regeneration in nature

Home » Cnidarians

Category Archives: Cnidarians

Slow-cycling stem cells and Hydra regeneration

Hydra (a diploblastic polyp of the phylum Cnidarian) has been a classical model of regeneration since Abraham Trembley first studied the enormous plasticity of these animals already in the 18th century. Hydra are not constantly renewing their cells but also are capable of regenerating a whole animal from a small piece of their bodies. Remarkably, they are even able to regenerate a well-patterned organism from the re-aggregation of dissociated cells. At the cellular level, Hydra contains three distinct stem cell populations: the ectodermal and endodermal myoepithelial cells are differentiated cells are also stem cells for those specific lineages, respectively, and interstitial stem cells. The interstitial stem cells are multipotent stem cells that give rise to nerve cells, gland cells, nematocytes and gametes. Epithelial stem cells continuously divide in the body column, every 3-4 days and get displaced towards the anterior (tentacles) and posterior (basal disk or foot) tips where they terminally differentiate and progressively get sloughed off. Interstitial stem cells divide also in the body column but at a higher rate, every 24-30 hours and then migrate towards the tips as progenitor cells before their final differentiation.

Now, a paper from the laboratory of Yashoda Ghanekar ( reports the existence of slow-cycling cells within these 3 different compartments of stem cells. In various mammalian stem cell systems, slow-cycling or quiescent cells that do not normally go through division under normal physiological conditions have been described. These cells normally rest in the G0 phase of the cell cycle and divide at a very slow rate or only as a response to injury. Here, the authors report on the presence of slow-cycling cells within Hydra stem cells.

To determine the presence of such slow-cycling cells the authors pulsed Hydra with EdU (a thymidine analog that gets incorporated into the DNA during cell division) for one week to ensure that all cells undergo cell division at least twice and then chased for several weeks in a fresh medium without EdU. Cells that keep dividing will “lose” a detectable EdU signal. On the contrary cells that do not divide any more such as differentiated cells or quiescent stem cells retain the Edu labeling. After one week of pulse, 94-98% of interstitial cell were labeled as well as the 54-80% and the 46-51% of the ectodermal and endodermal epithelial cells, respectively. After four weeks of pulse, these percentages increased to 100% in interstitial cells and more than 90% in epithelial cells. These results indicated that even after 4 weeks few epithelial cells remained undivided. After a long chase (4 weeks) after the EdU pulse, a small but significant number of EdU-positive cells were found in the body column. After one week of pulse and up to ten days of chase around 2.6% of undifferentiated interstitial cells showed complete EdU label. After ten days, only partial labeled interstitial cells were detected (indicating that they were dividing). Ectodermal and endodermal epithelial cells retained the EdU label for much longer. Thus, after a 4 weeks chase, 2.1 and 1.8% of ectodermal and endodermal epithelia cells, respectively, had complete EdU label. Considering the average cell-cycle time of these different lineages, these results suggest that in all three there were cells that did not divide from approximately 8-10 cell cycles after the pulse.

Next, the authors used BrdU (another thymidine analog) and an antibody against mitotic cells to determine that these slow-cycling cells were in fact capable of re-entering cell division. Previous studies have suggested that the extracellular matrix (ECM) could provide a niche for the interstitial stem cells. Interestingly, the authors report here that the percentage of label-retaining interstitial cells in contact with ECM was much higher that that of cells that retained only partial labeling (dividing cells). In other systems quiescent cells are held in G0/G1 phase of the cell cycle. Recently, a study from the laboratory of Brigitte Galliot has reported that interstitial stem cells are paused at G2 phase. After one week of chase for interstitial cells and 3.5 weeks for epithelial cells, most of the cells in these compartments that retained the EdU label were in G2 phase.

Finally, the authors checked the potential contribution of these slow-cycling cells during regeneration. The authors performed midgastric amputation and analyzed head regeneration in animals chased either for one week or 2.5-3.5 weeks. The regenerating tips were cut at 1 and 3 hours of regeneration, macerated and then the authors counted the number of EdU-retaining cells with complete and partial label. As control they used the same body region from animals in chase. The authors found a 50% decrease in the number of cells with complete label at 1h of regeneration and a concomitant increase of cells with partial label, indicating that slow-cycling cells had entered cell division during this time.

In summary, the authors describe here a sub-population of Hydra stem cells that divide infrequently. These slow-cycling cells were present in the 3 stem cell lineages, were capable of re-entering the cell cycle and were activated to divide as a response to amputation during the first hour of regeneration.

Notch signalling in Hydra head regeneration

Hydra is one of the most basal animals and a classical model of regeneration. These polyps display an amazing plasticity as they can regenerate a whole animal from a tiny piece of their bodies, can reproduce asexually by budding and, amazingly, dissociated cells can re-aggregate to form a new individual. Hydra has a well-established oral-aboral axis. In the head (oral) region there is a hypostome and a ring of tentacles. Many studies have reported that the tip of the head has an organizing activity. Thus, if this head tip is transplanted into the body column it is able to induce a new head and a secondary axis. This organizing activity appears to be controlled by the Wnt/Beta-catenin pathway. HyWnt3 is expressed at the tip of the head in intact and regenerating animals. Ectopic induction of Wnt signalling in the body column induces secondary axes all along it.

In a recent paper from the laboratory of Angelika Böttger the authors describe how Notch signalling plays a key role in maintaining and re-establishing the Hydra head ( Notch signalling is a well-known and highly conserved pathway that plays a role, among other, in establishing well-defined boundaries as it can specify different cell fates in two adjacent cells. Once Notch transmembrane receptor is bound by a ligand the intracellular domain of the Notch receptor (NICD) is, then, cleaved and can go to the nucleus where it works as a co-activator of target genes. DAPT is a well-known inhibitor of the Notch pathway as it prevents cleavage of the active NICD and it is usually used in different animal models.

Upon culturing intact Hydras in DAPT for 48h the authors saw how the tentacles were shortened (beginning after 24h of treatment). After 48h of treatment DAPT was removed and in the following days these polyps displayed different levels of abnormalities in the patterning of their tentacles and heads, and even a new secondary hypostome appeared in some of them.  At the molecular level, the expression pattern of HyWnt3 at the tip of the hypostome was not affected by DAPT treatment. However, the expression of HyAlx (a marker or tentacle boundary) was severely affected. Previous studies had suggested that HyAlx might play a role in the specification of tissue for the formation of the tentacles. The results presented here support this view and point out Notch signalling as required for proper HyAlx expression and, therefore, differentiation of tentacles.

Next, the authors analysed the function of Notch signalling during head regeneration. In most of their experiments the animals were treated with DAPT for 24h prior to amputation. After 24h of regeneration in DAPT, this inhibitor was then removed from the medium and the animals were analyzed at different time points of further regeneration. After 60h only 18% of the animals had regenerated their heads. This lack of head regeneration can be explained because of the lack of re-establishment of a new head organizer. Thus, whereas in controls HyWnt3 is initially expressed in the regenerating tip in a broad cap-like pattern and then it becomes restricted to the tip of the new hypostome (around 36hpa), DAPT prevented the expression of HyWnt3 for up to 48h of regeneration. A similar result was observed for HyB-catenin. To check that DAPT really inhibits the formation of a head organizer the authors transplanted 24h regenerating tips of DAPT treated animals into the body column of host polyps. Only in 25% of the transplants a secondary axis developed in sharp contrast to the 100% of secondary axes observed when transplanting the regenerating tips of control animals.

Finally, the authors checked the expression of HyAlx during regeneration. In controls HyAlx is initially expressed in the whole regenerating tip (at 24h). At 48h this gene gets restricted to 4 or 5 rings at the sites where the tentacles will emerge. In DAPT-treated polyps, the initial expression of HyAlx appears normal but it never gets confined to 4 or 5 rings and it remains in the whole regenerating tip and gets expanded, suggesting that the whole tip had characteristics of tentacle precursors. Finally, the authors checked that upon DAPT treatment the expression of HyNotch in the regenerating tip is not affected but that of HyHes, a target of Notch signalling is abrogated.

Whereas in intact Hydra the authors suggest that Notch signalling is required to define the tentacle boundary, during regeneration Notch would be required to separate tentacle and hypostome cell fate. In their model, the Notch signal-receiving cells suppress the tentacle fate and become hypostomal cells by stabilizing the expression of HyWnt3 and HyB-catenin. Consequently, the inhibition of Notch signalling would allow the observed increased in tentacle cell fate and the lack of the Wnt/beta-catenin-mediated re-establishment of the head organizer.

Epimorphic regeneration in Nematostella and the use of the terms “epimorphosis” and “morphallaxis” in regeneration

In a recent paper Passamaneck and Martindale show that cell proliferation is necessary for  regeneration in Nematostella ( By blocking cell proliferation the authors are able to block also regeneration, suggesting that contrary to what it has been described in Hydra, Nematostella does not appear to have any other compensatory mechanism to allow regeneration in a context of no proliferation. This finding is relevant because it shows how different cnidarian species may use very different modes of regeneration based on the classically used terms of “epimorphosis” and “morphallaxis”.

Originally, in 1901 Thomas H. Morgan wrote that “… there are known two general ways in which regeneration may take place, although the two processes are not sharply separated, and may even appear combined in the same form. In order to distinguish broadly these two modes I propose to call those cases of regeneration in which proliferation of material precedes the development of the new part, epimorphosis. The other mode, in which a part is transformed directly into a new organism, or part of an organism without proliferation at the cut surfaces, morphallaxis”. Based on this definition it appears that when regeneration requires proliferation then it would be epimorhic regeneration, whereas regeneration in the absence of proliferation would be morphallactic. A more updated use of the term “epimorphic” includes also the definition by Richard J. Goss (The natural history (and mistery) of regeneration, 1991. In A History of Regeneration Research. Milestones in the evolution of a science, Ed. C.E. Dinsmore) and that states that “Epimorphic regeneration refers to the regrowth of amputated structures from an anatomically complex stump”, and that “The first event in epimorphic regeneration is the development of a blastema, or regeneration bud, derived from dedifferentiated cells, out of which the new structure will take shape”. So, the consensus is that epimorphic regeneration requires proliferation and the formation of a blastema. But not all blastemas are derived from dedifferentiated cells as stated in Goss definition. That is valid, for instance, for amphibian limb regeneration. But in planarians, regeneration occurs mainly by cell proliferation and the formation of a blastema that is derived from adult pluripotent stem cells and not after dedifferentiation.

But whereas the idea of epimorphic regeneration is quite well-established it cannot be said probably the same for the term of “morphallaxis”. Thus, Hydra has been usually used as an example of “morphallactic” regeneration because it has been know for many years that they can regenerate in the absence of proliferation or without a significant contribution of this proliferation. However, in 2009 the laboratory of Brigitte Galliot showed that in Hydra, and after midgastric bisection, head regeneration depends on an initial apoptotic response below the wound that triggers a proliferative zone with “blastema-like” features that significantly contributes to oral regeneration ( So, in that particular context regeneration seems to be “epimorphic”. As Morgan already said in his definitions, both processes, epimorphosis and morphallaxis, are not mutually exclusive. A good example of that are freshwater planarians that have been considered to follow a mixed epimorphic/morphallactic mode of regeneration. The basis of that is that in addition to the fundamental role of pluripotent stem cells in giving rise a regenerative blastema where the missing structures are formed, there is also a remodeling of the pre-existing tissues far away from the wound that help to attain the proper body proportions during regeneration. This remodeling, considered as morphallactic, is more evident in for example in head and tail pieces regenerating a new whole planarian. Thus, if you start with a big head piece containing a big brain it regenerates not only the whole body posterior to this head (through proliferation, blastema formation and growth of the regenerated part) but at the same time the original head with its brain go through an extensive remodeling so they decrease significantly in size in order to form a smaller planarian with perfect body (head/trunk/tail) proportions. So, following Morgan’s definition of morphallaxis at the beginning of this post literally, this remodeling would not require proliferation. However, a paper from the laboratory of Alejandro Sánchez-Alvarado on the temporal and spatial dynamics of Wnt genes expression during planarian regeneration shows that “… although pre-existing cells can assess their new A/P position in the absence of stem cells, anatomical tissue remodeling in planarians depends on the presence of stem cells” ( Therefore, this data would point out that “morphallaxis” could depend somehow also on proliferation.

To solve this apparent conflict one possibility could be to restrict the term of “morphallaxis” to the first definition given by the same Morgan in 1898 in which he wrote: “Thus, the relative proportions of the planarian are attained by a remodeling of the old tissue. I would suggest that this process of transformation be called a process of morpholaxis”. Then, “morphallaxis” can be clearly associated to “remodeling of the pre-existing tissues” and this would be “proliferation dependent” or “proliferation-independent” depending on the organism (planarian/Hydra) or the specific context of regeneration.

%d bloggers like this: