Scar-free wound healing hides in an embryonic brake that can be released

The wounds healed so completely that the scientists lost track of them. Hannah Tam, working under a dissection microscope at Harvard, had punched tiny holes in the skin of mouse embryos and needed fluorescent beads and henna ink to mark where those injuries had been. The skin had regrown hair follicles, blood vessels, pigment cells, fat, nerves - everything. No scar. No trace.

That vanishing act is something adult skin cannot perform. Cut yourself, and the body will reseal the surface, but most of the roughly 10 to 50 cell types that make up healthy skin never come back. What fills in instead is dense collagen scar tissue - functional, but fundamentally altered. Hair follicles disappear. Sweat glands are absent. The repaired patch is stiffer, paler, and permanently different from the skin around it.

Eight days that change everything

Tam, a graduate of Harvard's Kenneth C. Griffin Graduate School of Arts and Sciences, spent five years mapping exactly when embryonic skin loses its regenerative powers. Using a biopsy punch tool to create full-thickness wounds at different developmental stages, she compared outcomes in embryonic and newborn mice across multiple timepoints.

The window turned out to be remarkably narrow. Mice wounded three days before birth regenerated diverse cell types and produced skin nearly indistinguishable from unwounded tissue. Mice wounded five days after birth healed with epithelial coverage, dense collagen, and abnormally packed nerve fibers and immune cells. That is a shift from full regeneration to scarring in just eight days.

The finding, published March 20 in Cell, set up the central question: what molecular switch flips during that eight-day span?

Fibroblasts recruit too many nerves

Tam and her senior author, Ya-Chieh Hsu, professor of stem cell and regenerative biology at Harvard, expected the answer to involve immune cells. Midway through the project, the team hit a wall pursuing that hypothesis. The real culprit turned out to be something different: a conversation between fibroblasts and nerves.

In postnatal wounds, fibroblasts ramp up production of a gene called Cxcl12. That gene encodes a signaling molecule that recruits nerves to the injury site in excessive numbers - a phenomenon the researchers call hyperinnervation. The flood of nerve fibers at the wound impairs the regrowth of other skin cell types, locking the tissue into a scarring pathway.

In embryonic wounds, this does not happen. Fibroblasts do not upregulate Cxcl12, nerves arrive in normal numbers, and the full complement of skin cells regenerates.

Releasing the brake with Botox - and genetics

The team tested their hypothesis by depleting Cxcl12 in wounds of postnatal mice. With the nerve-recruiting signal removed, hyperinnervation was curtailed and skin regrew diverse cell types - approaching the regenerative capacity of embryonic healing.

They achieved a similar result by blocking local nerve signaling with botulinum toxin A, commonly known as Botox. Injecting the toxin near postnatal wounds dampened the excess nerve activity and allowed broader cell type regeneration.

Both experiments pointed to the same conclusion: postnatal skin retains an inherent ability to regenerate. That ability is not lost - it is actively suppressed by a fibroblast-nerve signaling loop that emerges after birth.

A simpler solution than expected

Before this study, Hsu anticipated that restoring embryonic-like healing would require recreating a cocktail of regeneration-promoting factors - essentially building the regenerative program from scratch. The actual mechanism proved far simpler. Rather than adding something new, the team only needed to remove a block.

That distinction matters for potential clinical translation. Blocking a single signaling pathway is, in principle, more tractable than reconstructing a complex developmental program. The fibroblast-nerve interaction identified here provides a specific, targetable mechanism rather than a vague aspiration toward "regenerative medicine."

What this does not yet prove

Several important caveats apply. This work was done entirely in mice. Mouse skin differs from human skin in thickness, hair density, and healing dynamics. Whether the Cxcl12-nerve axis operates the same way in human wounds is unknown.

The study demonstrated improved regeneration of multiple cell types but did not show perfect reconstruction of every skin component in postnatal animals. The results are closer to embryonic healing, not identical to it.

Botox is already used clinically for other purposes, which might suggest a shortcut to human applications. But the dosing, timing, and delivery method for wound healing would need to be worked out independently. The fact that a drug is FDA-approved for one indication does not guarantee safety or efficacy for another.

The narrow developmental window identified in mice - eight days - also raises questions about whether adult human skin, which has been in its postnatal configuration for decades, would respond the same way as skin just days past the embryonic threshold.

The fibroblast-nerve axis as a new target

What the study does establish is a previously unrecognized relationship between fibroblasts and nerves in wound healing. The field has traditionally focused on immune cells, growth factors, and stem cell populations as the key players in skin repair. The finding that nerve recruitment by fibroblasts is a primary determinant of whether skin scars or regenerates opens a different avenue of investigation.

Tam, now a postdoc at Scripps Research in California, emphasized the broader significance of this cell-cell communication. Fibroblasts and nerves had not been studied as direct partners in wound healing, and identifying their interaction as the switch between regeneration and scarring reframes the field's approach to the problem.

For Hsu, the work carries a specific kind of optimism - not the overheated variety, but the kind that comes from learning that a problem may be more solvable than it appeared. The regenerative capacity is already there, sitting dormant in adult skin. The challenge now is figuring out whether releasing that brake works as cleanly in human tissue as it does in a newborn mouse.