Molecular determinants of symmetry breaking in gastruloids

Embryonic development, the formation of a complex and specialised organism from a single fertilised oocyte, has fascinated scientists since centuries. Yet especially the study of mammalian embryogenesis remained challenging: crucial steps like gastrulation and organogenesis remain hidden after the embryo implants into the uterine wall. Research in developmental biology has recently been enriched by a plethora of complex in vitro models, so-called embryoids, that faithfully recapitulate different stages and processes of the first days of embryogenesis. In combination with advances in imaging and omics techniques, we now have the means to study basic principles of pattern formation, self-organisation, symmetry breaking, and cellular behaviour on the tissue, cellular, and sub-cellular level. One member of the ever-growing family of embryoids are gastruloids. These three- dimensional aggregates of embryonic stem cells undergo symmetry breaking: the polarised expression of the primitive streak marker Brachyury leads to the formation of an elongated structure with an anterior-posterior axis and derivatives of the three germ layers. While it has been shown that Wnt and Nodal signalling regulate this process and that the original culture protocol can be adapted to give rise to more specialised structures, it remains unknown how a simple aggregate of cells in uniform growth conditions is able to break symmetry in a self- organised manner. Shedding light on this question will not only help to further improve the development of embryoids in general, it may also increase our understanding of cellular behaviours during early embryogenesis. In the work presented in this thesis, we performed an image-based screen to identify symmetry breaking events in gastruloids and their molecular regulators. We perturbed thousands of gastruloids with an annotated compound library at three different time points. By analysing their phenotypic signature over time, we uncovered a previously unknown symmetry breaking event preceding the expression of Brachyury. Furthermore, we determined pathways regulating this symmetry breaking as well as primitive streak induction and elongation. To better understand gastruloid development and to identify the mechanisms of the early symmetry breaking event, we performed an imaging time course, an experiment made possible by the high-throughput culture platform developed during my PhD. We uncovered that before gastruloids are pulsed with the Wnt signalling activator CHIR99021, they establish spatial variability in Sox2 expression in an inside/outside pattern. Differences in Sox2 levels affect a cell’s ability to undergo mesendodermal differentiation, restricting Brachyury expression to the periphery of the spherical gastruloid. Single cell RNA sequencing revealed that Sox2-high cells re-express naïve pluripotency genes upon Wnt activation and that the level of Sox2 reflects the developmental progression of the cell. Finally, we showed that the establishment of Sox2 variability is essential for symmetry breaking and thus successful gastruloid formation and that it depends on FGF signalling and the formation of an extra- cellular matrix niche. In summary, this study presents an insight into the first steps of gastruloid development. It shows how the combination of quantitative methods including high-content screening, imaging, and RNA sequencing allows to uncover self-organising behaviour in complex three- dimensional models. Furthermore, the developed high-throughput culture platform and automated feature extraction from gastruloid images enables the quantitative study of gastruloid formation and sets an example for other embryoid systems