Dissecting lineage specification in EpiSC and neuromesodermal progenitor cultures

Abstract

During mouse embryo gastrulation, the pluripotent epiblast gives rise to the three embryonic germ layers, the ectoderm, mesoderm and endoderm. After somitogenesis begins and pluripotency disappears from the epiblast, bipotent neuromesodermal progenitors (NMPs) drive axis elongation, contributing to the formation of the posterior nervous system, as well as the axial and paraxial mesoderm. Early NMPs arise in the E8.5 mouse embryo, in and near the primitive streak, while late NMPs are found in the tail bud (E9.5 - E13.5). NMP regions are characterized by coexpression of Tbra (Brachyury) and Sox2. Sox1, another neural related transcription factor, has also been detected in NMP regions. Importantly, it has been shown that Sox1 expression increases as NMPs transit from the primitive streak to the tail bud stages. Mouse epiblast derived stem cells (EpiSCs) recapitulate the properties of the post-implantation epiblast and therefore serve as a good in vitro system for the study of early lineage specification events. EpiSCs express pluripotency factors and early differentiation markers, including Sox2, Sox1 and Tbra. Based on studies reporting that EpiSC cultures contain distinct subpopulations that have progressed further into lineage specification, I analyzed the properties of the Tbra expressing EpiSCs and by dissecting their expression profile, I assess whether these cells are pluripotent or they have progressed further into lineage specification, possibly into an NM fate. I show that EpiSC cultures include a large fraction of Tbra/Sox2 double positive cells; however, Nanog expression was detected in the vast majority of Tbra+/Sox2+ EpiSCs suggesting that most of the Tbra+ cells are pluripotent rather than bipotent NMPs. Using a previously published Tbra-GFP reporter cell line, I present that Tbra-GFP+ cells constitute a dynamic fraction of the culture that has not exited pluripotency (as shown by expression of the pluripotency markers), but have adopted an early primitive streak-like character. Similar to the cells of the posterior epiblast, these EpiSCs are in a reversible state and they retain their ability to undergo neural differentiation. In contrast to the overlap of Tbra and Sox2 positivity in self-renewing EpiSCs, it has been shown that Tbra expression is mutually exclusive with expression of Sox1-GFP, that seems to mark a distinct subpopulation with neural-like characteristics. In vitro NMPs can be generated from EpiSCs upon treatment with Fgf2 and the Gsk- 3 antagonist/Wnt agonist CHIRON99021 (FGF/CHI). In these conditions, 80% of the culture becomes Tbra+/Sox2+. Given that Sox1 is present in NMP regions in vivo, I hypothesized that the NMP cultures could contain Tbra+Sox1+ NM bipotent cells. Most importantly, the upregulation of Sox1 at the tail bud stages drove the hypothesis that Sox1 expression could mark the transition from an early- to a late-like NMP state in vitro. In this study, using a Sox1-GFP reporter cell line, I show that Tbra/Sox2/Sox1-GFP triple positive cells emerge in FGF/CHI treated EpiSCs. Importantly, Sox1-GFP+ cells express NMP markers and are enriched in transcripts of Hox genes. The expression profile of Sox1-GFP+ cells resembles the alteration of Hox gene activation that takes place in the caudal progenitor regions during the transition from early NMPs (E8.5) to late NMPs (E9.5-10.5) and hence supports the hypothesis that Sox1-GFP marks NMPs that correspond to the axial progenitors found at tail bud stages. Although the gene activity observed in the Sox1-GFP+ subpopulation correlates with the NM developmental potential, these cells exhibit strong neurogenic capacity, while evidence for their ability to give rise to mesoderm differentiation products is still lacking. Since Tbra and Sox1/Sox2 are not expressed in NMP regions exclusively, but also in mesoderm and neural fated tissues respectively, double rather than single reporter cell lines would be more suitable tools for tracking and isolating bipotent NM progenitors in vivo and in vitro. Here, I present the CRISPR/Cas9-mediated generation of a reliable Tbra-GFP reporter ES cell line that in contrast to the one published before, contains both endogenous Tbra loci intact. By targeting the Sox2 locus in the Tbra-GFP ES cells, I generated a Tbra-GFP/Sox2-tdTomato double reporter ES cell line, that in the future, could help us to dissect the molecular mechanisms underlying the self-renewal and differentiation of NMPs.