The parts were there. The conditions weren't.
Cut off a salamander's leg and it grows back. Cut off yours and it doesn't. That gap has puzzled biologists for centuries, and for most of that time the working assumption was that mammals simply lack whatever genetic machinery amphibians carry. A new study from researchers at EPFL in Lausanne suggests the real answer is both simpler and stranger: it may come down to how a wounded cell reads oxygen in the first hours after injury.
The team, led by Can Aztekin, ran experiments on amputated limb tissue kept alive outside the body. Mouse tissue, held in ordinary air, stalled. Tadpole tissue kept rebuilding. That split, visible in a dish, became the entry point for a deeper question: what exactly is different between the two?
The answer pointed to a protein called HIF1A, which cells use to sense how much oxygen is available. In low-oxygen conditions, HIF1A stays stable and active, and that stability appears to open the door to regeneration — prompting wound closure, shifting the cell's energy production toward glycolysis, and loosening the gene-access patterns that allow new tissue to form. In higher oxygen, HIF1A breaks down quickly. In mammalian tissue, that breakdown happens fast enough to shut off the regenerative program before it can gain any momentum. The wound seals, but it seals with scar tissue instead of new limb.
When the researchers lowered oxygen around embryonic mouse limb tissue, something shifted. Skin cells became more mobile, wound closure accelerated, and the early cellular states associated with regrowth began to appear. No new leg formed — the researchers were careful not to overstate what they saw — but the tissue moved through the opening steps of a program that normally never starts. That finding reframes the problem considerably: mammals don't appear to be missing the parts. They appear to be missing the right starting conditions.
Frog tadpoles added a useful counterpoint. Their limbs kept regenerating even at 60 percent oxygen, a concentration that would shut down mouse tissue entirely. The reason, the team found, is that tadpole cells produce less of the molecular machinery that normally degrades HIF1A, so the protein stays active longer even when oxygen is plentiful. Axolotl data fit the same pattern, which meant the finding wasn't a quirk of one species but something that held across the amphibians most studied for their healing abilities.
Mammals aren't entirely without regenerative capacity. Fingertip injuries in humans and mice can sometimes result in partial regrowth of the digit tip — a narrow exception that turns out to be instructive. Earlier research found that softer tissue at the wound site favors regrowth while stiffer tissue favors scarring. Placed alongside the oxygen findings, that small success suggests the full picture involves multiple early conditions that either align or don't. When they don't, the wound closes the wrong way and the window closes with it.
Age complicates the story further. Tadpoles lose much of their regenerative ability as they mature into frogs, and Aztekin's earlier work traced that loss to changes in the wound's surface tissue — signals that push repair toward scarring as the animal ages. The oxygen result fits neatly into that timeline as one more way a wound that might have regenerated instead gets derailed.
When the team compared data from frogs, axolotls, mice, and humans, the same divide kept appearing. Human cells clustered with mouse cells, showing the stronger oxygen-sensing pattern associated with early shutdown. Aztekin noted that mammalian regeneration has rarely been examined experimentally alongside amphibian regeneration in a directly comparable way, and that the side-by-side approach changes what questions are even askable.
The study, published in the journal Science, doesn't deliver a therapy or a timeline. What it delivers is a target — a specific, testable cellular response that occurs in the first hours after injury and appears to determine which path a wound takes. If that oxygen-sensing window can be manipulated early enough, the researchers suggest, mammalian healing might be nudged away from scarring. The centuries-old gap between a salamander and a human may not be written in the genes. It may be written in the wound.
Notable Quotes
For a long time, regeneration research focused on amphibians, while mammalian regeneration was rarely examined experimentally side by side in a comparable manner.— Can Aztekin, EPFL
By directly comparing species that can and cannot regenerate, we bring a fresh perspective to a centuries-old question.— Can Aztekin, EPFL
The Hearth Conversation Another angle on the story
So the big claim here is that mammals actually have the regeneration machinery — they just never turn it on?
That's the thrust of it. When researchers lowered oxygen around mouse embryonic tissue, the early regenerative steps started happening. The parts were there. The conditions weren't.
Why does oxygen matter so much in the first place?
Because the protein HIF1A, which cells use to sense oxygen levels, seems to act like a gate. Keep it stable and the wound heads toward rebuilding. Let it break down quickly — which is what happens in mammalian tissue at normal oxygen — and the gate closes before anything gets started.
And amphibians just keep that gate open longer?
Essentially. Tadpoles produce less of the machinery that degrades HIF1A, so the protein stays active even when oxygen isn't particularly low. The gate stays open longer, and the wound has more time to commit to regeneration rather than scarring.
What about the fact that humans can sometimes regrow a fingertip?
That's actually one of the more interesting details. It shows the system isn't completely absent in mammals — it's just rarely reached. The fingertip exception suggests that when local conditions happen to align, the machinery can still run. It's a narrow window, but it's real.
Does age factor in at all?
It does. Tadpoles lose their regenerative ability as they mature, and earlier work by the same researcher traced that to changes in wound surface tissue. The oxygen finding adds another layer — one more way the early conditions can go wrong as an animal ages.
What's the honest limit of what this study actually showed?
No one grew back a mouse leg. What they triggered was the opening act — wound closure behavior, cell mobility, gene-access changes. The first chapter, not the whole book. But knowing where the story starts is genuinely useful if you want to change how it ends.