Non-invasive pre-implantation genetic testing of human embryos: an emerging concept

Abstract

The accurate genetic screening of pre-implantation embryos currently entails the use of technically challenging and biologically invasive biopsies of the human embryos. Investigating a more conservative sampling approach has emerged as a timely and desired alternative. Circulating cell-free embryonic DNA is present in the blastocoel fluid and spent culture media of blastocysts, and this has lately been sought as an attractive source of genetic information. The genetic analysis of cell-free embryonic DNA has been reported, to be useful in evaluating the genetic constitution of embryos; thus, providing a potential alternative to conventional biopsy-derived pre-implantation genetic testing (PGT). In this review, we have summarized these non-invasive alternative applications of PGT and discussed their current limitations and future clinical implications.

Introduction

It is estimated that about 20–80% of human embryos are chromosomally aneuploid (Hassold and Hunt, 2001; Vera-Rodriguez et al., 2015). This phenomenon is the result of meiotic and mitotic developmental errors and appears to occur at a much higher frequency in humans than in experimental animals (Bond and Chandley, 1983). Established means of pre-implantation embryo genetic profiling entail the collection of embryo-derived material for genetic testing. Many studies on Day-3 embryo biopsies showed reassuring long-term neonatal outcomes (Desmyttere et al., 2008; Liebaers et al., 2009), and no significant differences in developmental morbidities and major congenital abnormalities among children born from biopsied embryos compared to children normally conceived (Nekkebroeck et al., 2008; Harper et al., 2012). The safety profile of conventional pre-implantation genetic testing (PGT), however, has been contested by other investigators in view of reports suggesting that the invasive removal of cells from pre-implantation embryos may interfere with embryonic development. Besides from invasive disruptive handling, embryo biopsies also require dedicated equipment and highly trained personnel, which amount to a measurable commitment in time and cost. Taken together, while embryo biopsy remains the cornerstone for PGT, interest in investing non-invasive alternatives seems very timely, reasonable and could improve the cost-efficiency and safety of the procedure.

Although first detected in adult blood in 1948 (Mandel, 1948), cell-free DNA has recently gained scientific interest. The isolation of cell-free DNA of fetal origin in the peripheral blood of pregnant women (Lo et al., 1997) has revolutionized the field of prenatal aneuploidy screening and has been established as the primary source of genetic information for non-invasive prenatal testing (Huppertz and Kingdom, 2004). More recently, human embryos were demonstrated to release DNA fragments into their environment (Assou et al., 2014). Cell-free DNA has been recovered from blastocoel fluid (BF) and spent culture media, but their significance to the embryonic genome has not been fully elucidated (Assou et al., 2014; Wu et al., 2015). The recognition of these nucleic acids in combination with improved genetic sequencing technology could prove to be the key elements required for a major breakthrough in the concept of non-invasive pre-implantation genetic testing (NI-PGT).

BF sampling

Clinical studies

In 2013, Palini et al. reported the identification of cell-free DNA fragments by real-time PCR in 90% of BF samples obtained from Day-5 blastocysts cultured in vitro (Palini et al., 2013). The median volume of BF collected was 0.3–0.5 nl expelled into 4 μl of 1 mmol/l Tris-HCl and 0.1 mmol/l EDTA (pH 8.0) and the median genomic DNA content was 9.9 pg per sample, which was similar to that found in a single cell (6.5 pg) (Serth et al., 2000). The investigators demonstrated a successful amplification efficiency of 95% for the multicopy gene testis-specific protein Y-linked 1 (TSPY1) on the Y chromosome, enabling the gender identification of male embryos. Although preliminary, the findings of this study were very interesting in that they laid the foundations for a novel minimally invasive alternative to biopsy-derived PGT for X-linked disorders (Palini et al., 2013). In 2015, Tobler et al. performed array comparative genomic hybridization (aCGH) analyses on 96 donated cryopreserved embryos (Tobler et al., 2015). After whole-genome amplification (WGA), amplifiable embryonic DNA was successfully recovered from 63% of BF samples. They were found to be concordant to the karyotypes of matched inner cell mass-trophectoderm (TE) in 48.3% (29 of 60) of embryos analyzed. The findings of this study led the authors to recommend against the use of BF as an alternative biopsy modality for PGT before more advanced technical improvements are achieved. It should be noted nonetheless that embryos used in the study were supernumerary and were donated for research because of their poor clinical suitability. These facts, therefore, may have limited the generalizability of the findings and influenced adversely the parameters of diagnostic accuracy.

Meanwhile, the findings of other investigators were more encouraging. Gianaroli et al. (2014) detected embryonic DNA in 76.5% of BF samples studied. After WGA amplification and aCGH analyses, they reported a concordance rate of 97.4% on the ploidy condition of BFs when compared to matched TE biopsies. In addition, Magli et al. (2016) reported comparable DNA detection rates of 82% in BF samples, and ploidy concordance rates of 97.1% with TE cells using similar molecular methodologies. Both research groups supported the role of BF aspiration as a promising minimally invasive alternative to embryo biopsy for chromosomal testing. After obtaining high-quality DNA amplification from BF using WGA, Zhang et al. (2016a,b) benchmarked the genomic coverage of BF-derived DNA fragments against matching blastomeres. Sequencing reads coverage was found to be consistent across most regions of the genome. Gene annotations identified sequences of the majority of known genes, highlighting the suitability of the BF cell-free DNA template for monogenic disease screening. Studies evaluating cell-free DNA amplification and analysis methods in BF are summarized in Table I.

Limitations and future research

Although the biological significance of fragmented DNA is unknown, some have suggested it may play a role in cell-to-cell communication within the developing embryo and its surroundings (Hammond et al., 2017). Recovered cell-free DNA from BF is also believed by some researchers to be generated from apoptosis during normal embryo development (Liu et al., 2017). Should mosaic embryos shed their abnormal cells into the blastocoel cavity during development as part of a natural correction process, this could then result in a potential mismatch between the ploidy profile of the BF and its corresponding embryo. In line with the above, Tobler et al. (2015) could identify the presence of aneuploid nuclei in the BF of euploid blastocysts which harbored aneuploid blastomeres at the cleavage stage. They supported the presence of corrective mechanisms that act to marginalize aneuploid cells into the blastocoel cavity. Conversely, Zhang et al. (2016a,b) found no correlations between the amount of DNA recovered from the BF and the quality scores of corresponding embryos. It should be noted that the main limitation of published comparative studies is that they are likely to misdiagnose mosaicism, unless single-cell analyses of the whole embryo are performed individually.

It should be noted that the technical reliability of the procedure in obtaining the appropriate sample volume sufficient for successful DNA isolation and amplification has not been defined. The failed amplification rates of BF-derived DNA were shown to be disappointingly higher than the generally accepted 2% reported from TE cells (Handyside, 2016). A learning curve was proposed by some investigators (Magli et al. 2016), who demonstrated significant improvements in technical performance for BF aspiration and handling, leading to a significant reduction in amplification failure from 50% to 0% after 1 year of experience. The immediate transfer of collected samples into cold PCR tubes followed by prompt short centrifugation also were proposed to improve DNA recovery rates and optimize the amplification and molecular analysis (Magli et al., 2016).

The BF aspiration appears to represent an attractive and promising minimally invasive method for obtaining embryo-derived DNA, particularly when considering the practice of collapsing artificially expanded blastocysts before vitrification to improve success rates. It is not known, however, whether depletion of the BF would alter the cell-to-cell communication within the developing embryo and/or with its environment. Research remains inconclusive regarding whether the DNA material obtained is truly representative of the genetic constitution of the embryo as a whole.

Embryo-spent culture media

Clinical studies

In a proof-of-concept study using quantitative PCR (qPCR) (Assou et al. 2014), the investigators confirmed the presence of cell-free DNA in the culture media of Day-5/6 embryos, measuring up to 27 ng/ml DNA per sample. The researchers were successful in amplifying the multicopy gene TSPY1 on the Y chromosome enabling the segregation of embryos on the basis of gender. These findings paved the way for a novel and non-invasive approach to the pre-implantation management of sex-linked diseases. Soon after, numerous reports from different research groups followed. Wu et al. (2015) reported the successful diagnosis of α-thalassemia from 202 spent embryo culture media after qPCR analyses. The diagnostic efficiency of qPCR analyses in spent media was found to be comparable to embryo biopsies (88.6% vs 82.1%, respectively). The overall allele dropout ratio was also similar for spent media and embryo biopsies and was found to be 12%. By using multiple annealing and looping-based amplification cycles (MALBAC) for WGA, Xu et al. (2016) performed next-generation sequencing (NGS) for total 24-chromosome screening on Day-3–5-spent culture media of donated vitrified-warmed embryos. The assay demonstrated a sensitivity of 88.2% and a specificity of 84.0% in identifying chromosomal aneuploidies in culture media samples. NI-PGT was described by the investigators as being more effective in selecting chromosomally intact embryos than identifying affected ones.

Liu et al. (2017) successfully used embryo-spent culture media for aneuploidy screening and diagnosis of the beta thalassemia IVSII654 mutation. After MALBAC WGA, the DNA detection rate in spent media samples was 90.90% with a mean DNA concentration of 26.16 ng/μl. Comparing culture samples to embryo biopsies, the concordance efficiencies for euploidies and aneuploidies were 90.0 and 72.7%, respectively (Liu et al., 2017). In this study, the concordance efficiency was also calculated at a specific beta globin (HBB) gene (IVSII654) allele site in conjunction with single-nucleotide polymorphism (SNP) linkage analysis and was found to be satisfactory, albeit based on a limited number of samples.

In an attempt to improve cell-free DNA yield, Kuznyetsov et al. (2018) and Li et al. (2018) recently suggested combining embryo-spent culture medium and BF to improve embryonic DNA evaluation and demonstrated that pooled culture medium/BF samples were associated with 97.5 and 100% DNA amplification rates, respectively, after WGA. Evaluating donated embryos using NGS, Kuznyetsov et al. (2018) showed that concordance rates for whole-chromosome copy number and single chromosome ranged between 87.5% and 96.4%, when non-invasive PGT was compared with TE biopsy and whole blastocyst analyses. In contrast, Li et al. (2018), using a similar approach, found disappointedly lower concordance rates of 45 and 50% when comparisons were also made with TE biopsy and whole blastocyst. Although both investigator teams argued in favor of pooling spent culture medium and BF samples to improve DNA amplification, they strongly disagreed on the value of the technique in predicting embryo aneuploidy. A summary of studies evaluating cell-free DNA amplification and analysis methods in spent culture medium can be found in Table I.

Limitations and future research

Sample collection

It remains unclear whether the choice of a sequential or continuous culture medium system influences the yield and composition of cell-free DNA recovery. By exchanging culture media on Day-3 of in vitro development, Xu et al. (2016) were able to amplify cell-free DNA in all examined samples of culture media. The exchange of culture media during the course of in vitro incubation was deemed helpful in eliminating cumulus cells and reducing the impact of earlier maternal contamination. Lane et al. (2017), moreover, demonstrated a higher aneuploidy testing accuracy when the spent culture medium has been in contact with embryos from Day 4 to 5 compared to Day 3 to 5. This observation may be explained by the increase in the embryonic-to-maternal DNA ratio, which occurs coincidentally with the exponential rise in the embryonic cell number at blastulation. In contrast, Feichtinger et al. (2017) proposed culturing embryos in a continuous medium system until blastulation in order to improve the cumulative yield of embryo-derived cell-free DNA. These findings were further supported by Hammond et al. (2017) who detected more DNA copies in continuous culture media compared to sequential media . Nonetheless, future studies are required to investigate which sampling time points in relation to the stage of embryo development are associated with better amplification rates.

Origins of cell-free DNA

Hammond et al. (2017) quantified DNA in the spent culture media of embryos as well as in two negative controls obtained from three different commercial media brands. Controls consisted of media sampled directly from the bottle and media incubated under culture conditions similar to the embryo dishes. The findings of the study showed evidence of low-level DNA contamination in all control media, including mixed male and female nucleic acids. While Feichtinger et al. (2017) confirmed the presence of maternal contamination in spent culture media of 22 ICSI-derived blastocysts, they failed to obtain any amplifiable DNA from out-of-the-bottle culture media. More recently, Vera-Rodriguez et al. (2018) performed WGA and PGT for aneuploidies analysis on 56 human embryo TE biopsies and corresponding spent culture media. To quantify the maternal contribution to cell-free DNA, SNP sequencing was performed showing that maternal contamination was massive in most spent culture media samples. Embryonic DNA accounted for a mere 8% of the total amount of cell-free DNA isolated. Measures proposed to reduce contamination during NI-PGT include the imperative application of ‘ICSI-for-all’ to reduce paternal DNA contamination by supernumerary sperm. Other measures also comprise the scrupulous removal of all cumulus cells during the denudation process prior to ICSI to keep maternal DNA interference to a minimum.

Origins of embryonic DNA

It has been suggested that defective embryonic cells are more likely to be eliminated into the culture medium as a part of an active repair mechanism, which brings more balance to the final embryo ploidy state. Vera-Rodriguez et al. (2018) performed cell-free DNA quantification in relation to the ploidy status of embryos determined by TE biopsy findings. The contribution to cell-free DNA by cell shedding mechanisms was found to be comparable between euploid and aneuploid embryos, suggesting that DNA origin may not be the outcome of self-correction processes for aneuploidies in humans.

DNA amplification methods

WGA is the preferred method of amplification for embryonic cell-free DNA. Although multiple displacement amplification ensures a wider coverage of the genome (Zong et al., 2012), the technique is associated with non-linear amplification bias (Dean et al., 2001). While PCR-based amplification covers a narrower portion of the genome (up to 36%) (Voet et al., 2013), it is more accurate for DNA copy number analysis (Treff et al., 2010a,b). PCR-based amplification is associated nonetheless with sequence-dependent bias due to the exponential amplification of random primers (Carter et al., 1992). The use of MALBAC has been suggested as the preferred amplification approach to cell-free DNA analyses because of its higher coverage of the genome and its very low allele dropout ratio (Zong et al., 2012). The technique has been associated with high uniformity and fidelity. Compared to the normal genome coverage for single-diploid human cells (~72%), the genome coverage for spent culture media was determined by Xu et al. (2016) to be 24–65% following MALBAC amplification by high-sequencing depth of 30× reads. The copy number patterns in the spent culture samples were observed to be identical to the corresponding whole-embryo biopsies (Xu et al., 2016).

Several challenges remain to be addressed before cell-free DNA from spent culture media could become clinically useful for the genetic assessment of embryos. In the current stage of technological development, NI-PGT could be considered a screening assay for gene-level abnormalities, but not for chromosomal and segmental aneuploidies, which require additional validation.

Conclusions

Over the past few years, PGT has been increasingly utilized in ART in parallel with growing trends calling for the routine genetic testing of all embryos prior to uterine implantation. The rapid emergence of high-throughput high-fidelity molecular platforms has significantly increased the accuracy and efficiency of genetic testing. While studies have confirmed the presence of measurable cell-free DNA in the BF and spent culture media of human embryos, several technical challenges have yet to be addressed before accepting NI-PGT as a reliable source of embryonic genomic information. Hurdles include the poor representation of the embryonic genome by cell-free DNA present in the BF and spent culture media as a result of low DNA yield, poor nucleic acid integrity, maternal contamination and embryo mosaicism. Although embryonic and maternal DNA constitutional differences remain poorly characterized, these hurdles do not appear to constitute a threat to the diagnostic reliability of NI-PGT for monogenic and X-linked disorders In such cases, the only limitation could be the incomplete representation of the whole embryonic genome by cell-free DNA. The detection accuracy for chromosomal numeric abnormalities, however, is expected to be severely impaired by embryo mosaicism and maternal contamination.

Well-designed studies are required to determine how accurately cell-free DNA represents the genetic constitution of the whole embryo and to evaluate the clinical efficacy of these methods on pregnancy outcomes.

Authors’ roles

CF conceived and wrote the article. FC reviewed the literature and contributed to the writing of the article. JA contributed to the final editing and preparation of the article.

Funding

No funding was received for this paper.

Conflict of interest

None declared.

References