Salamander appendage regeneration was discovered in the mid 1700s by Spallanzani, who first documented appendage regeneration in frogs and salamanders (Spallanzani, 1768). This and the work of others at this time demonstrating regeneration in hydra and worms showed that regeneration is a phenomenon widespread in nature, an observation which contributed substantially to the debate on whether organism development occurs from pre-existing germs (preformation) or by the unfolding of a series of events (epigenesis) (Bodemer, 1964). Since then, biologists have been fascinated by the concepts, rules and molecules that govern this remarkable process. The high degree of conservation among tetrapods is reflected in the similarity of molecular programs governing limb development between salamanders and mammals. Given this similarity, this unique limb regeneration system provides a guiding light for what kind of tissue organization and injury response is necessary to achieve functional regeneration, which may be emulated in the future to induce regeneration in mammals.
Regenerating the correct part of the limb
Positional memories are at the basis of appropriate regeneration of the missing part. When a salamander arm is amputated at the wrist, only a hand is regenerated, while when the limb is amputated in the upper arm, the elbow, lower arm and hand are regenerated. Butler and others provided key insights into the nature of this problem (Butler, 1955). He was able to generate inversely oriented limbs by amputating a limb at the wrist and then suturing it back into the body. Blood vessels and presumably nerves repopulated the structure so that when the “circularized” limb was amputated in the upper arm, a normal stump as well as an inversely oriented stump was generated. The normal stump regenerated the lower arm and hand as expected. Interestingly, the inversely oriented limb similarly regenerated lower arm and hand.
These results revealed that the regenerating limb does not read directionality of limb tissue but rather the cells at the amputation plane have a memory of their positional identity, and newly formed blastema cells change their identity to more distal (direction fingertip) identities. This conclusion gave rise to the concept of “The rule of distal transformation”.
The limb contains many different tissue types including epidermis, dermis, muscle nerve, Schwann cells, ligaments, tendons, bone and an interesting question is whether all tissues harbor this positional identity or only some tissues. To test this question, tissue-specific transplantation of GFP-expressing hand cells into the upper arm, followed by amputation was implemented (Kragl et al., 2009; Nacu et al., 2013). The prediction of this experiment was that any tissue with positional memory would only contribute to the regenerated hand. Such work found that only cells from the lateral plate mesoderm lineage (connective tissue, namely dermis, tendons, ligaments, bone) showed hand determination. This meant that any molecular system controlling positional memory in the mature limb would be present in the connective tissue cells.
We investigated molecular correlates of positional memory. Upper arm, lower arm and hand development are strongly influenced by two clusters of genes called the HoxA and HoxD complexes. The genes HoxA/D9, HoxA/D11, HoxA/D13 are expressed sequentially as the upper arm, lower arm and hand, respectively are specified. We compared the expression characteristics of the HoxA proteins during upper arm versus hand regeneration (Roensch et al., 2013). During upper arm regeneration, the blastema cells initially at 6 days express HoxA9 and then HoxA13 is expressed at the tip of the blastema by day 8. In contrast in a hand blastema, the HoxA13 protein and HoxA9 protein are simultaneously expressed already at day 6. This strongly suggested that upper arm and hand cells start with a different setpoint for initiating HoxA gene expression, and suggests that HoxA proteins may be functionally involved in positional memory. Our current work examining the chromatin organization in upper versus lower arm cells indeed shows that mature upper arm and hand cells have differential chromatin organization at the HoxA locus (Kawaguchi, Wang and Tanaka, unpublished).
Which tissues and processes distalize the identity of cells at the amputation plane has been a topic of long speculation and models. Some models suggested that the wound epidermis may be the most distal identity, while others proposed a series of cellular transformations based on cell interactions (Maden, 1977; Slack, 1980). Interestingly, many of the early models hypothesized that the first blastema cells generated had the most distal identity, with subsequent intercalation of values, based on averaging identities as cell proliferation produced an increasing number of cells (Bryant et al., 1977; Maden, 1977). In contrast, Meinhardt proposed a model in which a border generated between anterior and posterior cells resulted in the generation of a signaling center that supported growth of cells that recapitulated the progressive sequence of upper arm, lower arm and then hand regeneration as seen in development rather than by intercalation (Meinhardt, 1983). Our experiments observing HoxA protein expression during upper arm regeneration strongly support this progressive mode of upper arm, lower arm, and hand sequence of distalization rather than the intercalation model (Roensch et al., 2013).
The molecular identity of factors that distalize blastema cells is not fully understood. Based on limb development studies, it may be expected that fibroblast growth factors and possibly bone morphogenetic factors are involved, yet they seem insufficient alone to yield cell identity distalization and some models have suggested that the lower arm to hand transition occurs cell autonomously, yet nonetheless, this transition must occur spatially at the tip of the blastema (Capdevila et al., 1999; Mercader et al., 2000; Rosello-Diez et al., 2014). How this occurs in this spatially defined domain is not fully understood.
Interestingly, positional identity can be experimentally reset during regeneration, by exposing the blastema to retinoic acid (Maden, 1982). Upon generation of a hand blastema, if the animal is exposed to retinoic acid, a complete arm is regenerated from the wrist. Molecular studies showed that this effect is mediated by the Meis transcription factor, whose function during development is to specify upper arm development (Mercader et al., 2005).
Specificity of regeneration to amputation – the role of anterior/posterior positional memory
Why does regeneration occur upon appendage amputation and not simply from wounding? The answer to this question also ultimately rests on the existence of positional memory in limb cells. Remarkably, transplantation of left blastemas to right limb stumps, results in the regeneration of two additional, ectopic limbs (for references see (Bryant et al., 1977; Nacu and Tanaka, 2011). These results were interpreted to mean that anterior and posterior cells in the mature limb harbor separate memories of their positional identity. Upon blastema formation, as anterior and posterior cells enter the blastema, it was hypothesized that either their interface or communication between anterior and posterior cells is required for limb outgrowth. The presence of both anteriorly and posteriorly-derived cells only occurs upon limb amputation and therefore explains why regeneration does not happen upon simple limb wounding.
Lheureux (Lheureux, 1977) followed by others, and most recently Endo and Gardiner (Endo et al., 2004), demonstrated this concept by generation of ectopic limbs. Ectopic limbs could be elicited by deviating nerves to an anterior lateral wound site in the upper limb, together with grafting of posterior full thickness skin to generate an ectopic anterior/posterior interface at that site. Nerve deviation alone resulted in formation of a blastema-like structure, but this tissue regressed after three weeks. With the combination of nerve and anterior/posterior skin interface a fully patterned limb grew from that site, demonstrating the central role of anterior/posterior interfaces in sustaining limb regeneration.
We recently defined the molecular nature of the anterior and posterior requirement for regeneration. Sonic hedgehog is a morphogen localized to the posterior limb bud required for limb development and is re-expressed in the regeneration limb blastema (Nacu et al., 2016). We hypothesized that the capability to express sonic hedgehog represents the posterior tissue requirement in regeneration. To test this hypothesis, we deviated nerves to the anterior side and treated animals with smoothened agonist to activate sonic hedgehog signalling in place of posterior skin. This treatment was sufficient to induce an ectopic limb on the anterior surface. Given the known developmental positive feedback loop between sonic hedgehog and Fgf8 in limb development, we asked whether anteriorly localized Fgf8 represents the anterior component for limb regeneration. To test this hypothesis, we deviated nerves to the posterior upper limb surface and then induced Fgf8 expression via baculoviral induction. This expression was sufficient to induce a limb-like outgrowth from the posterior surface.
Based on these results, the current model is that cells in the adult posterior limb have potential to express sonic hedgehog while cells in the adult anterior limb have the potential to express Fgf8. Amputation of the limb results in anterior and posterior cells migrating to the limb tip and forming blastema cells that then launch the expression of these developmental regulators. This initiates a positive feedback interaction between anterior and posterior cells reinforcing expression of sonic hedgehog and Fgf8 which are essential for outgrowth of the limb. An important future direction is how the genome and chromatin are organized differently in anterior and posterior cells to restrict the potential to express sonic hedgehog only posteriorly and Fgf8 only anteriorly.
Tissue interactions required for regeneration
Regeneration involves the coordination of many different cell types, and these interactions occur dynamically to promote changes in cell differentiation, migration, proliferation, positional identity and maturation. Immediately after amputation, the blood clots at the wound site followed by migration of epithelial keratinocytes over the limb stump surface. The detection of injury by epithelial cells likely relies on changes in osmolarity, and the epithelial migration process appears to be controlled in part by TGFB signalling (For references, see (Bassat and Tanaka, 2021)). Subsequently, underlying connective tissue cells migrate to the amputation surface in response to Platelet-Derived Growth Factor, presumably from the clotted blood and later from blastema cells that express the ligand themselves (Currie et al., 2016).
These connective tissue cells form the majority of the blastema and direct the positioning and patterning events of limb regeneration described above. Cells are stimulated to undergo their first proliferative division by MARCKS-Like Protein released from the epidermis (Sugiura et al., 2016). This special wound epidermis also secretes a number of other factors such as Wnt3 Wnt5, Anterior Gradient and retinoic acid and is essential for the process of regeneration (See (Bassat and Tanaka, 2021)).
Interestingly, recent work suggests that infiltrating macrophages expressing Midkine, and nerve fibers are required for the wound epithelium to acquire its full properties (Tsai et al., 2020). Nerves release not only Anterior Gradient but also BMPs and FGFs that are important for the initial growth of the blastema cells (Kumar et al., 2007; Makanae et al., 2014). Finally, the anterior Fgf8 and posterior sonic hedgehog feedback loop is initiated in the blastema which sustains growth and patterning.
Conclusions and outlook
This is an exciting time in stem cell biology where our ability to study organogenesis and recapitulate it in vitro is pointing our ambitions to elicit in vivo regeneration. So far, stem cell transplantation approaches have not taken the spatial map of a tissue very much into account. The salamander, which regenerates its limb with exquisite precision and functionality has shown that cellular memory states in many different parts of the limb are used as a template and starting place for appropriate regeneration. These observations point to the need to establish such landmarks in human scenarios of regeneration.
References
Bassat, E., and Tanaka, E.M. (2021). The cellular and signaling dynamics of salamander limb regeneration. Curr Opin Cell Biol 73, 117-123.
Bodemer, C.W. (1964). Regeneration and the Decline of Preformationism in Eighteenth Century Embryology. Bull Hist Med 38, 20-31.
Bryant, P.J., Bryant, S.V., and French, V. (1977). Biological regeneration and pattern formation. Sci Am 237, 66-76, 81.
Butler, E.G. (1955). Regeneration of the urodele forelimb after reversal of its proximo-distal axis. J Morphol 96, 265-281.
Capdevila, J., Tsukui, T., Rodriquez Esteban, C., Zappavigna, V., and Izpisua Belmonte, J.C. (1999). Control of vertebrate limb outgrowth by the proximal factor Meis2 and distal antagonism of BMPs by Gremlin. Molecular cell 4, 839-849.
Currie, J.D., Kawaguchi, A., Traspas, R.M., Schuez, M., Chara, O., and Tanaka, E.M. (2016). Live Imaging of Axolotl Digit Regeneration Reveals Spatiotemporal Choreography of Diverse Connective Tissue Progenitor Pools. Dev Cell 39, 411-423.
Endo, T., Bryant, S.V., and Gardiner, D.M. (2004). A stepwise model system for limb regeneration. Dev Biol 270, 135-145.
Kragl, M., Knapp, D., Nacu, E., Khattak, S., Maden, M., Epperlein, H.H., and Tanaka, E.M. (2009). Cells keep a memory of their tissue origin during axolotl limb regeneration. Nature 460, 60-65.
Kumar, A., Godwin, J.W., Gates, P.B., Garza-Garcia, A.A., and Brockes, J.P. (2007). Molecular basis for the nerve dependence of limb regeneration in an adult vertebrate. Science 318, 772-777.
Lheureux, E. (1977). [Importance of limb tissue associations in the development of nerve-induced supernumerary limbs in the newt Pleurodeles waltlii Michah (author’s transl)]. J Embryol Exp Morphol 38, 151-173.
Maden, M. (1977). The regeneration of positional information in the amphibian limb. J Theor Biol 69, 735-753.
Maden, M. (1982). Vitamin A and pattern formation in the regenerating limb. Nature 295, 672-675.
Makanae, A., Mitogawa, K., and Satoh, A. (2014). Co-operative Bmp- and Fgf-signaling inputs convert skin wound healing to limb formation in urodele amphibians. Dev Biol 396, 57-66.
Meinhardt, H. (1983). A boundary model for pattern formation in vertebrate limbs. J Embryol Exp Morphol 76, 115-137.
Mercader, N., Leonardo, E., Piedra, M.E., Martinez, A.C., Ros, M.A., and Torres, M. (2000). Opposing RA and FGF signals control proximodistal vertebrate limb development through regulation of Meis genes. Development 127, 3961-3970.
Mercader, N., Tanaka, E.M., and Torres, M. (2005). Proximodistal identity during vertebrate limb regeneration is regulated by Meis homeodomain proteins. Development 132, 4131-4142.
Nacu, E., Glausch, M., Le, H.Q., Damanik, F.F., Schuez, M., Knapp, D., Khattak, S., Richter, T., and Tanaka, E.M. (2013). Connective tissue cells, but not muscle cells, are involved in establishing the proximo-distal outcome of limb regeneration in the axolotl. Development 140, 513-518.
Nacu, E., Gromberg, E., Oliveira, C.R., Drechsel, D., and Tanaka, E.M. (2016). FGF8 and SHH substitute for anterior-posterior tissue interactions to induce limb regeneration. Nature 533, 407-410.
Nacu, E., and Tanaka, E.M. (2011). Limb regeneration: a new development? Annu Rev Cell Dev Biol 27, 409-440.
Roensch, K., Tazaki, A., Chara, O., and Tanaka, E.M. (2013). Progressive specification rather than intercalation of segments during limb regeneration. Science 342, 1375-1379.
Rosello-Diez, A., Arques, C.G., Delgado, I., Giovinazzo, G., and Torres, M. (2014). Diffusible signals and epigenetic timing cooperate in late proximo-distal limb patterning. Development 141, 1534-1543.
Slack, J.M. (1980). A serial threshold theory of regeneration. J Theor Biol 82, 105-140.
Spallanzani L. (1768). Prodromo di un’opera da imprimersi sopra le riproduzioni animali. Giovanni Montanari, Modena.
Sugiura, T., Wang, H., Barsacchi, R., Simon, A., and Tanaka, E.M. (2016). MARCKS-like protein is an initiating molecule in axolotl appendage regeneration. Nature 531, 237-240.
Tsai, S.L., Baselga-Garriga, C., and Melton, D.A. (2020). Midkine is a dual regulator of wound epidermis development and inflammation during the initiation of limb regeneration. eLife 9.