Germline epigenetic programming, including genomic imprinting, substantially influences offspring development. Polycomb Repressive Complex 2 (PRC2) plays an important role in Histone 3 Lysine 27 trimethylation (H3K27me3)-dependent imprinting, loss of which leads to growth and developmental changes in mouse offspring. In this study, we show that offspring from mouse oocytes lacking the PRC2 protein Embryonic Ectoderm Development (EED) were initially developmentally delayed, characterised by low blastocyst cell counts and substantial growth delay in mid-gestation embryos. This initial developmental delay was resolved as offspring underwent accelerated fetal development and growth in late gestation resulting in offspring that were similar stage and weight to controls at birth. The accelerated development and growth in offspring from Eed-null oocytes was associated with remodelling of the placenta, which involved an increase in fetal and maternal tissue size, conspicuous expansion of the glycogen-enriched cell population, and delayed parturition. Despite placental remodelling and accelerated offspring fetal growth and development, placental efficiency, and fetal blood glucose levels were low, and the fetal blood metabolome was unchanged. Moreover, while expression of the H3K27me3-imprinted gene and amino acid transporter Slc38a4 was increased, fetal blood levels of individual amino acids were similar to controls, indicating that placental amino acid transport was not enhanced. Genome-wide analyses identified extensive transcriptional dysregulation and DNA methylation changes in affected placentas, including a range of imprinted and non-imprinted genes. Together, while deletion of Eed in growing oocytes resulted in fetal growth and developmental delay and placental hyperplasia, our data indicate a remarkable capacity for offspring fetal growth to be normalised despite inefficient placental function and the loss of H3K27me3-dependent genomic imprinting.

Hematopoietic stem cells emerge in the embryo from an aortic-derived tissue called the hemogenic endothelium (HE). The HE appears to give birth to cells of different nature and fate but the molecular principles underlying this complexity are largely unknown. Here we show, in the zebrafish embryo, that two cell types emerge from the aortic floor with radically different morphodynamics. With the support of live imaging, we bring evidence suggesting that the mechanics underlying the two emergence types rely, or not, on apicobasal polarity establishment. While the first type is characterized by reinforcement of apicobasal polarity and maintenance of the apical/luminal membrane until release, the second type emerges via a dynamic process reminiscent of trans-endothelial migration. Interfering with Runx1 function suggests that the balance between the two emergence types depends on tuning apicobasal polarity at the level of the HE. In support of this and unexpectedly, we show that Pard3ba – one of the four Pard3 proteins expressed in the zebrafish – is sensitive to interference with Runx1 activity, in aortic endothelial cells. This supports the idea of a signaling cross talk controlling cell polarity and its associated features, between aortic and hemogenic cells. In addition, using new transgenic fish lines that express Junctional Adhesion Molecules and functional interference, we bring evidence for the essential role of ArhGEF11/PDZ-RhoGEF in controlling the HE-endothelial cell dynamic interface, including cell-cell intercalation, which is ultimately required for emergence completion. Overall, we highlight critical cellular and dynamic events of the endothelial-to-hematopoietic transition that support emergence complexity, with a potential impact on cell fate.