Mayumi Ito (left) and George Cotsarelis.

About three years ago George Cotsarelis, a dermatologist at the University of Pennsylvania School of Medicine, was studying whether stem cells in the hair follicle — the part of the skin that hair grows from — could help heal wounds. Working with mice, his team found that when the skin is cut, stem cells move quickly from the lowest portion of the follicle, or 'bulge', where they usually reside, to the surface of the skin to form new skin cells (M. Ito et al. Nature Med. 11, 1351–1354; 2005). But during these experiments, postdoc Mayumi Ito noticed something unusual. New hairs also seemed to be growing inside the wound. “We really thought we were seeing things,” says Cotsarelis.

The dogma in the field has long been that new hair follicles cannot form in adult mammals, even after damage or loss. But Cotsarelis quickly found hints in the literature that this long-held belief might be mistaken. In 1956, pathologist Robert Billingham at University College London noticed new hair forming in skin that had been wounded in rabbits. And the same year, Albert Kligman, also a dermatologist from the University of Pennsylvania, found that people who had had the top layer of their skin scraped away showed signs of forming new hair follicles. “Those studies were in the 1950s and no one has paid much attention to them since,” says Cotsarelis.

Ito and Cotsarelis and their colleagues decided to follow up on their intriguing observation and determine whether the hairs in the wound grow from new hair follicles and, if so, whether the hair follicles develop from the stem cells of neighbouring follicles. Over the past 15 years, Cotsarelis' group has developed several strains of transgenic mice, which has allowed the researchers to visualize the hair-follicle stem cells as they migrate from deep within the skin to the surface of the wound site. The group has also developed several antibodies that can distinguish between various types of skin cell.

Armed with these tools, the team showed (page 316) that mice indeed grow new hair follicles at the wound site. But the investigators had another surprise. The new hair follicles did not arise from the stem cells in nearby hair follicles, but rather from cells in the top layer of the skin. They somehow acquire a more embryonic character, which allows them to develop into different cell types. “One of the lessons is that wounding generates a window of opportunity for manipulating hair-follicle formation,” says Cotsarelis. Because the mammalian hair follicle is considered to be a tiny organ, this finding could have implications for organ regeneration in mammals, in addition to providing a way to ameliorate baldness.

One of the main questions that Cotsarelis and his team are now pursuing is whether the same process takes place in human skin. On the basis of Kligman's work of 50 years ago, the answer would seem to be yes. Cotsarelis also hopes to determine the 'molecular signature' of the skin straight after wounding. Knowing which genes become turned on might suggest ways to induce the same pathway in skin cells, and thus coax them into becoming follicles without having to resort to making wounds.