A novel mechanism of gland formation in zebrafish involving transdifferentiation of renal epithelial cells and live cell extrusion

Richard W Naylor, Alan J Davidson

Preprint posted on June 22, 2018

Article now published in eLife at

A change of identity drives organogenesis: a rare event of transdifferentiation in zebrafish.

Selected by Giuliana Clemente

Categories: developmental biology

Context and Background:

Can the differentiated state really be considered terminal and irreversible? This was the challenge that Helen Blau took up in the mid-80s when, inspired by the work of John Gurdon, she set up to unhinge the dogma of terminal differentiation according to which the gene expression profile of a specialised cell is irreversible.

Blau was the first researcher to talk about cell plasticity. She proposed that the differentiated state, albeit stable, is not fixed and that cells can potentially change identity by re-activating and/or silencing specific sets of genes in a dynamic fashion.

This new concept has two important implications. First, the differentiated state is actively controlled and a constant regulation holds it in place. Secondly, terminally differentiated cells can acquire a new fate, adapting to changes in the environmental conditions, in a process known as transdifferentiation (1,2).

From an experimental point of view, the idea of plasticity led to the reasoning that somatic cells can be forced to reverse back to a stem cell-like state or into a different cell lineage by playing around with the expression levels of transcription factors. For example, in 1987 Tapscott and colleagues reported that expression of MyoD in fibroblasts was enough to force cells directly into the muscle cell lineage (3). Almost 20 years later, Yamanaka was able to revert skin cells of adult mice into pluripotent stem cells (iPS) by using a cocktail of the 4 transcription factors Sox-2, Klf-4, Oct-3/4, c-Myc (4). This pioneering work, that won Yamanaka the Nobel Prize, revolutionised the field of regenerative medicine, paving the way to more personalised therapies, disease modelling and drug discovery.

Although it is possible to achieve transdifferentiation in vitro, examples of this process in an in vivo context are rare and they normally occur upon injury or in pathological settings such as cancer. Even more unusual are examples of transdifferentiation during development and the extent to which the process contributes to organogenesis is still somewhat elusive.

Key findings:

In this preprint, Naylor and Davidson describe a rare event of transdifferentiation involved in the formation of the endocrine gland Corpuscles of Stannius (CS) in zebrafish. They show that the CS originates from a subset of cells of the distal early (DE) renal tubular epithelium (Figure 1A). The process of CS formation can be divided into two steps: the initial transdifferentiation of the DE cells into CS cells and the subsequent extrusion of the CS cells from the renal epithelium. The change in cell fate is under the control of the Notch signalling pathway. Notch promotes the switch from DE to CS lineage by promoting the export of Hnf1b, master regulator of the renal fate, from the nucleus to the cytoplasm (Figure 1B). This is the first time that nuclear export and sequestration of a transcription factor are shown to drive transdifferentiation. This working model is supported by experimental data obtained by nicely combining genetic and chemical approaches. Specifically, the authors show that morpholino or CrispR-Cas9 for irx3b (downstream effector of Hnf1b) result in an increased number of stc1+ cells and decreased nuclear localisation of Hbn1b. This observation suggests that the Hbnf1-irx3b network negatively regulates transdifferentiation. Conversely, the researchers observed a reduction in the number of CS cells specified in mindbomb mutants, defective in Notch, as well as upon the treatment with the g–secretase inhibitor cpdE, a result that places Notch as positive regulator of the DE-CS transition. As a final proof, an epistasis experiment was performed which consisted in treating irx3b crispants with cpdE. The inhibition of Notch rescued the loss of irx3b, resulting in less CS cells and the retention of Hnb1f in the nucleus.

The second step of the process is the extrusion of the CS cells from the renal tubule. The extrusion, which is driven by acto-myosin dependent apical constriction of the CS cells as proved by Blebbistatin treatment, is later followed by cell proliferation and clonal expansion of the small group of CS cells.


Figure 1: Origin of the Corpuscles of Stannius (CS) in zebrafish
A) The endocrine gland forms from the most posterior portion of the DE tubule by a process of transdifferentiation. At 24 hours post fertilisation (hpf), a small group of cells in the DE downregulate the expression of the renal marker cdh17 and at the same time start expressing stanniocalcin-1 (stc1). 50 hpf, the stc1-positive cells form a discrete structure above each pronephric tubule. B) The process of transdifferentiation is positively regulated by Notch signaling. Notch promotes the shuttling (blue arrow) of the transcription factor Hnf1b (orange) from the nucleus to the cytoplasm. Picture in (A) was adopted from figure 1 of the preprint.

Why I chose this paper:

We came a long way down the original idea that cells at their latest point of differentiation are unable to reverse to a progenitor state or change into another cell type. While cells can be forced experimentally to acquire a new identity, we now know that transdifferentiation occurs in nature under physiological conditions as well, for example during morphogenesis or tissue repair as a response to cell loss. Moreover, this phenomenon has been implicated in the onset of various forms of cancer but it can also be associated to other pathologies such as liver fibrosis and cirrhosis.

How transdifferentiation occurs at a molecular level is still a matter of intense research. Clearly, a better characterisation of the process will benefit our understanding of development, regeneration and diseases aetiology. Moreover, transdifferentiation harbours a great therapeutic potential and being able to properly manipulate cell identity is very appealing in regenerative medicine.

In this work the authors uncover a previously uncharacterised event of transdifferentiation in the developing zebrafish embryo. As such, the system represents a new model to study the physiological relevance of transdifferentiation in vivo and will allow to better address the molecular basis of cell plasticity.

Future directions and Questions to the authors:

The preprint leaves the readers with a lot of interesting open questions on how the DE-DC transition is regulated in space and time.

  1. At what time point during development does the Notch signalling hit the threshold to overcome the Hnf1b-irx3b regulation?
  2. How is the initial number of CS cells set? The Notch signalling could promote the transdifferentiation and at the same time prevent it from happening in the neighbouring cells by lateral inhibition.
  3. Is it Notch controlling the extrusion of the CS cells from the renal epithelium?


  1. Blau, H.M., Pavleth, G.K., Hardeman, E.C. et al. (1985). Plasticity of the differentiated state. Science 230: 758–766.
  2. Blau, H.M. and Baltimore, D. (1991) Differentiation requires continuous regulation. Journal of Cell Biology 112: 781–783.
  3. R.L. Davis, et al.-Expression of a single transfected cDNA converts fibroblasts to myoblasts- Cell, 51 (1987), pp. 987-1000.
  4. Takahashi, K. and Yamanaka, S. (2006) Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126: 663–676.


Posted on: 17th July 2018

Read preprint (No Ratings Yet)

  • Author's response

    Richard Naylor shared

    1. At what time point during development does the Notch signalling hit the threshold to overcome the Hnf1b-irx3b regulation?

    Some of the Notch components that are expressed in the pronephros during development follow a profile whereby they are initially expressed throughout the intermediate mesoderm before restricting to a spotted pattern within the PST and DE tubule segments (see Supplementary figure 4 of our preprint or see Fig.2 Ma et al., 2007/ Fig.4 Liu et al., 2007). This spatio-temporal expression profile might suggest that a Notch lateral inhibition mechanism is established between 15 hpf and 22 hpf (the timeframe when this restriction of expression occurs and, intriguingly, also the time-point when we begin to witness pronephric phenotypes in our irx3b morphants (~22 hpf, Fig.4A of the preprint)). We hypothesise that it is the posterior-most Notch+ cells that become reprogrammed to a CS fate and Notch+ cells anterior to this will become multi-ciliated cells as previously shown by Ma et al., 2007 and Liu et al., 2007. However, this hypothesis is deduced solely from the expression profiles of Notch genes so needs further functional investigations in order to be confirmed.

    1. How is the initial number of CS cells set? The Notch signalling could promote the transdifferentiation and at the same time prevent it from happening in the neighbouring cells by lateral inhibition.

    Following on from the answer to the previous question, if it is the posterior-most Notch+ cells that transdifferentiate to a CS fate, a major question for us during this project was why do the Notch+ cells anterior become multi-ciliated cells and not CS? This question remains unresolved and difficult to understand. But the author of this preLight is correct in suggesting that the lateral inhibition mechanism would regulate CS cell number, this is the process that we believe is occurring to regulate how the initial number of CS cells is set.

    1. Is it Notch controlling the extrusion of the CS cells from the renal epithelium?

    In our preprint we show that the process of apical constriction within CS cells from 32 hpf onwards is the driving force behind extrusion of these cells from the tubule epithelium. Whether Notch is involved directly in regulating myosin II contraction of apical F-actin remains to be seen, but in other contexts Notch lateral inhibition is similarly a prelude to apical constriction. In a recent paper in Development that used the fly system (An Y., et al; 2017), the role of Notch lateral inhibition in distinguishing which neuroblasts delaminate from their neighbours was used to study if Notch affects apical constriction. Interestingly, when Notch was inhibited cells that did undergo apical constriction displayed abnormal apical myosin patterns suggesting that the Notch pathway might be involved directly in apical constriction as well as indirectly via determining cell fate.

    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