The African clawed frog Xenopus laevis: a model ...

2 downloads 0 Views 1MB Size Report
Sep 18, 2016 - Zebrafish and salamanders such as the newt and the axolotl are capable of CNS ...... the other hand, the Levin group has described that V-ATPase proton ..... [104] G.B. Whitworth, B.C. Misaghi, D.M. Rosenthal, E.A. Mills, D.J. ...
Accepted Manuscript Title: The African clawed frog Xenopus laevis: a model organism to study regeneration of the Central Nervous System Author: Dasfne Lee-Liu Emilio E. M´endez-Olivos Rosana Mu˜noz Juan Larra´ın PII: DOI: Reference:

S0304-3940(16)30737-6 http://dx.doi.org/doi:10.1016/j.neulet.2016.09.054 NSL 32332

To appear in:

Neuroscience Letters

Received date: Revised date: Accepted date:

26-7-2016 18-9-2016 28-9-2016

Please cite this article as: Dasfne Lee-Liu, Emilio E.M´endez-Olivos, Rosana Mu˜noz, Juan Larra´ın, The African clawed frog Xenopus laevis: a model organism to study regeneration of the Central Nervous System, Neuroscience Letters http://dx.doi.org/10.1016/j.neulet.2016.09.054 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

The African clawed frog Xenopus laevis: a model organism to study regeneration of the Central Nervous System Dasfne Lee-Liu1*, Emilio E. Méndez-Olivos1, Rosana Muñoz1 and Juan Larraín1*

1

Center for Aging and Regeneration, Millennium Nucleus in Regenerative Biology, Faculty of

Biological Sciences, P. Universidad Católica de Chile, Alameda 340, Santiago, Chile *

Corresponding authors at: Faculty of Biological Sciences, P. Universidad Católica de Chile,

Alameda 340, Santiago, 8331150, Chile. Fax +56 226862824. http://laboratoriolarrain.uc.cl

E-mail address: [email protected] ; [email protected]

HIGHLIGHTS:

- Xenopus brain and spinal cord regenerate in the larva, but not after metamorphosis. - Optic nerve regeneration is maintained throughout frog lifespan. - During metamorphosis, remodeling of brain stem supraspinal tracts takes place. - Sox2+ progenitor cells in the brain and spinal cord respond to injury. - Studying Xenopus can provide important insights into improving neural regeneration.

ABSTRACT:

While an injury to the central nervous system (CNS) in humans and mammals is irreversible, amphibians and teleost fish have the capacity to fully regenerate after severe injury to the CNS. Xenopus laevis has a high potential to regenerate the brain and spinal cord during larval stages (47-54), and loses this capacity during metamorphosis. The optic nerve has the capacity to regenerate throughout the frog’s lifespan. Here, we review CNS regeneration in frogs, with a focus in X. laevis, but also provide some information about X. tropicalis and other frogs. We start with an overview of the anatomy of the Xenopus CNS, including the main supraspinal tracts that emerge from the brain stem, which play a key role in motor control and are highly conserved across vertebrates. We follow with the advantages of using Xenopus, a classical laboratory model 1

organism, with increasing availability of genetic tools like transgenesis and genome editing, and genomic sequences for both X. laevis and X. tropicalis. Most importantly, Xenopus provides the possibility to perform intra-species comparative experiments between regenerative and nonregenerative stages that allow the identification of which factors are permissive for neural regeneration, and/or which are inhibitory. We aim to provide sufficient evidence supporting how useful Xenopus can be to obtain insights into our understanding of CNS regeneration, which, complemented with studies in mammalian vertebrate model systems, can provide a collaborative road towards finding novel therapeutic approaches for injuries to the CNS.

Keywords: Xenopus; central nervous system regeneration; spinal cord; optic nerve; brain; regenerative model organisms.

1. Introduction

An injury to the central nervous system (CNS) in humans has grim consequences: damage is mostly irreversible and accompanied by severe impairment of motor and sensory function [25, 97]. It is therefore astounding that a group of organisms, namely amphibians and teleost fish, are capable of full recovery after severe injury to the CNS. Most importantly, it raises important questions: Which are the cellular and molecular mechanisms that allow such high regenerative potential? Could we harness this potential to improve regeneration in mammals, especially in humans?

Zebrafish and salamanders such as the newt and the axolotl are capable of CNS regeneration throughout their lifespans, while tailless frogs (order Anura) only have this potential during larval stages. Included in the latter are the African clawed frog Xenopus laevis, and the western or tropical clawed frog from the same genus, Xenopus tropicalis [6, 19, 52, 67]. One of Xenopus' fundamental traits for the study of regeneration is that its regenerative potential is restricted to larval or tadpole stages (up to stages 50-54), and is lost during metamorphosis, when it turns into a froglet (developmental stages 56-66) [4, 28, 31, 33, 53, 68]. One exception is the optic nerve, which has the potential to regenerate throughout Xenopus’ lifespan [36, 65]. The mechanisms that explain why Xenopus larvae are capable of CNS regeneration but froglets cannot are not yet fully understood, although the last two decades have had an increase in the use of Xenopus to study regeneration [19, 52, 67, 86].

2

Here, we aim to review the work performed in Xenopus, mainly X. laevis, but also X. tropicalis, on spinal cord, brain and optic nerve regeneration, including a discussion on how the knowledge generated in these and other anurans can provide valuable knowledge to develop novel therapeutic approaches to treat CNS injuries in mammals. The field of spinal cord regeneration in particular has grown importantly in the past decades, and our knowledge of this process is increasing, for which the spinal cord will occupy a great part of this review. By presenting the knowledge gained from studying spinal cord, brain and optic nerve regeneration in Xenopus, we hope to provide convincing evidence on the contribution this model organism can be to advance our knowledge in the field. By understanding how regeneration-competent organisms achieve CNS regeneration, we can obtain important insights into how neural regeneration and plasticity can be improved in mammals.

2. Advantages of Xenopus as a model organism to study central nervous system regeneration 2.1. Xenopus as a laboratory model organism Before modern pregnancy tests, X. laevis frogs were used for this purpose. From the late 1940’s to the 1970’s, frogs were injected with the urine of possibly pregnant women, and the presence of human chorionic gonadotropin (hCG) induced the frogs to lay a large number of eggs within 4 to 12 hours, indicating a positive result for the test [22, 75]. Before frogs, immature female mice or rabbits were used. However, this test took longer and was more expensive, as ovary maturation indicated the positive result, for which animals needed to be sacrificed after each test [30]. The ease with which female frogs can be induced to lay eggs using commercially available hCG, which can then be fertilized in vitro for synchronized development, the relatively large size of the eggs and embryos (1.2 mm for X. laevis), and their ex-utero development has made X. laevis a classical model organism to study early vertebrate embryonic development [43]. Milestone scientific advances in cell and developmental biology have been performed in this organism. For example, in the late 1980’s, a system for in vitro nuclear and chromatin assembly was developed in X. laevis, allowing the isolation of important components of the cell cycle, including the Meiosis maturation-promoting factor (MPF) [56, 69]. During the 1990’s, the first cloning of key signaling molecules that determine cell fate during dorsal-ventral patterning such as Chordin, Noggin and Follistatin, all antagonists of the Bone Morphogenetic Pathway (BMP) was also performed in X. laevis [18, 43].

As a classical laboratory model organism, Xenopus poses the following experimental advantages: 1) Standard protocols for Xenopus laboratory breeding and husbandry have been available for several decades, which are comparably simpler and of a lower cost than those required 3

by rodents [12]; 2) The availability of techniques to modify gene expression, like antisense morpholino oligonucleotides, the generation of transgenic animals [43, 47], and recent developments on genome editing techniques like the CRISPR/Cas9 system [8, 9, 70]; 3) The availability of Xenopus genetic and genomic data to study genes, gene families and gene networks, including ESTs (expressed sequence tags), UniGene clusters and continually updated genomic sequences for both X. laevis and X. tropicalis [43, 48]. The latter have allowed the use of highthroughput technologies in Xenopus, including RNA-Seq [3, 15, 53] and quantitative proteomics [74, 90]. Furthermore, the National Xenopus Resource (NXR) and the European Xenopus Resource Centre (EXRC), among other institutes, have an increasing resource of transgenic lines [92]. Importantly, amphibians including Xenopus, diverged more recently from amniotes (360 million years ago) than fish (over 400 million years ago), and frog and human genomes share a high degree of synteny [43, 44].

Specifically for the study of CNS regeneration, Xenopus has the following advantages: 1) The possibility to raise hundreds of regenerative larvae (3 weeks) and non-regenerative froglets (2 months), which can be used in comparative studies [11, 33, 68, 72]; 2) The availability of reproducible CNS injury protocols, such as spinal cord transection, which dates back to 30 years ago [66], partial ablation of the brain [23, 107, 108], and optic nerve transection [36, 55]; and the availability of an early fate map that allows targeting of specific reagents to a particular tissue. As a non-mammalian vertebrate, Xenopus lacks a somatosensory cortex and direct connections from the forebrain to the spinal cord, making its CNS circuitry simpler than that in mammals, but not as simple as invertebrate models, like Drosophila melanogaster or Caenorhabditis elegans, and still more amenable to elucidate cellular and molecular mechanisms.

It is also important to mention that Xenopus has some disadvantages, such as the long generation time for X. laevis (sexual maturation within 7-12 months), and the fact that it has an allotetraploid genome, making genetics harder to perform. X. tropicalis, on the other hand, has a shorter generation time (3-4 months) and a diploid genome, which make it a better model for genetic studies. When compared with zebrafish embryos, Xenopus embryos are less transparent (although natural albinos exist and allow better imaging), which makes live imaging more difficult. There are also very few validated tissue specific promoters. In spite of these caveats, the advantages of Xenopus genus frogs are extensive, strongly supporting its use for the study of CNS regeneration. 2.2. Development and remodeling of Xenopus Central Nervous System After fertilization, Xenopus embryogenesis continues inside the chorion until hatching at stages 35/36. Embryos continue to develop until they start free feeding, when the larval stage starts 4

(stage 46) (approximately 1 week after fertilization) [79]. Tadpoles enter metamorphosis when the thyroid gland begins to function, as thyroid hormones (TH) are responsible for driving the metamorphic process [11, 38]. Metamorphosis is divided into three different phases: 1) Premetamorphosis, stages 46-54, characterized by the appearance of the limb buds; and a high rate of cell proliferation in both the CNS and limb buds; 2) Pro-metamorphosis, stages 54-58, during which limb growth and differentiation occur; and 3) Metamorphic climax, stages 59-66, which includes tail resorption, organ remodeling, and a 70% body weight loss [13, 72]. Animals reach stage 66 at approximately 8 weeks after fertilization [11, 72]. These changes allow the tadpole to transition from its tailed form, which uses undulatory movement for locomotion, into a tetrapodal aquatic animal [81]. This transition in locomotor strategy demands remodeling of the CNS to adapt to the transformed anatomy of the resulting frog [59].

The development of the Xenopus CNS takes place in a caudorostral direction. The spinal cord forms first, followed by the hindbrain, and the most anterior portions of the brain [72]. Figure 1 depicts the brain from stage 50, 58 and 66 animals, all drawn to scale. The brain can be divided into three sections: the prosencephalon divided in the olfactory bulb, telencephalon and diencephalon (forebrain, green); the mesencephalon divided in the optic tectum, and tegmentum (midbrain, orange) and the rhombencephalon divided in the cerebellum and medulla oblongata (hindbrain, yellow). Functionally, the prosencephalon receives sensory input from the olfactory nerve (which innervates the olfactory bulb in the most anterior part of the telencephalon), and has both efferents and afferents to and from the mesencephalon (Fig. 2). The hypothalamus is located in the diencephalon (Fig. 2). The mesencephalon and rhombencephalon comprise the brain stem, where main integration of the sensory and motor systems occurs. Relevant parts of the brain stem are the optic tectum, innervated by retinal ganglion cells through the optic nerve, the tegmentum, which sends motor output from the mesencephalon (Fig. 2, orange), and the cerebellum, involved in motor learning and coordination (Fig. 1) (reviewed in [62]).

Since non-mammalian vertebrates do not have corticospinal tracts or tracts directly connecting the prosencephalon with the spinal cord, as mammals do, all motor control is integrated in the brainstem (Fig. 2). Neuronal nuclei in the brain stem that project into the spinal cord form supraspinal nerve tracts, which are remarkably conserved across mammalian and non-mammalian vertebrates [94]. The main supraspinal nuclei and the names of the axonal tracts that originate from them are the following (Fig. 2, Table 1): 1) Reticular formation, (forms the reticulospinal tract); 2) Vestibular nuclei, including lateral, medial and inferior nuclei (vestibulospinal tract), as well as Mauthner cells (not shown in Fig. 2); 3) Locus coeruleus, which forms the coeruleospinal tract; 4) 5

Red nucleus or nucleus ruber, which forms the rubrospinal tract; and 5) the hypothalamus, which forms the hypothalamospinal tract.

Sensory and motor neurons present in the Xenopus spinal cord are born during two neurogenesis waves. The first occurs between stages 9-26 for sensory neurons, and 9-46 for motor neurons. The second wave finishes by stage 53, for which most spinal cord neurons are born by the end of pro-metamorphosis [79]. The brain also shows an advanced development by stage 53 [72]. Therefore, during metamorphosis, the main process undergone by the CNS to transition from undulatory to tetrapodal locomotion is the remodeling of supraspinal tracts [81]. While most supraspinal neurons are born by stage 53, their axonal tracts extend into the spinal cord at different stages of development (Table 1). In stage 50, reticulospinal and vestibulospinal tracts already innervate spinal segments extensively [17, 95]. Coerulospinal and rubrospinal tracts are less developed; the latter only start to invade the spinal cord at stage 50 [95, 96]. Importantly, a group of reticulospinal and vestibulospinal axons that innervate the tail in the tadpole continue to persist after metamorphosis, where they extend collaterals that will innervate the lumbar cord for limb control [17, 95]. During pro-metamorphosis, coerulospinal and rubrospinal tracts continue to develop. The latter, in particular, reach the lumbar cord at stage 58, when hindlimbs start to be used in locomotion, and hypothalamic fibers reach the spinal cord at stage 57 [95, 96]. Also, a second wave of reticulospinal and vestibulospinal axons arise from these nuclei during metamorphosis, which is specifically aimed at the lumbar spinal cord, also for limb innervation, marking one of the differences between stage 50 and stage 66 supraspinal tracts. Mauthner cells on the other hand, maintain their lumbar projections in adult anurans [95, 96].

3. Optic nerve regeneration throughout Xenopus lifespan: the role of extrinsic and intrinsic factors The optic nerve connects the retina with the optic tectum through retinal ganglion cells (RGC); their neuronal bodies are located in the eye and project to the tectum (section 2.2). Lesions to the optic nerve are performed by transection (clean incision that completely interrupts the optic nerve) or crush injury (where the optic nerve is damaged by compressing with forceps or a similar tool) [76, 110]. The first report on anuran regenerative potential of the optic nerve was performed in pre- and post-metamorphic stages of six different frog and toad species, and demonstrated its capacity to regenerate throughout these anurans’ lifespans [88]. Fifteen years later, this lifelong potential was also shown to be present in the X. laevis optic nerve [36]. Both Sperry and Gaze’s studies suggested through functional and histological methods that retinal orientation determined the connection pattern that would later be found in the innervated tectum. Later, Maturana and 6

colleagues used the electrophysiological registry of a single regenerating fiber in the optic nerve to show that axons regenerated to their original and intended target in the tectum [61]. In this section, we will address the following questions that arise from these reports: 1) Which factors allow optic nerve neurons to regenerate their axons? 2) How do they achieve navigation through their intended path during regeneration?

The timing of optic nerve regeneration has been shown in adult Rana temporaria. The first tectal response is observed 20 days after section of the optic nerve, and a normal retinotectal projection is achieved between 40 and 70 days after the injury [37]. Optic nerve regeneration requires remyelination after an injury, and in X. laevis tadpoles (stages 54-56), this process begins between 15 to 19 days after crush or transection. Additionally, a 50% myelination level (when compared to a control condition) is only achieved 95 days after the injury [76].

Axonal regeneration and plasticity can be affected by intrinsic or extrinsic factors. Intrinsic factors are genetic programs expressed by the regenerating neuron, while extrinsic factors include the environment that surrounds the neuron, such as the extracellular matrix, soluble factors and other cell types (reviewed in [2, 27]). The regenerative potential shown by optic nerve neurons could therefore be affected by either, or both, types of factors. Evidence supporting the role of intrinsic factors in anuran optic nerve regeneration was first obtained from studies using postmetamorphic Bufo marinus, in which a crush injury of the optic nerve was coupled with [35S]methionine labeling of proteins transported along the axon. Similar evidence in X. laevis also supports the relevance of transported proteins during axon regeneration in the optic nerve [91]. Two-dimensional electrophoresis of enriched proteins allowed the identification of four proteins, including Growth associated protein 43 (GAP43) [84], later described as a major component of the growth cone membrane in the developing rat brain [83]. GAP43 has also been shown to be a substrate for Protein kinase C (PKC), a major regulator of actin cytoskeleton dynamics that is present in lipid rafts in the plasma membrane of peripheral mammalian nerves [51]. A more recent study in juvenile (< 6 months) X. laevis froglets showed that after optic nerve crush injury, knocking down of hnRNP K (Heterogenous nuclear ribonucleoprotein K) with antisense morpholino oligonucleotides impairs axonal regeneration, and also decreases GAP43 protein levels, among other targets, also supporting a role for intrinsic neuronal factors [55]. Also, a recent report of the transcriptome in X. laevis RGCs using RNA-Seq shows that tubulin, gap43, klf6 and elements of the JAK/STAT pathway (regeneration-associated genes) are upregulated, while the regeneration inhibitor factor klf4 is downregulated after optic nerve crush. [104].

7

The regenerative potential of the spinal cord is limited to larval stages, which could, in part, be caused by inhibitory extrinsic factors. Glial cells and myelin derived molecules have been shown to inhibit axon regeneration (reviewed in [64]). A remarkable in vitro experiment using retinal explants derived from X. laevis froglets compared axonal growth when co-culturing with spinal cord or optic nerve-derived oligodendrocytes. This in vitro experiment showed that in explants grown in optic nerve or tectum-derived oligodendrocytes, 66% of axons grow successfully, while only 16% axons are capable of growing across spinal cord-derived oligodendrocytes, supporting an inhibitory role of extrinsic factors. Culture supplementation with the IN-1 monoclonal antibody increases the number of axons growing across spinal cord cells to 41% [50]. The IN-1 antibody had been described to neutralize the inhibitory effect of mammalian CNS myelin on axonal growth [78], and was later demonstrated to recognize Neurite outgrowth inhibitor (Nogo) [14, 80]. While at the time, Lang and colleagues reported that the IN-1 antibody only bound to spinal cord, but not optic nerve-derived Xenopus froglet tissue sections, a later study by the same group did find Nogo-A in the Xenopus genome, and detected its expression in different regions of the CNS, including the optic nerve [49]. Nogo-A was one of the earliest identified members belonging to the group of “myelin inhibitor elements”, which are described as extrinsic signals that modulate axonal growth and/or regeneration [106]. However, given that it is expressed in the both the optic nerve and the spinal cord in Xenopus, the answer to why it only has an inhibitory effect on the former, but not the latter, remains elusive.

A recent report from the Zuber group indicates that they could not find the gene for glial fibrillary acidic protein (GFAP) in the Xenopus genome, suggesting that this gene was lost during anuran evolution [60]. They show that after retinal ganglion cell axotomy, Müller glia react to injury with morphological changes like hypertrophy, which leads to glial scar formation, in spite of the intriguing lack of GFAP in Xenopus. This report suggests that other intermediate filament proteins, such as vimentin and/or peripherin, could be acting instead of GFAP in Xenopus glia.

Elements of the immune system have also been shown to play a role during optic nerve regeneration. In Xenopus, the 5F4 monoclonal antibody was raised to specifically recognize macrophages and microglia. Immunolabeling shows that both populations arrive at the injury site during the first 24 hours after the lesion in stage 54-56 tadpoles, reaching peak levels at day 5, after which they show a progressive decrease. Control levels are achieved 30 days after the injury [41, 105]. Interestingly, during the process of degeneration and reinnervation of the optic tectum after an optic nerve lesion, the number of synapses decreases dramatically within 4 days after the injury. Post-synaptic sites are then removed by phagocytosis, and when the newly regenerated axons reach 8

the tectum, they induce new post-synaptic membrane specializations. In contrast, in mammals, the original post-synaptic membrane specializations continue to be present for several months after denervation [73].

Regarding navigation and guidance during optic nerve regeneration, early experiments in stage 47-48 tadpoles in which the optic nerve was cut and deflected to enter through the oculomotor foramen instead of the optic foramen, led to axons that were still capable of reaching the chiasma, but instead of decussating there, reached the ipsilateral tectum innervating it, instead of the contralateral [46]. This result suggests an extrinsic axon guidance component, as axons do not return to their originally intended path, but are instead deflected into an incorrect target. Furthermore, Fawcett performed an interesting study in Xenopus in which optic nerve growth was evaluated in a ‘virgin tectum’. He achieved this by surgically removing one eye at stage 28, before optic nerve fibers enter the tectum. Animals were then allowed to develop, and 2 months after metamorphosis, an optic nerve section of the remaining nerve was performed. As it regenerated, a group of axons grew into the ipsilateral tectum (the ‘virgin tectum’) using pathways different to those observed in a control, innervated, tectum. These results suggest that the navigation system relies on fiber-fiber interactions, where growing axons follow the paths of previously existing axons [26].

Another experiment in X. laevis larvae and adults studied the relationship between glial cells and regenerating axons, demonstrating a close association of optic nerve glial cells and regenerating axons after injury, thus guiding their growth/navigation [10]. Furthermore, in vitro experiments in mouse retinal explants have shown that cells in the chiasma are a source of diffusible factors that reduce neurite outgrowth [103].

In adult Rana pipiens, RGC regeneration after optic nerve injury has been addressed extensively by the Blanco group. Animals treated with a single application of basic Fibroblast growth factor (bFGF) in the proximal stump of an amputated optic nerve promotes RGC survival. However, this effect is not observed after an intraocular bFGF injection [7]. They have shown that this effect is, in part, mediated by an increase in Brain-derived neurotrophic factor (BDNF) signaling [87], and by the activation of the Mitogen-activated protein kinases (MAPK) and Protein kinase A (PKA) [77]. Furthermore, the same group found that endogenous FGF signaling is upregulated after optic nerve injury, with an increase in FGF2, FGFR1 and FGFR3 protein levels in both retina homogenates and whole RGCs through immunofluorescence [21]. Two years later, they also demonstrated that the effect of treating the optic nerve after a crush injury with ciliary 9

neurotrophic factor (CNTF) in Rana pipiens leads to a 750 μm growth after 2 weeks, as opposed to the 300 μm growth shown by untreated axons. BDNF or FGF treatment lead to a 400 μm growth, and a combination of all three factors has an effect similar to CNTF-only treatment [102].

The field of optic nerve regeneration (summarized in Table 2) has not been studied as extensively as that of optic nerve development in Xenopus, but provides a valuable experimental paradigm to evaluate whether the signals that are used during optic nerve development are the same as those used during regeneration. As several questions remain unanswered in this field, the advances in the past decade in genetic tools (see section 2.1) available in Xenopus could lead to exciting avenues to answer them. The evidence provided so far indicates that axonal navigation (and regeneration) may rely on both intrinsic (e.g., local translation in the axonal growth cone) as well as extrinsic factors (e.g., glial cells and secreted factors such as FGF and CNTF) (Table 2). Further work will be required to completely elucidate the mechanisms that drive axonal regeneration and guidance in Xenopus, which could then be extrapolated to mammalian models.

4. Brain regeneration

Unlike the optic nerve, regeneration of the Xenopus brain is mainly observed until the end of premetamorphosis (stage 54) [28]. Brain regeneration has been studied in the telencephalon (mainly the olfactory bulb), and in the optic tectum (Section 2.2 and Fig. 1) (reviewed in [23], [63]). Lesion paradigms have been diverse, ranging from a non-lethal ablation of a portion of the brain [108], to necrosis-inducing methods such as mechanical injury, strong electrical stimulation, and laser ablation [45], as well as aspiration of neurons using gentle vacuum in the optic tectum [63]. The regenerative process in the brain has been studied in regenerative stages, as well as using comparative approaches between regenerative and non-regenerative stages [23]. Here, we will discuss the different cellular mechanisms and signaling molecules that have been described to contribute to regenerate the damaged portion of the brain. This includes proliferation of progenitor cells, differentiation and/or neurogenesis, the release of growth factors promoting these processes, and notably, the expulsion of damaged cells.

The ablation of the anterior telencephalon in larval stages 47-53 results in complete regeneration of the lost structure [23, 108]. After this type of injury, the continuity of the ventricles in the telencephalon becomes interrupted, but these are rapidly sealed by adjacent ependymal cells that proliferate to close the wound, and then replace the damaged tissue within 30 days from surgery [108]. While adult frogs also respond to brain injury with proliferation of ependymal cells, 10

closing of the ventricular lumen is not successfully achieved [28], possibly accounting for the failure of regeneration in adults [23, 108]. In fact, transplantation of larval telencephalon cells into juvenile non-regenerative froglets after injury have shown that grafted cells are able to reconstruct the ablated structure [108].

As mentioned above, the ablation of a portion of the brain tissue usually leads to a requirement to replace lost cells through cell proliferation. In the above example, ependymal cells proliferate to close the wound first and then give rise to the cells in the regenerated tissue [108]. Filoni and colleagues described a correlation between the size of the population of undifferentiated cells with proliferative capacity (which most likely correspond to neural stem and progenitor cells) in the brain, with the regenerative potential shown at different stages of metamorphosis. While early larval stages (stage 48) show a large and extensive area of actively proliferating cells, this number is much lower in stages 55-56 during pro-metamorphosis [28]. Finally, in post-metamorphic froglets, the rate of cell proliferation decreases until it becomes limited to specific areas of the postmetamorphic brain [28].

After a lesion, cellular damage will occur, which in several cases includes the release of intracellular content and signals that either induce more damage or apoptosis of neighboring cells. In stage 44-48 larvae, mechanical injury or injury by electrical stimulation leads to tissue necrosis and a consequent release of ATP from damaged cells. Remarkably in the larval brain, this leads to activation of purinergic receptors and activation of calcium waves in neural progenitor cells, causing reorganization of the cytoskeleton and activation of the actin/myosin contractile machinery. This process is dependent on Rho kinase and results in strong and fast contraction of the neuroepithelium in the apical-basal axis, leading to nuclear movement and cell form changes. These contractions cause a rapid expulsion of damaged cells into the ventricles, preventing the cascade of cell death signaling to reach adjacent cells and the worsening of the lesion, thus allowing wound healing and regeneration of the damaged tissue [45].

During development of the olfactory system, sensory neurons develop from the olfactory placodes, which develop independently from the primitive neural plate. In fact, olfactory sensory neurons develop first and project their axons into the prosencephalon, and then the olfactory bulb develops. Olfactory sensory neurons reach the prosencephalon at stage 32, and complete development of the telencephalon requires innervation of olfactory sensory neurons [42, 89]. In a similar manner, successful regeneration of the telencephalon after ablation in larval stages has also been shown to require innervation from the olfactory placode [107]. The authors of this work 11

performed ablation of the anterior portion of the telencephalon in stage 53 animals. Twelve days after injury, some animals already show reconnection of the olfactory nerve with the telencephalon. Remarkably, 8 months after the ablation injury, only animals that had achieved bilateral reconnection of the olfactory nerves with the telencephalon had successfully regenerated both lobes of the telencephalon. Animals that showed unilateral reconnection of the olfactory nerve had instead regenerated a single lobe of the telencephalon, and when no regeneration of the olfactory nerve was observed, the telencephalon was not regenerated either [107]. Furthermore, to strengthen the relationship between the olfactory epithelium in the nose with the olfactory bulb, chemical damage to the olfactory epithelium has been shown to induce an increase of the brain-derived neurotrophic factor (BDNF) in the granular cells of the olfactory bulb [32].

In order to address whether regeneration of the injured brain tissue leads to visual function recovery, the Cline group uses a system that evaluates a visual avoidance behavior dependent on the optic tectum [63] (Section 2.2). In stage 47 larvae, a unilateral focal ablation of cells in the optic tectum by gentle aspiration leads to an increase in tectal progenitor cell proliferation, which, by pulse-chase experiments, have been shown to differentiate into neurons. Furthermore, these tectal progenitor cells express Sox2 [63]. Lesioned animals are capable of recovering the visual avoidance behavior (which indicates functional recovery of the visual system) within 7-8 days after the lesion. On the other hand, complete ablation of the optic tectum does not lead to functional recovery of the visual system [20].

Brain injury leads to loss of tissue and also to cellular damage, which releases injury signals. The regenerative brain of larval Xenopus shows active proliferation in response to damage for rapid wound healing, followed by further progenitor proliferation and differentiation to replace lost cells in the damaged tissue. In addition, damaged cells are expelled to avoid further propagation of damage signaling. Finally, for functional recovery to occur, proper connections between the newly formed tissue and its targets is required and successfully achieved, leading to a remarkable structural and functional recovery of the injured brain (Table 3). This sequence of events that lead to brain regeneration gives us valuable insight into which are the cellular and molecular events that occur in a model that achieves successful regeneration of neural tissue.

5. Regeneration after spinal cord injury

Regeneration of the Xenopus spinal cord has been addressed using two main injury paradigms: the first is tail or caudal amputation in the Xenopus larva (stages 43-52), and the second, 12

spinal cord transection (stages 50-66). In the first model, an incision is performed to cut the tail, removing, for example the caudal half of it. Spinal cord regeneration can then be studied during the concomitant regrowth of all tissues that will form the new tail, which include the spinal cord, muscle, notochord, skin and fins [16, 57, 85, 86]. This model has shed light on signaling pathways required for tail regeneration, but has the disadvantage that it cannot be performed in postmetamorphic frogs for stage-comparative studies (with the exception of studies involving the refractory period, see next paragraph and Table 4). In the second model, spinal cord transection, a dorsal section of the spinal cord is surgically exposed in either larval or post-metamorphic stages, followed by an incision that completely severs the spinal cord, interrupting all ascending and descending axonal tracts [4, 33, 66, 68, 76, 82]. Due to the severity of the injury and the “cleanness” of the incision, both amputation and transection are highly reproducible, providing a solid ground to study the cellular and molecular mechanisms of spinal cord regeneration (Table 4). A variation of the transection injury is resection or ablation, wherein two incisions are made on the spinal cord, allowing the removal of a small segment that leaves a larger injury site between rostral and caudal stumps [29].

5.1. Regenerative capacity during metamorphosis

Evaluating tail regeneration is relatively straightforward, since restoration of the amputated tissue can be directly evaluated. Tissue sections can then be performed to determine whether their previous architecture is recovered. In Xenopus, regeneration after tail amputation occurs during all larval stages, except during the “refractory period” between stages 45-47, when no tissue regeneration is observed [5, 85]. Evaluating the extent of regeneration after a spinal cord transection is more complex because the regenerated tissue cannot be directly observed. Consequently, histological analyses of tissue sections and recovery of functional capacity are used to evaluate regenerative capacity after spinal cord transection.

Detailed histological analyses show that regeneration in stage 54 animals is slower than that observed in stage 50 animals, and that while in stage 50, most animals show a regenerated spinal cord and have recovered morphological aspects such as axonal reconnection and continuity of the ependymal canal, the morphology and tissue architecture of the regenerated spinal cord in stage 54 animals is less similar to the uninjured spinal cord. By stage 56, the number of axons crossing the lesion site is very limited, and is virtually inexistent in stage 66 [68].

13

Our laboratory has reported functional recovery of swimming ability at different stages of metamorphosis using a qualitative [33] and a quantitative approach [68]. In the qualitative analysis, swimming behavior after transection was classified into 4 categories: paraplegia, stimulated locomotion, circular swimming and coordinated swimming. By 20 days post transection (dpt), all transected stage 50 animals recover circular or coordinated swimming (out of 40). Stage 54 animals take longer to regenerate, with 9 out 13 animals recovering circular or coordinated swimming at approximately 30 dpt. Stage 58 animals remain mainly paraplegic or show only stimulated locomotion at 40 dpt, and only 2 out of 14 transected animals recover coordinated swimming by 40 dpt. Stage 66 froglets remain paraplegic throughout all measured time-points after transection [33]. In the quantitative approach, animals are left to rest for 5 minutes in a 15 cm diameter petri dish in 100 mL of buffer solution, and the distance they swim during the next 5 minutes is measured using video recording. Stages 50, 54, 56 and 66 were evaluated, and significant recovery of swimming ability was only observed for stages 50 and 54, and stage 54 animals regenerate at a slower rate than stage 50 animals. These results are in agreement with histological analyses [68].

Nonetheless, earlier studies have included observations regarding functional recovery after transection in metamorphic stages. Sims [82] evaluated functional recovery in stage 56 X. laevis tadpoles, and reported that animals regain coordinated swimming that involves the whole animal. Beattie and colleagues [4] observed successful functional recovery after transection in X. laevis between stages 50 and 62. Their observations state that behavioral recovery is inversely related to the state at which the injury is performed, with juvenile froglets being unable to recover normal swimming behavior. These results strongly support how useful a quantitative measure for functional recovery is to make results comparable across research groups.

5.2. Tissue and cellular origin of the regenerated spinal cord

At the tissue level, a cell fate study by Jonathan Slack's group used the tail amputation model to demonstrate that the origin of the regenerated spinal cord is the same tissue in the amputation stump, and that no dedifferentiation or transdifferentiation across tissue types occurs [34]. In this study, the neural plate from stage 14-15 wild type embryos was replaced by a transplant of the neural plate from an embryo that ubiquitously expressed GFP. Animals were allowed to grow into stage 50, and tail amputation was performed. Since GFP in the regenerated tail was restricted to the spinal cord, and similar experiments performed by labeling the notochord and the presomite plate showed that each tissue gave rise to its own regenerated part in the new tail [34]. Also in tail amputation experiments, we have shown that Sox2 is required for successful regeneration of the 14

whole tail, in spite of the fact that it is only expressed in spinal cord ependymal cells. Overexpression of a Sox2 dominant negative leads to reduced tail regeneration (of both the spinal cord and surrounding tissues), suggesting that the spinal cord could command tail regeneration [33].

After spinal cord injury, both ascending and descending axonal tracts are interrupted, extensive tissue damage caused by both the primary injury (traumatic event) and followed by secondary injury, in which recruitment of inflammatory cells and reactive astrocytes to the injury site leads to further tissue damage (reviewed in [19, 52, 67]). Two major processes required for successful spinal cord regeneration are axonal growth and neurogenesis. Growth of axons crossing the lesion site was first observed by Sims [82], followed by a study by Beattie and colleagues [4]. Axonal growth across the lesion site started at 5 days after transection in larvae, and through labeling using HRP (horseradish peroxidase), they determined that brainstem projecting neurons crossed the lesion site, and included serotonergic tracts. However, it was not until later studies performed by Ben Szaro’s group that specific supraspinal nuclei were identified to regenerate after spinal cord injury [38, 39].

Axonal growth and axonal regeneration are different concepts. Tuszynski & Steward [100] have provided a precise definition for these terms: “axonal growth” is the general term that may refer to the growing of a newly generated axon from a newborn neuron, as well as to the regrowth of an axon that has been severed. This latter process is defined as “axonal regeneration”, in which the axon of a pre-existing neuron that has been injured re-grows to innervate its previous target or a new one. While the work by Sims and Beattie showed that axonal growth across the injury site occurred after larval spinal cord injury, they did not determine whether this growth originated from newborn neurons in the brainstem or corresponded to axonal regeneration processes. Anterograde labeling using horseradish peroxidase (HRP) allowed the identification of brainstem projecting neurons growing across the injury site after transection in larval stages, including serotonergic fibers [4]. Axonal regeneration can be demonstrated using double retrograde labeling. Gibbs & Szaro used this technique in a spinal cord hemi-section model in stage 58-59 Xenopus animals [39]. They showed that supraspinal tracts from the reticular formation (superior, medial and inferior reticulospinal tracts, raphespinal tracts, and interstitiospinal tracts), can regenerate after a spinal cord hemi-section (Section 2.1). Reticulospinal tracts play a key role in motor control, and are highly conserved across vertebrates (Section 2.2) [94].

Neurogenesis has not been studied extensively after spinal cord injury. In the tail amputation model, the newly grown tail contains a new caudal spinal cord with a morphology similar to the 15

uninjured one, which suggests that neurogenesis takes place to reconstitute the new tail [35, 54]. However, detailed histological studies or lineage tracing experiments have not been performed in Xenopus. Recent work by our laboratory described the role of Sox2/3 expressing cells after spinal cord transection [68]. Sox2 is a key marker for neural stem and progenitor cells [24]. Histological analyses of tissue sections show that two days after transection of stage 50 tadpoles, an ablation gap of approximately 100 μm is observed between rostral and caudal stumps of the transected spinal cord [68]. Six days after the injury, groups of cells with a histological staining similar to ependymal cells populate the ablation gap, and eventually, the ependymal canal is reconstituted by 20 days after transection. The majority of ependymal canal lining cells are Sox2/3 positive, and BrdU incorporation experiments show that these cells proliferate in response to spinal cord transection in stage 50 tadpoles (in agreement with the results from histological analyses). While cell proliferation occurs in stage 66 juvenile froglets, these proliferating cells are mainly Sox2/3 negative cells and may give rise to non-neuronal cell types [68]. Furthermore, transient labeling of Sox3 expressing cells with green fluorescent protein (GFP), followed by two-photon microscopy, shows that cells with a radial glia-like morphology differentiate into cells with neuronal morphology at the injury site, by 5 days after transection, and that an increase in neural markers such as neurogenin 3 and neuroD (both at the mRNA level) is observed after transection in stage 50 animals. While double labeling and perdurable labeling experiments would provide definite evidence for neurogenesis after spinal cord transection in regenerative tadpoles, the evidence reported here suggests that Sox2/3 positive cells could be involved in neurogenesis. Furthermore, knock-down of Sox2 in the spinal cord or overexpression of a Sox2 dominant negative impairs spinal cord regeneration after transection [68].

A final appreciation from the cellular point of view is an observation made by Michel & Reier [66] and Beattie and colleagues [4], in which they observed an association of axons growing across the injury site in association with ependymoglial-like cells. Histological analyses performed by Muñoz et al. [68] also showed this association. These observations are in agreement with the process observed in the zebrafish and axolotl regeneration, where ependymoglial cells form bridges that act as a substrate for axonal growth [40, 111], suggesting that Sox2/3 positive cells could play a similar role during spinal cord regeneration in Xenopus tadpoles.

5.3. Molecular mechanisms Regarding the molecular mechanisms driving spinal cord regeneration, different approaches have been used to address this question. Before genome-wide approaches, the role of Xenopus as a 16

model for developmental biology served as a source for signaling pathways involved in early development, to evaluate whether they also played a role in regeneration. The Slack group has described several developmental pathways during tail regeneration: the BMP (Bone Morphogenetic Pathway) and Notch signaling cascades, both of which are involved in tail development [5]. They have also demonstrated that Wnt and FGF signaling are required for tail regeneration after amputation in stages 48-49, and that inhibitors of these pathways interrupt the regenerative process, while agents promoting Wnt activity lead to an accelerated growth of the regenerated tail [54]. On the other hand, the Levin group has described that V-ATPase proton pump-mediated changes in membrane potential are required during early stages of tail regeneration [1, 98], and also a requirement for histone deacetylases has been demonstrated [93, 99].

The advent of high-throughput approaches including microarrays, and then next-generation sequencing has allowed a high-scale approach into identifying novel mechanisms involved in tail and spinal cord regeneration. The Amaya group performed a microarray of the pre-metamorphic (stages 49-51) regenerating tail after amputation, by isolating the tip of the regenerating tail at different days after amputation [57]. This allowed the identification of several metabolic processes, which were further studied to identify a requirement for early ROS signaling for successful tail regeneration [58]. The Szaro group also obtained a transcriptomic profile using microarrays, evaluating gene expression in the hindbrain after spinal cord transection, under conditions in which thyroid hormone levels were manipulated to generate conditions permissive or inhibitory for functional spinal cord regeneration [38].

We have published the first transcriptome-wide profile using RNA-Seq of the response to spinal cord injury comparing a regenerative and a non-regenerative stage in Xenopus [53]. We isolated samples from a segment caudal to the injury site at 1, 2 and 6 days after transection or sham operation in stage 50 (regenerative) and stage 66 (non-regenerative) animals. We found extensive differences in the transcriptomes in response to transection when comparing the results from stage 50 and stage 66 animals. The most notable were the following. First, stage 50 tadpoles show the highest differential regulation of transcripts at 1 day after injury, while stage 66 froglets show a seemingly delayed regulation, with the highest number of differentially expressed transcripts detected at 6 days after injury, suggesting that the rate of the response is different between both stages and perhaps key for successful regeneration. Second, when comparing the repertoire of differentially regulated transcripts, a very different set was regulated in stage 50 when compared to stage 66. In fact, less than 20% of differentially expressed transcripts were regulated in both stages at all time points, while the rest were regulated in either stage in an exclusive time-dependent 17

manner. Third, we identified marked differences in the biological processes enriched among upand down-regulated transcripts from each stage. For example, more than half of the biological processes upregulated in tadpoles corresponded to metabolic processes at all time-points, whereas an upregulation of immune and inflammation processes was only observed in froglets. Other differentially regulated processes were neurogenesis, axonal regeneration, cell cycle, response to stress and developmental processes. This transcriptomic study has provided the first comparative profile between a regenerative and a non-regenerative stage of the same organism, and is now publicly available as a database at the molecular detail of which genes could explain why the tadpole can regenerate its spinal cord, and why this process fails in froglets. Amongst the identified transcripts, we found some conserved neurogenic transcription factors that show a different response in regenerative and non-regenerative stages. These include achaete-scute like 1 (ascl1) and neurod4, as well as genes expressed in the axonal growth cone, like the Netrin-1 receptor dcc.

6. Functional implications and conclusion

As shown by the previous sections, study of CNS regeneration in Xenopus has proven to be a fruitful source of information on the cellular and molecular mechanisms driving neural regeneration in different contexts. Regeneration of the optic nerve has shed light on axon regeneration, navigation and guidance mechanisms (Table 2), while the injury paradigms used in the brain shed more light on the role of neural progenitors, ependymal cells, mechanisms of damage control such as apoptotic cell expulsion, and differentiation of progenitor cells into functional cell types that restore visual function (Table 3). Furthermore, spinal cord regeneration models, including tail amputation and spinal cord transection have covered an even wider range of processes. Lineage tracing experiments have shown the origin of the cells in the regenerated tissue after tail amputation, and further characterization of the cell-types involved in regeneration highlight the importance of ependymal cells, which are mainly Sox2-positive. Furthermore, Sox2 expression is required for successful regeneration in the brain and the spinal cord [33, 63, 68]. Regarding the molecular mechanisms, several signaling pathways have been identified during tail [57] and spinal cord [53] regeneration transcriptomic databases have been generated and are publicly available for the study and finding of novel molecular mechanisms that may drive spinal cord regeneration (Table 4).

Therapeutic approaches to treat spinal cord injuries remain palliative and preventive, with no curative solution available yet [25]. In spite of this, preclinical research is moving in the direction of combinatory treatments which address different aspects of the spinal cord injury, 18

ranging from the use of neutralization of inhibitory factors, such as the antibody against Nogo-A [109], to physical therapy, which has been shown to be crucial for promoting plasticity after a CNS injury [101]. The need for combinatory treatments translates into the necessity of using diverse approaches to tackle the low regenerative capacity shown by humans and mammals in general. The use of rodents and higher vertebrates for spinal cord injury research has the advantage of being an anatomically more similar model to humans. As a drawback, the complexity of the CNS of higher vertebrates makes elucidating cellular and molecular mechanisms a harder task, for which a nonmammalian vertebrate like Xenopus arises as a model with a simpler CNS which is more amenable to approach. Furthermore, there is an increasing data overflow from findings generated in mammals which have been failed to be reproduced. This can be attributed to the injury model used, as well as the low number of animals that can be used when working with mammalian models. Xenopus larvae and froglets on the other hand, can be raised and used at much higher numbers, and as we have shown here, injury paradigms are relatively simple and highly reproducible. Cellular and molecular mechanisms have been elucidated by using Xenopus as a model that has an intermediate complexity level between cell culture, organoids, and mammalian vertebrates. While there are aspects we can only study in mammals, or learn in mammalian model organisms, there are others that we can only learn from regenerative organisms. We strongly believe that a collaborative and integrative approach to finding novel approaches to treat CNS injuries will lead to successful treatments.

CONFLICT OF INTEREST The authors have no financial conflicts or interests. ACKNOWLEDGEMENTS

We apologize to the authors whose work we couldn't cite because of the space constraints of this review. We thank Dr. Fernando Faunes, Víctor S. Tapia and the two anonymous reviewers for critical reading of our manuscript and valuable suggestions. Work in the authors’ laboratory is supported by research grants from: ICM-MINECON No. RC120003, CARE Chile UC-Centro de Envejecimiento y Regeneración PFB 12/2007, FONDECYT 1141162, and ICGEB (CRP/CHI-13– 01) (JL); L’Oréal Chile-UNESCO For Women in Science, and CONICYT Gastos Operacionales 21120769 (DLL). DLL and EMO are CONICYT PhD fellows.

19

REFERENCES [1]

[2]

[3]

[4] [5] [6] [7]

[8] [9] [10]

[11] [12] [13]

[14]

[15]

[16] [17] [18] [19] [20]

[21]

[22] [23]

D.S. Adams, A. Masi, M. Levin, H+ pump-dependent changes in membrane voltage are an early mechanism necessary and sufficient to induce Xenopus tail regeneration, Development 134 (2007) 1323-1335. F.T. Afshari, S. Kappagantula, J.W. Fawcett, Extrinsic and intrinsic factors controlling axonal regeneration after spinal cord injury., Expert reviews in molecular medicine 11 (2009) e37. N.M. Amin, P. Tandon, E. Osborne-Nishimura, F.L. Conlon, RNA-seq in the tetraploid Xenopus laevis enables genome-wide insight in a classic developmental biology model organism, Methods (San Diego, Calif) (2013) 1-12. M.S. Beattie, J.C. Bresnahan, G. Lopate, Metamorphosis alters the response to spinal cord transection in Xenopus laevis frogs, J. Neurobiol. 21 (1990) 1108-1122. C.W. Beck, B. Christen, J.M.W. Slack, Molecular Pathways Needed for Regeneration of Spinal Cord and Muscle in a Vertebrate, Developmental Cell 5 (2003) 429-439. C.G. Becker, T. Becker, Neuronal Regeneration from Ependymo-Radial Glial Cells: Cook, Little Pot, Cook!, Developmental Cell 32 (2015) 516-527. R.E. Blanco, A. López-Roca, J. Soto, J.M. Blagburn, Basic fibroblast growth factor applied to the optic nerve after injury increases long-term cell survival in the frog retina., The Journal of Comparative Neurology 423 (2000) 646-658. I.L. Blitz, J. Biesinger, X. Xie, K.W.Y. Cho, Biallelic genome modification in F(0) Xenopus tropicalis embryos using the CRISPR/Cas system., genesis 51 (2013) 827-834. I.L. Blitz, M.B. Fish, K. Cho, Leapfrogging: primordial germ cell transplantation permits recovery of CRISPR/Cas9-induced mutations in essential genes, Development (2016). R.C. Bohn, P.J. Reier, E.B. Sourbeer, Axonal interactions with connective tissue and glial substrata during optic nerve regeneration in Xenopus larvae and adults., Am. J. Anat. 165 (1982) 397-419. D.D. Brown, L. Cai, Amphibian metamorphosis, Dev. Biol. 306 (2007) 20-33. D.R. Buchholz, More similar than you think: Frog metamorphosis as a model of human perinatal endocrinology., Dev. Biol. 408 (2015) 188-195. D.R. Buchholz, S.-C.V. Hsia, L. Fu, Y.-B. Shi, A dominant-negative thyroid hormone receptor blocks amphibian metamorphosis by retaining corepressors at target genes., Mol. Cell. Biol. 23 (2003) 6750-6758. M.S. Chen, A.B. Huber, M.E. van der Haar, M. Frank, L. Schnell, A.A. Spillmann, F. Christ, M.E. Schwab, Nogo-A is a myelin-associated neurite outgrowth inhibitor and an antigen for monoclonal antibody IN-1., Nature 403 (2000) 434-439. C. Collart, N.D.L. Owens, L. Bhaw-Rosun, B. Cooper, E. De Domenico, I. Patrushev, A.K. Sesay, J.N. Smith, J.C. Smith, M.J. Gilchrist, High-resolution analysis of gene activity during the Xenopus mid-blastula transition, Development 141 (2014) 1927-1939. E.G. Contreras, M. Gaete, N. Sanchez, H. Carrasco, J. Larrain, Early requirement of Hyaluronan for tail regeneration in Xenopus tadpoles, Development 136 (2009) 2987-2996. G.R. Davis, P.B. Farel, Mauthner cells maintain their lumbar projection in adult frog., Neurosci. Lett. 113 (1990) 139-143. E.M. De Robertis, Y. Moriyama, The Chordin Morphogenetic Pathway, Elsevier Inc., 2016, 1-15 pp. J.F. Diaz Quiroz, K. Echeverri, Spinal cord regeneration: where fish, frogs and salamanders lead the way, can we follow?, Biochem. J. 451 (2013) 353-364. W. Dong, R.H. Lee, H. Xu, S. Yang, K.G. Pratt, V. Cao, Y.-K. Song, A. Nurmikko, C.D. Aizenman, Visual avoidance in Xenopus tadpoles is correlated with the maturation of visual responses in the optic tectum., J. Neurophysiol. 101 (2009) 803-815. M.V. Duprey-Díaz, J.M. Blagburn, R.E. Blanco, Changes in fibroblast growth factor-2 and FGF receptors in the frog visual system during optic nerve regeneration., J. Chem. Neuroanat. 46 (2012) 35-44. E.R. Elkan, The Xenopus Pregnancy Test., Br. Med. J. 2 (1938) 1253-1274.1252. T. Endo, J. Yoshino, K. Kado, S. Tochinai, Brain regeneration in anuran amphibians, Development, Growth & Differentiation 49 (2007) 121-129.

20

[24] [25] [26] [27] [28]

[29] [30] [31]

[32]

[33]

[34] [35] [36] [37]

[38] [39] [40]

[41] [42] [43] [44]

[45]

V. Episkopou, SOX2 functions in adult neural stem cells, Trends Neurosci. 28 (2005) 219221. V. Estrada, H.W. Müller, Spinal cord injury–there is not just one way of treating it, F1000prime reports (2014). J.W. Fawcett, Factors guiding regenerating retinotectal fibres in the frog Xenopus laevis, J. Embryol. Exp. Morphol. 90 (1985) 233-250. A.R. Ferguson, E.D. Stück, J.L. Nielson, Syndromics: A Bioinformatics Approach for Neurotrauma Research, Translational Stroke Research 2 (2011) 438-454. S. Filoni, S. Bernardini, S.M. Cannata, Differences in the decrease in regenerative capacity of various brain regions of Xenopus laevis are related to differences in the undifferentiated cell populations., Journal für Hirnforschung 36 (1995) 523-529. S. Filoni, L. Bosco, C. Cioni, Reconstitution of the Spinal Cord After Ablation in Larval Xenopus laevis, Acta Embryol. Morphol. Exp. 5 (1984) 109-129. H.S. Finkel, The Diagnosis of Pregnancy by the Aschheim-Zondek Test, N. Engl. J. Med. (1931). C.J. Forehand, P.B. Farel, Anatomical and behavioral recovery from the effects of spinal cord transection: dependence on metamorphosis in anuran larvae, The Journal of neuroscience : the official journal of the Society for Neuroscience 2 (1982) 654-652. J.L. Frontera, A.S. Cervino, L.D. Jungblut, D.A. Paz, Brain-derived neurotrophic factor (BDNF) expression in normal and regenerating olfactory epithelium of Xenopus laevis., Annals of anatomy = Anatomischer Anzeiger : official organ of the Anatomische Gesellschaft 198 (2015) 41-48. M. Gaete, R. Muñoz, N. Sánchez, R. Tampe, M. Moreno, E.G. Contreras, D. Lee-Liu, J. Larrain, Spinal Cord Regeneration in Xenopus Tadpoles Proceeds through Activation of Sox2 Positive Cells, Neural Dev 7 (2012) 13. C. Gargioli, Cell lineage tracing during Xenopus tail regeneration, Development 131 (2004) 2669-2679. C. Gargioli, J.M.W. Slack, Cell lineage tracing during Xenopus tail regeneration, Development 131 (2004) 2669-2679. R.M. GAZE, Regeneration of the optic nerve in Xenopus laevis., Q. J. Exp. Physiol. Cogn. Med. Sci. 44 (1959) 290-308. R.M. GAZE, M. Jacobson, A Study of the Retinotectal Projection during Regeneration of the Optic Nerve in the Frog, Proceedings of the Royal Society B: Biological Sciences 157 (1963) 420-448. K.M. Gibbs, S.V. Chittur, B.G. Szaro, Metamorphosis and the regenerative capacity of spinal cord axons in Xenopus laevis, Eur. J. Neurosci. 33 (2011) 9-25. K.M. Gibbs, B.G. Szaro, Regeneration of descending projections in Xenopus laevis tadpole spinal cord demonstrated by retrograde double labeling, Brain Res. 1088 (2006) 68-72. Y. Goldshmit, T.E. Sztal, P.R. Jusuf, T.E. Hall, M. Nguyen-Chi, P.D. Currie, Fgf-Dependent Glial Cell Bridges Facilitate Spinal Cord Regeneration in Zebrafish, J. Neurosci. 32 (2012) 7477-7492. I.A. Goodbrand, R.M. GAZE, Microglia in tadpoles of Xenopus laevis: normal distribution and the response to optic nerve injury., Anat. Embryol. (Berl). 184 (1991) 71-82. P. Graziadei, A.G. Monti-Graziadei, The influence of the olfactory placode on the development of the telencephalon inXenopus laevis, Neuroscience (1992). R.M. Harland, R.M. Grainger, Xenopus research: metamorphosed by genetics and genomics, Trends Genet. (2011). U. Hellsten, R.M. Harland, M.J. Gilchrist, D. Hendrix, J. Jurka, V. Kapitonov, I. Ovcharenko, N.H. Putnam, S. Shu, L. Taher, I.L. Blitz, B. Blumberg, D.S. Dichmann, I. Dubchak, E. Amaya, J.C. Detter, R. Fletcher, D.S. Gerhard, D. Goodstein, T. Graves, I.V. Grigoriev, J. Grimwood, T. Kawashima, E. Lindquist, S.M. Lucas, P.E. Mead, T. Mitros, H. Ogino, Y. Ohta, A.V. Poliakov, N. Pollet, J. Robert, A. Salamov, A.K. Sater, J. Schmutz, A. Terry, P.D. Vize, W.C. Warren, D. Wells, A. Wills, R.K. Wilson, L.B. Zimmerman, A.M. Zorn, R. Grainger, T. Grammer, M.K. Khokha, P.M. Richardson, D.S. Rokhsar, The genome of the Western clawed frog Xenopus tropicalis, Science 328 (2010) 633-636. L. Herrgen, O.P. Voss, C.J. Akerman, Calcium-Dependent Neuroepithelial Contractions Expel Damaged Cells from the Developing Brain, Developmental Cell 31 (2014) 599-613.

21

[46] [47] [48]

[49]

[50]

[51]

[52] [53]

[54] [55]

[56]

[57]

[58]

[59]

[60]

[61]

[62] [63]

[64] [65] [66]

E. Hibbard, Visual recovery following regeneration of the optic nerve through the oculomotor nerve root in Xenopus., Exp. Neurol. 19 (1967) 350-356. S. Ishibashi, K.L. Kroll, E. Amaya, A method for generating transgenic frog embryos., Methods in molecular biology (Clifton, NJ) 461 (2008) 447-466. J.B. Karpinka, J.D. Fortriede, K.A. Burns, C. James-Zorn, V.G. Ponferrada, J. Lee, K. Karimi, A.M. Zorn, P.D. Vize, Xenbase, the Xenopus model organism database; new virtualized system, data types and genomes., Nucleic Acids Res. 43 (2015) D756-763. M. Klinger, H. Diekmann, D. Heinz, C. Hirsch, S. Hannbeck von Hanwehr, B. Petrausch, T. Oertle, M.E. Schwab, C.A.O. Stuermer, Identification of two NOGO/RTN4 genes and analysis of Nogo-A expression in Xenopus laevis., Mol. Cell. Neurosci. 25 (2004) 205-216. D.M. Lang, B.P. Rubin, M.E. Schwab, CNS myelin and oligodendrocytes of the Xenopus spinal cord--but not optic nerve--are nonpermissive for axon growth, The Journal of … (1995). T. Laux, K. Fukami, M. Thelen, T. Golub, GAP43, MARCKS, and CAP23 modulate PI (4, 5) P2 at plasmalemmal rafts, and regulate cell cortex actin dynamics through a common mechanism, The Journal of cell … (2000). D. Lee-Liu, G. Edwards-Faret, V.S. Tapia, J. Larraín, Spinal cord regeneration: Lessons for mammals from non-mammalian vertebrates, genesis (2013) n/a-n/a. D. Lee-Liu, M. Moreno, L.I. Almonacid, V.S. Tapia, R. Munoz, J. von Marees, M. Gaete, F. Melo, J. Larrain, Genome-wide expression profile of the response to spinal cord injury in Xenopus laevis reveals extensive differences between regenerative and non-regenerative stages, Neural Development 9 (2014) 12. G. Lin, J.M.W. Slack, Requirement for Wnt and FGF signaling in Xenopus tadpole tail regeneration, Dev. Biol. 316 (2008) 323-335. Y. Liu, H. Yu, S.K. Deaton, B.G. Szaro, Heterogeneous Nuclear Ribonucleoprotein K, an RNA-Binding Protein, Is Required for Optic Axon Regeneration in Xenopus laevis, The Journal of neuroscience : the official journal of the Society for Neuroscience 32 (2012) 3563-3574. M.J. Lohka, J.L. Maller, Induction of nuclear envelope breakdown, chromosome condensation, and spindle formation in cell-free extracts., The Journal of Cell Biology 101 (1985) 518-523. N.R. Love, Y. Chen, B. Bonev, M.J. Gilchrist, L. Fairclough, R. Lea, T.J. Mohun, R. Paredes, L.A. Zeef, E. Amaya, Genome-wide analysis of gene expression during Xenopus tropicalis tadpole tail regeneration, BMC Dev Biol 11 (2011) 70. N.R. Love, Y. Chen, S. Ishibashi, P. Kritsiligkou, R. Lea, Y. Koh, J.L. Gallop, K. Dorey, E. Amaya, Amputation-induced reactive oxygen species are required for successful Xenopus tadpole tail regeneration, Nat Cell Biol 15 (2013) 222-228. N. Marsh-Armstrong, L. Cai, D.D. Brown, Thyroid hormone controls the development of connections between the spinal cord and limbs during Xenopus laevis metamorphosis., Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 165-170. R.I. Martinez-De Luna, R.Y. Ku, A.M. Aruck, F. Santiago, A.S. Viczian, D. San Mauro, M.E. Zuber, Müller glia reactivity follows retinal injury despite the absence of the glial fibrillary acidic protein gene in Xenopus, Dev. Biol. (2016). H.R. Maturana, J.Y. Lettvin, W.S. McCulloch, W.H. Pitts, Evidence That Cut Optic Nerve Fibers in a Frog Regenerate to Their Proper Places in the Tectum., Science (New York, NY) 130 (1959) 1709-1710. R.W. McDiarmid, R. Altig (Eds.), Integration: nervous and sensory systems, 1999, 149-169 pp. C.R. McKeown, P. Sharma, H.E. Sharipov, W. Shen, H.T. Cline, Neurogenesis is required for behavioral recovery after injury in the visual system of Xenopus laevis., The Journal of Comparative Neurology 521 (2013) 2262-2278. L. McKerracher, K.M. Rosen, MAG, myelin and overcoming growth inhibition in the CNS, Frontiers in molecular neuroscience 8 (2015) 967. V.M. McMurray, The development of the optic lobes in Xenopus laevis. The effect of repeated crushing of the optic nerve, J. Exp. Zool. (1954). M.E. Michel, P.J. Reier, Axonal-ependymal associations during early regeneration of the transected spinal cord in Xenopus laevis tadpoles, J. Neurocytol. 8 (1979) 529-548.

22

[67] [68] [69] [70]

[71] [72]

[73] [74]

[75] [76] [77]

[78]

[79] [80] [81]

[82] [83]

[84] [85] [86] [87]

[88] [89]

M. Moreno, K. Tapia, J. Larrain, Neural Regeneration in Xenopus Tadpoles during Metamorphosis, Xenopus Development (2014). R. Muñoz, G. Edwards-Faret, M. Moreno, N. Zuñiga, H. Cline, J. Larraín, Regeneration of Xenopus laevis spinal cord requires Sox2/3 expressing cells, Dev. Biol. (2015) 1-15. A.W. Murray, M.W. Kirschner, Cyclin synthesis drives the early embryonic cell cycle., Nature 339 (1989) 275-280. T. Nakayama, M.B. Fish, M. Fisher, J. Oomen-Hajagos, G.H. Thomsen, R.M. Grainger, Simple and efficient CRISPR/Cas9-mediated targeted mutagenesis in Xenopus tropicalis., genesis 51 (2013) 835-843. R. Nieuwenhuys, H.J. Ten Donkelaar, C. Nicholson, The Central Nervous System of Vertebrates, Springer-Verlag, Berlin Heidelberg, 1998, 2214 pp. P.D. Nieuwkoop, J. Faber, Normal table of Xenopus laevis (Daudin). A systematical and chronological survey of the development from the fertilized egg till the end of metamorphosis., Normal table of Xenopus laevis (Daudin) … (1956). A. Ostberg, J. Norden, Ultrastructural study of degeneration and regeneration in the amphibian tectum., Brain Res. 168 (1979) 441-455. L. Peshkin, M. Wühr, E. Pearl, W. Haas, R.M. Freeman, J.C. Gerhart, A.M. Klein, M. Horb, S.P. Gygi, M.W. Kirschner, On the Relationship of Protein and mRNA Dynamics in Vertebrate Embryonic Development., Developmental Cell 35 (2015) 383-394. S.S. POLACK, The xenopus pregnancy test., Can. Med. Assoc. J. 60 (1949) 159-161. P.J. Reier, H.F. Webster, Regeneration and remyelination of Xenopus tadpole optic nerve fibres following transection or crush, J. Neurocytol. 3 (1974) 591-618. W. Ríos-Muñoz, I. Soto, M.V. Duprey-Díaz, J. Blagburn, R.E. Blanco, Fibroblast growth factor 2 applied to the optic nerve after axotomy increases Bcl-2 and decreases Bax in ganglion cells by activating the extracellular signal-regulated kinase signaling pathway., J. Neurochem. 93 (2005) 1422-1433. B.P. Rubin, I. Dusart, M.E. Schwab, A monoclonal antibody (IN-1) which neutralizes neurite growth inhibitory proteins in the rat CNS recognizes antigens localized in CNS myelin., J. Neurocytol. 23 (1994) 209-217. G. Schlosser, N. Koyano-Nakagawa, C. Kintner, Thyroid hormone promotes neurogenesis in the Xenopus spinal cord, Dev. Dyn. 225 (2002) 485-498. R. Schweigreiter, The natural history of the myelin-derived nerve growth inhibitor Nogo-A., Neuron glia biology 4 (2008) 83-89. K.T. Sillar, D. Combes, S. Ramanathan, M. Molinari, J. Simmers, Neuromodulation and developmental plasticity in the locomotor system of anuran amphibians during metamorphosis., Brain research reviews 57 (2008) 94-102. R.T. Sims, Transection of the spinal cord in developing Xenopus laevis, J. Embryol. Exp. Morphol. 10 (1962) 115-126. J.H. Skene, R.D. Jacobson, G.J. Snipes, C.B. McGuire, J.J. Norden, J.A. Freeman, A protein induced during nerve growth (GAP-43) is a major component of growth-cone membranes., Science (New York, NY) 233 (1986) 783-786. J.H. Skene, M. Willard, Changes in axonally transported proteins during axon regeneration in toad retinal ganglion cells., The Journal of Cell Biology 89 (1981) 86-95. J.M.W. Slack, C.W. Beck, C. Gargioli, B. Christen, Cellular and molecular mechanisms of regeneration in Xenopus, Philos Trans R Soc Lond, B, Biol Sci 359 (2004) 745-751. J.M.W. Slack, G. Lin, Y. Chen, The Xenopus tadpole: a new model for regeneration research, Cell. Mol. Life Sci. 65 (2008) 54-63. I. Soto, J.J.C. Rosenthal, J.M. Blagburn, R.E. Blanco, Fibroblast growth factor 2 applied to the optic nerve after axotomy up-regulates BDNF and TrkB in ganglion cells by activating the ERK and PKA signaling pathways., J. Neurochem. 96 (2006) 82-96. R.W. Sperry, Optic nerve regeneration with return of vision in anurans, J. Neurophysiol. (1944). R.P. Stout, P.P. Graziadei, Influence of the olfactory placode on the development of the brain in Xenopus laevis (Daudin). I. Axonal growth and connections of the transplanted olfactory placode., Neuroscience 5 (1980) 2175-2186.

23

[90]

[91]

[92] [93] [94] [95]

[96]

[97] [98]

[99] [100] [101]

[102]

[103] [104]

[105]

[106]

[107]

[108]

[109]

L. Sun, M.M. Bertke, M.M. Champion, G. Zhu, P.W. Huber, N.J. Dovichi, Quantitative proteomics of Xenopus laevis embryos: expression kinetics of nearly 4000 proteins during early development., Scientific Reports 4 (2014) 4365. B.G. Szaro, Y.P. Loh, R.K. Hunt, Specific changes in axonally transported proteins during regeneration of the frog (Xenopus laevis) optic nerve., The Journal of neuroscience : the official journal of the Society for Neuroscience 5 (1985) 192-208. P. Tandon, F. Conlon, J.D. Furlow, M.E. Horb, Expanding the genetic toolkit in Xenopus: Approaches and opportunities for human disease modeling, Dev. Biol. (2016) 1-11. A.J. Taylor, C.W. Beck, Histone deacetylases are required for amphibian tail and limb regeneration but not development., Mech. Dev. 129 (2012) 208-218. H.J. ten Donkelaar, Organization of descending pathways to the spinal cord in amphibians and reptiles., Prog. Brain Res. 57 (1982) 25-67. H.J. ten Donkelaar, R. de Boer-van Huizen, Observations on the development of descending pathways from the brain stem to the spinal cord in the clawed toad Xenopus laevis., Anat. Embryol. (Berl). 163 (1982) 461-473. H.J. ten Donkelaar, R. de Boer-van Huizen, I. Bergervoet-Vernooy, Development, plasticity and regeneration of reticulospinal pathways: a case study in Xenopus laevis, Soc Neurosci Abstr, 1993, 617 pp. S. Thuret, L.D.F. Moon, F.H. Gage, Therapeutic interventions after spinal cord injury, Nat Rev Neurosci 7 (2006) 628-643. A.-S. Tseng, W.S. Beane, J.M. Lemire, A. Masi, M. Levin, Induction of vertebrate regeneration by a transient sodium current., The Journal of neuroscience : the official journal of the Society for Neuroscience 30 (2010) 13192-13200. A.-S. Tseng, K. Carneiro, J.M. Lemire, M. Levin, HDAC Activity Is Required during Xenopus Tail Regeneration, PLoS ONE 6 (2011) e26382. M.H. Tuszynski, O. Steward, Concepts and methods for the study of axonal regeneration in the CNS, Neuron 74 (2012) 777-791. R. van den Brand, J. Heutschi, Q. Barraud, J. DiGiovanna, K. Bartholdi, M. Huerlimann, L. Friedli, I. Vollenweider, E.M. Moraud, S. Duis, N. Dominici, S. Micera, P. Musienko, G. Courtine, Restoring voluntary control of locomotion after paralyzing spinal cord injury, Science (New York, NY) 336 (2012) 1182-1185. G.S. Vega-Meléndez, J.M. Blagburn, R.E. Blanco, Ciliary neurotrophic factor and fibroblast growth factor increase the speed and number of regenerating axons after optic nerve injury in adult Rana pipiens., J. Neurosci. Res. 92 (2014) 13-23. L.C. Wang, R.A. Rachel, R.C. Marcus, C.A. Mason, Chemosuppression of retinal axon growth by the mouse optic chiasm., Neuron 17 (1996) 849-862. G.B. Whitworth, B.C. Misaghi, D.M. Rosenthal, E.A. Mills, D.J. Heinen, A.H. Watson, C.W. Ives, S.H. Ali, K. Bezold, N. Marsh-Armstrong, F.L. Watson, Translational profiling of retinal ganglion cell optic nerve regeneration in Xenopus laevis, Dev. Biol. (2016) 1-14. M.A. Wilson, R.M. GAZE, I.A. Goodbrand, J.S. Taylor, Regeneration in the Xenopus tadpole optic nerve is preceded by a massive macrophage/microglial response., Anat. Embryol. (Berl). 186 (1992) 75-89. C.-J. Xu, J.-L. Wang, W.-L. Jin, The Neural Stem Cell Microenvironment: Focusing on Axon Guidance Molecules and Myelin-Associated Factors., Journal of molecular neuroscience : MN 56 (2015) 887-897. J. Yoshino, S. Tochinai, Functional regeneration of the olfactory bulb requires reconnection to the olfactory nerve in Xenopus larvae., Development, Growth & Differentiation 48 (2006) 15-24. J. Yoshino, S. Tochinai, Successful reconstitution of the non-regenerating adult telencephalon by cell transplantation in Xenopus laevis., Development, Growth & Differentiation 46 (2004) 523-534. A. Zemmar, O. Weinmann, Y. Kellner, X. Yu, R. Vicente, M. Gullo, H. Kasper, K. Lussi, Z. Ristic, A.R. Luft, M. Rioult-Pedotti, Y. Zuo, M. Zagrebelsky, M.E. Schwab, Neutralization of Nogo-A Enhances Synaptic Plasticity in the Rodent Motor Cortex and Improves Motor Learning in Vivo, The Journal of neuroscience : the official journal of the Society for Neuroscience 34 (2014) 8685-8698.

24

[110]

[111]

Y. Zhao, B.G. Szaro, The return of phosphorylated and nonphosphorylated epitopes of neurofilament proteins to the regenerating optic nerve of Xenopus laevis., The Journal of Comparative Neurology 343 (1994) 158-172. K.A. Zukor, D.T. Kent, S.J. Odelberg, Meningeal cells and glia establish a permissive environment for axon regeneration after spinal cord injury in newts, Neural Dev 6 (2011) 1.

25

FIGURE LEGENDS

Figure 1. Anatomy of the Xenopus brain. Diagrams showing a dorsal view of isolated brains from stage 50, 58 and 66 animals, depicting their comparative sizes in each stage (all drawn to scale), showing the different parts of the brain. a) Prosencephalon (green shades), or hindbrain, includes the olfactory bulb, telencephalon and diencephalon. The hypothalamus (not shown in the diagram), is located in the ventral part of the diencephalon. b) Mesencephalon (orange shades), or midbrain, includes the optic tectum (located dorsally), and tegmentum (located ventrally). c) Rhombencephalon (yellow shades), or hindbrain, includes the cerebellum and the medulla oblongata. Top-rostral; bottom-caudal. Scale bar: 1 mm.

26

Figure 2. Supraspinal nuclei and motor and sensory tracts in the brain. Lateral representation of the Xenopus post-metamorphic brain. Shown in light brown are the main supraspinal nuclei: 1) Reticular formation, which forms the reticulospinal tract; 2) Vestibular nuclei, which form the vestibulospinal tract and are also where Mauthner cells are located (not shown); 3) Locus coeruleus, which forms the coeruleospinal tract; 4) Red nucleus, or nucleus ruber, which forms the rubrospinal tract; 5) Hypothalamus, which forms the hypothalamospinal tract (represented by an arrowhead in the diagram). Motor tracts are shown in black, and sensory tracts in stipple. The entry point of the optic nerve is shown (and labeled) in the ventral part of the diencephalon. Left-rostral; right-caudal; Top-dorsal; bottom-ventral. This figure was adapted from The Central Nervous System of Vertebrates, by R. Niewenhuys, H.J. ten Donkelaar and C. Nicholson. Posters 4 and 5 from the “Extra Materials” (extras.springer.com) (pending permission) [71].

27

TABLES Table 1. Main supraspinal tracts in the brain stem. Supraspinal nucleus

Name of axonal tract

Reticular formation (mid/hindbrain)

Reticulospinal

main

Stage 50

Stage 58

Stage 66

References

Innervation fore hindlimbs.

Early arising axons innervate the tail.

Early arising axons remodel (extend collaterals) to innervate hindlimb motoneurons; new wave of axons arises specifically to innervate hindlimb motoneurons (lumbar cord).

Innervation of fore and hindlimbs. Mauthner cells persist innervating the lumbar spinal cord.

[17, 95]

of and

Vestibular nuclei (hindbrain)

Vestibulospinal (and Mauthner cells)

Locus coeruleus (midbrain)

Coeruleospinal

Less developed

Maturation

Mature

[95]

Red nucleus (midbrain)

Rubrospinal

Start of spinal cord invasion

Reach lumbar cord, when hindlimb locomotion starts

Mature

[96]

Hypothalamus (forebrain)

Hypothalamospinal

No innervation of spinal cord yet

Reach cord

Mature

[95]

the

spinal

A list of the main supraspinal nuclei present in the brain stem (and their location shown in parentheses), including the name of the main axonal tract that arises from each nucleus. Columns “stage 50, 58 and 66” indicate the stage of maturation in which each axonal tract is during metamorphosis.

28

Table 2. Optic nerve regeneration in anuran frogs Regenerative capacity

Factor description

Intrinsic factors

Axonally transported proteins Transcriptomics

Throughout complete lifespan [36, 88].

Immune system cells Extrinsic factors Glial cells

Other factors Signaling Functional recovery

Overall description 1) GAP43, a major component of the growth cone membrane, and substrate for PKC (role in actin cytoskeleton dynamics). [84] 2) hnRNP K is required for axonal regeneration (its knock-down leads to lower GAP43 levels) [55]. RNA-Seq of X. laevis RGCs after optic nerve crush shows differential regulation of: ⬆tubulin, gap43, klf6, jak/stat; ⬇klf4 [104]. 1) Macrophages and microglia are recruited within the first 24 hours after injury [41]. 2) Post-synaptic sites in the tectum are removed by phagocytosis after denervation [73]. 1) Optic nerve-derived oligodendrocytes are permissive for axonal growth [50]. 2) Spinal cord-derived oligodendrocytes are inhibitory [50]. 3) Nogo-A is a myelin-inhibitory element [14, 80]. 4) The Xenopus genome does not have GFAP [60]. Axon fiber-fiber interactions are required for navigation and guidance [26]. 1) bFGF promotes retinal ganglion cell survival (mediated by BDNF, MAPK and PKA) [7, 77, 87]. 2) CNTF enhances axonal growth [102]. The first tectal response is observed 20 days after optic nerve section in Rana temporaria [37].

29

Table 3. Brain regeneration in Xenopus Regenerative capacity

Larval stages stages 47-54 [23].

Summary

Regenerative stage

Non-regenerative stage

Damage control mechanisms

Ependymal cell proliferation and rapid closure of interrupted ventricles (stages 47-53) [108]. Expulsion of damaged cells (stages 4448) [45].

Ependymal cells proliferate but do not close interrupted ventricles (postmetamorphic) [108].

Neural stem/progenitor cells

Signaling

Functional recovery

Large and extensive area of actively proliferating undifferentiated cells (stage 48) [28]. Sox2+ progenitor cells differentiate into neurons after injury (stage 47) [63]. Successful telencephalon regeneration requires innervation from olfactory placode (stage 53) [42, 89]. Visual avoidance behavior is recovered within 7-8 days after unilateral optic tectum lesion, but not after complete ablation (stage 47) [20].

Less proliferation during prometamorphosis (stages 55-56), and progressive decrease in postmetamorphic froglets (stage 66 onwards) [28]. -

-

30

Table 4. Spinal cord regeneration in Xenopus Regenerative capacity

Summary

Tissue regeneration

Larval stages (functional recovery up to at least stage 54 [68]; refractory period for tail regeneration between stages 45-47 [5, 85]).

Functional recovery

Signaling

Transcriptomics

Regenerative stage Each tissue gives rise to its own regenerated part in the new tail after amputation (stage 50) [34]. Proliferation of Sox2+ cells is required for tail and spinal cord regeneration. Sox2+ cells differentiate into cells with neuronal phenotype [33, 68]. Axonal growth and regeneration of reticulospinal tracts (stages 58-59 after spinal cord hemisection) [39]. Quantitative report of functional recovery of swimming is only observed till stage 54 [68]. Qualitative reports describe decreasing functional recovery as metamorphic climax is achieved [4, 33, 82]. BMP, Notch [5], Wnt and FGF [54] signaling pathways are required for tail regeneration (stages 48-49). V-ATPase proton-pump and histone deacetylases are also required for tail regeneration (stages 40-41) [1, 98]. Microarray allows differential regulation of metabolic processes allows identification of ROS signaling for successful tail regeneration [57, 58]. Microarray of the hindbrain after spinal cord transection and thyroid hormone synthesis inhibitor treatment (methimazole) [38]. RNA-Seq of spinal cord 1, 2 and 6 days after transection in stage 50 tadpoles [53].

Non-regenerative stage Delayed cell proliferation is observed (at 6 days post-transection). These cells are mostly Sox2- [68]. Limited to no axonal growth across the lesion site (post-metamorphic) [53, 68]. No significant quantitative recovery of swimming is observed from stage 56 onwards [68]. No qualitative functional recovery is observed once metamorphosis concludes [4, 33, 82]. -

-

-

Microarray of the hindbrain after spinal cord transection and T3 hormone treatment [38]. RNA-Seq of spinal cord 1, 2 and 6 days after transection in stage 66 froglets [53].

31