Cloning and neuronal expression of a type III newt neuregulin and rescue of denervated, nerve-dependent newt limb blastemas by rhGGF2

Urodele amphibians are the only vertebrates that can regenerate their limbs throughout their life. The critical feature of limb regeneration is the formation of a blastema, a process that requires an intact nerve supply. Nerves appear to provide an unidentified factor, known as the neurotrophic factor (NTF), which stimulates cycling of blastema cells. One candidate NTF is glial growth factor (GGF), a member of the neuregulin (NRG) growth factor family. NRGs are both survival factors and mitogens to glial cells, including Schwann cells. All forms of NRGs contain an EGF-like domain that is sufficient to activate NRG receptors erbB2, erbB3, and erbB4. To investigate the involvement of neuregulin in newt limb regeneration, we cloned and characterized one neuregulin isoform, a neuregulin with a cysteine-rich domain (CRD-NRG), from newt (Notophthalmus viridescens) spinal cord. Results of in situ hybridization showed that the newt CRD-NRG is highly expressed in dorsal root ganglia and spinal cord neurons that innervate the limbs. We also demonstrated the biological activity of recombinant human GGF2 (rhGGF2) in urodele limb regeneration. When rhGGF2 was injected into denervated, nerve-dependent axolotl blastemas, the labeling index (LI) of blastema cells was maintained at a level near to that of control, innervated blastemas, whereas without rhGGF2 the LI decreased significantly. In another experiment, rhGGF2 was delivered into denervated, nerve-dependent blastemas either by direct infusion into blastemas or by injection into the intraperitoneal cavity. The denervated blastemas were rescued into a regeneration response.     

Ganglia implantation as a means of supplying neurotrophic stimulation to the newt regeneration blastema: cell-cycle effects in innervated and denervated limbs.

Regulation of blastema cell proliferation during amphibian limb regeneration is poorly understood. One unexplained phenomenon is the relatively low level of active cell cycling in the adult newt blastema compared to that of larval axolotls. In the present study, we used ganglia implantation as a means of "superinnervating" normally innervated adult newt blastemas to test whether blastema cell subpopulations are responsive to nerve augmentation. The effectiveness of implanted ganglia to provide neurotrophic stimulation was demonstrated in denervated blastemas. Blastemas implanted with 2 dorsal root ganglia and simultaneously denervated 14 days after amputation exhibited control levels of cell cycle activity 6 days later, as measured by 3H-thymidine pulse labeling. Denervated blastemas that were sham-operated or implanted with pituitary glands exhibited cell-cycle declines similar to those of denervated blastemas without implanted ganglia. Thus, 2 implanted ganglia provide neurotrophic stimulation equivalent to that of the normal nerve supply. Dorsal root ganglia implanted into normally innervated blastemas, which should effectively double neurotrophic activity to the blastema, had no effect on cell-cycle activity, innervated blastemas implanted with ganglia for 6 days exhibited pulse labeling indices similar to those of normally innervated blastemas. These data indicate that neurotrophic stimulation is not normally limiting in innervated limbs, and that some other factor, whether extrinsic or intrinsic to blastema cells, accounts for the relatively low level of active cell cycling in the adult newt blastema.     

Glial growth factor and nerve-dependent proliferation in the regeneration blastema of Urodele amphibians

After amputation of a limb from Urodele amphibians, division of the blastemal cells (the progenitor cells of the regenerate) depends on one or more unidentified growth factors provided by the nerve supply. Here we show that glial growth factor (GGF), a mitogenic protein previously purified from the bovine pituitary, is present in newt nervous system extracts. It is also detectable in extracts of the forelimb regeneration blastema, and its level there decreases after denervation. We have previously shown that blastemal cells dependent on the nerve for division are marked by a monoclonal antibody called 22/18. When denervated blastemas are cultured in the presence of partially purified GGF from newt brain, or pure GGF from the bovine pituitary, the thymidine labeling index of blastemal cells that are 22/18-positive is increased as much as sevenfold. These data indicate that GGF plays a role in nerve-dependent proliferation in the blastema.     

The effects of partial and complete denervation on adult newt forelimb blastema cell-cycle parameters

Abstract. The adult newt blastema cell-cycle time (cct) was measured by the percentage of labeled mitoses (PLM) method at the early-bud and mid-bud stages and was found to be 42.9 and 42.7 h, respectively. At both stages, the DNA synthetic phase (S) occupied the majority (75%) of the cct. However, the blastema labeling index (LI) after a 2-h pulse of ³H-thymidine was less than 30% i.e., considerably less than predicted from the ratio of the duration of S over the cct. Compared to that of controls, the PLM plot for partially denervated blastemas exhibited a coincident and equal-sized first peak of labeled mitoses and a coincident but smaller second peak of labeled mitoses. After 24 h of continuous labeling, the LI of control blastemas reached 53%, whereas the LIs of partially denervated and completely denervated blastemas reached only 33% and 20%, respectively. These results are consistent with the view that many cells of adult newt blastemas are not actively progressing through the cell cycle and that the number of noncycling cells is increased by partial or complete denervation. The noncycling cells are probably in the G₁ phase of the cell cycle.     

Image analysis of blastema cell proliferation in denervated limb regenerates of the newt, Pleurodeles waltlii M.

When denervated at the medium bud stage, limb blastemas of the newt, Pleurodeles waltlii Michah, stop growing. In order to better understand the role of nerves in the cell cycle in blastemas, we studied the distribution of mesenchymal cells in the G0-1, G1, S, G2 and M phases 48 and 96 h after denervation. The cell-cycle phases were determined by examining Feulgen-stained nuclei using a SAMBA 200 (System for Analytical Microscopy in Biological Applications) cell image processor. The cell nuclei were automatically analyzed by calculating 18 parameters related to the densitometry and texture of chromatin, and the shape of each nucleus. Cell-cycle phases were classified according to the unsupervised recognition method using a SAMBA 200 system as proposed by Moustafa and Brugal for cell-kinetics analysis. The classification obtained was tested against the results of stepwise linear discriminant analysis performed according to the method of Giroud. Our results show that, in blastemas 96 h after denervation, the percentage of cells in the S, G2, and M phases decreases significantly, while the percentage of G1 and G0-1 cells increases (+ 51% for G1 cells; + 30% for G0-1 cells). Thus, it appears that denervation of medium-bud-stage limb blastemas promotes the lengthening of G1 and premature exiting of cells from the cycle into the G0-1 phase. These results show that nerves (i.e., neurotrophic factor) regulate cell kinetics during newt limb regeneration by maintaining blastema mesenchymal cells in the cell-cycle.     

Regenerate epithelium and skin glands of the adult newt react to the same monoclonal antibody

A search for specific proteins involved in newt limb regeneration, using monoclonal antibodies against forelimb blastemas, led to the detection of an antigen in the regenerate epithelium. Fluorescent-antibody-labeled cells first appeared just prior to blastema outgrowth. From bud through early digit stages this antibody reacted with nearly all of the regenerate epithelial cells. Other tissues also reacted, including nerve, blood vessels, and gastrointestinal tract. The behavior of the reactive cells in the regenerate epithelium, and their close association with immediately adjacent skin glands, raises several new possibilities for the origin of the regenerate epithelium.     

Hyaluronidase activity and glycosaminoglycan synthesis in the amputated newt limb: comparison of denervated, nonregenerating limbs with regenerates.

Amputated, regenerating forelimbs have been compared with the contralateral, denervated non-regenerating limb stumps in the adult newt Notophthalmus viridescens, with respect to hyaluronidase activity and the incorporation of ³H-acetate into glycosaminoglycans (GAG). At 10 days after amputation, which is the time of maximum hyaluronate production in the early growing regenerate, incorporation of ³H-acetate into GAG (cpm/mg protein) in the denervated, nonregenerating limb stump was approximately 50% of that in the contralateral regenerating limbs. At this stage, hyaluronate was the major GAG being produced, but the ratio of incorporation into hyaluronate relative to chondroitin sulfate was reduced in the denervated limbs. In intact, nonamputated limbs, the incorporation into GAG was 5% of that in the regenerating limb 10 days after amputation, and 10% of that in the denervated stumps.At 25 days, cartilage is forming and chondroitin sulfate synthesis predominates in the normal regenerate whilst the contralateral, denervated limb stumps are forming scars. GAG synthesis in the latter was less than one-quarter the level seen in the regenerating limbs, mostly due to low incorporation into chondroitin sulfate.Hyaluronidase activity, which appears in the regenerating limb during differentiation of skeletal elements (20–45 days), was not detectable in limbs denervated early enough to prevent regeneration. However, limbs denervated after formation of the blastema will regenerate without nerve, and hyaluronidase activity in such limbs was normal. Thus, hyaluronidase activity appears when regeneration reaches the cartilage deposition stage, with or without nerve.     

Expression and biological effect of urodele fibroblast growth factor 1: relationship to limb regeneration

Fibroblast growth factors (FGFs) have been previously implicated in urodele limb regeneration. Here, we examined expression of FGF-1 by blastema cells and neurons and investigated its involvement in wound epithelial formation and function and in the trophic effect of nerves. Neurons innervating the limb and blastema cells in vivo and in vitro expressed the FGF-1 gene. The peptide was present in blastemas in vivo. Wound epithelium thickened when recombinant newt FGF-1 was provided on heparin-coated beads, demonstrating that the FGF-1 was biologically active and that the wound epithelium is a possible target tissue of FGF. FGF-1 did not stimulate accessory limb formation. FGF-1 was as effective as 10% fetal bovine serum in maintaining proliferative activity of blastema cells in vitro but was unable to maintain growth of denervated, nerve-dependent stage blastemas when provided on beads or by injection. FGF-1 had a strong stimulating effect on blastema cell accumulation and proliferation of limbs inserted into the body cavity that were devoid of an apical epithelial cap (AEC). These results show that FGF-1 can signal wound epithelium cap formation and/or function and can stimulate mesenchyme accumulation/proliferation in the absence of the AEC but that FGF-1 is not directly involved in the neural effect on blastema growth.     

Stimulation in cell culture of mesenchymal cells of newt limb blastemas by EDGF I or II (basic or acidic FGF)

After amputation of a newt limb, a blastema forms on the amputation plane and later differentiates to regenerate all the missing parts of the limb. Proliferation of blastema cells is under the control of severed nerves which deliver a 'neurotrophic factor' (NTF) of unknown nature. In order to characterize this factor we use a primary culture of blastema mesenchymal cells; changes in mitotic index after 48-h colchicine treatment indicate mitogenic activity of potential growth substances. These cells, which are stimulated by nerve extracts (mitotic index X 6), were tested with two purified growth factors extracted from bovine retina or brain (EDGF I = basic FGF and EDGF II = acidic FGF). We show that these two growth factors stimulate proliferation of blastema cell cultures in a dose-dependent manner. Maximal stimulation was obtained at 3 pM for EDGF I (mitotic index X 5.7) or 300 pM for EDGF II (mitotic index X 4.9). So it appears that these two growth factors have a mitogenic activity on blastema mesenchymal cells similar to that obtained with nerve extracts. The fact that two different growth factors can stimulate these cells raises the question of whether both are present in NTF and/or whether there are receptors to both EDGF I and EDGF II on mesenchymal cell membranes.     

Nerve–blastema interactions induce fibroblast growth factor-1 release during limb regeneration in Pleurodeles waltl

Previous studies have shown that both fibroblast growth factor (FGF)-1 and nerves play an important function during limb regeneration, but no correlation between these two regeneration factors has yet been demonstrated. In the present study we first establish that exogenous FGF-2, a member of the FGF family that binds to the same high-affinity receptors as FGF-1, is able to stimulate both [³H]-thymidine incorporation and the mitotic index in the mesenchyme and the epidermal cells of denervated blastemas. We then use cocultures of spinal cord and blastema on heparin-coated dishes, an in vitro system mimicking the in vivo interactions during limb regeneration, to show that interactions between nerve fibers from the spinal cord and the blastema enhance the release of bioactive FGF-1. Release of this growth factor seemed to correlate with nerve fiber regeneration, as it decreased in the presence of the dipeptide Leu-Ala, known to inhibit neurite outgrowth, while the inverse dipeptide Ala-Leu was inactive. Therefore, these results support our hypothesis that the interaction between nervous tissue and blastema is permissive for the release of FGF-1, which in turn stimulates blastema cell proliferation.     

Effects on adult newt limb regeneration of partial and complete skin flaps over the amputation surface.

The influence of the wound epithelium on the cellular events preceding blastema formation was examined by comparing dedifferentiation, DNA labeling indices, and mitotic indices of the distal mesodermal tissues in control regenerating newt forelimbs and in amputated forelimbs covered with a flap of full thickness skin. Three kinds of results were seen following the skin-flap graft operations. Epidermal migration across the amputation surface was completely inhibited in 22% (8) of the cases and these limbs repaired the amputation wound but did not form regeneration blastemas. In 11% (4) of the experimental limbs, essentially normal wound epithelia displaced the skin flaps and the limb stumps formed blastemas and regenerated. The majority of the skin grafts (67%) exhibited epidermal migration restricted to the free edges of the flaps. These limbs formed eccentric blastemas on the ventral side of the limb next to the dermis-free epidermis and regenerated laterally in that direction.     

Nerve-independence of limb regeneration in larval Xenopus laevis is related to the presence of mitogenic factors in early limb tissues.

Early limbs of larval Xenopus laevis can form a regeneration blastema in the absence of nerves. The nerve-independence could be due to the synthesis of neurotrophic-like factors by the limb bud cells. To test this hypothesis, two series of experiments were performed. Series A: the right hindlimbs of stage 57 larvae (acc. to Nieuwkoop and Faber. 1956. Normal table of Xenopus laevis [Daudin]. Amsterdam: North-Holland Pub. Co.), which are nerve-dependent for regeneration, were amputated through the tarsalia. The regenerating limbs were submitted to: sham denervation; denervation; denervation and implantation of a fragment of an early limb, or a late limb, or a spinal cord. Series B: froglets were subjected to amputation of both forelimbs. The cone blastemas were transplanted into denervated hindlimbs of stage 57 larvae, together with a fragment of an early or a late limb. The results in series A showed that the implantation of early limb tissue into the denervated blastema maintained cell proliferation at levels similar to those observed after the implantation of a spinal cord fragment or in sham denervated blastemas. However, the implantation of late limb tissues were ineffective. The results of series B showed that the implantation of early limb tissue, but not of late limb tissue prevented the inhibition of cell proliferation and the regression of denervated limb blastemas of juveniles. These results indicate that the nerve-independence is related to the synthesis of diffusible mitogenic neurotrophic-like factors in early limb tissues, and that nerve-dependence is established when differentiated cells of late limb tissues stop producing these factors.     

Blastema cell proliferation in vitro: effects of limb amputation on the mitogenic activity of spinal cord extracts

Primary cultures of mesenchymal cells of axolotl limb blastemas provide a very sensitive in vitro bioassay for studying nerve dependence of newt regeneration. These cells can be stimulated by crude spinal cord extracts of non-amputated animals in a dose-dependent manner up to 60 micrograms protein/ml of culture medium; at this concentration the mitotic index is increased 4-fold. Spinal cord extracts of axolotls 14 days after forelimb amputation (i.e., late bud stage) are more efficient in stimulating blastema cell proliferation (+50%) than extracts of axolotls 7 days after forelimb amputation (i.e., early bud stage) or of axolotls without amputation. In a similar manner, spinal cord extracts of young axolotls 14 days after forelimb amputation, are more stimulatory than older axolotls 14 d after forelimb amputation which regenerate only a very small blastema during the same time. It appears that spinal cord mitogenic activity is enhanced after limb amputation, probably in correlation with blastema cell requirements for limb regeneration.     

A stepwise model system for limb regeneration

The amphibian limb is a model that has provided numerous insights into the principles and mechanisms of tissue and organ regeneration. While later stages of limb regeneration share mechanisms of growth control and patterning with limb development, the formation of a regeneration blastema is controlled by early events that are unique to regeneration. In this study, we present a stepwise experimental system based on induction of limb regeneration from skin wounds that will allow the identification and functional analysis of the molecules controlling this early, critical stage of regeneration. If a nerve is deviated to a skin wound on the side of a limb, an ectopic blastema is induced. If a piece of skin is grafted from the contralateral side of the limb to the wound site concomitantly with nerve deviation, the ectopic blastema continues to grow and forms an ectopic limb. Our analysis of dermal cell migration, contribution, and proliferation indicates that ectopic blastemas are equivalent to blastemas that form in response to limb amputation. Signals from nerves are required to induce formation of both ectopic and normal blastemas, and the diversity of positional information provided by blastema cells derived from opposite sides of the limb induces outgrowth and pattern formation. Hence, this novel and convenient stepwise model allows for the discovery of necessary and sufficient signals and conditions that control blastema formation, growth, and pattern formation during limb regeneration.     

Promotion of spermatogonial proliferation by neuregulin 1 in newt (Cynops pyrrhogaster) testis

We have previously shown that mammalian follicle-stimulating hormone (FSH) promotes the proliferation of spermatogonia and their differentiation into primary spermatocytes in organ culture of newt testis. In the current study, we performed microarray analysis to isolate local factors secreted from somatic cells upon FSH treatment and acting on the germ cells. We identified neuregulin 1 (NRG1) as a novel FSH-upregulated clone homologous to mouse NRG1 known to control cell proliferation, differentiation and survival in various tissues. We further isolated cDNAs encoding two different clones. Amino acid sequences of the two clones were 75% and 94% identical to Xenopus leavis immunoglobulin (Ig)-type and cysteine-rich domain (CRD)-type NRG1, respectively, which had distinct sequences in their N-terminal region but identical in their epidermal growth factor (EGF)-like domain. Semi-quantitative and quantitative PCR analyses indicated that both clones were highly expressed at spermatogonial stage than at spermatocyte stage. In vitro FSH treatment increased newt Ig-NRG1 (nIg-NRG1) mRNA expression markedly in somatic cells, whereas newt CRD-NRG1 (nCRD-NRG1) mRNA was only slightly increased by FSH. To elucidate the function of newt NRG1 (nNRG1) in spermatogenesis, recombinant EGF domain of nNRG1 (nNRG1-EGF) was added to organ and reaggregated cultures with or without somatic cells: it promoted spermatogonial proliferation in all cases. Treatment of the cultures with the antibody against nNRG1-EGF caused remarkable suppression of spermatogonial proliferation activated by FSH. These results indicated that nNRG1 plays a pivotal role in promoting spermatogonial proliferation by both direct effect on spermatogonia and indirect effect via somatic cells in newt testes.     

Neurotrophic and neuroprotective effects of the neuregulin glial growth factor-2 on dopaminergic neurons in rat primary midbrain cultures

Glial growth factor-2 (GGF2) and other neuregulin (NRG) isoforms have been shown to play important roles in survival, migration, and differentiation of certain neural and non-neural cells. Because midbrain dopamine (DA) cells express the NRG receptor, ErbB4, the present study examined the potential neurotrophic and/or neuroprotective effects of GGF2 on cultured primary dopaminergic neurons. Embryonic day 14 rat mesencephalic cell cultures were maintained in serum-free medium and treated with GGF2 or vehicle. The number of tyrosine hydroxylase-positive (TH+) neurons and high-affinity [3H]DA uptake were assessed at day in vitro (DIV) 9. Separate midbrain cultures were treated with 100 ng/mL GGF2 on DIV 0 and exposed to the catecholamine-specific neurotoxin 6-hydroxydopamine (6-OHDA) on DIV 4. GGF2 treatment significantly increased DA uptake, the number of TH+ neurons, and neurite outgrowth when compared to the controls in both the serum-free and the 6-OHDA-challenged cultures. Furthermore, three NRG receptors were detected in the midbrain cultures by western blot analysis. Immunostaining for glial fibrillary acidic protein revealed that GGF2 also weakly promoted mesencephalic glial proliferation in the midbrain cultures. These results indicate that GGF2 is neurotrophic and neuroprotective for developing dopaminergic neurons and suggest a role for NRGs in repair of the damaged nigrostriatal system that occurs in Parkinson's disease.     

Stimulation of lens regeneration from the newt dorsal iris when implanted into the blastema of the regenerating limb

Dorsal iris from the eyes of adult Notophthalmus viridescens was transplanted into the blastema of regenerating limbs, subcutaneously in the limb or shoulder region, into the dorsal fin of larval newts and into the hindbrain of larval Ambystoma maculatum. The iris implants into the blastema regenerated lens vesicles or lenses with fibers in 40–75% of the cases. Multiple lenses were found in a few instances. No lenses developed from iris implants into the dorsal fin. Twenty percent of subcutaneous implants of iris formed lenses or lens vesicles, but lens regeneration from implants into the brain occurred only rarely. Denervation of the limb at the time of iris transplantation into the blastema greatly reduced the number of lenses regenerated. Studies on nerve fiber distribution in dorsal fin, subcutaneous areas, and denervated and innervated regenerating limbs, using the Bodian method, showed a general correlation between density of nerve fibers in the implant site and the incidence of lens regeneration from iris implants into that site. These results provide some evidence for a trophic action of nerve fibers on lens regeneration from the iris.