Early mammalian development remains poorly understood, mainly due to the inaccessibility of embryos in utero. To overcome this limitation, recent studies have demonstrated that embryonic stem cells can aggregate and self-organize in vitro into multicellular systems resembling the post-implantation embryo during gastrulation. In particular, I focus on models known as gastruloids, derived from mouse embryonic stem cells. Gastruloids transition from a spherical and homogeneous stem cell aggregate into a 3D structured tissue, breaking their initial symmetry through tissue elongation and polarization of gene expression.
In this doctoral work, I investigated how and when spatial heterogeneities emerge during symmetry breaking. The three-dimensional geometry and high cell density of gastruloids hinder deep imaging and quantification of cell-scale processes, and the development displays high phenotypic variability. To address these issues, I developed a pipeline for 3D multiscale analysis that integrates whole-mount two-photon imaging, image processing and 3D nuclei segmentation. The aim of this high-throughput pipeline is to quantify how gene expression interplays with physical properties (cell density, proliferation, nuclei deformation) at both the cell and tissue scales. I then applied these tools to investigate the role of cell divisions during symmetry breaking, specifically testing whether: (i) proliferation correlates with spatial position or gene expression and (ii) cell divisions are oriented within the tissue.
Through the development of experimental and computational tools, this study establishes gastruloids as a relevant model for quantifying the emergence of tissue organization from homogeneous pluripotent cells and for deepening our understanding of early mammal development.