Pluripotent stem cell derived embryo models (SEMs) are an emerging research platform that has demonstrated the ability to capture the developmental architecture and dynamics of mammalian embryonic development. Human and mouse naïve and naïve-like ESCs (nESCs) have both proven their capacity to generate post-implantation SEMs, containing the embryonic and extraembryonic compartments present in post-implantation embryos: the primitive endoderm (PrE) that gives rise to the yolk-sac, and the trophectoderm (TE) compartment that gives rise to the placenta.
Mouse complete SEMs are generated following the separate induction of PrE from nESCs via transient ectopic expression of Gata4 (major PrE promoting transcription factor), and TEs that are generated following Cdx2 (major TE promoting transcription factor) transgenic overexpression in nESCs. The separate induction of PrE and TE from nESCs, is followed by a cell mixing step in which unperturbed nESCs fraction (that gives rise to the epiblast (Epi) and embryo proper) is mixed with the two induced TE and PrE cell fractions at an optimal ratio, which kickstarts the nucleation of cell-sorting events, leading to self-organization of structures mimicking post-implantation mouse embryos. Mouse complete SEMs were able so far to reach a developmental stage equivalent to E8.5 following exposure to optimized ex utero static and dynamic culture devices and growth conditions that were first optimized to support natural mouse embryos at equivalent developmental stages. These studies highlighted the notion that mouse naive pluripotent stem cells can retain sufficient plastic potential, and with the help of ectopic transgene expression, can be used to build an entire model of a developing embryo that can proceed through and beyond gastrulation.
Human complete SEMs could also be induced from human naïve and naïve-like ESCs, however this could be achieved without the use of transgenes. The separate induction of nESCs towards three different extra-embryonic lineages was done merely by signaling pathway modulation with cytokines and small molecule inhibitors. Once the TE and PrE/ExEM (extra-embryonic mesoderm) cell fractions were separately induced, they were mixed at an optimized ratio with the Epi cells and subjected to optimized ex utero culture conditions that can support their growth to developmental stages equivalent to day-14 post fertilization. The fact the mouse SEM generation involves transgene ectopic expression, poses a technical challenge, as well as a scientific one, raising the question whether mouse nESCs can be coaxed without the use of transgenes to generate competent extra-embryonic cell fractions capable of generating mouse SEMs, equivalent to those previously obtained, which we sought to explore in this study.
Moreover, when considering human and non-human primate protocols to generate blastoids from nESCs, which represent embryo models for the pre-implantation blastocyst stage, those protocols do not involve separate induction of extra-embryonic lineages followed by mixing the different induced cell fractions. But rather, the starting nESCs are exposed to a mix of signaling molecules that preserves an Epi fraction, yet induces a portion of the cells into PrE or TE fate within the starting same cell aggregate originating from nESCs (we refer to this as “in aggregate induction”). Thus, the latter protocols do not involve an active mixing stage of different and separately induced cell lineages, which is technically more cumbersome and possibly compromises the quality and efficiency of embryo models obtained. Thus, in this study we set out to explore whether and how an alternative mouse SEM protocol can be generated while achieving (i) in-aggregate multilineage induction (i.e. avoiding separate induction of the different extra-embryonic lineages followed by their re-mixing) and (ii) circumventing the use of transgenes to induce PrE and TE from mouse nESCs, thus leading to the generation of transgene-free SEMs (TF-SEMs).
