Seeing the unseen: Live imaging of late-stage preimplantation human embryos

Recent advances in live imaging have enabled real-time visualisation of human embryogenesis using light-sheet microscopy, a gentle technique that illuminates embryos one plane at a time. These tools not only reveal fundamental aspects of early development, but also have the potential to inform clinical practice in reproductive medicine.

In our recent study, my colleagues and I introduce a novel pipeline for labelling and long-term 3D fluorescence imaging of human embryos, revealing mitotic errors during the final stages of preimplantation development. A striking finding was the presence of micronuclei – small extranuclear structures containing whole chromosomes, or chromosome fragments – in cells that nevertheless remained viable, continued to divide and contributed to the developing blastocyst.

We wondered which cells tolerate these errors, and what this might mean for embryo development. By tracking individual cells across development, we discovered that some abnormalities are confined to cells destined to become the trophectoderm, which will give rise to extraembryonic tissue (specifically the placenta). Despite mis-segregated chromosomes and micronuclei, these cells continued to proliferate and contribute to blastocyst formation.

This suggests that chromosomal instability is tolerated in extraembryonic lineages, and may not reflect the chromosomal status of the entire embryo. These findings suggest that preimplantation genetic testing for aneuploidy (PGT-A) may overestimate chromosomal abnormalities (for example, in the case of mosaic aneuploidy), a consideration that could influence clinical decisions about embryo transfer (see BioNews 1312). An important next step will be to label cells fated to give rise to the embryo proper, to determine whether similar rates of late-stage chromosome segregation errors arise, and whether these mis-segregated cells are tolerated or selectively eliminated.

Tracking chromosome behaviour in live human embryos is challenging. Prior studies have relied upon microinjecting messenger RNA (mRNA) to transiently express fluorescent proteins. This approach is limited to either the one-cell (zygote) stage of development (as described in this 2022 study), or the two-cell stage (as described in this 2024 study), because each cell must be injected individually (see BioNews 1239). Moreover, the fluorescent signal from injected mRNA fades before the embryo reaches the blastocyst stage, restricting long-term observations.

Chemical dyes offer a possible alternative, allowing structures such as DNA, actin, microtubules, mitochondria, lysosomes and the plasma membrane to be labelled simply by adding the dye to the culture medium. While useful for short-term imaging, these dyes are not protein-specific and they can be toxic with prolonged exposure, making them unsuitable for longer-term imaging.

Our method overcomes these limitations by electroporation of blastocyst-stage embryos, where brief low-voltage electrical pulses are used to create tiny pores in the cell membrane in order to allow the introduction of mRNA encoding a histone H2B-fluorescent protein fusion. This method – combined with light-sheet microscopy – allowed high-resolution imaging of human embryos every 15 minutes for up to 48 hours, while maintaining embryo viability. The combination of electroporation and low-phototoxicity light-sheet imaging enables comprehensive visualisation of developing human blastocysts, with minimal disturbance.

Segmentation and tracking of nuclei in human embryos permitted analysis of cell cycle dynamics and division orientation. Human and mouse embryos exhibited comparable mitotic durations, but interphase in human embryos lasted approximately twice as long. Trophectoderm nuclei aligned in parallel to the embryo surface, ensuring that subsequent divisions were perpendicular to the surface, maintaining the stability and proper function of epithelial layers as well as how the different cell types within a tissue are arranged and maintained.

Across hundreds of recorded mitoses, our study identified a higher frequency of chromosome misalignment in human embryos compared to mouse embryos (approximately eight percent vs four percent). Abnormalities in the structure or organisation of the mitotic spindle likely underlies these errors.

Mis-segregated chromosomes formed micronuclei, which were typically not reincorporated into the nucleus in the next division. Instead, micronuclei were inherited by one daughter cell, and occasionally remained adjacent to the main nucleus.

Trophectoderm cells containing micronuclei continued to divide and contribute to blastocyst formation, suggesting a remarkable tolerance for chromosomal instability (where a cell's chromosomes are not stable, and as such are frequently gained, lost or rearranged when the cell divides) in this extraembryonic tissue. This could, in turn, buffer potential developmental defects in embryonic tissues. Extending the culture and time of imaging will be important to determine the extent of this contribution, and its persistence in the developing placenta.

There are limitations to our approach. Electroporation at later stages results in uneven labelling of trophectoderm cells, which take up the fluorescent marker more effectively than cells in the inner cell mass. This variability currently restricts detailed analysis of mitoses in inner cells.

Nevertheless, the platform is adaptable. It can be expanded to label other structures such as the nuclear envelope, cytoskeletal components, adhesion molecules and polarity proteins. These capabilities open the door to exploring fundamental questions about chromosome segregation, micronuclei formation and early lineage specification in the human embryo.

In summary, combining mRNA electroporation with low-phototoxicity light-sheet imaging allows unprecedented visualisation of human blastocyst development, enhancing our understanding of human embryogenesis. This in turn can potentially inform clinical practice and embryo selection strategies, illuminating the earliest events of human development.

---

The usefulness of PGT-A will be debated at this year's PET Annual Conference – What Does Genomics Mean for Fertility Treatment? – at a session entitled 'PGT-A as an IVF Add-On: 25 Years of Controversy'.

This is an in person only conference, taking place in central London on Wednesday 10 December 2025. Other sessions will explore issues including expanded carrier screening, polygenic risk scores and donor conception. Find out more and register here.