The aim of this study was for multiple laboratories—Takara Bio’s in-house laboratory and independent clinical laboratories A, B, and C—to test the performance of the Embgenix PGT-A Kit in detecting copy number variations (CNVs), such as whole chromosome, segmental, and mosaic aneuploidies in TE biopsy and gDNA samples.

NGS is the method of choice for PGT-A of in vitro fertilization (IVF) embryos (Vendrell et al. 2017). The overall goal of PGT-A is to improve rates of embryo implantation and live birth by enabling assessment and prioritization of embryos for transfer on the basis of copy number status in addition to other established criteria (Kemper et al. 2019). With integrated workflows from sample preparation to analysis, NGS-based PGT-A offers a simpler, more scalable solution for aneuploidy screening as compared to preceding methods.

Leveraging Takara Bio’s technologies for WGA and library preparation and a customized bioinformatics pipeline, Embgenix PGT-A is an end-to-end solution for detection of whole chromosome, segmental, and mosaic aneuploidies in human embryos. The Embgenix PGT-A Kit has been optimized to analyze DNA obtained from embryo biopsies or genomic DNA. The workflow starts with cell lysis and PicoPLEX WGA from template DNA followed by library preparation and incorporation of unique dual indexes (UDI). Following purification, quantification, and pooling, libraries are ready for multiplex sequencing on Illumina NGS platforms (Figure 1).

PicoPLEX WGA, used to generate enough DNA for NGS library preparation, is shown in Figure 2. PicoPLEX is the gold standard for PGT-A applications by virtue of its superior reproducibility, fidelity, and uniformity as compared to other WGA methods (Zhang et al. 2017). WGA products are enzymatically fragmented and then stem-loop adapters and UDI are incorporated to yield high-quality Illumina-ready libraries (Figure 3). Sequencing and analysis of the libraries enables accurate screening of all 24 chromosomes for whole- and sub-chromosomal CNVs.

Following sequencing, cloud-based Embgenix Analysis Software (Figure 4) is used to analyze sequencing data from each sample. With Human genome build 38 (hg38) as an alignment reference, CNVs are identified and displayed in CNV plots as well as idiogram and list formats. Via a web-browser, users can easily perform CNV analysis, interactively review and amend CNV calls, and generate downloadable PDF reports with a variety of customizable features.

Detection of full chromosome and segmental aneuploidies in gDNA samples 

The first study, using previously characterized gDNA samples, evaluated the feasibility and reliability of the Embgenix PGT-A Kit in detecting full-chromosome and segmental aneuploidies. The Embgenix software detected chromosomal abnormalities affecting whole chromosomes with high accuracy and confidence. An example plot resulting from this study is shown in Figure 5.

The Embgenix PGT-A Kit was applied to analyze 22 reference specimens with segmental aneuploidies. Embgenix PGT-A successfully detected 100% of the segmental aneuploidies, which ranged in size from 7.1 Mb to 91.4 Mb, with no false positives called. For 20 of the 22 segmental aneuploidies, Embgenix PGT-A provided highly accurate estimates (within 15%) of CNV size (Table 1). Figure 6 shows example plots of accurate calls by Embgenix PGT-A for relevant CNV structures requiring high resolution of segmental detection (Coriell samples NA13019 and NA14485). In some rare cases, the CNV lies across a region of the genome where it is difficult to perform read mapping, the process of aligning reads to the reference genome. These “low mappability regions” may be excluded from the reported CNV size, resulting in significant underestimation of the reported CNV size. The underestimation of size for one CNV in Coriell sample NA20556 may have arisen due to this issue, as that CNV extends into a low mappability region.

Table 1. Study with 22 samples, showing accurate reporting of segmental aneuploidies including size prediction.  

Detection of mosaicism

Next, we tested the accuracy of mosaic detection by Embgenix PGT-A. Two clinical laboratories, A and B, performed studies using mosaic samples created using mixtures of two samples with different aneuploidies. Laboratory A used mixtures of Coriell gDNA samples NA05966 (21.2 Mb trisomy in Chr14) and NA10925 (16.1 Mb monosomy in Chr7). The ratios tested are shown in Figure 7, which presents the results of the study as a comparison between Embgenix PGT-A and another NGS PGT-A method, Reproseq. Embgenix PGT-A successfully detected the segmental mosaicisms represented in the gDNA mixtures at 50% and 75%. Reproseq did not detect the mosaicisms at 50% and 75%.

The second Clinical Laboratory, B, performed a study using two gDNA samples with known CNVs. The first of these samples contained a 14 Mb segmental monosomy of Chr11. The second gDNA sample contained trisomy of Chr18 (78 Mb). Seven mixtures of these two gDNA samples were prepared. The resulting mixtures were processed using Embgenix PGT-A and Embgenix Analysis Software.

As shown in Table 2, Embgenix PGT-A successfully detected the Chr18 trisomy in all mosaic mixtures tested. Embgenix PGT-A successfully detected the 14 Mb Chr11 aneuploidy in mosaics where the sample with that aneuploidy made up at least 40% of the mixture.

Table 2. Study of mosaic mixture demonstrating detection of full chromosome aneuploidy (78 Mb) and 14 Mb segmental aneuploidy.

Multiple IVF clinics collaborated to validate the performance of Embgenix PGT-A with TE biopsy samples. The table in Figure 8, Panel A summarizes two independent datasets generated in laboratories at two separate clinics, B and C. In the study performed by Clinical Laboratory B, 34 samples (including 26 TE samples and 8 gDNA samples) were analyzed using Embgenix PGT-A with Embgenix Analysis Software and VeriSeq PGS Kit with BlueFuse Multi Software. Embgenix results were 91% concordant with results obtained from the clinic's established NGS PGT-A analysis method. The three samples generating non-concordant results were from the same embryo. Clinical laboratory B determined these samples probably came from parts of the embryo with different genetic content.

In the study performed by Clinical Laboratory C, multiple biopsies from the trophectoderm (TE) or inner cell mass (ICM) of 71 embryos were analyzed using Embgenix PGT-A and other NGS PGT-A methods and software. Embgenix results were 100% concordant with results obtained from the established NGS PGT-A analysis for each embryo. Figure 8, Panels B, C, and D  show a representative example comparing analysis of two TE samples and one ICM sample from Embryo 9. The results of analysis with Embgenix PGT-A were 100% concordant with data obtained by sample analysis with established methods and software. Figure 9 shows results for another example from this study, Embryo 47.

We have shown that the Embgenix PGT-A Kit, by incorporating PicoPLEX WGA, Illumina library preparation, and Embgenix Analysis Software for copy number assessment, provides a complete solution for PGT-A.

Our study data demonstrate that Embgenix PGT-A rapidly and accurately calls whole chromosome and segmental aneuploidies greater than 10 Mb, as well as low (30–50%) and high (50–70%) mosaicisms for defined mixtures of samples with known aneuploidies.

The results also show that Embgenix PGT-A accurately determines the size of CNVs in segmental aneuploidies involving a wide array of different chromosomal regions. Testing in independent clinical laboratories using both gDNA and TE and ICM biopsies has shown concordance between results from Embgenix PGT-A and results from established approaches for NGS-based PGT-A.

Taken together, the results clearly show the feasibility of application of Embgenix PGT-A and establish the kit as a viable solution for current and future PGT-A applications.

Sample collection

For experiments performed by Takara Bio's in-house laboratory, cell lines with defined segmental aneuploidies were purchased from Coriell Biorepository (USA) as reported in Table 3. 26 TE biopsies were obtained by Clinical Laboratory B. A total of 71 embryos were included in the study conducted by Clinical Laboratory C. For each embryo, multiple biopsies were taken from the trophectoderm (TE) or inner cell mass (ICM).

Table 3. Panel of Coriell gDNA samples containing segmental aneuploidies

Whole genome amplification

Samples were processed following the Embgenix PGT-A protocol. The kit employs a single-tube protocol for the amplification of genomic DNA starting from samples stored in liquid N2, a standard practice for WGA starting material in this field.

Sequencing

Sequencing-ready libraries were pooled and sequenced using a MiSeq® instrument with 24 samples per flow cell or using a NextSeq® instrument with 96 samples per flow cell.

Genomic data analysis and aneuploidy screening

Sequencing data were analyzed using Embgenix Analysis Software.

Kemper, J. M. et al. Preimplantation Genetic Testing for Aneuploidy: A Review. Obstet Gynecol Survey 74(12), 727–737 (2019).

Vendrell, X. et al. New protocol based on massive parallel sequencing for aneuploidy screening of preimplantation human embryos. Systems Biology in Reproductive Medicine 63(3), 162–178 (2017).

Zhang, X. et al. The comparison of the performance of four whole genome amplification kits on ion proton platform in copy number variation detection. Bioscience Reports 37 BSR20170252 (2017)