This week I am not going to comment on any specific paper on regeneration but rather publicize the launch of a new journal dedicated to this fascinating field of regeneration. The new journal is called Regeneration and will be published by John Wiley & Sons (http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2052-4412). The Editor-in-Chief is Susan V. Bryant a prestigious researcher in the field of limb regeneration in amphibians. The associate editors include Kiyokazu Agata, Enrique Amaya, Cheng-Ming Chuong, Ken Muneoka, Ken Poss and Elliot Meyerowitz. Regeneration is an open access journal and will be the first journal to be dedicated exclusively to regeneration and tissue repair in animals and plants. Therefore the regeneration community have now an excellent tool to disseminate our results in a more compact way that should make more visible our advances and research and, at the same time, attract more people to this field. Hopefully this journal will be useful also to establish bridges between naturally occurring regeneration and the need to enhance assisted regeneration. I really hope that we can all contribute to the success of this journal. I will keep you posted when the first issue is available.
microRNAs (miRNAs) are small non-coding RNA molecules (21-23 nucleotides) that have been identified as important conserved regulators of gene expression. In animals miRNAs usually bind to a complementary region within the 3’UTR of target mRNA molecules resulting either in target degradation o translational repression. miRNAs have been associated to a variety of diseases, cancer and the development of the nervous system, among other processes. Related to regeneration, miRNAs have been involved in liver and cardiac regeneration.
Now a recent paper from the laboratory of Jamie Ian Morrison has characterized the role of miRNAs during heart regeneration in newts (http://www.ncbi.nlm.nih.gov/pubmed/24055866). Contrary to mammals, newts (similarly to zebrafish) are capable of regenerating their heart. The authors divide newt heart regeneration into three main stages: early wounding response (7-14 dpi), hyperplasia (21-30 dpi) and reverse remodelling (45-60 dpi). They performed a microarray screen to identify miRNAs differentially expressed during heart regeneration. Initially, they identified 37 miRNAs that showed a significant expression pattern between 7 and 21 dpi. Then they decided to concentrate in miR-128 that showed a very high expression at a time-point when cardiac hyperplasia is elevated. In situ hybridizations experiment showed that in uninjured hearts miR-128 was expressed at very low levels in sporadic cardiomyocytes lining the epicardial border. In contrast, at 21 dpi there was a significant increase of miR-128 expression around the regenerative zone, not only in cardiomyocytes but also in non-cardiomyocytes. This peak of expression of miR-128 coincides with the peak of hyperplasia during heart regeneration.
In other models miR-128 has been described as a tumour suppressor gene but it has not been related to heart biology. Using an antagomir (miRNAs inhibitors that prevent the binding of miRNAs to their targets) knockdown strategy the authors found first a significant increase in the proliferation of non-cardiomyocyte cells in the regenerating heart, whereas no clear effect on the proliferation of cardiomyocytes was observed. These results suggest that during heart regeneration miR-128 would act as an inhibitor of proliferation of non-cardiomoycyte cells. Another consequence of miR-128 inhibition was that as regeneration proceeds there is a persistent presence of extracellular matrix and fibrin tissue within the regeneration zone. This excessive fibrin and collagen matrix deposition after miR-128 inhibition (in clear contrast to the little scarring observed in control regenerating hearts), resulted in a delayed regeneration. Therefore, miR-128 appears to have an important role regulation extracellular matrix deposition during heart regeneration.
Finally, the authors followed a bioinformatics approach to identify putative targets of miR-128 and found the transcription factor islet1 as a potential gene to be regulated by miR-128. These finding agree with previous results of the same laboratory that had identified islet1 as a factor differentially expressed during newt heart regeneration. Here, the authors report that, compared to uninjured hearts, the expression of islet1 significantly decreases at 21 dpi when the expression of miR-128 is maximum. Importantly, the expression of islet1 increased by a 60-fold when miR-128 was inhibited. As islet1 has been described as a chordate cardiac progenitor cell marker these results would fit with a model in which miR-128 would act, by inhibiting islet1, as a negative regulator of progenitor cell activity and would promote cell differentiation during heart regeneration.
In summary, this is the first report of miRNAs being involved in heart regeneration in newts and provide new actors and tools that can help to understand in the near future how these animals regenerate their heart.
A regenerative response can be triggered after several kinds of wounds. In some cases injuries result in loss of tissues or structures whereas in others not. As the regenerative responses that need to be activated in these two situations are expected to be quite different a question to answer is how is this controlled. That is, how the body senses how much it has lost and needs to be rebuilt? Which specific molecular mechanisms are activated to regenerate in tissue absence?
A recent paper from the laboratory of Peter Reddien tackles these questions in freshwater planarians (http://www.ncbi.nlm.nih.gov/pubmed/24040508). Planarians are amazing animals as they can efficiently regenerate after any kind of injury. In the last years it has been shown how upon amputation planarians respond in a quite stereotypical pattern that distinguishes a response driven by small wounds without almost no loss of tissue from that triggered by the loss of large amount of tissue, as it can be for example, head amputation. Thus, upon amputation there is a first general mitotic peak at 6 hours throughout the regenerating fragment that is also observed after any small injury such as a puncture or incision. Then, there is a second mitotic peak at 48 hours just at the wound region adjacent to the blastema. This peak is specific after injuries that result in tissue loss. Similarly, amputation results in a first apoptotic peak at 4 hours mainly concentrated around the wound that, again, is a general response to any kind of wounding. Then, there is a second apoptotic peak at 72 hours more uniformly distributed throughout the regenerating fragment and that is specific for injuries with tissue loss. Finally, the laboratory of Peter Reddien has also identified in the past a collection of genes that are rapidly induced after wounding and whose expression persists in those cases in which regeneration occurs in tissue absence.
In this paper the authors characterize the function of a follistatin homologue in planarians. Follistatins are well known inhibitors of the TGF-beta signalling pathway. Upon silencing follistatin (fst) by RNAi, no defects were observed in intact planarians suggesting that this gene was not required for homeostatic cell turnover. However, treated animals were incapable of regenerating. At the cellular and gene expression level the authors show how after fst silencing and amputation the 1st mitotic peak normally occurred at 6 hours but then the 2nd peak (specific for regeneration in tissue absence) did not take place. Also, the first apoptotic peak was normal in those treated animal whereas the second peak (again the one specific for regeneration in tissue absence) was inhibited. Finally, the authors show how the expression of some of the genes that are wound-induced is not maintained over time as it normally happens in control animals. Upon amputation planarians does not only produce a blastema in which the missing structures will regenerate but the pre-existing tissues are remodelled in order to accommodate the novel proper body proportions. This remodelling is also inhibited after fst silencing. Overall, these results suggest that fst is required to activate a proper regeneration program in the absence of tissue. This is further supported by the fact that there is strong correlation between the kind of injury produced and the level and persistence of fst upregulation. Injuries such as incisions induced fst expression but this did not persist after 48 hours; however, after the excision of lateral tissue wedges, fst expression persisted. Also, a greater fst expression was seen when larger amounts of tissue were missed.
As follistatin is a conserved inhibitor of the TGF-beta superfamily the authors looked for putative TGF-beta ligands that could be regulated by fst. The rationale was to do double silencing of fst and several TGF-beta ligands and see whether any of those double knockdowns was able to suppress the fst RNAi phenotype. The authors found that the co-silencing of fst with planarian either activin-1 (act-1) or activin-2 (act-2) suppressed the fst phenotype; that is, those animals regenerated normally. These results suggest that fst would promote a regenerative response in the context of tissue absence by inhibiting the action of act-1 and act-2. Therefore, it seems that activins would normally inhibit the triggering of a major regenerative response after small injuries that does not imply tissue loss.
In a previous paper from the laboratory of Phil Newmark (commented in this blog on January 24th 2013) the authors had also characterized fst as an essential factor for regeneration and reported that fst would inhibit activin signalling. In that study the authors focussed on a possible role of fst together with notum in defining an organizing center during anterior regeneration. Now, this new paper confirms some of those previous results on fst and activin signalling but expands fst function to a more general role in activating a regenerative program in tissue absence regardless of the polarity of the regenerating fragment.
In summary, this paper describes a pathway that seems to work specifically in activating regeneration in tissue absence and opens the door to further investigate the role of activin (and other TGF-beta ligands) signalling in different models and injury repairing/regenerating contexts.
It is well known that limb regeneration in amphibians is nerve-dependent. Several neural-derived factors, including FGFs or substance P, have been suggested to play this role. But it was not until few years ago that the laboratory of Jeremy Brockes identified nAG, an important protein expressed in Schwann cells and with the capacity to rescue, for the first time, regeneration up to the digit stage in a denervated newt limb. A different paradigm in which to study nerve-dependency in amphibian limb regeneration is the so-called accessory limb model (ALM). After a skin wound, amphibians regenerate it normally (without any scar). However, if at the same time, a nerve is deviated to the center of this skin wound a blastema-like outgrowth (bump) is formed. Over time most of these bumps regress, but if a piece from the skin of the opposite site (anterior/posterior) is grafted to the site of the skin wound to which a nerve has been deviated then most of the bumps continue growing and eventually can form an ectopic limb (ALM). In this model, an ALM blastema is considered to display many of the features of a regular limb blastema.
In a recent paper by the laboratory of Akira Satoh the authors continue their research to find which factors can be important to induce blastema formation and regeneration in this context (http://www.ncbi.nlm.nih.gov/pubmed/23769980). To do that they investigated the action of possible factors that can substitute the function of a deviated nerve in inducing a blastema-like outgrowth in a skin wound. From a deep sequencing analysis they decided to investigate the Growth and differentiation factor-5 (Gdf5), a member of the TGF-beta superfamily. After the application of Gdf5 in a skin wound they observed the development of a bump similar to an ALM blastema. Thus, it contained small undifferentiating-looking cells, the expression of tenascin-C and fibronectin was upregulated whereas type I collagen appeared downregulated, resulting in an extracellular matrix typical of a regenerative environment and, finally, the blastema marker Msx2 was also detected. However, the expression of Prrx-1, another marker for an ALM blastema was not observed. Also, when fibroblast dedifferentiate to give rise to blastema cells they acquire the capability to differentiate into cartilage, which can be assessed if those cells are grafted to a bone healing region. Whereas grafted cells from an ALM blastema can end up making cartilage, cells from a Gdf5-induced blastema never did it. Therefore, even Gdf5 was able to induce a blastema-like bump it was not exactly the same as a regeneration blastema.
However, if Gdf5 was applied together with Fgf2 and Fgf8 then the blastema-like bump induced appeared quite similar to a regeneration blastema: Prxx1 was induced and those bump cells could be differentiated into cartilage. Remarkably if a piece of skin from the opposite site was grafted to the original skin wound, the application of Gdf5, Fgf2 and Fgf8 was able to induce the regeneration of a limb-like structure displaying digit-like formation, as it happens in a normal ALM context, but without the requirement of a deviated nerve.
Therefore, it seems that Gdf5, Fgf2 and Fgf8 can cooperate to substitute for the requirement of nerve to induce limb regeneration, at least in the ALM paradigm. In the future it would be interesting to characterize better Gdf5 as it does not seem to be expressed by nerve cells and it is not clear what could be its receptor. Finally, the authors describe how Gdf5 appears to activate phosphorylated ERK (pERK), as it also do Fgf2 and Fgf8, which opens the door to investigate the function of pERK in blastema induction and formation, as it has been described in other models.
In previous posts I have discussed the importance of comparing similar developmental processes during embryogenesis and regeneration. This is important in order to determine up to what extend regeneration is a mere recapitulation of embryonic development or, on the contrary, it implies different actors. The answer to this question is not probably unique (or white or black) and will depend on the model used and the specific context analysed. The laboratory of Thomas Becker and Catherine Becker has recently published how dopamine promotes the formation of motor neurons both during embryogenesis and adult regeneration in zebrafish (http://www.ncbi.nlm.nih.gov/pubmed/23707737).
As axons descending from the brain to the spinal cord may control proliferation and differentiation of neural progenitors in that spinal cord they focussed on the putative role of brain-derived dopamine. In fact, the only source of dopamine in the spinal cord comes from axons projecting from the diencephalon. First, the authors addressed the function of dopamine in the generation of motor neurons during embryonic development by three independent approaches. In all cases, they found that by decreasing the number of dopaminergic neurons or dopamine production there was a significant decrease in the number of motor neurons in the developing spinal cord. On the contrary, the uses of a dopamine agonist lead to an increase in the number of these motor neurons. They also found that this action of dopamine is mediated by the D4a receptor. At the cellular level, dopamine action seems to expand the pool of motor neuron progenitors. At the molecular level, the activation of the D4a receptor appears to inhibit the cAMP/protein kinase A pathway, which in turn inhibits the Hedgehog (Hh) pathway. The Hh has been shown also to be important for motor neuron differentiation.
Once determined the role of dopamine from descending axons on motor neuron development the authors sought to determine whether dopamine was also required for the regeneration of this neuronal population. It is well known that zebrafish can regenerate their spinal cord upon transection and that Hh pathway is required for it. As in embryos, the only sources of dopamine in adults are descending axons from the brain. Upon complete transection dopaminergic projections and activation of the D4a receptor are mainly detected rostral to the lesion site, which results in many more new motor neurons generated rostral than caudal to the lesion. When these dopaminergic axonal projections were ablated the number of new motor neurons regenerated rostral to the lesion was significantly reduced, pointing out a conserved role of dopamine on regeneration. Remarkably, the application of a dopamine agonist in the caudal site of the lesion (a region in which very little neurogenesis is normally observed) was able to induce the expression of D4a as well as to promote a high increase in the number of motor neurons regenerated. Moreover, the effects of dopamine on motor neuron regeneration appeared to be also mediated through the Hh pathway as happens during development.
In summary, these results suggest that dopamine activates Hh signalling (through inhibiting cAMP-dependent PKA activity) to promote the expansion of the pool of progenitor cells and allow motor neuron generation in two different contexts: embryogenesis and adult regeneration. More importantly, this data opens the possibility that this pathway can be modulated with agonists in order to promote or enhance neural regeneration in other vertebrates.