The quest to model and modulate embryonic development became a recent cornerstone of stem cell and developmental biology. Mammalian developmental timing is adjustable in vivo by preserving preimplantation embryos in a dormant state called diapause. Inhibition of the growth regulator mTOR (mTORi) pauses mouse development in vitro, yet constraints to pause duration are unrecognized. By comparing the response of embryonic and extraembryonic stem cells to mTORi-induced pausing, we identified lipid usage as a bottleneck to developmental pausing. Enhancing fatty acid oxidation (FAO) boosts embryo longevity, while blocking it reduces the pausing capacity. Genomic and metabolic analyses of single embryos point toward a deeper dormant state in FAO-enhanced pausing and reveal a link between lipid metabolism and embryo morphology. Our results lift a constraint on in vitro embryo survival and suggest that lipid metabolism may be a critical metabolic transition relevant for longevity and stem cell function across tissues.

To ensure dosage compensation for X-linked genes between the sexes, one X chromosome is silenced during early embryonic development of female mammals. This process of X-chromosome inactivation (XCI) is initiated through upregulation of the RNA Xist from one X chromosome shortly after fertilization. Xist then mediates chromosome-wide gene silencing in cis and remains expressed in all cell types except the germ line and the pluripotent state, where XCI is reversed. The factors that drive Xist upregulation and thereby initiate XCI remain however unknown. We identify GATA transcription factors as potent Xist activators and demonstrate that they are essential for the activation of Xist in mice following fertilization. Through a pooled CRISPR activation screen we find that GATA1 can drive ectopic Xist expression in murine embryonic stem cells (mESCs). We demonstrate that all GATA factors can activate Xist directly via a GATA-responsive regulatory element (RE79) positioned 100 kb upstream of the Xist promoter. Additionally, GATA factors are essential for the induction of XCI in mouse preimplantation embryos, as simultaneous deletion of three members of the GATA family (GATA1/4/6) in mouse zygotes effectively prevents Xist upregulation. Thus, initiation of XCI and possibly its maintenance in distinct lineages of the preimplantation embryo is ensured by the combined activity of different GATA family members, and the absence of GATA factors in the pluripotent state likely contributes to X reactivation. We thus describe a form of regulation in which the combined action of numerous tissue-specific factors can achieve near-ubiquitous expression of a target gene.