An embryonic chick wing, stained with an antibody to reveal the normal pattern of innervation. Courtesy of Jonathan Clarke and Paul Martin, University College London, UK.

Healing of a skin wound can be a long and complicated process, and it seems that the problem is made even worse if the cutaneous nerve supply is impaired. In conditions such as diabetic neuropathy, complications associated with skin wounds have even led to amputation in extreme cases. Do nerves play an active role in the healing process? The fact that skin often becomes hyperinnervated after wounding would certainly support such a theory. However, some people have argued that the loss of pain sensation after denervation makes tissue more susceptible to further injury, thereby slowing the healing process.

As reported in Developmental Biology, Harsum et al. have set out to resolve this conundrum by studying the relationship between innervation and wound healing in the developing chick embryo. The skin of the early embryo shows a remarkable capacity for wound healing, and wounds made at embryonic day 4 (E4) heal within 24 hours, leaving no scar. However, the rate of healing becomes progressively slower as development proceeds, and the skin becomes increasingly prone to scarring, indicating that the mechanisms of wound healing change with time.

By irradiating the E2 neural tube between somites 12 and 20, Harsum et al. generated chick embryos with nerveless wings. At E4, the skin healed normally in these embryos, but around the time that the skin would normally become innervated (E7), healing became impaired, and was significantly slower than in wild-type skin at the same stage. The effects of continual injury due to lack of sensation could be ruled out, because the embryos were grown in a liquid environment, which was unlikely to cause damage.

The results of this experiment indicate a role for nerve-derived signals in wound healing, but only from the time that the skin normally becomes innervated. Dependence on these signals is acquired during development and, as it also seems to be acquired by nerveless skin, it clearly does not require contact with nerves. Interestingly, wounded fetal skin does not become hyperinnervated, so this does not seem to be a prerequisite for healing. Rather, the authors argue that hyperinnervation is a consequence of the inflammatory response, which does not occur in fetal tissue.

The next step will be to identify the signals released by nerves that are beneficial for tissue repair. Substance P and fibroblast growth factors are likely candidates, and both have been shown to be released by damaged nerves. It will also be important to understand why the response of the skin to these signals changes during development. With this knowledge, it should be possible to devise ways to improve the efficiency of wound healing in adult skin.