Limb regrowth takes two

Article metrics

Salamanders can regenerate several of their organs, including amputated limbs. Analysis of a Mexican salamander shows that crosstalk between two signalling molecules regulates limb regeneration. See Letter p.407

The ability of salamanders to regenerate their limbs following amputation was first described in the eighteenth century by the Italian naturalist Lazzaro Spallanzani1. Researchers have been trying to identify the formula for regeneration ever since, in the hope of reawakening this lost ability in mammals. Experiments performed 40 years ago2 showed that limb regeneration requires the participation of tissues from both sides of the limb — the side closer to the thumb (the anterior side) and the opposite, posterior side. However, the mechanisms involved in this process have remained a mystery. On page 407, Nacu et al.3 show that two diffusible signalling molecules underlie this requirement for anterior and posterior tissue.

During limb regeneration, cells from the limb stump form a mass of progenitor cells called the blastema, which gives rise to the new structure4. There are three requirements for this process. First, wounding must occur. This may seem obvious, but wounding triggers several tissue-repair responses that are essential for regeneration. The second requirement is proper innervation, because limbs from which nerves have been removed do not regenerate5. Finally, the blastema must contain abutting domains of cells derived from the anterior-most and posterior-most positions in the limb stump2. This last observation indicates that crosstalk between cells of differing origins triggers mechanisms that are essential for regeneration.

If wounding on the anterior side of the limb is combined with redirection of nerves from the limb proper into the wounded site, and with transplantation of a patch of posterior limb tissue onto the wound, an extra (accessory) limb can form6. By contrast, wounding and nerve redirection without transplantation result in a 'frozen' blastema — regeneration does not occur owing to the lack of posterior tissue. Using highly regenerative Mexican salamanders called axolotls, Nacu and colleagues took this 'anterior-only' experimental set-up as a starting point to precisely test the molecular pathways involved in antero-posterior crosstalk. The authors found that treating frozen blastemas with a molecule called smoothened agonist (SAG) reawakened the tissue, leading to production of an accessory limb from exclusively anterior tissue.

SAG mimics the activity of a protein called Sonic hedgehog (Shh), which is normally secreted by the posterior-most cells of the blastema. Shh diffuses across the tissue, binding to receptors on the surface of receiving cells to trigger intracellular signalling cascades. Nacu et al. found that stimulation of Shh signalling in anterior blastema cells by SAG activated another secreted diffusible protein, fibroblast growth factor 8 (FGF8). Sustained FGF8 activity was necessary for anterior-only limb production (Fig. 1a). Furthermore, activation of FGF8 could induce anterior limb growth even in the absence of Shh signalling.

Figure 1: Complementary signals promote limb regeneration.

a, In salamanders, if a limb is wounded and nerves are redirected to the wound site, an extra limb can arise from a cell cluster called a blastema. However, the blastema generates a new limb only if it contains cells from both the anterior and posterior sides of the original limb. Nacu et al.3 report that blastemas that contain only anterior tissue cannot form limbs because they do not receive a signalling protein called Shh, which is secreted by the posterior cells. Anterior cells normally produce the signalling protein FGF8 in response to Shh signalling, triggering limb formation. b, By contrast, posterior-only blastemas receive Shh, but are unable to activate FGF8 in response. c, Following limb amputation, both posterior and anterior cells contribute to the blastema, and regeneration is promoted by cross-induction between Shh and FGF8.

Like anterior-only blastemas, posterior-only blastemas cannot regenerate, despite containing Shh-expressing cells. Investigating this seeming contradiction, the authors found that posterior tissue could not activate strong and sustained FGF8 signalling in response to Shh (Fig. 1b). Forced FGF8 expression in the posterior tissue led to the production of accessory limbs — but, in this case, only when Shh signalling was also active, because inhibition of Shh blocked regeneration. Finally, Nacu et al. confirmed that signalling by Shh and FGF8 was required for physiological limb regeneration after amputation (Fig. 1c).

In summary, the authors have neatly dissected the antero-posterior crosstalk required for regeneration, and have defined the factors that induce blastema development. Anterior cells cannot regenerate alone because they lack Shh signalling from posterior cells. And posterior cells cannot regenerate alone because they are unable to respond to Shh with sustained activation of FGF8. However, why forced FGF8 expression in the absence of Shh signalling cannot promote posterior-only limb growth remains unexplained.

In future analyses, it will be important to determine whether this finding reflects the fact that anterior and posterior tissues have different requirements for regeneration. An alternative possibility is that it reflects experimental design limitations — for example, nonspecific effects of the drug that the researchers used to inhibit Shh in the posterior cells. Similarly, the reason for anterior-specific FGF8 activation by Shh remains unexplained, and should be a focus of future research.

The mechanisms at work during limb regeneration have long been debated, and models fall into two broad classes. Those in the first class posit that every cell contains positional information, and that regeneration involves only local cellular interactions. By contrast, models in the second class invoke the action of diffusible signals produced by specific regions in the blastema called organizers, which direct development of the surrounding tissue.

Nacu and colleagues' work favours organizer models, but does not exclude some degree of positional encoding. Consistent with organizer models, the authors clearly demonstrate that anterior cells do not produce Shh, and that anterior-only accessory limbs do not acquire posterior characteristics (for example, they lack digits that normally arise on the posterior side of the limb). But the authors' data also indicate that blastema cells tend to form sharply defined anterior and posterior domains, suggesting that cells identify themselves according to their position within broad domains — as anterior or posterior, for instance, and, along a perpendicular axis, as dorsal or ventral. In several ways, these observations fit the boundary model7, which states that organizer regions that produce diffusible signalling molecules are induced at the boundaries between broad limb domains.

Shh and FGF8 have well-defined roles in embryonic limb development in various vertebrate species. A positive feedback loop between the two signalling pathways has been described, and involves relay signalling by Gremlin, a third secreted diffusible protein8,9. Nacu et al. identified robust Gremlin expression associated with the interaction between anterior and posterior tissue, suggesting that this feedback loop probably also operates during limb regeneration in salamanders.

Given the evolutionary conservation of the Shh and FGF pathways, the current study brings us a step closer to the possibility of using defined factors to promote tissue regeneration in mammals — provided that a blastema can be induced. This advance, together with the same group's recent identification10 of a diffusible molecule called MLP that is responsible for the cell proliferation leading to blastema formation, represents substantial progress in the quest for a general formula for organ regeneration. Footnote 1


  1. 1.

    See all news & views


  1. 1

    Dinsmore, C. E. Int. J. Dev. Biol. 40, 621–627 (1996).

  2. 2

    Bryant, S. V. Nature 263, 676–679 (1976).

  3. 3

    Nacu, E., Gromberg, E., Oliveira, C., Drechsel, D. & Tanaka, E. M. Nature 533, 407–410 (2016).

  4. 4

    Nacu, E. & Tanaka, E. M. Annu. Rev. Cell Dev. Biol. 27, 409–440 (2011).

  5. 5

    Singer, M. Q. Rev. Biol. 27, 169–200 (1952).

  6. 6

    Lheureux, E. J. Embryol. Exp. Morphol. 38, 151–173 (1977).

  7. 7

    Meinhardt, H. J. Embryol. Exp. Morphol. 76, 115–137 (1983).

  8. 8

    Niswander, L., Jeffrey, S., Martin, G. R. & Tickle, C. Nature 371, 609–612 (1994).

  9. 9

    Zúñiga, A., Haramis, A.-P. G., McMahon, A. P. & Zeller, R. Nature 401, 598–602 (1999).

  10. 10

    Sugiura, T., Wang, H., Barsacchi, R., Simon, A. & Tanaka, E. M. Nature 531, 237–240 (2016).

Download references

Author information

Correspondence to Miguel Torres.

Related links

Related links

Related links in Nature Research

Developmental biology: A cellular view of regeneration

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Torres, M. Limb regrowth takes two. Nature 533, 328–330 (2016) doi:10.1038/nature17889

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.