If tadpoles of the African clawed toad Xenopus laevis lose their tail, they can usually make another — but not always. So say Caroline W. Beck and colleagues, who have identified a 'refractory period' during which tails cannot regenerate (Dev. Cell 5, 429–439; 2003). Studies of this down-time have provided insight into the molecular mechanisms that underlie tail regeneration.

The ability to regenerate limbs is scattered throughout the animal kingdom, even occurring in the adults of some species — certain lizards, for instance. But the phenomenon is not well understood at the molecular level. Beck et al. have used a variety of new molecular techniques to tackle this question.

The authors started by investigating tail regeneration at different stages of tadpole development. They found that 3 and 7 days into development, at stages 42 and 48, around two-thirds of the tadpoles could regrow their tails after amputation (see picture; the arrowheads represent the position at which the tail was amputated). At stage 49 and beyond, that figure was near 100%. But at stages 46 and 47, only a very few of the tails regenerated (although the tadpoles developed into froglets normally).

What's going on during this refractory period? A look at the regenerating tails showed that a thin skin formed over the cut stump, and that unspecialized cells accumulated beneath the skin; these later began to produce the spinal cord, muscle cells and other cells required for tail formation. This did not occur in the non-regenerating tails.

Proteins involved in the bone morphogenetic protein (BMP) and Notch signalling pathways are required for embryonic tail development. Might they also be expressed in regenerating tails? Beck et al. show that they are, but that they are not expressed in tails amputated during the refractory period. If, however, the BMP receptor is forcibly expressed during this period, tail regeneration occurs in nearly all cases. Further findings hint that the BMP pathway activates the Notch pathway to regenerate the spinal cord, but works independently of Notch to regenerate muscle. Whether these molecular events underlie the remarkable phenomenon of regeneration more generally remains to be seen.