SDF-1α/CXCR4 signaling mediates digit tip regeneration promoted by BMP-2

Previously we demonstrated that BMP signaling is required for endogenous digit tip regeneration, and that treatment with BMP-2 or -7 induces a regenerative response following amputation at regeneration-incompetent levels (Yu et al., 2010 and Yu et al., 2012). Both endogenous regeneration and BMP-induced regeneration are associated with the transient formation of a blastema, however the formation of a regeneration blastema in mammals is poorly understood. In this study, we focus on how blastema cells respond to BMP signaling during neonatal digit regeneration in mice. First, we show that blastema cells retain regenerative properties after expansion in vitro, and when re-introduced into the amputated digit, these cells display directed migration in response to BMP-2. However, in vitro studies demonstrate that BMP-2 alone does not influence blastema cell migration, suggesting a requirement of another pivotal downstream factor for cell recruitment. We show that blastema cell migration is stimulated by the cytokine, SDF-1α, and that SDF-1α is expressed by the wound epidermis as well as endothelial cells of the blastema. Blastema cells express both SDF-1α receptors, CXCR4 and CXCR7, although the migration response is inhibited by the CXCR4-specific antagonist, AMD3100. Mice treated with AMD3100 display a partial inhibition of skeletal regrowth associated with the regeneration response. We provide evidence that BMP-2 regulates Sdf-1α expression in endothelial cells but not cells of the wound epidermis. Finally, we show that SDF-1α-expressing COS1 cells engrafted into a regeneration-incompetent digit amputation wound resulted in a locally enhanced population of CXCR4 positive cells, and induced a partial regenerative response. Taken together, this study provides evidence that one downstream mechanism of BMP signaling during mammalian digit regeneration involves activation of SDF-1α/CXCR4 signaling by endothelial cells to recruit blastema cells.     

Hyperbaric Oxygen Promotes Proximal Bone Regeneration and Organized Collagen Composition during Digit Regeneration

Oxygen is critical for optimal bone regeneration. While axolotls and salamanders have retained the ability to regenerate whole limbs, mammalian regeneration is restricted to the distal tip of the digit (P3) in mice, primates, and humans. Our previous study revealed the oxygen microenvironment during regeneration is dynamic and temporally influential in building and degrading bone. Given that regeneration is dependent on a dynamic and changing oxygen environment, a better understanding of the effects of oxygen during wounding, scarring, and regeneration, and better ways to artificially generate both hypoxic and oxygen replete microenvironments are essential to promote regeneration beyond wounding or scarring. To explore the influence of increased oxygen on digit regeneration in vivo daily treatments of hyperbaric oxygen were administered to mice during all phases of the entire regenerative process. Micro-Computed Tomography (μCT) and histological analysis showed that the daily application of hyperbaric oxygen elicited the same enhanced bone degradation response as two individual pulses of oxygen applied during the blastema phase. We expand past these findings to show histologically that the continuous application of hyperbaric oxygen during digit regeneration results in delayed blastema formation at a much more proximal location after amputation, and the deposition of better organized collagen fibers during bone formation. The application of sustained hyperbaric oxygen also delays wound closure and enhances bone degradation after digit amputation. Thus, hyperbaric oxygen shows the potential for positive influential control on the various phases of an epimorphic regenerative response.     

Regenerative ability of double-half and half upper arms in the newt, Notophthalmus viridescens

The upper arms of adult newts (Notophthalmus viridescens) were surgically manipulated to create double-half dorsal, double-half ventral, double-half anterior, and double-half posterior upper arms, and longitudinal half-dorsal, half-ventral, half-anterior, and half-posterior upper arms. Amputation through the double-half upper arms usually failed to elicit normal distal regeneration, despite the fact that an apparently normal regeneration blastema was initially formed. Instead, regeneration in these cases was limited to the formation of a variable number of small cartilage elements. On the basis of these results it is concluded that a complete limb circumference is required for distal transformation in newts, in addition to the well-established requirements for a wound epidermis, adequate innervation and dedifferentiation leading to blastema formation. A model for the sequential generation of new parts of the limb pattern during distal transformation from a complete circumference is presented. This model can also account for the occurrence of normal early stages of regeneration in double-half upper arms. Half upper arms which were amputated immediately were shown to develop single, complete regenerates. If amputation of half upper arms was delayed three or more weeks to permit complete wound healing, a supernumerary limb from the lateral wound surface sometimes developed in addition to a complete, single limb from the distal amputation surface.     

Rethinking the blastema

The phenomenon of tissue regeneration has been well documented across many species. Although some possess the capacity to completely restore an entire amputated limb, others are limited to just the distal digit tip. Initiation of limb regeneration has been described to start with the formation of a blastema, the composition of which has long been thought to consist of undifferentiated pluripotent cells derived through the process of dedifferentiation. Competing theories have been proposed, however, including cellular contributions through transdifferentiation and tissue-specific stem cells. Recent studies have now begun to shed light on this controversy, demonstrating tissue resident stem cells to be an evolutionarily conserved measure for limb regeneration.     

Higher Vertebrates Do Not Regenerate Digits and Legs Because the Wound Epidermis Is Not Functional

The necessity of injury, nerves, and wound epidermis for urodele limb regeneration is well accepted. Whether one or more of these three factors is limiting in amputated nonregenerating limbs of other vertebrates is a problem area in need of resolution. One view, that higher vertebrates possess inadequate innervation for limb regeneration to occur, is not strongly supported by experimental results. Superinnervation of lizard and mammalian limbs fails to elicit limb regeneration. Furthermore, in the well-known cases of mammalian regeneration, deer antlers and rabbit ears, a nerve requirement has not been demonstrated.
In urodeles, the wound epidermis has recently been shown to have the role of maintaining dedifferentiated cells of the amputated limb stump in the cell cycle. The result of this wound epidermal stimulus is a sufficient number of cell divisions such that blastema formation occurs.
We postulate that in amputated limbs of higher vertebrates, the wound epidermis is nonfunctional. Dedifferentiated or undifferentiated cells are not maintained in the cell cycle and blastema formation therefore does not occur. Instead, tissue regeneration occurs precociously due to lack of a cycling stimulus. The scar tissue which forms at the limb tips of nonregenerating vertebrates is the result of a nonfunctional wound epidermis.     

Axonal regrowth is impaired during digit tip regeneration in mice

Mice are intrinsically capable of regenerating the tips of their digits after amputation. Mouse digit tip regeneration is reported to be a peripheral nerve-dependent event. However, it is presently unknown what types of nerves and Schwann cells innervate the digit tip, and to what extent these cells regenerate in association with the regenerative response. Given the necessity of peripheral nerves for mammalian regeneration, we investigated the neuroanatomy of the unamputated, regenerating, and regenerated mouse digit tip. Using immunohistochemistry for β-III-tubulin (β3T) or neurofilament H (NFH), substance P (SP), tyrosine hydroxylase (TH), myelin protein zero (P0), and glial fibrillary acidic protein (GFAP), we identified peripheral nerve axons (sensory and sympathetic), and myelinating- and non-myelinating-Schwann cells. Our findings show that the digit tip is innervated by two digital nerves that each bifurcate into a bone marrow (BM) and connective tissue (CT) branch. The BM branches are composed of sympathetic axons that are ensheathed by non-myelinating-Schwann cells whereas the CT branches are composed of sensory and sympathetic axons and are ensheathed by myelinating- and non-myelinating-Schwann cells. The regenerated digit neuroanatomy differs from unamputated digit in several key ways. First, there is 7.5 fold decrease in CT branch axons in the regenerated digit compared to the unampuated digit. Second, there is a 5.6 fold decrease in myelinating-Schwann cells in the regenerated digit compared to the unamputated digit that is consistent with the decrease in CT branch axons. Importantly, we also find that the central portion of the regenerating digit blastema is aneural, with axons and Schwann cells restricted to peripheral and distal blastema regions. Finally, we show that even with impaired innervation, digits maintain the ability to regenerate after re-amputation. Taken together, these data indicate that nerve regeneration is impaired in the context of mouse digit tip regeneration.     

Responses to amputation of denervated ambystoma limbs containing aneurogenic limb grafts

The developing neural tubes and associated neural crest cells were removed from stage 30 Ambystoma maculatum embryos to obtain larvae with aneurogenic forelimbs. Forelimbs were allowed to develop to late 3 digit or early 4 digit stages. Limbs amputated through the mid radius-ulna regenerated typically in the aneurogenic condition. Experiments were designed to test whether grafts of aneurogenic limb tissues would rescue denervated host limb stumps into a regeneration response. In Experiment 1, aneurogenic limbs were removed at the body wall and grafted under the dorsal skin of the distal end of amputated forelimbs of control, normally innervated limbs of locally collected Ambystoma maculatum or axolotl (Ambystoma mexicanum) larvae. In Experiment 1, at the time of grafting or 1, 2, 3, 4, 5, 7, or 8 days after grafting, aneurogenic limbs were amputated level with the original host stump. At 7 and 8 days, this amputation included removing the host blastema adjacent to the graft. The host limb was denervated either one day after grafting or on the day of graft amputation. These chimeric limbs only infrequently exhibited delayed blastema formation. Thus, not only did the graft not rescue the host, denervated limb, but the aneurogenic limb tissues themselves could not mount a regeneration response. In Experiment 2, the grafted aneurogenic limb was amputated through its mid-stylopodium at 3, 4, 5, 7, or 8 days after grafting. By 7 and 8 days after grafting, the host limb stump exhibited blastema formation even with the graft extending out from under the dorsal skin. The host limb was denervated at the time of graft amputation. When graft limbs of Experiment 2 were amputated and host limbs were denervated on days 3, 4, or 5, host regeneration did not progress and graft regeneration did not occur. But, when graft limbs were amputated on days 7 or 8 with concomitant denervation of the host limb, regeneration of the host continued and graft regeneration occurred. Thus, regeneration of the graft was correlated with acquisition of nerve-independence by the host limb blastema. In Experiment 3, aneurogenic limbs were grafted with minimal injury to the dorsal skin of neurogenic hosts. When neurogenic host limbs were denervated and the aneurogenic limbs were amputated through the radius/ulna, regeneration of the aneurogenic limb occurred if the neurogenic limb host was not amputated, but did not occur if the neurogenic limb host was amputated. Results of Experiment 3 indicate that the inhibition of aneurogenic graft limb regeneration on a denervated host limb is correlated with substantial injury to the host limb. In Experiment 4, aneurogenic forelimbs were amputated through the mid-radius ulna and pieces of either peripheral nerve, muscle, blood vessel, or cartilage were grafted into the distal limb stump or under the body skin immediately adjacent to the limb at the body wall. In most cases, peripheral nerve inhibited regeneration, blood vessel tissue sometimes inhibited, but other tissues had no effect on regeneration. Taken together, the results suggest: (1) Aneurogenic limb tissues do not produce the neurotrophic factor and do not need it for regeneration, and (2) there is a regeneration-inhibiting factor produced by the nerve-dependent limb stump/blastema after denervation that prevents regeneration of aneurogenic limbs.     

Temperature-Sensitive Mutations That Cause Stage-Specific Defects in Zebrafish Fin Regeneration

When amputated, the fins of adult zebrafish rapidly regenerate the missing tissue. Fin regeneration proceeds through several stages, including wound healing, establishment of the wound epithelium, recruitment of the blastema from mesenchymal cells underlying the wound epithelium, and differentiation and outgrowth of the regenerate. We screened for temperature-sensitive mutations that affect the regeneration of the fin. Seven mutations were identified, including five that fail to regenerate their fins, one that causes slow growth during regeneration, and one that causes dysmorphic bumps or tumors to develop in the regenerating fin. reg5 mutants fail to regenerate their caudal fins, whereas reg6 mutants develop dysmorphic bumps in their regenerates at the restrictive temperature. Temperature-shift experiments indicate that reg5 and reg6 affect different stages of regeneration. The critical period for reg5 occurs during the early stages of regeneration before or during establishment of the blastema, resulting in defects in subsequent growth of the blastema and failure to differentiate bone-forming cells. The critical period for reg6 occurs after the onset of bone differentiation and during early stages of regenerative outgrowth. Both reg5 and reg6 also show temperature-sensitive defects in embryonic development or in ontogenetic outgrowth of the juvenile fin.     

Regeneration of Soft Tissues Is Promoted by MMP1 Treatment after Digit Amputation in Mice

The ratio of matrix metalloproteinases (MMPs) to the tissue inhibitors of metalloproteinases (TIMPs) in wounded tissues strictly control the protease activity of MMPs, and therefore regulate the progress of wound closure, tissue regeneration and scar formation. Some amphibians (i.e. axolotl/newt) demonstrate complete regeneration of missing or wounded digits and even limbs; MMPs play a critical role during amphibian regeneration. Conversely, mammalian wound healing re-establishes tissue integrity, but at the expense of scar tissue formation. The differences between amphibian regeneration and mammalian wound healing can be attributed to the greater ratio of MMPs to TIMPs in amphibian tissue. Previous studies have demonstrated the ability of MMP1 to effectively promote skeletal muscle regeneration by favoring extracellular matrix (ECM) remodeling to enhance cell proliferation and migration. In this study, MMP1 was administered to the digits amputated at the mid-second phalanx of adult mice to observe its effect on digit regeneration. Results indicated that the regeneration of soft tissue and the rate of wound closure were significantly improved by MMP1 administration, but the elongation of the skeletal tissue was insignificantly affected. During digit regeneration, more mutipotent progenitor cells, capillary vasculature and neuromuscular-related tissues were observed in MMP1 treated tissues; moreover, there was less fibrotic tissue formed in treated digits. In summary, MMP1 was found to be effective in promoting wound healing in amputated digits of adult mice.     

Temporal distribution of 5BrdU-labelled cells suggests that most injured tissues contribute proliferating cells for the regeneration of the tail and limb in lizard

After tail and limb amputation in lizard, injection of 5BrdU for 6 days produces immunolabelled cells in most tissues of tail and limb stumps. After further 8 and 16 days, and 14 and 22 days of regeneration, numerous 5BrdU-labelled cells are detected in regenerating tail and limb, derived from most stump tissues. In tail blastema cone at 14 days, sparse-labelled cells remain in proximal dermis, muscles, cartilaginous tube and external layers of wound epidermis but are numerous in the blastema. In apical regions at 22 days of regeneration, labelled mesenchymal cells are sparse, while the apical wound epidermis contains numerous labelled cells in suprabasal and external layers, indicating cell accumulation from more proximal epidermis. Cell proliferation dilutes the label, and keratinocytes take 8 days to migrate into corneous layers. In healing limbs, labelled cells remain sparse from 14 to 22 days of regeneration in wound epidermis and repairing tissues and little labelling dilution occurs indicating low cell proliferation for local tissue repair but not distal growth. Labelled cells are present in epidermis, intermuscle and peri-nerve connectives, bone periosteum, cartilaginous callus and sparse fibroblasts, leading to the formation of a scarring outgrowth. Resident stem cells and dedifferentiation occur when stump tissues are damaged.     

Morphological stages of regeneration in the planarian Dugesia tigrina: A light and electron microscopic study

A precise sequence of four morphological stages of head regeneration in the planarian Dugesia tigrina has been determined by light and electron microscopy. Each stage is identified by a particular morphogenetic process: I, wound healing; II, blastema development; III, growth; IV, differentiation. A wound epidermis consisting of a thin, sheet-like layer of cells, rapidly forms from undamaged epidermal cells at the wound margin. The early blastema is comprised of neoblasts which mature into regeneration cells. The maturational changes include the appearance of a nucleolus, nuclear pores, and perinuclear dense aggregates of granulofibrillar material in these cells. These elements are not evident in the neoblasts of the younger blastema. No mitotic cells are encountered in the blastema or wound epidermis. Cytoplasmic expansion of the regeneration cells is correlated with the formation of numerous microtubules radiating from a juxtanuclear centrosphere. During differentiation of muscle cells, distended, granule-studded cisternae, having moderately fibrillar contents, are regularly disposed adjacent to small patches of myofilaments. Presumptive epidermal cells are recognized by prominent “islands” of finely fibrillar cytoplasm. These cytoplasmic zones persist for a time during definitive differentiation when Golgi bodies, vacuoles, mucous droplets, and rhabdites become evident. The newly formed epidermal cells become inserted among the cells of the wound epidermis. Thus, cells of both the blastema and of the wound epidermis contribute to the reconstituted epidermis.     

Msx1‐2 immunolocalization in the regenerating tail of a lizard but not in the scarring limb suggests its involvement in the process of regeneration

The immunolocalization of the muscle segmental homoeobox protein Msx1‐2 of 27–34 kDa in the regenerating tail blastema of a lizard shows prevalent localization in the apical ependyma of the regenerating spinal cord and less intense labelling in the wound epidermis, in the apical epidermal peg (AEP), and in the regenerating segmental muscles. The AEP is a micro‐region of the regenerating epidermis located at the tail tip of the blastema, likely corresponding to the AEC of the amphibian blastema. No immunolabelling is present in the wound epidermis and scarring blastema of the limb at 18–21 days of regeneration, except for sparse repairing muscles. The presence of a proximal–distal gradient of Msx1‐2 protein, generated from the apical ependyma, is suggested by the intensity of immunolabelling. The AEP and the ependyma are believed to induce and maintain tail regeneration, and this study suggests that Msx1‐2 proteins are components of the signalling system that maintains active growth of the tail blastema. The lack of activation and production of Msx1‐2 protein in the limb are likely due to the intense inflammatory reaction following amputation. This study confirms that, like during regeneration in fishes and amphibians, also the blastema of lizards utilizes common signalling pathways for maintaining regeneration.     

Plasticity of photoreceptor-generating retinal progenitors revealed by prolonged retinoic acid exposure

Background

Epimorphic regeneration results in the restoration of lost tissues and structures from an aggregation of proliferating cells known as a blastema. Among amniotes the most striking example of epimorphic regeneration comes from tail regenerating lizards. Although tail regeneration is often studied in the context of ecological costs and benefits, details of the sequence of tissue-level events are lacking. Here we investigate the anatomical and histological events that characterize tail regeneration in the leopard gecko, Eublepharis macularius.

Results

Tail structure and tissue composition were examined at multiple days following tail loss, revealing a conserved pattern of regeneration. Removal of the tail results in a consistent series of morphological and histological events. Tail loss is followed by a latent period of wound healing with no visible signs of regenerative outgrowth. During this latent period basal cells of the epidermis proliferate and gradually cover the wound. An additional aggregation of proliferating cells accumulates adjacent to the distal tip of the severed spinal cord marking the first appearance of the blastema. Continued growth of the blastema is matched by the initiation of angiogenesis, followed by the re-development of peripheral axons and the ependymal tube of the spinal cord. Skeletal tissue differentiation, corresponding with the expression of Sox9, and muscle re-development are delayed until tail outgrowth is well underway.

Conclusions

We demonstrate that tail regeneration in lizards involves a highly conserved sequence of events permitting the establishment of a staging table. We show that tail loss is followed by a latent period of scar-free healing of the wound site, and that regeneration is blastema-mediated. We conclude that the major events of epimorphic regeneration are highly conserved across vertebrates and that a comparative approach is an invaluable biomedical tool for ongoing regenerative research.     

Epimorphic vs. tissue regeneration in Xenopus forelimbs.

Postmetamorphic froglets of Xenopus laevis regenerate hypomorphic unbranched spikes from amputated arm stumps. These are composed primarily of cartilage, produced from blastemalike structures sparsely populated with cells and rich in connective tissue. Some consider these outgrowths to be an example of epimorphic regeneration produced from blastemas, albeit deficient ones. Others interpret them as a case of tissue regeneration derived from fibroblastemas augmented by chondrocytes and periosteal and perichondrial fibroblasts. To resolve these alternatives, forelimbs were amputated proximal to the wrist, skinned, and inserted through the body wall into the abdominal cavity. In the absence of skin, epidermal wound healing failed to occur and blastemas could not develop. After 2 months, by which time controls had regenerated spikes averaging 3.38 mm long, the denuded stumps had not given rise to outgrowths. They typically developed cartilaginous caps on the severed ends of the radius-ulna, and in rare cases formed amorphous growths of cartilage. If blastema formation is considered diagnostic of epimorphic regeneration and tissue regeneration can proceed in the absence of epidermal wound healing and blastema formation, these findings lead to the conclusion that Xenopus limb regeneration is epimorphic.     

Nerve-induced ectopic limb blastemas in the Axolotl are equivalent to amputation-induced blastemas

Adult urodeles (salamanders) are unique in their ability to regenerate complex organs perfectly. The recently developed Accessory Limb Model (ALM) in the axolotl provides an opportunity to identify and characterize the essential signaling events that control the early steps in limb regeneration. The ALM demonstrates that limb regeneration progresses in a stepwise fashion that is dependent on signals from the wound epidermis, nerves and dermal fibroblasts from opposite sides of the limb. When all the signals are present, a limb is formed de novo. The ALM thus provides an opportunity to identify and characterize the signaling pathways that control blastema morphogenesis and limb regeneration. Our previous study provided data on cell contribution, cell migration and nerve dependency indicating that an ectopic blastema is equivalent to an amputation-induced blastema. In the present study, we have determined that formation of both ectopic blastemas and amputation-induced blastemas is regulated by the same molecular mechanisms, and that both types of blastema cells exhibit the same functions in controlling growth and pattern formation. We have identified and validated five marker genes for the early stages of wound healing, dedifferentiation and blastema formation, and have discovered that the expression of each of these markers is the same for both ectopic and amputation-induced blastemas. In addition, ectopic blastema cells interact coordinately with amputation-induced blastema cells to form a regenerated limb. Therefore, the ALM is appropriate for identifying the signaling pathways regulating the early events of tetrapod limb regeneration.     

Epidermal downgrowths in regenerating rabbit ear holes.

Rabbits are unique among mammals in that their ears can regenerate tissues from the margins of full thickness holes which grow in and completely fill the opening in about two months. The circular blastema that forms around the edges of the hole differentiates a new sheet of cartilage as it regenerates in a centripetal direction. Similar holes in other mammals fail to regenerate and form scar tissue instead of a blastema. Histological studies of the healing around the edges of rabbit ear holes reveal that during the second week, when the epidermis is completing its migration across the wound from the opposite sides of the ear, conspicuous tongues of epidermal cells grow down into the underlying tissues at the edges of the wound. These epidermal downgrowths are situated between the original intact dermis of the skin and the more central tissues which give rise to the blastema. Such downgrowths are of a transient nature, and are no longer found once the blastema rounds up toward the end of the second week. Since they are not found in the healing of similar wounds in rabbit ears prevented from regenerating by prior removal of their cartilaginous sheets, nor in the naturally nonregenerating ears of sheep and dogs, it is considered that these downgrowths of healing epidermis may play a role in the unusual regenerative response of ear tissues in the rabbit.     

The regeneration of axolotl limbs covered by frog skin

Forearm skin of Stage XXIV Rana pipiens, which cannot regenerate limbs, was removed and placed upon the skinned forearms of young axolotls. The axolotl limbs were amputated immediately through the level of the grafts. Frog epidermis migrated to cover the amputation surface. Dedifferentiation and early blastema formation occurred beneath the frog wound epidermis. Limb regeneration continued, but in time axolotl epidermis overgrew the frog epidermis. The experiment shows that epidermis from nonregenerating frog limbs is still capable of supporting typical epimorphic regeneration.     

Retinoic acid signaling controls the formation, proliferation and survival of the blastema during adult zebrafish fin regeneration

Adult teleosts rebuild amputated fins through a proliferation-dependent process called epimorphic regeneration, in which a blastema of cycling progenitor cells replaces the lost fin tissue. The genetic networks that control formation of blastema cells from formerly quiescent stump tissue and subsequent blastema function are still poorly understood. Here, we investigated the cellular and molecular consequences of genetically interfering with retinoic acid (RA) signaling for the formation of the zebrafish blastema. We show that RA signaling is upregulated within the first few hours after fin amputation in the stump mesenchyme, where it controls Fgf, Wnt/β-catenin and Igf signaling. Genetic inhibition of the RA pathway at this stage blocks blastema formation by inhibiting cell cycle entry of stump cells and impairs the formation of the basal epidermal layer, a signaling center in the wound epidermis. In the established blastema, RA signaling remains active to ensure the survival of the highly proliferative blastemal population by controlling expression of the anti-apoptotic factor bcl2. In addition, RA signaling maintains blastema proliferation through the activation of growth-stimulatory signals mediated by Fgf and Wnt/β-catenin signaling, as well as by reducing signaling through the growth-inhibitory non-canonical Wnt pathway. The endogenous roles of RA in adult vertebrate appendage regeneration are uncovered here for the first time. They provide a mechanistic framework to understand previous observations in salamanders that link endogenous sources of RA to the regeneration process itself and support the hypothesis that the RA signaling pathway is an essential component of vertebrate tissue regeneration.