Research

Fate decisions and organization in the preimplantation embryo

Embryonic development is a fascinating process during which different cell types are made and organized into complex and functional structures, ultimately building an entire organism. When observing embryonic development, patterns emerge with remarkable reproducibility. However, focusing on the level of individual cells, especially during mammalian development, often reveals a seemingly chaotic scene. Cells divide, shift positions, change shapes, interact with different neighbors and environments, and exhibit fluctuations in gene expression. Yet from this messy situation, robust and reproducible patterns emerge. How, then, are the correct cell types produced at the right time, in the right location, and in the right proportions?

To address these questions, we take advantage of the simplicity and accessibility of preimplantation development, during which three cell types, the epiblast (EPI), the trophectoderm (TE) and the primitive endoderm (PE) arise from two consecutive decisions and self-organize to construct the blastocyst. EPI cells will give rise to the fetus, while TE and PE cells will make extraembryonic tissues, such as the placenta and the yolk sac endoderm, respectively, to support fetal development following implantation. We establish genetically encoded fluorescent reporter mouse lines for key factors involved in these fate decisions and use a combination of cutting-edge live imaging and computational image processing to visualize and probe the mechanisms of cell fate acquisition with unprecedented spatial and temporal resolution.

Using these powerful tools, we investigate the mechanisms driving initial symmetry-breaking in the embryo, how cells interpret geometrical cues and signaling inputs to instruct cell fate specific gene expression patterns and how cell fates are propagated over space and time to achieve reproducible cell-type proportions (Chalifoux and Avdeeva Dev Biol 2025; Avdeeva and Chalifoux Development 2025; Kim-Yip and Denberg Current Biology 2025).

Overview of preimplantation development in mouse.

Live imaging epiblast and primitive endoderm segregation during 2 days of development. Embryo expressing H2B-miRFP720; Gata6-GFP (PE transcription factor); Nanog-mCherry (EPI transcription factor). Merge of two transcription factor reporters (top right). Kim-Yip and Denberg Current Biology 2025.

Live imaging germline cyst divisions and fractures in the embryonic ovary. Levy and Leite Current Biology 2024.

Germline cyst formation and function

The formation of animal gametes commonly proceeds through a stage of cysts where future germ cells are connected by intercellular bridges (ICBs), derived from arrested cytokinetic furrows. These connections enable sharing of cytoplasmic components between germ cells and, in the female germline, enrich select cells in the cyst to become the oocyte(s). In some organisms, like Drosophila, each germline cyst has the same number of cells and the same pattern of cell-cell connections. Mammalian germline cysts, on the other hand, have variable size and structure. While the formation of invariant germline cysts is relatively well understood, the origins of variable clusters remain largely unknown. We have developed a novel high-resolution live imaging approach to study the dynamics of germline cyst formation in mice (Levy and Leite, Current Biology 2024). Using quantitative image analysis, genetic and pharmacological perturbations and modeling we revealed that cyst formation is accompanied by dynamic and undirected protrusions of germ cells comprising the cyst. As a consequence, the size of a mammalian cysts is dictated by the tug- of-war between intercellular bridges and random cellular protrusions. This work not only reveals the mechanistic basis for generating variability in germline cysts, but also introduces a novel live imaging approach to study fundamental questions in germ cell biology and oocyte formation in mammals.

Development of live imaging and genome engineering tools

Quantitative live imaging is a technique revolutionizing developmental biology and has recently gained momentum by advancements in microscopy. The current bottlenecks for this approach, especially in mammals, are a) the generation of genetically modified animal models suitable for visualizing and manipulating different factors during live imaging studies and b) suitable computational tools to quantitatively analyze the vast datasets generated.

We have contributed towards solutions for both these limitations by a) developing two CRISPR-based genome editing methods; the first allowing us to generate knock-in reporter lines to visualize developmental dynamics in embryos (Gu and Posfai, Nature Biotechnology 2018) and the second enabling us to introduce small programmed edits into the embryo genome in a highly efficient and precise manner that allows for phenotype analysis in the same generation (Kim-Yip and McNulty, Nature Biotechnology 2024), and by b) establishing a semi-automated image analysis pipeline with collaborators that enables the extraction of quantitative measurements of time-lapse datasets (Nunley et al., Development 2024).

In collaboration with the Shvartsman group we establish robust 3D nuclear instance segmentation and lineage tracking for time lapse datasets of developing preimplantation embryos. Nunley et al., Development 2024.

2C-HR-CRISPR microinjections. Gu and Posfai, Nature Biotechnology 2018.