When Salamanders Regrow Arms and Humans Form Scars: The 300-Million-Year Divergence

In 1768, Lazzaro Spallanzani published something that sounded like science fiction: salamanders could regrow amputated limbs. Not just heal the wound—actually regenerate a complete, functional limb with bones, muscles, nerves, and blood vessels. Over 250 years later, humans still cannot do this. Lose a finger, and it is gone forever. But dig beneath this apparent biological unfairness, and you find a story of evolutionary trade-offs, molecular complexity, and a surprising fact: the genes for regeneration never left us.

The Blastema: Nature’s 3D Printer

The axolotl, a salamander native to Mexico’s Lake Xochimilco, represents the gold standard of vertebrate regeneration. When an axolotl loses a limb, something remarkable happens within hours. The wound does not simply close with fibrous tissue. Instead, cells near the injury site begin a process called dedifferentiation—they revert from specialized adult cells back to a more primitive, embryonic-like state.

These dedifferentiated cells migrate to the wound site and aggregate into a structure called a blastema. Think of it as a biological construction zone: a mass of progenitor cells that will divide, differentiate, and rebuild the missing structure piece by piece. The blastema is not just a lump of stem cells—it contains cells with positional memory, knowing whether they should build a shoulder, an elbow, or a fingertip.

Image source: PMC - The axolotl limb blastema: cellular and molecular mechanisms

This process follows a precise choreography. First, wound epithelial cells migrate to cover the amputation surface within hours. This wound epidermis thickens to form the apical epithelial cap (AEC), which signals to underlying tissues to begin dedifferentiation. By day 7-10, the blastema is visible as a cone-shaped bulge. Over the next 30-60 days, depending on limb size and temperature, the blastema proliferates and redifferentiates, restoring the limb with perfect fidelity.

What Humans Do Instead: The Scar Barrier

When a human suffers a limb injury, the body’s priority is survival, not restoration. Within minutes, platelets and fibrin form a clot. Inflammatory cells flood the site, clearing debris and fighting infection. Then fibroblasts arrive—and this is where the human and axolotl paths diverge dramatically.

Fibroblasts in human wounds deposit collagen rapidly and abundantly, forming a dense, disorganized matrix. This scar tissue acts as a biological patch: it seals the wound quickly and provides mechanical strength. But it comes at a cost. The scar lacks the original tissue’s architecture, blood vessels, nerves, and functional cells. It is a dead end, not a starting point.

The extracellular matrix (ECM) plays a crucial role in this divergence. In regenerative species like axolotls, the wound ECM remains permissive—rich in hyaluronic acid and low in mature collagen, allowing cell migration and reprogramming. In mammals, the ECM rapidly becomes restrictive, with cross-linked collagen fibers that physically trap cells and signal them to stop proliferating.

Research published in 2025 showed that inhibiting YAP (Yes-associated protein) signaling immediately after wounding can prevent scarring and promote regeneration in mice. Verteporfin, a YAP inhibitor, shifted the healing response from fibrotic repair toward a more regenerative phenotype. This suggests the scar pathway is not inevitable—it is a molecular choice our cells make.

The Nerve Connection: A Required Signal

One of the most surprising discoveries in regeneration biology is the absolute requirement for nerve signaling. If you amputate a salamander limb and also cut the nerves supplying it, regeneration fails completely. The blastema forms initially but cannot sustain proliferation.

The molecular mediator of this nerve dependence is neuregulin-1 (NRG1). Nerves secrete NRG1, which binds to ErbB2 receptors on blastema cells, driving their proliferation. In 2016, researchers demonstrated that delivering NRG1 to denervated limbs can rescue regeneration, even without nerve input.

This nerve dependency has profound implications for human medicine. Spinal cord injuries not only cause paralysis but also impair regeneration throughout the body by disrupting nerve-derived signals. The salamander has evolved a system where nerves and tissues are locked in a regenerative partnership—break the partnership, and regeneration stops.

Positional Identity: How Cells Know What to Build

Perhaps the most puzzling aspect of limb regeneration is positional memory. When an axolotl loses a hand, it regrows a hand—not a shoulder or an elbow. The blastema somehow “knows” what is missing and builds exactly that.

The answer lies in a gradient of cell surface proteins and gene expression patterns that encode positional identity along the proximal-distal axis (shoulder to fingertip). Prod1, a salamander-specific cell surface protein related to mammalian CD59, is expressed in an exponential gradient: high in proximal (shoulder) regions, low distally (fingertips).

Retinoic acid (RA) plays a central role in this positional encoding. A 2024 study published in Nature Communications revealed that the enzyme CYP26B1, which degrades retinoic acid, determines segment identity during regeneration. When CYP26B1 is blocked, blastema cells receive excess RA signaling and shift their identity proximally—they start building structures they should not build at that location.

Hox genes, the ancient architects of body patterning, are also reactivated during regeneration. MEIS homeodomain proteins, which regulate Hox gene expression, help establish proximal identity. The regenerating limb essentially recapitulates embryonic development, using the same genetic toolkit that originally built the limb.

The Evolutionary Trade-off: Regeneration vs. Cancer

Why would evolution abandon such a useful ability? The leading hypothesis centers on cancer.

Regeneration requires cells to dedifferentiate and proliferate rapidly—precisely the behaviors that define cancer cells. The molecular machinery that enables a salamander to regrow a limb could, if misregulated, drive tumor formation. Mammals may have traded regeneration for enhanced cancer suppression.

This trade-off is not theoretical. Axolotls and other highly regenerative species do have higher baseline rates of cell proliferation. Yet they rarely develop tumors. How they achieve this balance remains unclear, but it likely involves tightly controlled signaling windows—blastema cells dedifferentiate only in specific contexts and for limited durations.

The tumor suppressor p53 plays a key role in this balance. In mammals, p53 responds to DNA damage by arresting the cell cycle or triggering apoptosis. This protects against cancer but also limits the sustained proliferation needed for regeneration. Some regenerative species appear to have modified p53 pathways that allow controlled proliferation without compromising genomic stability.

Mammals That Break the Rules

Not all mammals have lost regeneration entirely. The African spiny mouse (Acomys) can regenerate skin, cartilage, and even punctured ear tissue without scarring. When researchers compared healing in spiny mice versus standard lab mice, they found dramatic differences in inflammatory response and ECM composition.

Spiny mice produce less dense collagen at wound sites and maintain higher levels of hyaluronic acid. Their macrophages show a different activation profile, favoring tissue remodeling over fibrosis. This suggests the mammalian regenerative capacity was not completely lost—it was suppressed, and in some species, partially reawakened.

Even humans retain limited regeneration. Children under age 10 can regenerate amputated fingertips if the wound is left uncovered and not surgically closed. The nail bed contains Wnt-active nail stem cells that orchestrate this process, recruiting nerves and mesenchymal cells to rebuild the distal phalanx.

A 2025 study published in Regenerative Medicine mapped the clinical phases of human fingertip regeneration, identifying distinct molecular signatures at each stage. The process involves blastema-like structures and recapitulates developmental pathways—evidence that the machinery for regeneration exists in humans, dormant but not destroyed.

Reactivating Dormant Potential

If regeneration genes still exist in humans, can we reactivate them? Research suggests cautious optimism.

In 2013, scientists at Harvard discovered that Lin28a, an RNA-binding protein involved in developmental timing, could enhance tissue repair in mice. Mice engineered to overexpress Lin28a showed improved regeneration of digit tips, ear tissue, and hair. Lin28a appears to shift cellular metabolism toward a more embryonic state, favoring proliferation over quiescence.

Gene therapy approaches have also shown promise. Researchers have identified sets of “master genes” that, when delivered via viral vectors, can induce regeneration-like responses in mammalian tissues. The challenge is precision—activating these genes systemically would risk uncontrolled proliferation.

The Wnt/β-catenin pathway remains a central target. This signaling cascade drives blastema formation across regenerative species, from zebrafish to salamanders. Pharmacological activation of Wnt signaling can enhance wound repair in mammals, but must be carefully controlled to avoid oncogenic effects.

The Future: Borrowing from Salamanders

The goal of regenerative medicine is not to make humans into salamanders. It is to understand the principles that enable regeneration and apply them selectively. This might involve:

• Modulating the immune response to reduce fibrosis
• Delivering nerve-derived factors like neuregulin to injury sites
• Creating permissive ECM environments using biomaterials
• Temporarily activating developmental genes in specific tissues

The axolotl genome, sequenced in 2018, contains about 32 billion base pairs—ten times the size of the human genome. Within that expansive genetic landscape lie the instructions for regeneration. Decoding those instructions, and learning to apply them safely, represents one of biology’s grand challenges.

Regeneration is not magic. It is molecular choreography, a dance of signals and responses that salamanders perform effortlessly. Humans traded this dance for other advantages—larger brains, longer lifespans, lower cancer risk. But the music never stopped. The question is whether we can learn to hear it again.

References

1. McCusker, C., Gardiner, D. M. (2011). The axolotl limb blastema: cellular and molecular mechanisms driving blastema formation and limb regeneration. Developmental Biology, 332(2), 269-281. https://pmc.ncbi.nlm.nih.gov/articles/PMC4895312/

2. Seifert, A. W., et al. (2012). Skin shedding and tissue regeneration in African spiny mice (Acomys). Nature, 489, 561-565. https://pmc.ncbi.nlm.nih.gov/articles/PMC3480082/

3. Farkas, J. E., et al. (2016). Neuregulin-1 signaling is essential for nerve-dependent axolotl limb regeneration. Development, 143(16), 2724-2731. https://pubmed.ncbi.nlm.nih.gov/27317805/

4. Kumar, A., et al. (2007). Molecular basis for the nerve dependence of limb regeneration in an adult vertebrate. Science, 318(5851), 772-777. https://pmc.ncbi.nlm.nih.gov/articles/PMC2696928/

5. Takeo, M., et al. (2013). Wnt activation in nail epithelium couples nail growth to digit regeneration. Nature, 499(7457), 228-232. https://www.nature.com/articles/nature12214

6. Godwin, J. W., et al. (2013). Macrophages are required for adult salamander limb regeneration. PNAS, 110(23), 9415-9420. https://pmc.ncbi.nlm.nih.gov/articles/PMC5469729/

7. Echeverri, K., et al. (2024). Retinoic acid breakdown is required for proximodistal positional identity during axolotl limb regeneration. Nature Communications. https://www.nature.com/articles/s41467-025-59497-5

8. Nguyen, Q. H., et al. (2025). Inhibiting mechanotransduction prevents scarring and yields wound regeneration. Science Translational Medicine. https://www.science.org/doi/10.1126/scitranslmed.adt6387

9. Gawriyski, L., et al. (2025). Human fingertip regeneration follows clinical phases with distinct molecular signatures. Regenerative Medicine, 20(1). https://www.nature.com/articles/s41536-025-00441-y

10. Baddar, N. W. A., et al. (2024). A chromatin code for limb segment identity in axolotl limb regeneration. Developmental Cell, 61(4), 521-535. https://www.cell.com/developmental-cell/pdf/S1534-5807(24)00300-9.pdf

11. Shyh-Chang, N., et al. (2013). Lin28 enhances tissue repair by reprogramming cellular metabolism. Cell, 155(4), 778-792. https://www.cell.com/cell/fulltext/S0092-8674(13)01280-5

12. Bely, A. E., & Nyberg, K. G. (2010). Evolution of animal regeneration: A tangle story. Frontiers in Ecology and Evolution. https://www.frontiersin.org/journals/ecology-and-evolution/articles/10.3389/fevo.2021.621686/full