Mouse embryonic stem cells self-organize into trunk-like structures with neural tube and somites

Background: 

Stem cell based in vitro differentiation models have emerged as powerful tools to dissect genetic networks and signaling pathways involved in the development of various cell types during embryogenesis. Protocols have been developed to differentiate pluripotent stem cells (PSCs) into several embryonic and adult cell types, by providing the cocktails of growth factors or their agonists and antagonists to mimic the in vivo signaling environment. Historically, most of the differentiation methods were established by culturing PSCs on specific substrate in 2D or by formation of 3D embryoid bodies. Though largely beneficial, these systems have limited applications. 2D systems lack the complexities of an embryo with respect to cellular organization. Though embryoid bodies are multicellular, they lack embryo like organization and are prone to experimental variation. Over the last decade, advances in 3D culture technologies have led to the emergence of several organoids developed from tissue resident stem cells or PSCs. For example, in a recent study, mouse embryonic stem cells have been used to develop multi-axial structures called gastruloids, in vitro structures exhibiting several features of mammalian gastrulation. In gastruloids, mouse ESCs differentiate into multi-lineage derivatives which organize into different compartments, thereby partially mimicking embryonic axis formation and multicellular organization. Gastruloids also undergo elongation in rostro-caudal axis, which is a major hallmark of gastrulation dependent posterior growth in the embryos. However, a major limitation in the existing gastruloid model is the lack of morphogenetic events following fate segregation[1]. For example, existing gastruloids lack the capability to form neural tube and somites. In this preprint, Veenvliet and Bolondi et al., show that embedding the gastruloids in extra-cellular matrix compounds efficiently induces several morphogenetic events to form an embryonic trunk-like structure.

 

Major findings: 

Gastruloids embedded in extra-cellular matrix induces embryo-like morphogenesis 

In this study, the authors embed 96hrs gastruloids in Matrigel and culture them for additional 24hrs. This induces embryo-like morphogenesis in the gastruloids to form trunk-like-structures (TLS). Marker and reporter-based expression analyses reveal that TLS exhibit cell types derived from all the three germ layers as previously observed in gastruloids. These include presomitic mesoderm, spinal cord progenitors and gut progenitors from mesoderm, ectoderm and endoderm, respectively. Strikingly, TLS also exhibit several embryo-like morphological features such as formation of the neural tube and segmented blocks of somites. Notably, these morphological features are absent in previously reported gastruloids. Comparative transcriptome analysis between gastruloids and TLS reveal the global similarity between gastruloids and TLS at the level of transcription. However, computational analysis revealed major differences in expression of genes involved in embryonic morphogenesis and other related processes. With in-depth analysis, the authors demonstrate that embedding of gastruloids in Matrigel induces the expression of genes involved in cell-cell adhesion and cell-matrix adhesion, suggesting the potential mechanism underlying the induction of morphogenesis and formation of TLS.

Cells in TLS follow embryonic developmental trajectories to acquire post-occipital trunk-like identity 

In developing embryos, the germ layer derivatives located at different position along the anterior- posterior axis follow distinct developmental trajectories and obtain compartment-specific characteristics. To get further insights into characteristics of cell types in the TLS, the authors analyze the expression profile of a wide panel of developmental genes. This analysis finds high correlation between TLS and post-occipital stage (trunk region) of E8.5 mouse embryos. Of note, in the embryos, post-occipital somites and neural tube emerge from a subset of axial progenitor population called neuromesodermal progenitors (NMPs). Single-cell sequencing of TLS identifies the presence of NMPs in them, which further supports their post-occipital identity. Furthermore, deep computational analysis of single cell sequencing data, at three different time points of differentiation of TLS, reveals the striking overlap in the genetic trajectory of differentiation of NMP into neural or somitic lineage between TLS and the E8.5 embryos. To validate the developmental authenticity of the model, authors have tested the ability of TLS to recapitulate the developmental phenotype caused by the mutation of a key mesoderm specification gene called Tbx6. Mutation of Tbx6 leads to fate transformation of the paraxial mesoderm into the ectopic neural tube in the mouse embryo. Similar to the embryos, TLS generated with Tbx6-/- cells form ectopic neural tubes at the expense of paraxial mesoderm and somites. Overall, these analyses suggest that the formation of cell lineages in TLS follow embryonic developmental pathways, thus providing a powerful experimentally amenable and physiologically relevant model to address in depth research questions.

What I like about this preprint: 

Inducing embryonic morphogenetic events in a reproducible manner in vitro has been a tough challenge. This study achieves to recapitulate complex morphogenetic events associated with mouse axis elongation and trunk formation with a relatively simple alteration to the existing in vitro model of gastruloid culture. Another strength of this study lies in the developmental authenticity of the cell lineage specification. Considering these two points, TLS provides a simple yet powerful, easy-to- handle and experimentally amenable tool to address complex research questions associated with mouse gastrulation. For example, this model would allow to decipher the physical and molecular basis of complex cellular interactions associated with morphogenesis using live imaging, high resolution genetic and physical manipulations.

Questions to the authors: 

1. Does segmentation of somites in the TLS follow similar oscillatory mechanisms as that of embryos? Do somites have intrinsic size control mechanism?

2. The transitions from early to late NMPs have been associated with major change in their developmental potential. It has been suggested that it correlates with trunk-to-tail transition, which in turn associates with larger scale events such as regression of endoderm and formulation of neural rube via secondary neurulation[2]. Since there is a shift from early to late NMPs by 108hrs TLS, do authors observe the formation of tail like structures as well?

References: 

1. Beccari, Leonardo, Naomi Moris, Mehmet Girgin, David A. Turner, Peter Baillie- Johnson, Anne-Catherine Cossy, Matthias Lutolf, Denis Duboule, and Alfonso Martinez Arias. “Multi-axial self-organization properties of mouse embryonic stem cells into gastruloids.” Nature 562, no. 7726 (2018): 272-276.

2. Aires, Rita, Luisa de Lemos, Ana Nóvoa, Arnon Dias Jurberg, Bénédicte Mascrez, Denis Duboule, and Moisés Mallo. “Tail bud progenitor activity relies on a network comprising Gdf11, Lin28, and Hox13 ” Developmental cell 48, no. 3 (2019): 383- 395.

 

doi: https://doi.org/10.1242/prelights.17778

Read preprint (4 votes)