Menu

Close

WNT signaling memory is required for ACTIVIN to function as a morphogen in human gastruloids

Anna Yoney, Fred Etoc, Albert Ruzo, Jakob J Metzger, Iain Martyn, Shu Li, Christoph Kirst, Thomas Carroll, Eric D Siggia, Ali H Brivanlou

Preprint posted on May 30, 2018 https://www.biorxiv.org/content/early/2018/05/30/333948.1

Article now published in eLife at http://dx.doi.org/10.7554/elife.38279

How can the same signalling pathway drive both pluripotency and differentiated state during development? A new study uses human gastruloids to unravel the pleiotropic nature of Activin signalling pathway.

Selected by Sundar Naganathan

Background

Precise spatiotemporal regulation of diverse signalling pathways is critical in patterning a developing embryo. A temporal signalling hierarchy was recently described in geometrically-confined human embryonic stem cell (hESC) colonies, termed as human gastruloids, where the presentation of BMP ligand led to the activation of Wnt and Nodal pathways, resulting in mesendoderm differentiation (Martyn et al., 2018). The final step in this cascade that involves the Nodal/Activin pathway signals through SMAD2/3 transcription factors. The Activin pathway is highly pleiotropic in nature such that it regulates contrasting functions during development. While before gastrulation SMAD2/3 signalling achieves maintenance of pluripotency, during gastrulation it drives differentiation and fate specification. Disentangling the pleiotropic nature of developmental pathways is a long-standing problem in embryogenesis. Yoney et al. addressed this problem in human gastruloids and showed that Activin signalling pathway drives differential outcomes depending on previous exposure to Wnt signalling.

Key findings

The authors first characterized the response of human gastruloids to different signalling ligands. While presentation of BMP or Wnt ligands to gastruloids resulted in mesendoderm differentiation, presentation of Activin failed to induce differentiation and the cells remained in a pluripotent state. Through live imaging of SMAD2 signalling as well as through RNA-seq and motif enrichment analysis they uncover two possible reasons. First, SMAD2 and downstream expression of mesendoderm differentiation markers were observed to transiently increase (for 1-2 hrs), followed by a slow decrease over the next 6 hrs. Second, even though the SMAD2 signal declined, a weak signal was observed to persist for long periods of time, which resulted in a stable induction of pluripotency markers. Together, these observations suggested that upon presentation of Activin, hESCs briefly switched on the differentiation program before returning back to their pluripotent state.

If stimulation of Activin pathway leads cells to maintain pluripotency, how can the same pathway drive mesendoderm differentiation? The authors performed interesting follow-up experiments, where they showed that if hESCs were exposed transiently to Wnt ligands, presentation of Activin robustly induced mesendoderm differentiation and pluripotency was lost. Importantly, Wnt priming did not alter the transient signalling dynamics of SMAD2, however a stable induction of mesendodermal differentiation markers was detected. Moreover, this priming effect of Wnt was abolished when hESCs were treated with an inhibitor of b-catenin that acts downstream of the Wnt pathway. These results suggest that initial activation of the b-catenin pathway through Wnt changes the response of hESCs to Activin stimulation. This study uncovers that differential outcomes from the same signalling pathway occur via lasting memories of exposure to upstream signalling pathways.

Why I chose this preprint

1. This article provides a critical breakthrough in establishing a temporal relationship between the Wnt and Activin signalling pathways that ultimately leads to mesendoderm differentiation. It convincingly shows that only cells endowed with a Wnt memory can respond to Activin and undergo differentiation. Also, these results further our understanding of pleiotropic nature of developmental signalling pathways.

2. The experimental approach adopted in this study brings out the importance of following signalling dynamics in a quantitative fashion to understand embryonic pattern formation.

3. The experiments performed in this study also establish human gastruloids derived from human embryonic stem cells as a powerful model system to study long-standing problems in embryogenesis.

Open questions

A critical observation from the authors is that upon presentation of Activin, a weak SMAD2 signal persists for long periods of time, which is important to shuttle cells back to a pluripotent state. What are the mechanisms by which SMAD2 continues to signal weakly? How do hESCs respond differently to such a small change in SMAD2 signalling?

There seems to be a competition between expression of mesendoderm differentiation markers and pluripotency markers. Activin signalling transiently increases differentiation markers but stably induces pluripotency markers. If exposed to Wnt, however, this balance is shifted with differentiation markers being stably induced. How does this competition work at the level of transcription factors that regulate expression of these contrasting sets of genes?

It is straightforward to expect that an upstream signalling pathway prepares cells to respond differently to a downstream signalling pathway, say by activating specific transcription factors. The article discussed here further confirms this idea and has conclusively established a temporal relationship between Wnt and Activin pathways during development. However, the molecular mechanisms by which the b-catenin pathway prepares cells to become differentiation competent is unclear. Furthermore, does exposure to BMP ligand lead to an outcome similar to Wnt priming? One would expect so, as BMP activates Wnt expression, but this remains to be shown.

References
1. Martyn I. et al., Self-organization of a human organizer by combined Wnt and Nodal signaling, Nature, 2018.

Tags: cellular memory, micropatterned cell culture, pattern formation, pluripotency, transcription factor dynamics

Posted on: 27th June 2018

Read preprint (2 votes)




  • Have your say

    Your email address will not be published. Required fields are marked *

    This site uses Akismet to reduce spam. Learn how your comment data is processed.

    Sign up to customise the site to your preferences and to receive alerts

    Register here
    Close