I read a paper in Cell yesterday morning that stopped me mid-coffee. Before lunchtime I was texting a my buddy, Prof. Jon Russin, a neurosurgeon (now Chief in Vermont) about resurrecting a grant from a few years ago.
The paper is from Ya-Chieh Hsu’s lab at Harvard: Hyperinnervation Inhibits Organ-Level Regeneration in Mammalian Skin (Tam et al., Cell, March 2026). And it fundamentally reframes how we think about wound healing.
The Setup
Here’s what we’ve always assumed: adult mammalian skin can’t regenerate. It can repair — close the gap, lay down scar — but it can’t rebuild the original architecture. No hair follicles. No fat cells. No sweat glands. Just collagen and epithelium. Scar. We’ve treated this as a biological law.
Turns out it’s not a law. It’s a brake.
The Discovery
Hsu’s team showed that three days before birth, a mouse can be wounded and will regenerate its skin completely — hair follicles, adipocytes, blood vessels, lymphatics, nerves, melanocytes — the works. Indistinguishable from unwounded tissue. (They had to mark the wound sites with fluorescent beads because the healing was so perfect they couldn’t find them otherwise.)
But by five days after birth, that capacity vanishes. The wound fills with scar. Dense collagen. Abnormally thick nerve fibers. Missing cell types.
What changes in that eight-day window?
A specific population of fibroblasts appears after birth and starts expressing a chemokine called CXCL12. This signal recruits excessive sensory nerve fibers into the wound bed via the CXCR4 receptor. And those nerves — the hyperinnervation — are what blocks regeneration. The nerves are the brake.
The Punchline
When the team injected botulinum toxin A into postnatal mouse wound sites — yes, Botox — it reduced nerve ingrowth and restored full skin regeneration. Complete. Multilineage. Indistinguishable from unwounded tissue.
They got the same result three different ways: genetic deletion of Cxcl12 in fibroblasts, injection of BTX-A, and selective ablation of sensory neurons. The conclusion is airtight: nerve signaling activity — not just nerve presence — is what suppresses regeneration. Block the signal, and the regenerative capacity that was always there is released.
As Hsu put it: “I didn’t think that we’d have to retract a brake, which actually is good news — it’s a lot easier.”
Why This Matters for Diabetic Foot Ulcers
Now here’s where my coffee went cold.
Every patient with a diabetic foot ulcer has peripheral neuropathy. That’s why they ulcerate — they can’t feel the repetitive trauma. We’ve always thought of the neuropathy as the cause of the problem: no sensation, no protection, tissue breakdown.
But the Hsu paper forces a different question. Neuropathic patients have lost protective sensation, yes. But have they lost all nerve fiber signaling activity in the wound bed? Almost certainly not. Diabetic neuropathy impairs sensory transmission, but the nerve fibers don’t disappear — they remodel, sprout aberrantly, and critically, they still release neuropeptides and still participate in chemokine signaling.
Which means: the very nerve fibers that can’t protect our patients from ulcerating may still be suppressing their ability to heal.
That’s the paradox. Patients ulcerate because of neuropathy. And they may heal poorly in part because residual nerve signaling in the wound bed is still jamming the regeneration program. Like noise-cancelling headphones that block the music but let the static through.
BTX-A could be the tool that cancels the static.
We’ve Been Here Before — Almost
This isn’t the first time botulinum toxin has crossed paths with our world. Several years ago, we submitted an NIH R01 proposing a BTX-A trial for DFU healing. The reviewers liked the team. They liked the concept. But they wanted a stronger mechanistic foundation. Why would a nerve blocker help heal a wound?
The Hsu Cell paper is that foundation.
Separately, our group served as PI on an Ipsen-sponsored Phase 2 randomized controlled trial of abobotulinumtoxinA (Dysport) for hallux valgus pain (Armstrong et al., J Foot Ankle Surg 2023), giving us published safety and tolerability data for BTX-A injection into foot structures. And just this past January, a JAMA Dermatology meta-analysis (Zhu et al., 2026) pooled 30 studies of BTX-A for ischemic digital ulcers and found complete healing in over 85% of patients.
The mechanism from Cell. The clinical signal from JAMA Dermatology. The foot safety data from our own trial. The pieces are lining up.
And We Can See It Now — Literally
One more piece. Our colleagues/partners at Caltech and USC — Lihong Wang, Charles Liu, Jonathan Russin, and Tze-Woei Tan — just published a new imaging modality called RUS-PAT (Rotational Ultrasound and Photoacoustic Tomography) in Nature Biomedical Engineering (Zhang et al., 2026). It produces 3D panoramic images of soft tissue, microvasculature, and tissue oxygenation at depths up to 4 cm in under a minute. No contrast. No radiation. No sedation.
And here’s the kicker: Figure 6 of that paper shows RUS-PAT imaging of a patient with diabetic foot ulcers. It’s already been done. The vascular architecture around the wound bed — visible, quantifiable, and trackable over time.
Screenshot
If the Hsu mechanism operates in human DFU tissue the way it operates in mice, RUS-PAT gives us the tool to watch it happen — to image the nerve and vascular changes before and after BTX-A treatment, non-invasively, at every visit.
What’s Next
We’re considering putting together a multi-site clinical trial — a randomized, double-blind study of perilesional BTX-A as an adjunct to standard wound care in neuropathic DFUs. RUS-PAT imaging to track the wound bed biology. Tissue biopsies to measure the CXCL12/CXCR4 axis in human DFU tissue for the first time. A test of the neuropathy paradox: among patients who all lack sensation, does the degree of residual nerve fiber signaling in the wound bed predict who responds?
We’re pursuing this through both an industry partnership and a federal grant. The conversations started today. Literally today.
The last FDA-approved drug for DFU healing was becaplermin in 1997. That’s almost thirty years with nothing new on the shelf. If this mechanism translates to humans, botulinum toxin — an inexpensive, FDA-approved, widely available agent that every physician knows how to inject — could become the first mechanistically grounded pharmacologic therapy for wound regeneration.
Not repair. Regeneration.
Stay tuned.
Reference: Tam K, et al. Hyperinnervation inhibits organ-level regeneration in mammalian skin. Cell. 2026. DOI: 10.1016/j.cell.2026.02.027
#DiabeticFoot #WoundHealing #Botox #BotulinumToxin #Regeneration #Cell #Neuropathy #DFU #LimbPreservation #ActAgainstAmputation #CXCL12 #CXCR4 #RUSPAT #PhotoacousticImaging #WoundRegeneration #AmputationPrevention
Leave a Reply