1 s2.0-s0015028214020809-main

14

Click here to load reader

Transcript of 1 s2.0-s0015028214020809-main

Page 1: 1 s2.0-s0015028214020809-main

ORIGINAL ARTICLE: GENETICS

Blastocentesis: a source of DNA forpreimplantation genetic testing.Results from a pilot study

Luca Gianaroli, M.D., M. Cristina Magli, M.Sc., Alessandra Pomante, Ph.D., Anna M. Crivello, B.Sc.,Giulia Cafueri, B.Sc., Marzia Valerio, B.Sc., and Anna P. Ferraretti, M.D.

Reproductive Medicine Unit, Societ�a Italiana Studi di Medicina della Riproduzione, Bologna, Italy

Objective: To investigate the presence of DNA in blastocyst fluids (BFs) and to estimate whether the chromosomal status predicted byits analysis corresponds with the ploidy condition in trophectoderm (TE) cells, the whole embryo, and that predicted by polar bodies(PBs) or blastomeres.Design: Prospective study.Setting: In vitro fertilization unit.Patient(s): Seventeen couples undergoing preimplantation genetic screening with the use of array comparative genomic hybridizationon PBs (n ¼ 12) or blastomeres (n ¼ 5).Intervention(s): BFs and TE cells were retrieved from 51 blastocysts for separate chromosomal analysis.Main OutcomeMeasure(s): Presence of DNA in BFs and assessment of the corresponding chromosome condition; correlation with theresults in TE cells and those predicted by the analysis done at earlier stages.Result(s): DNAwas detected in 39 BFs (76.5%). In 38 of 39 cases (97.4%) the ploidy condition of BFs was confirmed in TE cells, and therate of concordance per single chromosome was 96.6% (904/936). In relation to the whole embryo, the ploidy condition corresponded inall cases with a per-chromosome concordance of 98.1%. The testing of PBs and blastomeres had 93.3% and 100% prediction of BFploidy condition with a concordance per chromosome of 93.5% and 94%, respectively.Conclusion(s): Blastocentesis could represent an alternative source of material for chromosomal testing, because the BF is highly

Use your smartphone

predictive of the embryo ploidy condition and chromosome content. Our data confirm the rele-vance of the oocyte and of the early-cleavage embryo in determining the ploidy condition of theresulting blastocyst. (Fertil Steril� 2014;-:-–-. �2014 by American Society forReproductive Medicine.)Key Words: Blastocyst, blastocele, blastomere, polar bodies, preimplantation genetic screening

Discuss: You can discuss this article with its authors and with other ASRM members at http://fertstertforum.com/gianarolil-blastocentesis-preimplantation-genetic-testing/

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.

he first sign of morphologic dif- creating an osmotic gradient that is preventing leakage of the liquid from

T ferentiation in the mammalianembryo happens at the morula

stage, when some cells are diverted tobecome part of the inner cell mass(ICM) and others to generate the tro-phectoderm (TE) lineage, organized asan outer layer surrounding the ICM.

After the formation of the two lin-eages, Naþ ions are accumulated on thebasolateral side of the TE epithelium,

Received June 8, 2014; revised and accepted AugustL.G. has nothing to disclose. M.C.M. has nothing to d

nothing to disclose. G.C. has nothing to discloseto disclose.

Reprint requests: Luca Gianaroli, M.D., Reproductivedella Riproduzione, V. Mazzini 12, 40138 Bolog

Fertility and Sterility® Vol. -, No. -, - 2014 0015-Copyright ©2014 American Society for Reproductivehttp://dx.doi.org/10.1016/j.fertnstert.2014.08.021

VOL. - NO. - / - 2014

partially originated by different iso-forms of Naþ/Kþ adenosine triphos-phatases (1). This gradient promotesthe accumulation of water across theepithelium through the activity oftransmembrane channels (2). The accu-mulated fluid merges and expands as asingle entity, forming the cavity knownas the blastocele. The subsequent devel-opment of tight junctions makes a seal,

13, 2014.isclose. A.P. has nothing to disclose. A.M.C. has. M.V. has nothing to disclose. A.P.F. has nothing

Medicine Unit, Societ�a Italiana Studi di Medicinana, Italy (E-mail: [email protected]).

0282/$36.00Medicine, Published by Elsevier Inc.

the cavity (3). These processes in thehuman embryo normally happen 4–5 days after fertilization, when the blas-tocyst stage begins. The progressiveaccumulation of the fluid inside thecavity and the constant duplication ofcells lead to the enlargement of theblastocyst and the thinning of thezona pellucida (ZP). These steps culmi-nate in the hatching of the blastocystfrom a natural breach in the ZP. Atthis stage, in vivo, the embryo is readyto implant in a receptive uterus.

In assisted reproductive technol-ogy, culture to blastocyst has been pro-posed as a method for an efficientselection of the best embryo to transfer.Nevertheless, top-quality embryos,

1

Page 2: 1 s2.0-s0015028214020809-main

ORIGINAL ARTICLE: GENETICS

including blastocysts, frequently fail to implant, mainlybecause of aneuploidy which is recognized as one of themain factors affecting embryo implantation (4–8). For thisreason, preimplantation genetic screening (PGS) techniques,through the analysis of the embryo chromosome status,should help in improving the pregnancy rates by avoidingthe transfer of aneuploid embryos.

According to recently published data, blastocystsrepresent the stage providing the most reliable results forPGS (9–11). Biopsy at earlier stages, polar bodies (PBs) andblastomeres, seems to be inadequate owing to the additionalabnormalities contributed by sperm and initial mitoses (forPB testing), and to mosaicism (especially for blastomereanalysis). Biopsy of TE cells has several advantages: Morethan one cell is available, the embryonic mass is nottouched, and the results obtained provide the mostcomplete representation of the preimplantation embryo'schromosomal constitution. At this stage, correctionmechanisms of meiotic reciprocal aneuploidies, if any, havealready occurred, preventing the exclusion from transfer ofblastocysts resulting from meiotic aneuploidy rescue (12).Though to a lower extent compared with the embryo at thecleavage stage, mosaicism can still be detected inblastocysts (4, 13). Therefore, even if it is not clear to whatextent TE cells are representative of the ICM, clinical resultsof the application of PGS on blastocysts seem to bereassuring and very promising (9–11).

Very recently, it was reported that the blastocyst fluid(BF) contains DNA, possibly representing another sourceof DNA for genetic analysis (14). Should these findings beconfirmed, the aspiration of the BF, a procedure that weterm blastocentesis, could easily become the preferredmeans of blastocyst biopsy. The aims of the present studywere: 1) to verify the presence of DNA in BFs; 2) to estimatewhether the chromosomal status predicted by its analysiscorresponds with the ploidy condition of PBs, blastomeres,and TE cells; and 3) to estimate whether the chromosomalstatus predicted by PBs, blastomeres, BFs, and TE cell anal-ysis corresponds with the ploidy condition of the wholeembryo.

MATERIALS AND METHODSPlan of the Study

This was a prospective study including 17 couples (maternalage 37.6 � 3.5 y) undergoing PGS by array comparativegenomic hybridization (CGH) in PBs (12 patients) or blasto-meres (5 patients) because of advance maternal age (n ¼ 14)or repeated IVF failures (n¼ 3). In all, these couples generated71 blastocysts, of which 51 were investigated for the presenceof DNA in the BF.

Each of the patients signed an informed consent to havefurther chromosomal analysis performed on supernumeraryembryos. Identification of these embryos took into consider-ation the following aspects:

� Development to blastocyst.� Aneuploid status predicted by chromosome assessment

performed on PBs or blastomeres.

2

� Euploid status predicted by chromosome assessmentperformed on PBs or blastomeres in blastocysts destinedto cryopreservation.

According to the study plan, the BF was aspirated andTE cells were biopsied for separate chromosomal analysis.When discarded from possible clinical use, and after the BFand TE biopsies were performed, whole blastocysts classifiedas nonviable (15) also underwent chromosomal analysis.

The study was approved by our Institutional ReviewBoard (IRB no. 20110503).

Biopsy Procedures

Biopsy of PB1 and PB2 was sequentially performed immedi-ately before intracytoplasmic sperm injection (ICSI; PB1biopsy)and 6–9 hours after ICSI (PB2 biopsy). The ZP was opened me-chanically after meticulous removal of all adhering cumuluscells. PBs were transferred to polymerase chain reaction (PCR)tubes to be processed separately for chromosomal analysis (16).

Blastomere biopsy was performed at�62 hours after ICSIin embryos at the 6–8-cell stage presenting with regularmorphology (15). An infrared diode laser was used to opena small breech in the ZP from where all adhering cumuluscells had been removed. A nucleated blastomere was gentlyaspirated, transferred to a PCR tube, and processed for chro-mosome analysis.

BFs from expanded blastocysts were aspirated by an ICSIpipette that was inserted in the point of contact between twoTE cells to minimize the amount of crossed cytoplasm(Supplemental Fig. 1, available online at www.fertstert.org).Great attention was paid to avoid the aspiration of anycellular material. The aspirated fluid was directly transferredto a PCR tube kept on ice without the addition of any buffer.The tubes were immediately spun and stored at �80�C untilfurther processing. A single aspiration was performed, andthe volume of the aspirated fluid was �1 mL.

TE biopsy was performed by removing 3–5 cells that hadherniated through the breach previously opened in the ZP atthe time of PB or blastomere biopsy. When necessary, the bi-opsy procedure was completed by laser pulses. The retrievedcells were transferred to a PCR tube for further processing.

In case of blastocysts not considered for clinical use (15),the whole embryo, after the above biopsies, was transferred toa PCR tube and processed.

All biopsies were performed in Hepes-buffered mediumsupplemented with protein supplement (5 mg protein/mL)under oil (Lifeglobal Media). The PCR tubes containing thebiopsies were stored at �80�C until further processing.

Whole-genome Amplification and ArrayComparative Genomic Hybridization

Amplification of all biopsies was performed in a class IIlaminar flow cabinet with the use of the Sureplex kit(Rubicon; Bluegnome). The quality of the DNA amplificationwas determined by loading 5 mL of the final reaction onto a1.5% agarose gel. An aliquot of the amplified DNA waslabeled for array CGH (24sure; Bluegnome). After arrayCGH, each sample was analyzed for the presence of gains

VOL. - NO. - / - 2014

Page 3: 1 s2.0-s0015028214020809-main

Fertility and Sterility®

and losses with the use of Blue-Fuse Multi software and theeuploid/aneuploid status of the corresponding oocyte or em-bryo predicted (17). Visualization and reporting of aneuploidywas on a per-chromosome basis.

For the evaluation of the results from the different stages,two indicators were used: the ploidy condition and the chro-mosome concordance. The first defines the correspondenceamong the different stages of the studied embryo in termsof euploidy or aneuploidy. This indicator has a clinicalimpact, because an embryo is assessed as transferrable ornot transferrable based on its ploidy condition. The term chro-mosome concordance defines the percentage of correspon-dence of all studied chromosomes between the differentstages of the analyzed embryos. Full concordance indicatescases where the ploidy condition was confirmed and all singlechromosomes corresponded. Partial concordance refers tocases where the ploidy condition was confirmed, but not allsingle chromosomes corresponded. Null concordance in-cludes cases where the ploidy condition was discordant.

RESULTSThe patients included in the study generated 71 blastocysts,corresponding to a 69% blastocyst rate calculated per numberof fertilized oocytes. This figure was similar after PB and blas-tomere biopsy. The chromosome status of the embryos thatfailed to develop to blastocysts was mostly aneuploid (74%).

The BF was aspirated from 51 blastocysts, of which 37were from oocytes having had PB biopsy and 14 fromembryos biopsied at the cleavage stage.

Analysis of the Blastocelic Fluid for the Presence ofDNA

After whole-genome amplification (WGA), DNA was detectedin 39 BFs out of 51 (76.5%), providing 39 complete sets withthe chromosome status available for PBs or blastomeres, BFs,

TABLE 1

Overall chromosome concordance by stage at biopsy calculated per ploid

Concordanc

Full Par

BFs vs. PBEmbryos 21 (70) 7Chromosomes 483/483 (100) 126/161

BFs vs. blastomeresEmbryos 8 (88.9) 1Chromosomes 192/192 (100) 11/24

BFs vs. TE cellsEmbryos 32 (82) 6Chromosomes 768/768 (100) 121/144

TE cells vs. PBsEmbryos 21 (70) 8Chromosomes 483/483 (100) 152/184

TE cells vs. blastomeresEmbryos 7 (77.8) 2Chromosomes 168/168 (100) 38/48

Note: BF ¼ blastocyst fluid; PBs ¼ polar bodies; TE ¼ trophectoderm.a P< .01.

Gianaroli. Blastocentesis as a source of DNA. Fertil Steril 2014.

VOL. - NO. - / - 2014

and TE cells (Supplemental Tables 1 and 2, available online atwww.fertstert.org). The mean amount of DNA detected,amplified from the BF, was 900.38 ng/mL (range 876.3–939.5 ng/mL).

In the 12 BF samples with no results (23.5%), failed ampli-fication occurred, as revealed by the absence of a specificband in the agarose gel run for 5 minutes. However, whenthe gel was electrophoresed for a longer time, a smearappeared, suggesting possible DNA degradation in the orig-inal sample (lane 6 in Supplemental Fig. 2, available onlineat www.fertstert.org). The subsequent labeling and hybridiza-tion led to inconclusive results. Ten of these blastocystsderived from aneuploid oocytes (n ¼ 6) or blastomeres(n ¼ 4), whereas the remaining two were predicted to beeuploid by PB (n¼ 1) or blastomere (n¼ 1) testing. The wholeembryo could be analyzed in seven cases, and the resultsconfirmed those obtained in the previous biopsies.

Chromosomal Complement of the Blastocelic Fluidin Relation to theData Obtained by Polar Bodies orBlastomere and Trophectoderm Cells

The data from the 39 BFs with informative results were eval-uated in relation to the data obtained with the use of the chro-mosomal testing in PBs (n ¼ 30) or blastomeres (n ¼ 9;Table 1). In all, BFs reflected the ploidy condition predictedby PBs in 93.3% of cases (28/30) and by blastomeres in100% of cases (9/9), accounting for a total ploidy concor-dance of 94.9% of cases (37/39), with two cases (5.1%)discordant.

In more detail, for PBs there was full chromosomeconcordance in 21 samples (70% of 30), partial concordancein seven samples (23.3%), and null concordance in the re-maining two samples (6.7%). For blastomeres, full chromo-some concordance was found in eight samples (88.9% of 9)and partial concordance in one sample (11.1%).

y condition (euploid vs. aneuploid) and per single chromosome.

e, n (%)

Totaltial Null

(23.3) 2 (6.7) 30(78.3) 36/46 (78.3) 645/690 (93.5)a

(11.1) 0 9(46) 203/216 (94.0)

(15.4) 1 (2.6) 39(84.0) 15/24 (62.5) 904/936 (96.6)a

(26.7) 1 (3.3) 30(82.6) 22/23 (95.7) 657/690 (95.2)

(22.2) 0 9(79.2) 206/216 (95.4)

3

Page 4: 1 s2.0-s0015028214020809-main

ORIGINAL ARTICLE: GENETICS

Altogether, a full chromosome concordance was detectedin 29 BF samples with the corresponding PBs (n ¼ 21) andblastomeres (n ¼ 8) (Supplemental Table 1). A euploid condi-tion had been predicted in six cases (three after PB testing andthree after blastomere testing) and confirmed in the BF of theresulting blastocysts, including one case where all theaneuploidies detected in PB1 were compensated by reciprocalaneuploidies in PB2 (Fig. 1). The remaining 23 samples wereconfirmed to be aneuploid (see an example in Fig. 2).

Partial concordance was detected in 8 samples, whichwere all predicted to be aneuploid by PBs (n ¼ 7) or blasto-mere (n ¼ 1). Here, the ploidy condition was concordant,but BFs showed aneuploidies that were only partially pre-dicted by the analysis at the previous stages (SupplementalTable 2). An example is shown in Supplemental Figure 3(available online at www.fertstert.org).

Null concordance was found in two samples, both definedas aneuploid by PB testing. Here, the ploidy condition foundin the BFs was discordant with the prediction made by PBs(Supplemental Table 2). In one of these cases, chromosomalabnormalities were found only in the BF, whereas PBs andTE cells were euploid. In the other case, PBs predicted an em-bryo trisomic for chromosome 6, but the resulting blastocystwas euploid according to the analysis of BF, TE cells, and thewhole embryo.

When comparing the results from BFs and the corre-sponding TE cells, in 97.4% of sets (38/39) the ploidy condi-tion of BFs was confirmed in TE cells (Table 1). In moredetail, there was full chromosome concordance in 32 cases(82% of 39), partial concordance in six cases (15.4%), andnull concordance in one case (2.6%). The discordantcase was classified as aneuploid by BF analysis and

FIGURE 1

The aneuploidies in polar body 1 (PB1) were compensated by reciprocal anewas euploid according to analysis of the blastocelic fluid (BF) and trophectGianaroli. Blastocentesis as a source of DNA. Fertil Steril 2014.

4

euploid by TE analysis with nine discordant chromosomes(Supplemental Table 2).

Looking at single chromosomes in the whole dataset, therate of concordance between BF and the other biopsies was93.5% with PBs, 94% with blastomeres, and 96.6% withTE cells (96.6% vs. 93.5%; P< .01; Table 1).

No statistically significant difference was detected whenthe results from TE cell analysis were compared with thedata obtained by PBs and blastomeres (Table 1).

Chromosomal Complement of the BlastocystAssessed by the Analysis of the Whole Embryo inRelation to the Data Obtained by Polar Bodies,Blastomeres, Blastocyst Fluid, and TrophectodermCells

To add further information regarding the comparisons of thechromosome results obtained at each stage, the whole embryowas analyzed in 26 sets, 20 having had PB biopsy and six blas-tomere biopsy (Table 2; Supplemental Table 3, available onlineat www.fertstert.org). All of them had had BF and TE biopsies.

The ploidy condition of the whole embryo was predictedin 95% of cases by PBs (19/20), in 100% of cases by blasto-meres (6/6), and in 100% of cases by both BF and TE cells(26/26; Table 2). The only discordant case was predicted asaneuploid for chromosome 6 by PBs and found to be euploidin the whole embryo as well as in the corresponding BF andTE cells (Supplemental Table 3).

Full chromosome concordance was similar throughout allstages, namely, 80%with PBs (16/20), 100%with blastomeres(6/6), 80.8% with TE cells (21/26), and 84.6% with BF (22/26).Partial concordance was 15% with PBs (3/20), 19.2% with

uploidies in polar body 2 (PB2). As a result, the corresponding blastocystodem cells (TE).

VOL. - NO. - / - 2014

Page 5: 1 s2.0-s0015028214020809-main

FIGURE 2

A gain for chromosome 8 in PB1 and for chromosomes 15 and 22 in PB2 predicted a blastocyst monosomic for chromosomes 8, 15, and 22. Thiscondition was confirmed in both the BF and the TE cells. Abbreviations as in Figure 1.Gianaroli. Blastocentesis as a source of DNA. Fertil Steril 2014.

Fertility and Sterility®

TE cells (5/26), and 15.4% with BFs (4/26). Finally, nullconcordance was found with one PB case (5% of 20).

Calculated per single chromosome, the grade of concor-dance was 98.9% with PBs, 100% with blastomeres, 97.9%with BFs, and 98.1% with TE cells.

DISCUSSIONThe most relevant findings in our study were the detection ofDNA in the majority of BFs and the strong prediction value ofPB and blastomere testing on the blastocyst chromosomalstatus.

In 23.5% of the investigated BFs, no informative DNAcould be detected. No morphologic factors in the studied blas-tocysts could be recognized as able to predict the finding ofDNA in BFs, although a common feature in the blastocystswith failed amplification was the presence of a poor-qualityTE. On the other hand, this characteristic did not preventthe finding of DNA in the BFs of other cases. After attemptingWGA, we observed that the samples that failed to amplifyprobably had highly fragmented DNA, as suggested by thesmear observed in the agarose gel when it was electrophor-esed for 15 minutes (Supplemental Fig. 2). More cases areneeded to see if there is a connection between these observa-tions and the absence of informative DNA in the BF. Technicalproblems could also have occurred, especially during thetubing of the BF, and this is a step that we are carefully revis-iting. We are also trying to evaluate if the time of biopsy couldhave an effect on the subsequent outcome. As a preliminaryfinding, we retrieved the BF twice from two blastocysts at24-hour intervals. In one case, WGA failed in the first biopsyand was successful in the second sample; in the other case,

VOL. - NO. - / - 2014

both samples were successful and provided the samediagnosis.

The presence of DNA in BFs, detected and successfullytested in 76.5% of the studied samples, is actually not surpris-ing. All cells surrounding the blastocele are metabolicallyactive and secrete different substances, including severaltypes of proteins (18, 19). It is well known that cell deathnormally occurs even in good-quality blastocysts, in bothICM and TE compartments. A study evaluating blastocystsoriginating from embryos with different levels of fragmenta-tion demonstrated higher levels of apoptosis in embryos ofexcellent morphology, suggesting a possible role of apoptosisin the regulation of cell number (20). Because the aspiration ofthe fluid was done by carefully avoiding the contaminationby any cellular material, we are confident that the extractedDNA is a true reflection of the BF content. In addition, thedetection of some differences in the chromosome status ofBFs compared with that of TE cells supports a different originof the two samples.

Based on the above considerations, the presence of DNAin BFs could be consequent to its release from dead cells intothe blastocelic cavity, with its actual quantity being possiblyrelated to the rate of cell death. Therefore, there could bethree different scenarios in the analysis of the BF: 1) TheDNA in the BF is too fragmented or too scarce to result insuccessful genome amplification; 2) the DNA in the BF am-plifies and appears to be a reflection of the ploidy conditionof the blastocyst, taking TE cells as the reference (TE biopsybeing the conventional procedure for PGS of blastocysts);and 3) the DNA in the BF is discordant with the chromo-somal complement of the blastocyst, taking TE cells as thereference.

5

Page 6: 1 s2.0-s0015028214020809-main

TABLE 2

Overall chromosome concordance by stage at biopsy over the whole embryo calculated per ploidy condition (euploid vs. aneuploid) and per singlechromosome.

Concordance, n (%)

TotalFull Partial Null

Whole embryo vs. PBsEmbryos 16 (80) 3 (15) 1 (5) 20Chromosomes 368/368 (100) 65/69 (94.2) 22/23 (95.7) 455/460 (98.9)

Whole embryo vs. blastomereEmbryos 6 (100) 0 0 6Chromosomes 144/144 (100) 144/144 (100)

Whole embryo vs. TE cellsEmbryos 21 (80.8) 5 (19.2) 0 26Chromosomes 504/504 (100) 107/120 (89.2) 611/624 (97.9)

Whole embryo vs. BFEmbryos 22 (84.6) 4 (15.4) 0 26Chromosomes 528/528 (100) 84/96 (87.5) 612/624 (98.1)

Note: Abbreviations as in Table 1.

Gianaroli. Blastocentesis as a source of DNA. Fertil Steril 2014.

ORIGINAL ARTICLE: GENETICS

Regarding the first scenario, this happened in our datasetat a frequency of 23.5%, with no informative DNA found in12 of the 51 studied samples. As already mentioned, no fac-tors were identified that could be associated with thenegative or positive finding of DNA.

For the second scenario, the analysis of the BF was inagreement with that of the corresponding TE cells in 38samples (74.5% of 51 samples) with full chromosomeconcordance per single chromosome in 32 of them.

Finally, the remaining set represented the third scenario(2% of 51 samples), with BF showing multiple chromosomeanomalies in a blastocyst that was predicted to be euploidby PB and TE analysis (Table 1).

In summary, the ploidy condition between BFs andTE cells was found to be concordant in all cases except one(38/39, 97.4%), suggesting that BF could represent a valuablematerial for clinical use, provided that the proportion of sam-ples with informative DNA is raised to levels similar to thoseobtained by conventional biopsy procedures.

Looking at single chromosomes, the global evaluation ofthe data in the 39 samples with successful amplification anddiagnosis of DNA in the BF demonstrated full concordancebetween BFs and TE cells in 82% of cases (32/39), with theDNA retrieved by blastocentesis reflecting the exact chromo-somal content of TE cells. In the remaining sets, some chro-mosome variations were detected between BFs and TE cells,which is not surprising if we consider the possible origin ofDNA in BF, which could also contain DNA released fromthe ICM (20). As presented in Table 1, the BF was extremelysensitive, similarly to TE cells, in revealing all of the chromo-some variations that occurred during embryo development.As expected, the concordance per single chromosome wassignificantly lower with PBs compared with TE cells, confirm-ing that the BF collects the DNA produced by TE cells as wellas the ICM.

In view of our results, because the artificial collapse ofthe blastocele is routinely applied in the vitrification ofexpanded and hatching blastocysts (21), it would be advis-able to store the BF and to have it at least as a back-up

6

sample until further refinement of its use as alternativeDNA source is reached.

The second relevant finding coming from our study is thehigh level of prediction of the blastocyst chromosomal statusmade by the analysis done at the previous stages. Taking TEcells as the reference (Table 1), the prediction of ploidymade by PB or blastomere testing was confirmed in 97.4%of cases (96.7% and 100%, respectively). When consideringthe number of abnormalities per single chromosome, PBsand blastomeres had similar prediction power of the chromo-some complement of TE cells (95.2% and 95.4% respectively).A similar trend was observed when taking the whole embryoas the reference (Table 2), with the ploidy condition confirmedin 95% of PBs and 100% of blastomeres but the concordanceper single chromosome rising to 98.9% and 100%, respec-tively. In other words, TE cells are only partially representa-tive of the whole blastocysts, because the ICM could havesome differences regarding chromosome condition. It canbe speculated that the BF collecting DNA from both TE andICM could be more sensitive in defining the whole blastocystchromosome status. The high prediction level of PBs of thewhole embryo chromosome condition is in contrast withother authors claiming that testing at the PB stage is theless accurate method compared with blastomeres, mainlyowing to the high incidence of post-zygotic events (22). Inthis respect, it must be mentioned that in that work theauthors assumed that TE cells are representative of the blasto-cyst chromosome complement, an assumption that theycorroborated in another work with the use of fluorescencein situ hybridization (23). This is probably true in most cases,especially when considering our data as well as the positiveclinical outcome associated with PGS on blastocysts (10, 11,24). However, when we analyzed the whole embryo andcompared the results with those derived from TE cells, theconcordance per single chromosomes was 97.9% (Table 2;Supplemental Table 3). Based on these findings, takingTE cells as the point of reference to evaluate the power ofprediction of PGS made at previous stages could be nottotally correct. On the other hand, the capacity of BF in

VOL. - NO. - / - 2014

Page 7: 1 s2.0-s0015028214020809-main

Fertility and Sterility®

defining the whole embryo ploidy condition was 100% and98.1% per single chromosome, suggesting that aninformative BF represents a true reflection of the blastocystchromosome condition. Blastocentesis being a less invasivemethod of biopsy, this strategy could become the preferredchoice for PGS.

We observed that the lowest rates of concordance in BFsper single chromosome were frequently related to the pres-ence of multiple abnormalities in PBs or blastomeres. Itcould be speculated that highly aneuploid zygotes or earlyembryos are predisposed to develop mosaic blastocysts,possibly owing to defective sister chromatid cohesion thatcan result in chromosome missegregation and aneuploidy(25). Misaligned chromosomes in a metaphase aneuploidcell or lagging chromosomes between anaphase cells alsopromote the formation of mosaicism. It has been reportedthat aneuploidy induces an increase in DNA recombinationand a decrease in DNA damage repair efficiency, which areboth forms of genomic instability (26). In other words, aneu-ploidy per se seems to be able to induce chromosomal insta-bility, and this could explain the type of results obtained insome cases of partial concordance (Supplemental Table 2).More on the subject is added by an experimental mousemodel showing that mosaic aneuploid embryos can developand implant in female uterine tissue and initiate the gastru-lation process, but quickly degenerate (27). Searching for themechanism responsible for the demise of these embryos, itwas found that it is caused by the activation of a spatiallyand temporally controlled p53-independent apoptotic mech-anism and does not result from a failure to progress throughmitosis (27). Therefore, the initial state of primary aneu-ploidy resulted in a rapid evolution of mosaicism and earlyembryonic death. This gestational loss due to aneuploidmosaicism could account for the large proportion of humanpregnancy loss before clinical recognition. The per-chromosome discordances detected in our study betweenBFs and TE cells, and between them and the whole embryo,could be a reflection of this mechanism which could beespecially relevant in the case of oocytes carrying multipleaneuploidies (Supplemental Table 2).

In conclusion, the present data confirmed the relevance ofthe oocyte and of the early-cleavage embryo in determiningthe ploidy of the resulting blastocyst. Therefore, althoughthe incidence of post-zygotic events can not be disregarded,the oocyte and the initial stages of embryogenesis do providereliable results for PGS (and preimplantation genetic diag-nosis). The finding of DNA in BFs, besides representing apossible alternative source of material for chromosomal anal-ysis, could contribute additional information on the study ofchromosome segregation in the early embryo.

REFERENCES1. Barcroft LC, Moseley AE, Lingrel JB, Watson AJ. Deletion of the Na/K-ATPase

alpha1-subunit gene (Atp1a1) does not prevent cavitation of the preimplan-tation mouse embryo. Mech Dev 2004;121:417–26.

2. Barcroft LC, Offenberg H, Thomsen P, Watson AJ. Aquaporin proteins inmurine trophectoderm mediate transepithelial water movements duringcavitation. Dev Biol 2003;256:342–54.

VOL. - NO. - / - 2014

3. Watson AJ, Natale DR, Barcroft LC. Molecular regulation of blastocyst for-mation. Anim Reprod Sci 2004;82:583–92.

4. Magli MC, Jones GM, Gras L, Gianaroli L, Korman I, Trounson AO. Chro-mosome mosaicism in day 3 aneuploid embryos that develop tomorphologically normal blastocysts in-vitro. Hum Reprod 2000;15:1781–6.

5. Sandalinas M, Sadowy S, Alikani M, Calderon G, Cohen J, Munne S. Devel-opmental ability of chromosomally abnormal human embryos to develop tothe blastocyst stage. Hum Reprod 2001;16:1954–8.

6. Li M, Marin DeUgarte C, Surrey M, Danzer H, DeCherney A, Hill DL. Fluores-cence in-situ hybridization reanalysis of day-6 human blastocysts diagnosedwith aneuploidy on day 3. Fertil Steril 2005;84:1395–400.

7. Hassold T, Hall H, Hunt P. The origin of human aneuploidy: where we havebeen, where we are going. Hum Mol Genet 2007;16:R203–8.

8. Fragouli E, Lenzi M, Ross R, Katz-Jaffe M, Schoolcraft WB, Wells D. Compre-hensive molecular cytogenetic analysis of the human blastocyst stage. HumReprod 2008;23:2596–608.

9. 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;5:2.

10. Schoolcraft WB, Katz-Jaffe MG. Comprehensive chromosome screening oftrophectoderm with vitrification facilitates elective single-embryo transferfor infertile women with advanced maternal age. Fertil Steril 2013;100:615–9.

11. Scott RT Jr, Upham KM, Forman EJ, Hong KH, Scott KL, Taylor D, et al.Blastocyst biopsy with comprehensive chromosome screening and freshembryo transfer significantly increases in vitro fertilization implantationand delivery rates: a randomized controlled trial. Fertil Steril 2013;100:697–703.

12. Forman EJ, Treff NR, Stevens JM, Garnsey HM, Katz-Jaffe MG, Scott RT Jr,et al. Embryos whose polar bodies contain isolated reciprocal chromo-some aneuploidy are almost always euploid. Hum Reprod 2013;28:502–8.

13. Fragouli E, Wells D. Aneuploidy in the human blastocyst. CytogenetGenome Res 2011;133:149–59.

14. Palini S, Galluzzi L, de Stefani S, Bianchi M, Wells D, Magnani M, et al.Genomic DNA in human blastocoele fluid. Reprod Biomed Online 2013;26:603–10.

15. Alpha Scientists in Reproductive MedicineESHRE Special Interest Group ofEmbryology. The Istanbul consensus workshop on embryo assessment: pro-ceedings of an expert meeting. Hum Reprod 2011;26:1270–83.

16. Magli MC, Montag M, Koester M, Muzii L, Geraedts J, Collins J, et al. Polarbody array CGH for prediction of the status of the corresponding oocyte.Part II: technical aspects. Hum Rep 2011;26:3181–5.

17. Geraedts J, Montag M, Magli MC, Repping S, Handyside A, Staessen C,et al. Polar body array-CGH for prediction of the status of thecorresponding oocyte. Part I: clinical results. Hum Reprod 2011;26:3173–80.

18. d’Alessandro A, Federica G, Palini S, Bulletti C, Zolla L. A mass spectrometry-based targeted metabolomics strategy of human blastocoele fluid: a prom-ising tool in fertility research. Mol Biosyst 2012;8:953–8.

19. Jensen PL, Beck HC, Petersen J, Hreinsson J, W�anggren K, Laursen SB, et al.Proteomic analysis of human blastocoel fluid and blastocyst cells stem cellsand development. Stem Cells Dev 2013;22:1126–35.

20. Hardy K, Stark J, Winston RML. Maintenance of the inner cell mass inhuman blastocysts from fragmented embryos. Biol Reprod 2003;68:1165–9.

21. Mukaida T, Oka C, Goto T, Takahashi K. Artificial shrinkage of blastocoelesusing either a micro-needle or a laser pulse prior to the cooling steps of vitri-fication improves survival rate and pregnancy outcome of vitrified humanblastocysts. Hum Reprod 2006;21:3246–52.

22. Capalbo A, Bono S, Spizzichino L, Biricik A, Baldi M, Colamaria S, et al.Sequential comprehensive chromosome analysis on polar bodies, blasto-meres and trophoblast: insights into femalemeiotic errors and chromosomalsegregation in the preimplantation window of embryo development. HumReprod 2013;28:509–18.

7

Page 8: 1 s2.0-s0015028214020809-main

ORIGINAL ARTICLE: GENETICS

23. Capalbo A, Wright G, Elliott T, Ubaldi FM, Rienzi L, Nagy ZP. FISH reanalysisof inner cell mass and trophectoderm samples of previously array-CGHscreened blastocysts shows high accuracy of diagnosis and no major diag-nostic impact of mosaicism at the blastocyst stage. Hum Reprod 2013;28:2298–307.

24. Forman EJ, Upham KM, ChengM, Zhao T, Hong KH, Treff NR, et al. Compre-hensive chromosome screening alters traditional morphology-based embryoselection: a prospective study of 100 consecutive cycles of planned fresheuploid blastocyst transfer. Fertil Steril 2013;100:718–24.

8

25. Watrin E, Prigent C. Sister chromatid cohesion and aneuploidy. In: StorchovaZ, editor. Aneuploidy in health and disease. Intech. Accessed May 16, 2012.Available from: www.intechopen.com/books/aneuploidy-in-health-and-disease/sister-chromatid-cohesion-and-aneuploidy.

26. Sheltzer JM, Blank HM, Pfau SJ, Tange Y, George BM, Humpton TJ, et al.Aneuploidy drives genomic instability in yeast. Science 2011;333:1026–30.

27. Lightfoot DA, Kouznetsova A, Mahdy E, Wilbertz J, H€o€og C. The fate ofmosaic aneuploid embryos during mouse development. Dev Biol 2006;289:384–94.

VOL. - NO. - / - 2014

Page 9: 1 s2.0-s0015028214020809-main

SUPPLEMENTAL FIGURE 1

Aspiration of the blastocyst fluid from an expanded blastocyst. Anintracytoplasmic sperm injection pipette was inserted in the point ofcontact between two trophectoderm cells to minimize the amountof crossed cytoplasm. Attention was paid to avoid the aspiration ofcellular material.Gianaroli. Blastocentesis as a source of DNA. Fertil Steril 2014.

VOL. - NO. - / - 2014

Fertility and Sterility®

8.e1

Page 10: 1 s2.0-s0015028214020809-main

SUPPLEMENTAL FIGURE 2

Amplification product of seven blastocelic fluid (BF) samples on 1.5%agarose gel. (A) Gel electrophoresed for 5 minutes at 100 V, and (B)the same gel run for 15 minutes. The lane contents are labeled asfollows: one kb ladder, followed by seven BF samples. The last twolanes are positive and negative control samples, respectively. Theamplification band is strong in all samples and absent in sample 6(A). When the gel was run for 15 minutes, a smear appeared insample 6 (B).Gianaroli. Blastocentesis as a source of DNA. Fertil Steril 2014.

8.e2 VOL. - NO. - / - 2014

ORIGINAL ARTICLE: GENETICS

Page 11: 1 s2.0-s0015028214020809-main

SUPPLEMENTAL FIGURE 3

A gain for chromosome 16 in the polar body 2 (PB2) predicted a blastocyst monosomic for chromosome 16. This condition was found in both theblastocelic fluid (BF) and trophectoderm cells (TE), but with the addition of monosomy 9.Gianaroli. Blastocentesis as a source of DNA. Fertil Steril 2014.

VOL. - NO. - / - 2014 8.e3

Fertility and Sterility®

Page 12: 1 s2.0-s0015028214020809-main

SUPPLEMENTAL TABLE 1

Chromosomal status in blastocelic fluid (BF) samples showing full correspondence with the results predicted by polar bodies (PBs) orblastomeres.

SampleID Age (y) PB1 PB2 Blastomere BF TE

3 38 euploid loss 2 – gain 2 gain 25 38 gain 8 gain 15, 22 – loss 8,15,22 loss 8,15,227 36 – loss 14 loss 14 loss 149 36 – euploid euploid euploid13 36 gain 4, 5, 6, 7, 9, 11, 12,

15, 19, 20, Xloss 1, 2, 3, 8, 10, 13, 14,

16, 18

gain 1, 2, 3, 8, 10, 13, 14,16, 18

loss 4, 5, 6, 7, 9, 11, 12, 15,19, 20, X

– euploid euploid

19 42 euploid euploid – euploid euploid21 38 – – euploid euploid euploid22 38 – – euploid euploid euploid24 41 – – loss 16 loss 16 loss 1628 42 euploid loss 21 – gain 21 gain 21,

loss 129 41 euploid euploid – euploid euploid33 35 euploid gain 15 – loss 15 loss 1557 41 euploid gain 19, 21

loss 18, 22– gain 18, 22

loss 19, 21gain 18, 22loss 19, 21

58 41 euploid loss 15 – gain 15 gain 1585 36 loss 22 loss 14 – gain 14, 22 gain 14, 2286 36 gain 9q

loss 9p, 15loss 9 – gain 9q, 15 gain 9q, 15

87 36 gain 16loss 13

loss 16 – gain 13 gain 13

88 36 loss 15 gain 15 (chromosome) – loss 15 loss 1589 36 gain 4q (chromosome) gain 16

loss 4q– loss 16, 4q loss 16, 4q

90 36 euploid gain 13 – loss 13 loss 13111 37 – – loss 8, 16 loss 8, 16 loss 8, 16112 37 – – gain 2, 7, 21

loss 10gain 2, 7, 21loss 10

gain 2, 7, 21loss 10, 11, 16

114 37 – – gain 8, 16 gain 8, 16 gain 8, 16116 40 gain 21

loss 13gain 22 – gain 13

loss 21, 22gain 13loss 21, 22

118 40 gain 15, 21 gain 22 loss 15, 21, 22 gain 10loss 15, 21, 22

119 40 gain 7qloss 4, 7p

loss 7 – gain 4, 7p(chromosome)

gain 4, 7p(chromosome)

121 40 loss 9 euploid – gain 9 gain 9124 40 gain 18

loss 16gain 16loss 18, 22

– gain 22 gain 22

126 40 gain 21loss 18

gain 18loss 8

– gain 8loss 21

gain 8loss 21

Note: In all samples except three (samples 28, 112, and 118) the full concordance also included cells biopsied from the trophectoderm (TE). All gains and losses observed in PBs were due tochromatid predivision except where indicated.

Gianaroli. Blastocentesis as a source of DNA. Fertil Steril 2014.

8.e4 VOL. - NO. - / - 2014

ORIGINAL ARTICLE: GENETICS

Page 13: 1 s2.0-s0015028214020809-main

SUPPLEMENTAL TABLE 2

Chromosomal status in BF samples and corresponding TE showing partial or null correspondence with the results predicted by PBs orblastomeres.

SampleID Age (y) PB1 PB2 Blastomere BF TE

1 41 loss 16 loss 12 – gain 12,16loss 15

gain 12,16loss 15

4 38 euploid gain 22 – loss 17,22 loss 17,2211 36 euploid gain 16 – loss 9, 16 loss 9, 166 37 gain 9

loss 1gain 1, 16, 22loss 9

– gain 12loss 15, 16

loss 16,22

8 36 – – gain 5, 10, 17, 21loss 1, 6, 14, 15, 22

gain 8, 13, 21loss 1, 5, 17, 18, X

gain 21

10 36 gain 1, 2, 4, 8, 11, 12, 13,15, 20, 22

loss 3, 5, 9, 10, 16, 18, 19,21, X

euploid – loss 12 loss 12

18 42 gain 4, 19–

gain 5loss 4

– gain 2, 4, 8, 12, 20loss 5, 13, 14

gain 2,8,17,18loss 10,13,19

70 38 gain 20loss 21

euploid – gain 16loss 20

gain 16loss 20

2 38 euploid euploid – gain 5, 8, 11, 12, 15, 19loss 3, 9, 16

euploid

117 40 loss 6, 22 gain 22 – euploid euploidNote: Partial concordance was found in eight sets. In the first three samples (samples 1, 4, and 11), BFs corresponded with the predictions made by PBs and blastomeres with the addition of newaneuploidies. In the subsequent five samples (samples 6, 8, 10, 18, and 70), only some of the anomalies predicted by PBs and blastomeres appeared in BFs, and newoneswere detected. In two cases(samples 10 and 70), BFs and TE cells fully matched. Null concordancewas detected in the last two samples (2 and 117), where the chromosomal status of the DNA retrieved by the BF did not matchwith that predicted by PBs. All gains and losses observed in PBs were due to chromatid predivision. Abbreviations as in Supplemental Table 1.

Gianaroli. Blastocentesis as a source of DNA. Fertil Steril 2014.

VOL. - NO. - / - 2014 8.e5

Fertility and Sterility®

Page 14: 1 s2.0-s0015028214020809-main

SUPPLEMENTAL TABLE 3

Chromosome status in PBs or blastomeres compared with the results obtained in TE cells and the whole embryo.

Sample ID Age (y) PB1 PB2 Blastomere BF TE Whole embryo

19 42 euploid euploid - euploid euploid euploid21 38 – – euploid euploid euploid euploid22 38 – – euploid euploid euploid euploid24 41 – – loss 16 loss 16 loss 16 loss 1633 35 euploid gain 15 – loss 15 loss 15 loss 1557 41 euploid gain 19,21

loss 18,22– gain 18,22

loss 19,21gain 18, 22loss 19, 21

gain 18, 22loss 19,21

58 41 euploid loss 15 – gain 15 gain 15 gain 1586 36 gain 9q

loss 9p,15loss 9 – gain 9q,15 gain 9q, 15 gain 9q, 15

87 36 gain 16loss 13

loss 16 – gain 13 gain 13 gain 13

88 36 loss 15 gain 15 (chromosome) – loss 15 loss 15 loss 1589 36 gain 4q (chromosome) gain 16

loss 4q– loss 16,4q loss 16, 4q loss 16, 4q

90 36 euploid gain 13 – loss 13 loss 13 loss 13111 37 – – loss 8, 16 loss 8, 16 loss 8, 16 loss 8, 16114 37 – – gain 8, 16 gain 8, 16 gain 8, 16 gain 8, 16116 40 gain 21

loss 13gain 22 – gain 13

loss 21, 22gain 13loss 21, 22

gain 13loss 21, 22

119 40 gain 7qloss 4, 7p

loss 7 – gain 4, 7p (chromosome) gain 4, 7p (chromosome) gain 4, 7p (chromosome)

121 40 loss 9 euploid – gain gain 9 gain 9124 40 gain 18

loss 16gain 16loss 18, 22

– gain 22 gain 22 gain 22

126 40 gain 21loss 18

gain 18loss 8

– gain 8loss 21

gain 8loss 21

gain 8loss 21

18 42 gain 4, 19 gain 5loss 4

– gain 2, 4, 8, 12, 20loss 5, 13, 14

gain 2, 8, 17, 18loss 10, 13, 19

loss 5, 19

28 42 euploid loss 21 – gain 21 gain 21loss 1

gain 21

70 38 gain 20loss 21

euploid – gain 16loss 20

gain 16loss 20

gain 7loss 20

85 36 loss 22 loss 14 – gain 14, 22 gain 14, 22 gain 4, 14, 22112 37 – – gain 2, 7, 21

loss 10gain 2, 7, 21loss 10

gain 2, 7, 21loss 10, 11, 16

gain 2, 7, 21loss 10

117 40 loss 6 (chromosome), 22 gain 22 – euploid euploid euploid118 40 gain 15,21 gain 22 – loss 15, 21, 22 gain 10

loss 15,21,22gain 10loss 15,21,22

Note: The samples 19 to 126 are those with full concordance throughout all the studied stages. All gains and losses observed in PBs were due to chromatid predivision except where indicated. Abbreviations as in Supplemental Table 1.

Gianaroli. Blastocentesis as a source of DNA. Fertil Steril 2014.

8.e6

VOL.-

NO.-

/-

2014

ORIG

INALARTIC

LE:GEN

ETICS