O-060 Single-cell DNA sequencing reveals that current bulk DNA sequencing methods lead to an underestimation of chromosomal mosaicism in human blastocysts

Abstract

Abstract Study question What is the incidence of chromosomal mosaicism in trophectoderm and inner cell mass from good quality human blastocysts after NGS-based single-cell molecular karyotyping? Summary answer Single-cell analysis showed that almost all human blastocysts are mosaic, involving low level mosaicism and/or abnormal cells resulting from reciprocal error events. What is known already Chromosome segregation in early human embryos is considered to be error prone. While meiotic errors affect all cells within embryos, mitotic errors lead to chromosomal mosaicism, the presence of cytogenetically different cells within an embryo. The reported incidence of mosaicism in blastocysts after bulk DNA sequencing of multicellular trophectoderm (TE) biopsies performed for preimplantation genetic testing for aneuploidy (PGT-A) varies between 2-19%. However, bulk DNA analysis can only assess net average chromosome gains or losses with a detection limit of 20-30%, and it makes distinguishing true mosaicism from technical artefact problematic. Study design, size, duration Observational study in human good quality surplus embryos which were donated for research purposes (CCMO, NL82597.000.22). The embryos were thawed and cultured until the blastocyst stage. From the blastocysts with at least a morphology grade 3BB, the ICM was dissociated from the TE. From 55 embryos, both samples were successfully disaggregated into single cells and manually placed in 384 well plates. For the cytogenetic analysis, a validated method for single-cell molecular karyotyping was used (scKaryo-seq). Participants/materials, setting, methods Embryos with at least two cytogenetically different cells were considered mosaic. To investigate if bulk DNA sequencing would have detected this mosaicism, we performed an in-silico reanalysis on our single-cell data. For each embryo, TE cells with the same mitotic abnormality were quantified. Next, we determined the number of mosaic embryos in which these abnormalities were present in at least 20% of TE cells, as a cut-off for what PGT-A methods are expected to detect. Main results and the role of chance On average 42% of the total number of cells that could be isolated per embryo were successfully karyotyped, giving a total of 1057 karyotyped cells (522 normal, 535 abnormal) from 55 embryos. Six embryos (11%) were normal, four (7%) were uniformly abnormal and 45 (82%) were mosaic. From these, 14 (26%) embryos were aneuploid mosaic with cytogenetically different abnormal cells and 31 (56%) were diploid-aneuploid mosaic with normal and abnormal cells. Here cytogenetically different abnormal cells were also frequently observed, with products of reciprocal events detected in 20 embryos. The in-silico reanalysis was performed on 38 embryos with a mosaic TE. It predicted that only 11% of the mitotic abnormalities observed in TE cells through scKaryo-seq would have been detected by bulk DNA sequencing. In 19 out of 38 embryos at least one mitotic abnormality affected more than 20% of the TE cells, meaning that only 50% of the embryos would have been recognized as mosaic with bulk DNA analysis. Bulk DNA sequencing methods used for PGT-A lead to an underestimation of mosaicism, since low-level mosaicism that affects a few cells within an embryo and products of reciprocal events without a net gain or loss remain undetected. Limitations, reasons for caution Isolation and cytogenetic analysis of viable single cells at the blastocyst stage is technically challenging and only 42% of the cells per embryo could be successfully karyotyped. Furthermore, it is unknown to what extent in vitro culture conditions and the freezing-thawing process influenced our findings. Wider implications of the findings These single cell observations support the notion that mosaicism is a common biological phenomenon in human blastocysts. Any selective mechanism eliminating embryos or cells with chromosomal abnormalities is still not fully active at this stage. The underestimation of mosaicism by bulk DNA analysis has potential implications for current PGT-A practices Trial registration number not applicable