute at 37�C. Themicroarray slide was then scanned at 5-mm resolution on anAgilent DNA microarray scanner. Feature extraction wasundertaken with the use of Agilent's Feature Extraction Soft-ware (v. 10.7.3.1) and data analysis carried out with the use ofa customized version of Cytosure Interpret Software (OGT).The whole process from DNA amplification to results couldbe completed in �16 hours and at a cost similar to existingaCGH methods.

Validation of Aneuploidy Detection with the Useof Clinical Material

Ninety-seven Picoplex-amplified clinical samples, previouslyscreened for aneuploidies with the use of an alternative mi-croarray CGH platform (24sure; Bluegnome), were reanalyzedwith the use of the new microarray and the results comparedfor concordance. These samples were derived from actualclinical cases performed previously for aneuploidy screeningof polar bodies and biopsied embryonic cells. These samplesconsisted of 27 polar bodies (PBs), 50 blastomeres, and 20trophectoderm samples. Diagnosis of aneuploidy was madein a blinded fashion, without any knowledge of the originalscreening results. An independent scientist was responsiblefor assessing the two sets of data for concordance.

A subset of 20 embryos (19 cleavage stage and one blas-tocyst) was also analyzed with the use of a third, independentcytogenetic method, fluorescent in situ hybridization (FISH).These whole embryos were donated to research, and all oftheir cells were spread on microscope slides for FISH analysis.

VOL. - NO. - / - 2014

Fertility and Sterility®

The methods used for cell fixation and FISH were carried outas described previously (17, 18). Chromosomes 8, 13, 14, 15,16, 17, 18, 20, 21, 22, X, and Y were tested in three roundsof hybridization applied to all cells. An extra hybridizationwas performed to test specific embryos for any additionalchromosomes found to be abnormal via microarray analysisbut not included in the first three rounds of FISH.

Initial evaluation of the new microarray revealed prob-lems in detecting losses and gains of chromosome 19. Newprobes for chromosome 19 were selected from a large poolon a separate microarray platform. Probes were chosen onthe basis of consistently displaying excellent signal intensityand providing reliable copy number analysis for chromosome19 across multiple single-cell samples of known chromo-somal status. The selected probes were added to the microar-ray, this new version was tested against five new samples(single blastomeres) derived from embryos known to be aneu-ploid for chromosome 19, and correct results were obtained inall cases.

Evaluation of mtDNA Quantification

To assess whether the new microarray was able to accuratelymeasure the relative quantity of mtDNA in a sample, four celllines were used (A549, Red, B2342B, and 206F). Picoplex-amplified DNA and unamplified genomic DNA from each ofthe cell lines were applied to the microarray to determinewhether WGA introduced any distortions in the relativequantities of mtDNA. Additionally, the amount of mtDNAin each of these DNA samples was quantified with the useof real-time polymerase chain reaction (PCR, using standardtechniques), confirming whether the results obtained fromthe microarray were accurate.

Assessment of SNP Probes for Inclusion on theMicroarray

As a proof of principle, a number of SNP probes were evaluatedfor potential inclusion on the microarray. SNP probes wereselected from the commercially available International Stan-dard Cytogenomic Array uniparental disomy array v. 1.0(OGT) on the basis of a preliminary experiment during whichPicoplex-amplified products were applied to the microarray. Atotal of 1,282 different SNPs were selected from an initial poolof 6,186. The SNP probes were placed on the array in triplicate.

To test the new microarray, aliquots from surplus Pico-plex products, amplified from 14 single blastomeres fromtwo previous clinical cases (preimplantation genetic diagnosisfor single-gene disorders) were used. The embryo samples,along with genomic DNA extracted from the parents, wereanalyzed with the use of the same microarray protocol asoutlined above.

Statistical Analysis

Fisher exact test (Graphpad) was used to make comparisonsregarding aneuploidy detection between the different sampletypes. The same test was also used for analysis of SNP results.The Mann-Whitney U test and independent-samples t test

VOL. - NO. - / - 2014

(SPSS v. 19.0.0) were used for analysis of data obtainedfrom relative quantitation of mtDNA in PB/embryo samples.

Ethics

Necessary ethical approvals were obtained for usage ofresearch samples. Chromosomal and mitochondrial analysesinvolved informed patient consent and were carried out afterappropriate licencing and with approval from the local EthicsCommittee [National Research Ethics Service CommitteeSouth Central].

RESULTSDevelopment of Customized Array and Selectionof Best WGA Method

Forty-eight WGA products were produced with the use ofthree different methods (MDA, Genomeplex, Picoplex)applied to isolated single cells. The most successful WGAmethod was determined to be Picoplex, and consequentlythis was the method used during all subsequent work. Anal-ysis of Picoplex products provided a correct diagnosis for16/16 cells, with detection of 16/16 individual aneuploidies.In contrast, only a minority of aneuploidies was detected inMDA samples and none in the Genomeplex products. In addi-tion, artefactual losses and gains of some chromosomes wereobserved for MDA and Genomeplex samples.

Validation of the New Oligonucleotide Microarrayfor Aneuploidy Screening

Samples affected by 164 aneuploidies, including examples oflosses and gains for all 24 types of chromosome, wereassessed with the use of the first iteration of the new microar-ray. Thirty aneuploidies were tested in PBs, 118 in single blas-tomeres, and 16 in trophectoderm samples (SupplementalTable 1, available online at www.fertstert.org). Scoring ofsamples was carried out in a blinded fashion (i.e., no informa-tion was given to personnel carrying out the test regarding theoriginal diagnosis that these samples had previouslyreceived). The microarray successfully detected 156/164 an-euploidies (95.1% detection rate) and detected three extraaneuploidies not detected with the use of the 24sure microar-rays. There was no significant difference in the degree ofconcordance for the three different types of samples analyzed.

Four of the eight aneuploidies not observed with the useof the new microarray and two of the three extra aneuploidiesdetected involved chromosome 19, suggesting a potentialproblem for the accurate determination of the copy numberof this chromosome. It is well known that interpretation ofchromosome 19 aneuploidies can be problematic when usingCGH, as noted in many earlier studies (19–21).

Although the microarray was developed primarily fordetection of whole chromosome aneuploidies, 14 of the 164aneuploidies tested were segmental abnormalities. The sizeof deleted/duplicated chromosomal segments ranged from19.3 Mb to 133.5 Mb. The new microarray successfully de-tected 12/14 of these aneuploidies. The only two segmentalabnormalities not detected both affected chromosome 19(size of segments not detected: 21.6 Mb and 23.3 Mb), again

3

ORIGINAL ARTICLE: REPRODUCTIVE BIOLOGY

highlighting a problem with the diagnosis of aneuploidyaffecting that chromosome.

All of the embryos tested during the course of this studyunderwent a single biopsy. The cells sampled were then sub-jected to WGA and tested with the use of the new microarrayas well as the 24sure microarray. Additionally, 20 of the em-bryos were later disaggregated and all of their remaining cellsreassessed with the use of FISH. These embryos ultimatelyyielded three pieces of data: 1) aCGH of biopsied materialwith the use of the new microarray; 2) aCGH of biopsiedmaterial with the use of the 24sure microarray; and 3) FISHanalysis of multiple cells from the embryo. Included in thiscohort were the embryos for which the new microarray haddetected three extra aneuploidies not detected with the useof the 24sure microarray.

Overall, 26 aneuploidies were detected with the use ofFISH. The 24sure microarray platform succeeded in detectingall of these aneuploidies, and it detected one extra aneuploidy(Supplemental Table 1). The new microarray detected 25 ofthe aneuploidies detected with the use of FISH and four extraaneuploidies. One of the extra aneuploidies was the one alsodetected with the use of the 24sure microarray. The rest of theextra aneuploidies detected by the novel microarray were notdetected with the use of FISH or the 24sure microarray, indi-cating that they were likely to have been artefacts, althoughthe possibility that they were caused by low-level mosaicismcan not be entirely ruled out. Again, analysis of chromosome19 was shown to be the site of most discrepancies.

Because chromosome 19 was responsible for more thanone-half of the discordant results (Supplemental Table 1),new probes for this chromosome were selected and added tothe microarray. The improved version of the microarray wastested against five additional samples derived from embryosknown to be affected by chromosome 19 aneuploidies. Theabnormalities were successfully detected in all cases (Fig. 1).

Taking into consideration all results obtained from vali-dation of the revised microarray, the aneuploidy concordancerate was determined to be 97.7% at the level of the individualchromosome (171/175 aneuploidies). In terms of clinicaldiagnosis (i.e., normal or abnormal), the second iteration ofthe microarray agreed with the established array in 99% ofsamples (101/102).

Validation of Microarray-based mtDNAQuantification

The novel microarray was assessed also for its ability tocorrectly determine the relative quantity of mtDNA in thesamples tested. Unamplified genomic DNA samples fromfour cell lines were applied to microarrays, and the averagefluorescence intensity for the 261 probes specific for the mito-chondrial genome was calculated. This process was repeatedwith the use of DNA derived from the same cell lines andamplified with the use of Picoplex. Results for Picoplex-amplified DNA were concordant with those obtained fromunamplified DNA, indicating that WGA did not introducedistortions in relative amount of mtDNA. The cell lines wereplaced in the following order according to their mtDNAcontent, regardless of whether WGA was used:

4

A549> Red> B2342B> 206F. This order was also confirmedby comparing Ct values after amplification of a mitochondrialDNA sequence with the use of quantitative real-time PCR.

Analysis of mtDNA Quantity in Clinical Samples

After confirmation that relative mtDNA quantity can bedetermined with the use of the microarray approach, 89 sam-ples (27 PBs, 42 blastomeres, and 20 trophectoderm samples)were analyzed. Not surprisingly, given the large difference incytoplasmic volume, blastomeres were found to containsignificantly more mtDNA than PBs and trophectodermsamples (P< .001), and trophectoderm samples had moremtDNA than PBs (P< .001). Results were analyzed accordingto maternal age (Fig. 2). Two groups were considered foreach sample type: those derived from younger patients(30–37 years old), and samples from patients of advancedreproductive age (ARA; 38–45 years old). In PB samples,mtDNA quantity in the ARA group was found to be signifi-cantly lower than in the younger age group (P¼ .04). Incontrast, single-blastomere and trophectoderm samples ofthe ARA group had a tendency toward larger quantities ofmtDNA than the samples derived from younger patients.However, differences observed for blastomeres and trophec-toderm samples did not reach statistical significance.

Assessment of SNP Probes for Inclusion on theDeveloped Microarray

Figure 3A shows a comparison of the SNP calls obtained froman embryo and its parents. When all chromosomes wereexamined, the informative SNPs giving a correct call (i.e., agenotype consistent with inheritance of parental alleles)were found to represent 40.1%. The SNPs giving mendelianinconsistencies (i.e., SNPs showing genotypes incompatiblewith inheritance from the specific parents because of geno-typing errors) constituted 13.6% of the overall informativeSNPs. The rest of the informative SNPs displayed homozygos-ity, indicative of phenomena such as allele dropout.

It was clear that the number of SNPs giving accurate datawas insufficient to allow reliable determination of genotypesat individual loci. However, when taken together, the datawere adequate to confirmwhether or not anembryowas derivedfrom a specific couple. To verify this, results from embryos werecompared with data from mismatched parents as well as theirtrue parents. Figure 3B shows results of SNP analysis whenembryo DNA was compared with nonmatching parents. Ofthe informative SNPs tested, the proportion with a genotypecompatible with inheritance from the nonmatched couple wasfourfold fewer (9.8%) than seen when embryo genotypes werecompared with the correct parents (P< .001). Additionally, theproportion of SNPs displaying mendelian inconsistenciesalmost doubled (to 25.1%) formismatched couples and embryos(P< .01). Results from all 14 samples tested with the use of themicroarray were similar to those illustrated in Figure 3.

DISCUSSIONThere is growing evidence that aneuploidy is one of the mostimportant causes of embryo implantation failure and that

VOL. - NO. - / - 2014

FIGURE 1

Chromosome profile of an aneuploid blastomere processed with two microarray comparative genomic hybridization platforms. (A) Red circlesindicate gain and loss of chromosomal material as detected by the 24sure microarray (v. 2). The same microarray indicated the embryo to befemale (relative excess of X chromosome material and deficiency of Y chromosome compared with a normal male reference sample). (B) Resultsfrom the customized microarray were entirely concordant. Error bars represent the minimum and maximum values obtained from thefluorescence intensity of the probes attached on the microarray after applying the whole-genome amplification samples and the reference sample.Konstantinidis. Comprehensive assessment of human embryos. Fertil Steril 2014.

VOL. - NO. - / - 2014 5

Fertility and Sterility®

FIGURE 2

Normalized fluorescence intensity values obtained for mtDNA fromanalysis of polar bodies (PBs), blastomeres, and trophectodermsamples for different age categories. The error bars represent theminimum and maximum values obtained from the fluorescenceintensity of the mtDNA specific probes. The black lines inside thebox plots represent the median values.Konstantinidis. Comprehensive assessment of human embryos. Fertil Steril 2014.

ORIGINAL ARTICLE: REPRODUCTIVE BIOLOGY

ongoing pregnancy rates can be improved in assistedreproductive cycles if chromosomally normal embryos areidentified and selected for transfer (4–7). The newmicroarray described in the present study demonstrated aconcordance rate of 97.7% compared with the most widelyused method for the comprehensive cytogenetic analysis ofpreimplantation embryos. Given that many embryoscontain more than one aneuploidy, the clinical diagnoses(i.e., normal or abnormal) showed even closer agreement(99%) between the two platforms.

Earlier aCGH studies of polar bodies, blastomeres, andtrophectoderm biopsies have revealed that diagnostic errorscan occur, although in most cases the incidence is <5%(22–24). A combination of biological and technicalchallenges means that no aneuploidy screening methodapplied to embryos can be considered to be infallible, a factthat makes it difficult to place an absolute value onaccuracy. What can be said of the new microarray is that itsperformance appears to be similar to existing methods ofaCGH in current use for the purpose of embryo analysis.

As well as providing information about aneuploidy, thenew microarray was shown to be capable of determiningthe relative amount of mtDNA. Although a single organellecan contain more than one copy of the mitochondrialgenome, it is likely that the quantity of mtDNA is reflectiveof the relative number of mitochondria in the sample (i.e.,when comparing samples of the same type, more mtDNAgenerally means a larger number of mitochondria per cell).The number of mitochondria may provide an insight intothe bioenergetic capacity of the cells analyzed and thereforeof the oocyte/embryo fromwhich they were derived, althoughultimately experiments measuring metabolic activity wouldbe needed for confirmation.

Results obtained during this study revealed that PBsderived from ARA women tend to have lower quantities ofmtDNA compared with those derived from younger women.

6

Other studies carried out on human oocytes have also indi-cated that mtDNA copy number decreases with advancingmaternal age (25, 26). Although much of the decline inoocyte competence seen with age can be attributed toincreasing aneuploidy rates (27), it is conceivable thatmitochondrial abnormalities might also play a role (28).

Interestingly, at the blastocyst stage a nonsignificanttrend in the opposite direction was seen: embryos from oldermothers tending to have increased levels of mtDNA. This isconcordant with a recent study that measures the amountof mtDNA in blastocysts with the use of an alternate method(quantitative PCR) (10). That study also demonstrated that25% of chromosomally normal blastocysts that fail to implantafter transfer to the uterus contain an excessive quantity ofmtDNA, confirming that measurement of this feature hasthe potential to provide clinically relevant informationadditional to that provided by aneuploidy testing (10).

In addition to aneuploidy screening and relative quanti-fication of mtDNA, the possibility of including SNP probeson the microarray was examined. SNPs have previouslybeen assessed in human embryos with the use of microarrays,but never in combination with aCGH (29–32). As with earlierstudies that have used microarrays to assess SNPs, a largenumber of genotyping errors were observed, attributable toproblems such as allele dropout. Nevertheless, it was clearthat when results from an embryo were cross-referencedwith those of its parents, the number of SNP loci with plau-sible genotypes (i.e., consistent with inheritance of one allelefrom each parent) was significantly higher than when SNPresults from an embryo were compared with those of an un-related couple (P< .001). These results indicate that althoughassessment of polymorphisms cannot avoid samples beingmixed up in the IVF laboratory, it does have the potential toallow detection of such errors before embryo transfer.

Aswell as confirmationof parental origin, analysis of SNPshas the potential to reveal which of the transferred embryos (incases of multiple embryo transfer) actually resulted in birth,providing a powerful tool for research studies assessing factorsthat might affect the implantation potential of embryos (3, 31,33). Analysis of SNPs also has the potential to indicate samplescontaminatedwith extraneous DNA (e.g., an excessive numberof alleles might be detected that are not present in eitherparent). Reducing the risk of errors caused by contaminationwith extraneous genetic material is a major consideration forany genetic diagnosis requiring DNA amplification.

CONCLUSIONA novel microarray was developed, capable of aneuploidyidentification in varieties of cells typically used for the pur-pose of preimplantation diagnosis of aneuploidy. Chromo-some abnormalities of the type detected are known to beextremely common in human oocytes and embryos and arelikely to be one of the most significant factors influencingembryo implantation. In addition to aneuploidy detection,the microarray developed during this study allows relativequantitation of mtDNA copy number; a factor of potentialclinical and scientific significance. The possibility that mea-surement of mtDNA could assist in the selection of the most

VOL. - NO. - / - 2014

FIGURE 3

Results obtained from application of amplified embryonic DNA on selected single-nucleotide polymorphism (SNP) probes. (A) Embryo comparedwith matching parents. (B) Embryo compared with nonmatching parents. Results are shown for chromosomes 1, 2, and 3 only; the otherchromosomes (4–22, X, and Y) provided similar data. Many SNPs were assessed, but only those that were informative for determining parentalinheritance are shown in the figure. Black dots indicate SNPs displaying genotypes consistent with inheritance of one allele from each of theparents, green dots indicate inheritance of a single paternal allele, and yellow dots indicate the presence of a single maternal allele (expected ifa SNP is affected by allele dropout or in cases of monosomy or uniparental isodisomy). Brown dots indicate SNPs displaying an apparentinheritance of both paternal alleles with no maternal contribution, and blue indicates the presence of both maternal alleles and an absence ofpaternal alleles. Pink and purple dots indicate SNPs with mendelian inconsistency (i.e., a pattern of inheritance that is impossible given thegenotypes of the two parents—presence of alleles that are not possessed by the parents).Konstantinidis. Comprehensive assessment of human embryos. Fertil Steril 2014.

Fertility and Sterility®

appropriate embryos for transfer—and thereby enhance IVFsuccess rates—should be investigated further. In the presentresearch, links between mtDNA content and age wereobserved, suggesting that mitochondria may have a role,direct or indirect, in reproductive aging.

This study confirmed that it is technically feasible tocombine interrogation of SNP loci with aCGH on a single mi-croarray. This offers exciting possibilities in terms of embryoidentification and avoidance of misdiagnoses due to DNAcontamination. However, further optimization and validationare required (e.g., selection of SNP probes that consistentlyprovide correct calls) before such an approach can be usedfor diagnosis of specific mutations. Taken together, theunique combination of genotyping, aneuploidy detection,and mtDNA quantification offered by this microarray pro-vides a valuable tool for scientific research and offers newpossibilities for the clinical evaluation of embryos andoocytes.

VOL. - NO. - / - 2014

Acknowledgments: The authors thank the staff of Reproge-netics at New Jersey (USA) for their assistance with this work,particularly Drs. Pere Colls and Xue Zhong Zheng and Mr.Tomas Escudero. The authors also thank Ms. Lorna Macleod(University of Oxford) for her help in validation performedregarding mtDNA probes included on the developed array.Furthermore, they thank Prof. Joaquima Navarro (UniversitatAutonomadeBarcelona) for provision of certain cell lines usedin this study. Dagan Wells is funded by the Oxford NationalInstitute for Health Research Biomedical Research Centre.

REFERENCES1. Munn�e S, Lee A, Rosenwaks Z, Grifo J, Cohen J. Diagnosis of major chromo-

some aneuploidies in human preimplantation embryos. Hum Reprod 1993;8:2185–91.

2. Verlinsky Y, Cieslak J, Freidine M, Ivakhnenko V, Wolf G, Kovalinskaya L,et al. Pregnancies following pre-conception diagnosis of common aneu-ploidies by fluorescent in-situ hybridization. Hum Reprod 1995;10:1923–7.

7

ORIGINAL ARTICLE: REPRODUCTIVE BIOLOGY

3. Wells D, Alfarawati S, Fragouli E. Use of comprehensive chromosomalscreening for embryo assessment: microarrays and CGH. Mol Hum Reprod2008;14:703–10.

4. Yang Z, Liu J, Collins GS, Salem SA, Liu X, Lyle SS, et al. Selection of singleblastocysts for fresh transfer via standard morphology assessment aloneand with array CGH for good prognosis IVF patients: results from a random-ized pilot study. Mol Cytogenet 2012;5:24.

5. Forman EJ, Hong KH, Ferry KM, Tao X, Taylor D, Levy B, et al. In vitro fertil-ization with single euploid blastocyst transfer: a randomized controlled trial.Fertil Steril 2013;100:100–7.

6. Schoolcraft WB, Katz-Jaffe MG. Comprehensive chromosome screening oftrophectoderm with vitrification facilitates elective single-embryo transferfor infertilewomenwith advancedmaternal age. Fertil Steril 2013;100:615–9.

7. Scott RT Jr, Upham KM, Forman EJ, Hong KH, Scott KL, Taylor D, et al. Blas-tocyst biopsy with comprehensive chromosome screening and fresh embryotransfer significantly increases in vitro fertilization implantation and deliveryrates: a randomized controlled trial. Fertil Steril 2013;100:697–703.

8. Harton GL, Munn�e S, SurreyM, Grifo J, Kaplan B, McCulloh DH, et al. Dimin-ished effect of maternal age on implantation after preimplantation geneticdiagnosis with array comparative genomic hybridization. Fertil Steril 2013;100:1695–703.

9. van Blerkom J. Mitochondria in early mammalian development. Semin CellDev Biol 2009;20:354–64.

10. Fragouli E, Spath K, Alfarawati S, Wells D. Quantification of mitochondrialDNA predicts the implantation potential of chromosomally normal embryos.Fertil Steril 2013;100(Suppl):S1.

11. Reynier P, May-Panloup P, Chr�etien MF, Morgan CJ, Jean M, Savagner F,et al. Mitochondrial DNA content affects the fertilizability of human oocytes.Mol Hum Reprod 2001;7:425–9.

12. May-Panloup P, Chr�etien MF, Jacques C, Vasseur C, Malthi�ery Y, Reynier P.Low oocyte mitochondrial DNA content in ovarian insufficiency. HumReprod 2005;20:593–7.

13. Santos TA, el Shourbagy S, St John JC. Mitochondrial content reflects oocytevariability and fertilization outcome. Fertil Steril 2006;85:584–91.

14. Zeng HT, Ren Z, Yeung WS, Shu YM, Xu YW, Zhuang GL, et al. Low mito-chondrial DNA and ATP contents contribute to the absence of birefringentspindle imaged with Polscope in in vitro matured human oocytes. HumReprod 2007;22:1681–6.

15. Van Blerkom J, Davis PW, Lee J. ATP content of human oocytes and devel-opmental potential and outcome after in-vitro fertilization and embryotransfer. Hum Reprod 1995;10:415–24.

16. Zhao J, Li Y. Adenosine triphosphate content in human unfertilized oocytes,undivided zygotes and embryos unsuitable for transfer or cryopreservation.J Int Med Res 2012;40:734–9.

17. Munn�e S, M�arquez C, Magli C, Morton P, Morrison L. Scoring criteria forpreimplantation genetic diagnosis of numerical abnormalities for chromo-somes X, Y, 13, 16, 18 and 21. Mol Hum Reprod 1998;4:863–70.

18. Colls P, Goodall N, Zheng X, Munn�e S. Increased efficiency of preimplanta-tion genetic diagnosis for aneuploidy by testing 12 chromosomes. ReprodBiomed Online 2009;19:532–8.

8

19. Moore DH 2nd, Pallavicini M, Cher ML, Gray JW. A t-statistic for objectiveinterpretation of comparative genomic hybridization (CGH) profiles. Cytom-etry 1997;28:183–90.

20. Voullaire L, Slater H, Williamson R, Wilton L. Chromosome analysis ofblastomeres from human embryos by using comparative genomic hybridiza-tion. Hum Genet 2000;106:210–7.

21. Guti�errez-Mateo C, Wells D, Benet J, S�anchez-García JF, Berm�udez MG,Belil I, et al. Reliability of comparative genomic hybridization to detect chro-mosome abnormalities in first polar bodies and metaphase II oocytes. HumReprod 2004;19:2118–25.

22. Fragouli E, Alfarawati S, Daphnis DD, Goodall NN, Mania A, Griffiths T, et al.Cytogenetic analysis of human blastocysts with the use of FISH, CGH andaCGH: scientific data and technical evaluation. Hum Reprod 2011;26:480–90.

23. Guti�errez-Mateo C, Colls P, S�anchez-García J, Escudero T, Prates R,Ketterson K, et al. Validation of microarray comparative genomic hybridiza-tion for comprehensive chromosome analysis of embryos. Fertil Steril 2011;95:953–8.

24. Handyside AH, Montag M, Magli MC, Repping S, Harper J, Schmutzler A,et al. Multiple meiotic errors caused by predivision of chromatids in womenof advanced maternal age undergoing in vitro fertilisation. Eur J Hum Genet2012;20:742–7.

25. de Boer KA, Jansen RP, Leigh DA, Mortimer D. Quantification of mtDNAcopy number in the human secondary oocyte. Hum Reprod 1999;14:91–2.

26. Chan CC, Liu VW, Lau EY, Yeung WS, Ng EH, Ho PC. Mitochondrial DNAcontent and 4977 bp deletion in unfertilized oocytes. Mol Hum Reprod2005;11:843–6.

27. Fragouli E, Alfarawati S, Spath K, Jaroudi S, Sarasa J, Enciso M, et al.The origin and impact of embryonic aneuploidy. Hum Genet 2013;132:1001–13.

28. Barritt JA, Cohen J, Brenner CA. Mitochondrial DNA point mutation inhuman oocytes is associated with maternal age. Reprod Biomed Online2000;1:96–100.

29. Schoolcraft WB, Treff NR, Stevens JM, Ferry K, Katz-Jaffe M, Scott RT Jr. Livebirth outcome with trophectoderm biopsy, blastocyst vitrification, andsingle-nucleotide polymorphism microarray–based comprehensive chromo-some screening in infertile patients. Fertil Steril 2011;96:638–40.

30. Johnson DS, Gemelos G, Baner J, Ryan A, Cinnioglu C, Banjevic M, et al.Preclinical validation of microarray method for full molecular karyotypingof blastomeres in a 24-h protocol. Hum Reprod 2010;25:1066–75.

31. Treff NR, Su J, Tao X, Miller KA, Levy B, Scott RT Jr. A novel single-cell DNAfingerprinting method successfully distinguishes sibling human embryos.Fertil Steril 2010;94:477–84.

32. Treff NR, Northrop LE, Kasabwala K, Su J, Levy B, Scott RT Jr. Single nucleo-tide polymorphism microarray–based concurrent screening of 24-chromo-some aneuploidy and unbalanced translocations in preimplantationhuman embryos. Fertil Steril 2011;95:1606–12.

33. Treff NR, Su J, Kasabwala N, Tao X, Miller KA, Scott RT Jr. Robust embryoidentification using first polar body single nucleotide polymorphismmicroarray–based DNA fingerprinting. Fertil Steril 2010;93:2453–5.

VOL. - NO. - / - 2014