An endodermal subpopulation gives rise to neuromesodermal progenitors in the posterior chick embryo

Background

The progressive lineage restriction of cells on their way to becoming part of a functional tissue was considered a core principle in developmental biology. During gastrulation, pluripotent cells decide whether to become ectoderm, mesoderm or endoderm, and from there each germ layer differentiates into progressively more restricted cell types.

However, the discovery of neuromesodermal progenitors (NMPs), residing at the posterior end of the elongating body axis, challenged this idea (1). NMPs are bipotent cells that contribute to both neural and mesodermal tissues. They are usually defined by co-expression of the mesodermal marker TBXT and the neural marker SOX2, but their exact definition – and in some cases, their existence – remains hotly debated. The existence of a truly bipotent mesendoderm cell is also likewise debated, and a consensus is yet to emerge (2,3).

In both cases here, the question is whether bipotent cells are specified during gastrulation. Far less attention has been paid to the capacity of the endoderm to cross germ layer boundaries after its initial specification. The preprint highlighted here provides surprising evidence that endoderm-derived cells contribute to the NMP population and may help drive body axis elongation.

Key findings

Endoderm cells acquire mesenchymal properties

At the ventral surface of Hensen’s node lie progenitors of the definitive endoderm, thought to contribute primarily to the gut and associated organs. During body axis elongation, the node regresses from anterior to posterior, eventually becoming continuous with the tail bud, which then drives further elongation through posterior growth. Using electroporation to selectively tag the ventral surface of the chick embryo with GFP, the authors observed that once node regression is complete, cells of the ventral node migrate dorsally into the mesenchyme, settling into sites outside the gut. They then tested the cells for several hallmarks of EMT, finding that they indeed become mesenchymal as they ingress.

Next, the authors asked whether these cells are truly of endodermal origin. They again tagged the ventral tissue with GFP, but now combined this with single-cell RNA-seq at three stages. At early stages, the ventral node cells were found to express endodermal genes such as SOX17. By stage HH10, 97% of GFP+ cells fell within a distinct endoderm cluster, but had already lost expression of SOX17, suggesting that endodermal identity is lost as cells begin their ingression.

Endoderm to NMP transformation

But what identity do they adopt? At stage HH15, some GFP+ cells were found within clusters identified as NMPs and paraxial mesoderm. Co-staining for SOX2 and TBXT revealed that most of the GFP+ cells in the tailbud expressed both markers, the classic signature of NMPs. Together with their finding that the SOX2 N1 enhancer – another marker used to distinguish NMPs – is active in these cells, the case that ventral node endoderm cells adopt an NMP identity becomes quite convincing.

The authors then showed that there are more than just transcriptional similarities. By labelling the ventral node endoderm with GFP and the caudal lateral epiblast – where traditional NMPs derive from – with RFP, they demonstrated that by HH15 these populations converge at the chordoneural hinge – where NMPs are known to reside. At later stages, descendants of both cell populations were found in the somites, as expected for NMPs, as well as a few in the neural tube and notochord. Using a dominant-negative FGFR targeted to the ventral node endoderm, the authors showed that FGF signalling is required for the ingression of these cells, and that blocking it slowed down axial elongation.

Endoderm cells give rise to multipotent progenitors

Next, they asked whether individual ventral node endoderm cells are producing descendants that span across germ layers. They used TrackerSeq lineage barcoding to tag individual cells in the posterior endoderm with GFP and a unique, heritable barcode at stage HH10, then sequenced at HH18. Crucially, they identified node endoderm clones with descendants in the mesoderm, endoderm, ectoderm or NMP clusters, with some clones spanning more than one germ layer – providing good evidence that the node endoderm contains both fate-restricted and bipotent progenitors.

Finally, since both node endoderm and traditional NMPs contribute to the somites, the authors asked if they take a similar molecular route to get there. Using computational trajectory inference, they identified putative lineage drivers of somitogenesis unique to either the node endoderm or the NMP trajectory, suggesting they arrive at the same fate through different routes. The node endoderm-specific drivers were active earlier, leading the authors to speculate that this may reflect the erasure of endodermal identity before the transition towards somite fate begins.

Why is this work important?

This work is valuable because it provides more evidence that challenges the preconceived notions of lineage restriction. Differentiation is traditionally seen as a progressive narrowing of potential, with NMPs an exception to the rule, but this study suggests that perhaps germ layers shouldn’t be thought of as lines that can’t be crossed, and that there may be more examples of seemingly lineage-restricted cells crossing germ layer boundaries.

Why did I choose to highlight this preprint?

As a scientist with an interest in the organiser, this paper conjures up many questions and ideas. During my PhD work, I did a lot of organiser ablation experiments, and I was stunned by the embryo’s ability to regulate, and also by the lack of true definition between ‘cell type’ clusters in single-cell datasets at early stages. This preprint builds on this plasticity of cell identities during early development, which could be an important evolutionary adaptation. Having multiple contributions to an essential process like axial elongation could improve its robustness against perturbation; however, this would require the NMP pool to compensate when the endoderm-derived pool is lacking.

Recent work has revealed the existence of a stem cell niche in the posterior part of Hensen’s node that generates notochord, somite and floor plate progenitors as the axis extends (4). The organiser is often described as a position in the embryo, rather than a fixed cell population (5) that can impart resident, self-renewing behaviour onto cells that enter into it – through normal cell movements, or via a graft – allowing them to contribute to axial tissues (4). At later stages, self-renewing, resident axial progenitor cells have also been found in the chordoneural hinge (6,7). Oikonomou et al. don’t describe their cell population as stem cells, probably because their cells ingress and leave, rather than remaining resident, and their single timepoint barcoding cannot show that individual node endoderm cells self-renew while contributing daughters to the axis. However, it would be interesting to see whether the ventral node environment is instructive enough to impose an endodermal identity and a subsequent endoderm-to-NMP transition, mirroring how the node imparts resident behaviour to cells grafted into it.

Questions for the authors:

1. What happens to overall axis length when ventral node endoderm ingression is blocked?
2. How do the multipotent cells in the node endoderm differ from node stem cells?
3. What do you think these results tell us about the node: might it have a broader function in promoting plasticity and/or stemness?
4. What proportion of the NMP pool do you think derives from node endoderm cells? Do you think there could be more contributing cells coming from elsewhere?
5. Do endoderm-derived NMPs contribute to specific somites, or specific parts of the somite? And do you think that these cells are any different once they have formed the mature somite, compared to those that developed the ‘normal’ way?

Bibliography:

1. Tzouanacou E, Wegener A, Wymeersch FJ, Wilson V, Nicolas JF. Redefining the Progression of Lineage Segregations during Mammalian Embryogenesis by Clonal Analysis. Developmental Cell. 2009 Sep 15;17(3):365–76. doi:10.1016/j.devcel.2009.08.002 PubMed PMID: 19758561.
2. Probst S, Sagar, Tosic J, Schwan C, Grün D, Arnold SJ. Spatiotemporal sequence of mesoderm and endoderm lineage segregation during mouse gastrulation. Development. 2021 Jan 7;148(1):dev193789. doi:10.1242/dev.193789
3. Masamsetti VP, Salehin N, Kim HJ, Santucci N, Weatherstone M, McMahon R, et al. Lineage contribution of the mesendoderm progenitors in the gastrulating mouse embryo. Developmental Cell. 2025 Jul 21;60(14):1991-2006.e9. doi:10.1016/j.devcel.2025.02.015 PubMed PMID: 40132585.
4. Solovieva T, Lu HC, Moverley A, Plachta N, Stern CD. The embryonic node behaves as an instructive stem cell niche for axial elongation. Proceedings of the National Academy of Sciences. 2022 Feb;119(5):e2108935119. doi:10.1073/pnas.2108935119
5. Joubin K, Stern CD. Molecular Interactions Continuously Define the Organizer during the Cell Movements of Gastrulation. Cell. 1999 Sep 3;98(5):559–71. doi:10.1016/S0092-8674(00)80044-6 PubMed PMID: 10490096.
6. McGrew MJ, Sherman A, Lillico SG, Ellard FM, Radcliffe PA, Gilhooley HJ, et al. Localised axial progenitor cell populations in the avian tail bud are not committed to a posterior Hox identity. Development. 2008 Jul 1;135(13):2289–99. doi:10.1242/dev.022020
7. Solovieva T, Wilson V, Stern CD. A niche for axial stem cells – A cellular perspective in amniotes. Developmental Biology. 2022 Oct 1;490:13–21. doi:10.1016/j.ydbio.2022.06.015

Tags: axis elongation, body axis, cell fate, chick, chicken, development, embryo, endoderm, gallus gallus, lineage tracing, nmps