Semaphorin3A induces nerve regeneration in the adult cornea-a switch from its repulsive role in development

Semaphorin three A has one of the clearest reputations in developmental neuroscience. It is a stop signal. It tells growing nerve fibers to collapse their tips, pull back, and go elsewhere. That reputation was earned over decades of careful work, and it holds up in embryos. But Zhang and colleagues just showed that the same molecule, in an adult cornea that has been injured, does the opposite. The tissue floods itself with Semaphorin three A, and the nerves grow back faster. The molecule didn't change. The neurons did. To understand why that matters, you need to know what corneal nerves actually do. The cornea is the most densely innervated tissue in the human body— not the brain, not the fingertips, but the clear dome at the front of the eye. Those sensory fibers are not just there for pain. They supply trophic support, the kind of ongoing biochemical maintenance that keeps the epithelium healthy and clear. When the nerves are gone, the tissue breaks down. Wounds don't heal. Ulcers form. Zhang and colleagues describe a condition called neurotrophic keratopathy— chronic corneal damage driven by nerve loss— and it's caused by insults most of us have encountered or will: trauma, infection, surgery, diabetes, chemical burns. After LASIK, corneal nerves are routinely severed. In diabetic patients, slow nerve degeneration is a recognized complication. Nerves can grow back after these injuries, but they never fully recover their original density and branching pattern. That incomplete return is the clinical problem this paper is trying to solve. So the search for molecules that could push regeneration further has focused largely on growth factors— nerve growth factor, Vascular Endothelial Growth Factor A, Vascular Endothelial Growth Factor B. These are the molecules the field already knew could help. What Zhang and colleagues found when they looked at gene expression after corneal injury was that all six members of the Semaphorin class three family are expressed in the cornea under normal conditions. But after epithelial injury, one of them— Semaphorin three A— spikes quickly. It increases threefold within a single day. That's the puzzle. A tissue trying to regenerate its nerves is turning up the volume on a molecule famous for repelling them. To test whether this was signal or noise, the team went back to basics. In embryonic dorsal root ganglion neurons, or DRG neurons, which are the sensory cells that connect peripheral tissue to the spinal cord, Semaphorin three A behaved exactly as expected. The researchers cultured embryonic neurons, stimulated them with fifty nanograms per milliliter of nerve growth factor until long neurites with large growth cones had formed, then added Semaphorin three A. Within thirty minutes, the growth cones collapsed. The axons retracted. Time-lapse imaging captured the regression in real time, proceeding at roughly thirty to forty micrometers per hour until the neurites had pulled back almost entirely. Semaphorin three A abolished nerve growth factor-induced growth in embryonic cells. That's the textbook result. Then the team did the same experiment in adult neurons. Adult DRG cells, treated with nerve growth factor, were then challenged with Semaphorin three A at the same concentration. Nothing collapsed. No retraction. No inhibition at any dose tested. The percentage of neurons extending neurites matched the nerve growth factor-only controls. They repeated this in trigeminal ganglion neurons— TG neurons, which are the sensory cells that actually innervate the cornea— and got the same result. Semaphorin three A did not block growth. These experiments were run in triplicate, evaluating an average of two hundred neurons per DRG dish and sixty per TG dish, so the sample sizes were substantial. But the finding that really lands is this: when Semaphorin three A was applied to adult neurons with no nerve growth factor at all, the neurons grew anyway. Semaphorin three A alone, at twenty nanograms per milliliter and above— with the strongest effect at fifty nanograms per milliliter— induced neurite outgrowth in both DRG and TG cells. The length of those neurites, the number of branches, the proportion of long fibers— all of it was statistically comparable to what nerve growth factor produced on its own. The molecule that collapses embryonic growth cones drove adult axon growth as effectively as the field's go-to nerve growth factor. From the dish, the team moved to a living animal. They used thy1-YFP mice— a strain engineered so that neurons express a yellow-green fluorescent protein, making nerve fibers directly visible under fluorescence imaging without any staining. That means you can watch nerves grow back in real time, in the intact tissue. The injury model was a two millimeter circular debridement of the central corneal epithelium, which removes both the epithelial cells and the sub-basal nerve plexus beneath them. Immediately after the debridement, a small pocket was created in the stroma and a sucralfate-hydron pellet was implanted— either a vehicle pellet containing phosphate-buffered saline or a Semaphorin three A pellet loaded with one hundred nanograms of recombinant protein. The pellet releases Semaphorin three A slowly into the surrounding tissue. Seven days later, the corneas were harvested, mounted flat, and the regenerated nerve fibers were traced and summed using Neurolucida software. The result was clear. Corneas treated with Semaphorin three A showed faster extension of superficial nerves and higher fiber density across the debrided zone compared with vehicle controls. Quantitatively, Semaphorin three A produced nearly three times the nerve regeneration of the vehicle— the same roughly threefold ratio that appeared in the injury-triggered gene expression surge. The experiments were repeated three times with five animals per treatment group each time. There was no corneal neovascularization, no inflammation, and no infection in any treated mouse. The effect appeared to be directly neuronal. Taken together, the in vitro and in vivo results tell a coherent story. Adult peripheral sensory neurons respond to Semaphorin three A as a growth signal. The injured cornea turns Semaphorin three A up threefold. Delivering Semaphorin three A exogenously after injury accelerates regeneration threefold in a living eye. The tissue may be deploying this molecule deliberately— not as a vestigial developmental brake that got switched on by mistake, but as part of an endogenous repair response. Why the switch happens between development and adulthood is a question the paper opens rather than closes. Zhang and colleagues note that prior reports have suggested adult dorsal root ganglion neurons can engage Semaphorin three A through signaling pathways distinct from those in embryonic cells, and that inhibitory effects in some adult populations may be restricted to small-diameter neurons. The authors also place Semaphorin three A among a broader class of guidance cues— Netrins, Slits, Ephrins— that have documented dual roles depending on context. But they are direct about the limits: "how this switch in function from development to adulthood occurs is a matter of further studies." The receptor landscape— neuropilins, plexins, co-receptors— likely differs between embryonic and adult neurons, but which differences drive the reversal remains unresolved. What the paper does resolve is the functional question, and that resolution has clinical weight. Neurotrophic keratopathy currently has few treatment options. Cenegermin— a recombinant nerve growth factor eye drop— exists but is expensive and not universally effective. Semaphorin three A offers a different molecule from a different class, one that the cornea itself already reaches for after injury. The absence of angiogenic or inflammatory effects in treated mice is an early but encouraging sign that the molecule acts on neurons specifically rather than triggering a broader tissue response. The next steps Zhang and colleagues call for— mechanistic clarity on the developmental switch, longer-term safety and efficacy studies— are the right ones before any clinical application. The broader implication is worth sitting with for a moment. For decades, axon guidance molecules like Semaphorin three A have been framed primarily as barriers— obstacles that misdirected regenerating nerves away from their targets and contributed to the failure of peripheral and central nerve repair. This paper suggests that framing may be incomplete. In adult peripheral neurons, at least, a canonical repellent acts like a growth factor. The cornea already knew that. It just took a careful set of experiments to catch up. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.