During limb regeneration adult tissue is converted into a zone of undifferentiated progenitors called the blastema that reforms the diverse tissues of the limb. Previous experiments have led to wide acceptance that limb tissues dedifferentiate to form pluripotent cells. Here we have reexamined this question using an integrated GFP transgene to track the major limb tissues during limb regeneration in the salamander Ambystoma mexicanum (the axolotl). Surprisingly, we find that each tissue produces progenitor cells with restricted potential. Therefore, the blastema is a heterogeneous collection of restricted progenitor cells. On the basis of these findings, we further demonstrate that positional identity is a cell-type-specific property of blastema cells, in which cartilage-derived blastema cells harbour positional identity but Schwann-derived cells do not. Our results show that the complex phenomenon of limb regeneration can be achieved without complete dedifferentiation to a pluripotent state, a conclusion with important implications for regenerative medicine.

The development of transgenesis in axolotls is crucial for studying development and regeneration as it would allow for long-term cell fate tracing as well as gene expression analysis. We demonstrate here that plasmid injection into the one-cell stage axolotl embryo generates mosaic transgenic animals that display germline transmission of the transgene. The inclusion of SceI meganuclease in the injections (Thermes, V., Grabher, C., Ristoratore, F., Bourrat, F., Choulika, A., Wittbrodt, J., Joly, J.S., 2002. I-SceI meganuclease mediates highly efficient transgenesis in fish. Mech. Dev. 118, 91-98) resulted in a higher percentage of F0 animals displaying strong expression throughout the body. This represents the first demonstration in the axolotl of germline transmission of a transgene. Using this technique we have generated a germline transgenic animal expressing GFP ubiquitously in all tissues examined. We have used this animal to study cell fate in the dorsal fin during development. We have uncovered a contribution of somite cells to dorsal fin mesenchyme in the axolotl, which was previously assumed to derive solely from neural crest. We have also studied the role of blood during tail regeneration by transplanting the ventral blood-forming region from GFP+ embryos into unlabeled hosts. During tail regeneration, we do not observe GFP+ cells contributing to muscle or nerve, suggesting that during tail regeneration blood stem cells do not undergo significant plasticity.

Complete regeneration of the spinal cord occurs after tail regeneration in urodele amphibians such as the axolotl. Little is known about how neural progenitor cells are recruited from the mature tail, how they populate the regenerating spinal cord, and whether the neural progenitor cells are multipotent. To address these issues we used three types of cell fate mapping. By grafting green fluorescent protein-positive (GFP + ) spinal cord we show that a 500 m region adjacent to the amputation plane generates the neural progenitors for regeneration. We further tracked single nuclear-GFP-labeled cells as they proliferated during regeneration, observing their spatial distribution, and ultimately their expression of the progenitor markers PAX7 and PAX6. Most progenitors generate descendents that expand along the anterior/posterior (A/P) axis, but remain close to the dorsal/ventral (D/V) location of the parent. A minority of clones spanned multiple D/V domains, taking up differing molecular identities, indicating that cells can execute multipotency in vivo. In parallel experiments, bulk labeling of dorsally or ventrally restricted progenitor cells revealed that ventral cells at the distal end of the regenerating spinal cord switch to dorsal cell fates. Analysis of PAX7 and PAX6 expression along the regenerating spinal cord indicated that these markers are expressed in dorsal and lateral domains all along the spinal cord except at the distal terminus. These results suggest that neural progenitor identity is destabilized or altered in the terminal vesicle region, from which clear migration of cells into the surrounding blastema is also observed.

The oral cavity of vertebrates is generally thought to arise as an ectodermal invagination. Consistent with this, oral teeth are proposed to arise exclusively from ectoderm, contributing to tooth enamel epithelium, and from neural crest derived mesenchyme, contributing to dentin and pulp. Yet in many vertebrate groups, teeth are not restricted only to the oral cavity, but extend posteriorly as pharyngeal teeth that could be derived either directly from the endodermal epithelium, or from the ectodermal epithelium that reached this location through the mouth or through the pharyngeal slits. However, when the oropharyngeal membrane, which forms a sharp ecto/endodermal border, is broken, the fate of these cells is poorly known. Here, using transgenic axolotls with a combination of fate-mapping approaches, we present reliable evidence of oral teeth derived from both the ectoderm and endoderm and, moreover, demonstrate teeth with a mixed ecto/endodermal origin. Despite the enamel epithelia having a different embryonic source, oral teeth in the axolotl display striking developmental uniformities and are otherwise identical. This suggests a dominant role for the neural crest mesenchyme over epithelia in tooth initiation and, from an evolutionary point of view, that an essential factor in teeth evolution was the odontogenic capacity of neural crest cells, regardless of possible 'outside-in' or 'inside-out' influx of the epithelium.

Cartilage of the vertebrate jaw is derived from cranial neural crest cells that migrate to the first pharyngeal arch and form a dorsal "maxillary" and a ventral "mandibular" condensation. It has been assumed that the former gives rise to palatoquadrate and the latter to Meckel's (mandibular) cartilage. In anamniotes, these condensations were thought to form the framework for the bones of the adult jaw and, in amniotes, appear to prefigure the maxillary and mandibular facial prominences. Here, we directly test the contributions of these neural crest condensations in axolotl and chick embryos, as representatives of anamniote and amniote vertebrate groups, using molecular and morphological markers in combination with vital dye labeling of late-migrating cranial neural crest cells. Surprisingly, we find that both palatoquadrate and Meckel's cartilage derive solely from the ventral "mandibular" condensation. In contrast, the dorsal "maxillary" condensation contributes to trabecular cartilage of the neurocranium and forms part of the frontonasal process but does not contribute to jaw joints as previously assumed. These studies reveal the morphogenetic processes by which cranial neural crest cells within the first arch build the primordia for jaw cartilages and anterior cranium.

An increasing number of genes are known to show expression in the cranial neural crest area. So far it is very difficult to analyze their effect on neural crest cell migration because of the lack of transplantation techniques. This paper presents a simple method to study the migratory behavior of cranial neural crest cells by homo- and heterotopic transplantations: Green fluorescent protein (GFP) RNA was injected into one blastomere of Xenopus laevis embryos at the 2-cell stage. The cranial neural crest area of stage 14 embryos was transplanted into the head or trunk region of an uninjected host embryo, and the migration was monitored by GFP fluorescence. The transplants were further examined by double immunostaining and confocal microscopy to trace migratory routes inside the embryo, and to exclude contaminations of grafts with foreign tissues. Our results demonstrate that we developed a highly efficient and reproducible technique to study the migratory ability of cranial neural crest cells. It offers the possibility to analyze genes involved in neural crest cell migration by coinjecting their RNA with that of GFP.

INTRODUCTIONAdult salamanders are best known for their capacity to regenerate an astounding range of body structures including the whole limb and tail, the central nervous system, and tissues of the eye and heart. The axolotl (Ambystoma mexicanum) represents the salamander species that is most easily bred in the laboratory, and for which the most comprehensive genetic, genomic, and transgenesis tools have been developed. As such, it serves as an important vertebrate model for studying regeneration and tissue repair. Beyond regeneration, axolotls have a deep and rich history as primary amphibian models, especially in research areas concerning embryonic development--most notably the inductive mode of germ cell formation. The easily obtained oocytes, high quantities of embryos produced by each spawning, large size of the embryo, and ability to graft tissues from individual to individual at any stage without rejection make the axolotl an advantageous model system for the study of development, electrophysiology, and regeneration.

No abstract