Freshwater planarians are among the very few animals capable of fully regenerating a new functional brain from a tiny piece of their bodies. Despite being animals relatively simple from a morphological point of view, several studies have reported the complexity of their central nervous system (CNS). This complexity can be seen by the high degree of molecular compartmentalization based on the expression patterns of many neural-specific genes. In addition, a large number of distinct neuronal populations expressing different neurotransmitters and neuropeptides have been identified.
An important aspect of planarian CNS regeneration is that even though the available tools allow us to determine quite precisely when and where the different neuronal populations regenerate and how the new brain is formed again at the morphological level, much less is known about when this new complex CNS is fully functional again and how the animals recover their normal behaviours. The main reason for this is that few behavioural assays have been established in these animals. So, even in intact non-regenerating planarians there is no much information about which genes might be regulating different behaviours controlled by the planarian sensory system.
Now a recent paper from the laboratory of Kiyokazu Agata describes some genes related to thermosensory signalling in these animals and reports how thermotaxis is re-established during regeneration (http://www.ncbi.nlm.nih.gov/pubmed/25411498). First, they showed that planarians of the species Dugesia japonica displayed normal locomotor activity and morphological shape at a temperature range from 15ºC to 25ºC. At lower temperatures (5-10ºC) their bodies crumpled and lost their motility; above 30ºC they displayed hyperkinesia and became lethargic or even died just after 1 hour. Then, the authors developed a thermotaxis assay based on a method previously used in C. elegans. To put it simple, they created a radial thermal gradient from the centre of a Petri dish towards its edges: the Tª at the centre being 17ºC and at the edges, 25ºC. When the temperature was uniform all throughout the Petri dish planarians moved towards the edges. However, when a thermal gradient was created planarians tended to move towards the centre of the Petri dish and stay there, where the Tª was lower. Decapitated planarians put in a dish with a thermal gradient did not seem to recognize it and moved randomly towards the edges, as they were in a dish with a uniform Tª. In contrast, the small amputated head pieces were able to recognise the gradient and moved towards the centre of the dish, indicating that the head region was necessary and sufficient for thermotaxis in planarians. In all these experiments and assays all the animals displayed the same locomotor activity independently of being intact, headless or head pieces in Petri dishes with a uniform Tª or under a thermal gradient.
Next, the authors investigated when this thermotactic behaviour was recovered during head regeneration. During the first 4 days of regeneration no thermotactic response was seen even though at this stage planarians had already regenerated the eyes and a new small brain. However, a strong recovery of the thermotactic behaviour was observed at day 5 of regeneration. Previous studies had shown that negative phototaxis is also recovered after 5 days of regeneration.
To try to identify genes involved in planarian thermotaxis the authors focussed on the family of Transient Receptor Potential (TRP) ion channels as they have been involved in regulating a variety of sensory systems including thermosensation in different animals. They identified seven genes homologs to TRPs that displayed different expression patterns in planarians. One of them DjTRPMa was expressed in a scattered pattern throughout the body, although there were more cells in the head region that in the rest of the body. Remarkably, the silencing of DjTRPMa by RNAi resulted in a clear defect in thermotaxis. Thus, and contrary to controls, the animals in which DjTRPMa was silenced never rested in the coolest centre area of the Petri dish. Importantly, RNA treatment did not cause any locomotor defect in those animals. That TRP genes are also involved in thermotaxis in planarians was further supported by the results obtained after treating the planarians with AMTB that specifically antagonizes a thermosensitive TRPM family protein in mammals. These animals moved randomly in the Petri dish and did not tend to go the cold central area. Overall, these results suggested that DjTRPMa would be expressed in thermosensory neurons and might be required for thermotaxis.
During regeneration, DjTRPMa-expressing cells appeared de novo by day 2 whereas a normal thermotactic behaviour was not recovered until day 5, suggesting that thermotaxis would depend on something else other than these thermosensory neurons. Then, the authors analysed the effect on thermotaxis of silencing some genes related to different neurotransmitters expressed in specific neuronal populations. Interestingly, the silencing of DjTPH (a marker of serotonergic neurons) resulted in an abnormal thermotactic response (as DjTRPMa did), without disturbing other behaviours as negative phototaxis or proper locomotor activity. Double staining indicated that DjTRPMa and DjTPH were not coexpressed in the same cells. However, neural projections from serotonergic neurons extended towards DjTRPMa-positive cells, pointing out the possibility that serotonergic neurons could somehow transduce to the brain the temperature signals received by DjTRPMa neurons.
In summary, the authors have characterized for the first time a gene that regulates thermotaxis in planarians and analysed how this behaviour is recovered during regeneration. Further studies should help to better characterize the neural circuit that transduces those external signals to the brain and trigger the proper behaviour.
Among invertebrates some annelid species show also remarkable regenerative abilities. Within this group of animals, however, it is not rare to find species that can regenerate posterior regions but not anteriorly. Also, regeneration has not been extensively studied in all taxa. Now, a recent paper from the laboratory of Michael Weidhase and Conrad Helm (http://www.ncbi.nlm.nih.gov/pubmed/25392962) describes for the first time the process of anterior regeneration in the annelid Cirratulus cirratus, focussing mainly in the musculature and central nervous system (CNS). The authors used phalloidin to label the muscle and antibodies against serotonin and FMRFamide to observe the regenerating CNS.
As for the body wall musculature, it is mainly composed of an outer layer of circular fibers and an inner longitudinal muscle layer. The dorsal longitudinal muscles form a plate that covers the whole dorsal body. Ventrally, there is a medial and two ventro-lateral strands of longitudinal muscle. On the other hand, the CNS is a rope-ladder-like system. The ventral nerve cord consists of two main strands and each of them exhibits one ganglion per body segment. From these ganglia there are three major nerves that extend laterally. At the first chaetigerous segment both strands of the ventral nerve cord separate from each other to form the circumesophageal connective surrounding the mouth opening. This connective splits into a dorsal and a ventral root. The brain is located at the anterior end of the circumesophageal connective. Anti-serotonin immunostaining revealed two brain commissures: a stronger stained anterior commissure connected to the ventral root of the circumesophageal connective and a slighter stained posterior commissure connected with the dorsal root.
After amputation, the wound was closed and a blastema was formed within the first week. Then this blastema elongated anteriorly and about 10 days after decapitation the new mouth opening was evident. By the end of the second week the first outgrowing tentacles appeared. The whole regeneration process took place in about 1 month, although the regenerated head showed some differences with the original respect to the number and length of tentacles and branchiae.
By day 6 of regeneration a blastema was formed and contained outgrowths from the pre-existing longitudinal muscle fibers. From that time the longitudinal musculature kept growing and organized into muscular strands. By day 10 the ventral longitudinal strands reached their most anterior position, surrounding the regenerated musculature of the new mouth opening. At this stage the first circular muscle fibers appeared.
Concerning the nervous system, by day 6 after decapitation, a structure represented by three nerve loops was visible inside the blastema, one median loop and two laterals. At this stage the median loop was folded backwards and was connected to the inner bundles of both strands of the ventral nerve cord. On the other hand, the lateral loops were already oriented anteriorly and were connected to the outer bundles of the ventral nerve cord. At day 8 of regeneration the median loop was also oriented anteriorly. This tripartite loop-like structure has not been described in other regenerating annelids. Whereas the authors favour that the median loop will originate the ventral root of the regenerated circumesophageal connective (visible by day 12), they did not obtain clear data about the transition from the lateral loops to the dorsal root. Finally, the brain commissures were visible by day 14 of regeneration. At this stage, the regenerated nervous system elongated together with the blastema; however, distinct ganglia and lateral processes in the new nerve cords were still missing by day 18. Later, the first signs of body segmentation and differentiation of distinct ganglia and lateral nerves were observed.
In summary, this paper describes for the first time anterior regeneration in the annelid Cirratulus cirratus. Further studies are necessary to describe in more detail this process as well as to determine the origin of the regenerative cells that form the blastema and the new muscle fibers and central nervous system, among all the other anterior tissues.
In contrast to the central nervous system the axons of peripheral nerves are able to regenerate in many vertebrates, including mammals. How this regeneration is achieved and the cellular processes involved have been described in many studies. An important point here is that the cellular environment plays an important role in promoting this axonal regrowth. Among these elements, Schwann cells have been described as required for peripheral axons regeneration. Now, a recent paper from the laboratory of Michael Granato reports on the observation in vivo of axonal regeneration in zebrafish as well as the requirement of Schwann cells to direct those axons towards their original paths (http://www.ncbi.nlm.nih.gov/pubmed/25355219).
Here, the authors use transgenic zebrafish expressing GFP in motor neurons and RFP in Schwann cells to follow in vivo the regeneration of fully transected motor nerves in larvae at 5dpf (days post fertilization). The first thing they describe is how upon transection the amputated distal axons were degraded and fragmented leaving behind axonal debris and denervated Schwann cells. Then, Schwann cells suffered dramatic changes from a tube-like morphology to a shorter and more rounded one. As motor axons regrew into their original paths, Schwann cells membranes also reverted to their normal thinner appearance.
In order to define in vivo the role of Schwann cells the authors used different mutant strains that lack differentiated Schwann cells (mutants for sox10, erbb2, erbb3 and nrg1 type III). In all these mutants, motor axons developed normally through 5 dpf. However, upon transection, axon regeneration was severely affected in all mutants lacking Schwann cells. In them, in 50-80% of the transected nerves the regenerated axons failed to target their original paths and instead ectopically grew into more lateral territories. These results suggest that Schwann cells directed regenerating axons. One possibility could be that proper axonal regeneration would depend on some factor produced by Schwann cells during development. To address this question the authors used a transgenic line in which Schwann cells develop normally but can be eliminated at the requested time point of development. Then, the authors killed the Schwann cells at 3,5 dpf and transected the motor nerves at day 5 dpf. In this context, the regenerated axons also failed to find their original targets as previously described for the sox10 mutants, suggesting that Schwann cells must be present during regeneration to direct re-growing axons. In another set of experiments the authors show how in sox10 mutants the regenerative growth cones initially sprouted and extended from the proximal stump in multiple directions as it happens in wild-type larvae. However, and in contrast to wild-type, axons continued to extend in all directions without going back to their original paths.
Distal denervated Schwann cells could exert this guiding role either by providing a physical substrate for the regenerating axons or by producing some factors that would direct the re-growing axons towards their original paths. To try to distinguish between these two possibilities the authors used a transgenic line in which the degeneration of the amputated distal nerves was delayed by more than 1 week. They crossed this line with the sox10 mutant line so an axonal scaffold distal to the transection gap completely lacking Schwann cells was formed. In mutants for axonal degeneration but with normal Schwann cells, motor axons regenerated normally indicating that those distal amputated axons had no inhibitory action on regeneration. However, in the double mutant with sox10, the regenerating axons failed to find their original paths. In other experiments, the authors transected only half of the axons in the motor nerve to create a continuous axonal scaffold devoid of Schwann cells (in sox10 mutants). There, the presence of this axonal scaffold could only partially compensate for the role of Schwann cells as 50% of the transected nerves showed aberrant abnormal projections compared to the 90% observed in fully transected nerves lacking Schwann cells. Overall, these results suggest that Schwann cells provide more than just a permissive substrate for regeneration.
Finally, the authors analysed the role of dcc (deleted in colorectal cancer) a receptor for netrin, a well-known attractive guidance cue for axonal growth. At 5 dpf dcc was expressed in motor neurons as well as in Schwann cells, whereas netrin1b was expressed in Schwann cells before and after transection. Also, dcc was expressed in motor neurons during the first stages of regeneration. Then the authors used a mutant line for dcc in which, interestingly, motor axons and Schwann cells develop normally. However, upon transection, about 40% of the nerves extended axons not only to their original paths but also ectopically into more lateral territories.
In summary, the authors show here the process of motor axon regeneration in vivo in zebrafish. Schwann cells are required for guiding the regenerated axons to their original paths and do it probably not only by providing some physical substrate but actively producing some factor. Future experiments should determine if netrin could be one of such factors.
A long-standing question within the field of planarian regeneration relates to the nature of the neoblasts, planarian totipotent stem cells. These cells are the only cells with proliferative capabilities in these animals and are absolutely indispensable for regeneration. Based on morphological criteria (size, shape, nucleus/cytoplasm ratio) about 20-30% of the planarian cells have been considered as neoblasts. However, it has not been clear for many years whether all these cells represent a uniform neoblast population with identical o similar proliferative abilities and potentiality or, alternatively, neoblast comprise a quite heterogeneous cell population respect to these criteria. Recent studies from several laboratories have identified several genes and transcription factors that are expressed in different populations of neoblasts and are required for the differentiation of specific cell types. Based on these data Peter Reddien proposed two models for planarian neoblasts: 1) the naïve model in which neoblasts are a rather homogeneous cell population with the same potentiality and in which fate specification occurs only in the non-dividing neoblast progeny; in contrast, 2) the specialized model predicts that neoblasts are an heterogeneous population containing many different lineage-committed dividing cells.
Based on several recent studies planarian neoblasts seem to follow the specialized model as several cell type-specific genes are co-expressed with Smedwi-1, a planarian piwi homologue, considered as a neoblast marker. Now, a recent study from the laboratory of Peter Reddien further supports the specialized neoblast model (http://www.ncbi.nlm.nih.gov/pubmed/25254346). In this work, the authors first purified by FACS neoblast that were in the S or G2/M phases of the cell cycle (dividing neoblasts) and analysed their expression profiles by RNAseq. Then, they focussed on transcription factors as these proteins regulate cell differentiation. Here, they identified a list of transcription factors that were upregulated in dividing neoblasts during regeneration. To validate these results they next carried out in situ hybridizations on purified dividing neoblasts and all of the transcription factors tested were expressed within these cells although, evidently, with different percentages of positive cells.
They extended then these analyses to other previously characterized transcription factors specific for different cell types. Again, here, they identified a percentage of dividing neoblasts in which these factors were also expressed. Next, they focussed on the planarian nervous system as a target to test the specialized neoblast model as many different neuronal subpopulations have been identified. From the analyses of the transcription factors identified in their RNAseq experiments as well as for the search of conserved factors with conserved functions on the development of the nervous system they found 26 neural transcription factors that were expressed in dividing neoblasts. One of these factors, a klf homologue was coexpressed with cintillo in mechanosensory cells located around the head periphery. Remarkably, klf RNAi lead to the absence of these sensory cells during regeneration, despite these treated animals were capable to normally differentiate many other neuronal cell types. Similarly, a pax3/7 homologue was found to be expressed in the medial region of the brain in cells some of which also expressed a dopamine β-hydroxylase (DBH) gene. Animals in which pax3/7 was silenced by RNAi regenerated a significant reduced number of DBH positive cells. Altogether, these results suggest that these transcription factors would be necessary for the differentiation of specific neural lineages from distinct progenitor neoblast subpopulations.
Importantly, the authors also found that transcription factors associated to specific cell lineages (pharynx, central nervous system, eye, protonephridia or muscle) were not co-expressed in the same dividing neoblasts. These strongly suggest that distinct neoblast populations expressed specific combinations of transcription factors associated to different differentiated cell types.
In summary, this study reports 36 transcription factors expressed in dividing neoblasts from regenerating planarians. These factors are expressed in different cell types and tissues in adult planarians, which suggests that they may specify distinct subpopulations of lineage-committed dividing neoblasts, further supporting the specialized model.