Neuromesodermal Progenitors As the Cellular Source of Vertebrate Segments
The continuous posterior addition of somites requires an influx of progenitor cells to sustain the somitogenesis process. Work in several vertebrate model systems, including mouse, chick, Xenopus, and zebrafish has established that cells within the posteriormost embryonic structure called the tailbud act as the progenitor pool for new presomitic mesoderm production, and subsequent somite addition (Henrique et al., 2015; Kimelman, 2016; Martin, 2016). These cells are called neuromesodermal progenitors (NMPs) and are unique in that they have not made a germ-layer decision even after gastrulation. Local signaling cues direct the fate of neuromesodermal progenitors into either neural ectoderm, where they give rise to the spinal cord, or mesoderm, the majority of which gives rise to the somites with minor contribution to the endothelial tissue of the vascular system (Henrique et al., 2015; Kimelman, 2016; Martin, 2016) (Figure 5.3).
Retrospective lineage analysis in the mouse embryo showed that single cells can give rise to both spinal cord and somites as the body axis extends, thus confirming a long-standing hypothesis that germ-layer naive cells exist in the tailbud
FIGURE 5.3 Neuromesodermal progenitors give rise to the somites. NMPs, here shown in the tailbud, make a germ-layer decision between ectoderm (spinal cord) and mesoderm. The mesoderm is further patterned into paraxial and lateral fates. The paraxial mesoderm gives rise to the somites, whereas the NMP-derived lateral mesoderm gives rise to vascular endothelium. The initial germ-layer decision is based on canonical Wnt signaling activity, with Wnt signaling promoting mesodermal fate. Continued Wnt and FGF signaling induces paraxial fate in the mesoderm, while BMP signaling induces the endothelial fate.
(T/.ouanacou et al., 2009). This work changed our understanding of lineage relationships in the early mouse embryo, demonstrating that spinal cord cells are more closely related to somite cells than to other ectodermal derivatives such as the epidermis. Several years later, NMPs were identified in the zebrafish embryo (Martin and Kimelman, 2012). Although single zebrafish NMPs rarely give rise to daughter cells in both the spinal cord and somite lineages (Attardi et al., 2018; Martin, 2016), they are competent to differentiate into either cell type, and make that decision based on local canonical Wnt signaling (Martin and Kimelman, 2012). High Wnt signaling induces NMPs to become mesoderm, whereas low Wnt signaling causes NMPs to adopt a neural fate. NMPs are defined as cells that express both the neural-inducing transcription factor sox2 as well as the mesoderm-inducing transcription factor Brachyury (discussed in more detail in Section 5.6), and a posteriorly localized Brachyury/Sox2 population of cells has also been identified in human, chick, Xenopus, and axolotl (Garriock et al., 2015; Henrique et al., 2015; Martin and Kimelman, 2012; Olivera-Martinez et al., 2012; Taniguchi et al., 2017; Tsakiridis and Wilson, 2015). Thus, NMPs are evidently a conserved cell type present in all vertebrates that contribute cells to the somites throughout axis extension.