In 2014, scientists compared EPAS1 DNA from 40 Tibetans and 40 Han Chinese and linked a high-altitude adaptation to Denisovan-like ancestry. 12 years later, the finding traced the genetic gift to an interbreeding event about 48,700 years ago

At roughly 4,000 meters above sea level, the air has not suddenly lost its oxygen. Oxygen still makes up about 21% of dry air. What has changed is the pressure squeezing those oxygen molecules together. The result is a thinner supply of oxygen with every breath, creating a physiological problem that humans cannot simply solve by breathing harder. For Tibetans, however, evolution appears to have found an unusual answer, and part of that answer may have arrived from an extinct human population.

The clue lies near EPAS1, a gene involved in the body's response to low oxygen. Tibetan populations carry a distinctive stretch of DNA around this gene at unusually high frequency. Genetic evidence shows that this haplotype is strikingly similar to DNA recovered from Denisovans, an archaic human group known largely through fragments of ancient remains and their genomes. The important story is not that Tibetans inherited a special “altitude gene.” They inherited a particular version of a gene region that natural selection later favored.

How Denisovan DNA gave Tibetans a high-altitude super-gene

The basic chemistry of the atmosphere remains remarkably stable with altitude. Roughly one-fifth of dry air is oxygen whether a person is standing near sea level or high on the Tibetan Plateau. But atmospheric pressure falls as the weight of air above decreases. That means the partial pressure of oxygen also falls, reducing the pressure gradient that normally drives oxygen from the lungs into the bloodstream.


For a person arriving from low altitude, the body responds quickly. Breathing becomes deeper or faster, and over longer periods the kidneys increase signals that stimulate red blood cell production. More red blood cells mean more hemoglobin, and more hemoglobin means greater capacity to carry oxygen through the circulation. This response is useful, but it has a cost because excessively concentrated blood becomes more viscous and places additional demands on the cardiovascular system.

Tibetans evolved along a different physiological route. At around 4,000 meters, many Tibetan highlanders have substantially lower hemoglobin concentrations than Andean highlanders living at comparable elevations. Earlier physiological studies found that Tibetan men and women could have hemoglobin concentrations comparable to people living at sea level, despite breathing air with much lower oxygen pressure. Their bodies are not ignoring hypoxia. They are managing it without relying as heavily on erythrocytosis, the expansion of red blood cell mass.

That distinction matters because simply making more red blood cells is not a perfect solution. More hemoglobin can increase oxygen-carrying capacity, but excessive erythrocytosis can raise blood viscosity and contribute to chronic mountain sickness and cardiovascular strain. Tibetan adaptation appears to place greater emphasis on ventilation, blood flow and oxygen delivery to tissues while avoiding an excessive rise in hemoglobin.

The gene that senses a shortage of oxygen

EPAS1 belongs to the machinery cells use to detect and respond to oxygen availability. It encodes hypoxia-inducible factor 2-alpha, or HIF-2α, a transcription factor that helps regulate genes involved in the response to low oxygen. When oxygen becomes scarce, this molecular system can influence processes including erythropoietin signaling and red blood cell production.

This makes EPAS1 an especially logical place to look for high-altitude adaptation. Genome-wide studies found one of the strongest signatures of natural selection ever reported in Tibetans around the EPAS1 region. Variants in this region were also associated with hemoglobin concentration, connecting a genetic signal to a measurable physiological trait. But there was an even stranger clue hidden in the DNA itself.

The Tibetan sequence did not merely look unusual. It looked archaic.

A genetic signature from Denisovans

In 2014, researchers compared DNA around EPAS1 in Tibetan and Han Chinese individuals with the available Denisovan genome. They found a distinctive Tibetan haplotype, meaning a linked set of DNA variants, that closely matched Denisovan sequence. Its unusual structure and distribution made ordinary inheritance from a shared ancient ancestor an unlikely explanation. The evidence instead pointed toward introgression, the movement of DNA between populations through interbreeding.

The finding was remarkable for another reason. Tibetans do not have exceptionally large amounts of Denisovan ancestry across their entire genomes. Some populations in Oceania carry considerably more Denisovan-derived DNA overall. Yet those populations do not carry the same Tibetan EPAS1 haplotype at comparable frequency. This suggests that the important evolutionary event was not simply having Denisovan ancestry. It was possessing a particular archaic DNA segment that happened to become valuable under extreme high-altitude conditions.

That is the essence of adaptive introgression. Interbreeding introduces genetic variation into a population, and the environment determines whether some of that variation becomes useful. If a particular inherited sequence improves survival or reproduction, natural selection can increase its frequency generation after generation. An ancient genetic contribution can therefore become a modern evolutionary advantage without requiring a brand-new mutation to arise from scratch.

The “Denisovan gene” description is too simple

It is tempting to call EPAS1 a Denisovan gene, but that description gets the biology wrong. Modern humans already possess EPAS1. The Tibetan adaptation involves a particular haplotype around the gene, containing multiple linked variants. The key variants identified in the Tibetan population are largely non-coding, meaning they do not change the amino-acid sequence of the HIF-2α protein itself. Their importance appears to involve regulation of how the gene is expressed or processed.

Researchers have proposed that Tibetan-enriched EPAS1 variation reduces expression or alters processing of the gene's product. Experimental work has reported reduced EPAS1 expression and physiological responses consistent with a blunted reaction to chronic hypoxia. But the precise causal variants and the complete molecular chain connecting the archaic haplotype to the Tibetan phenotype remain an active area of research.

That uncertainty is important. Population genetics can show that a DNA segment was inherited from an archaic population and that natural selection strongly favored it. It is harder to identify exactly which nucleotide changes are responsible for the resulting physiology. A long inherited haplotype can contain many variants traveling together, making the final causal mutation difficult to isolate experimentally.

Why lower hemoglobin can actually help

The Tibetan strategy makes more sense when oxygen delivery is viewed as a whole system rather than a simple hemoglobin equation. Oxygen has to enter the lungs, cross into blood, circulate through vessels and finally diffuse into tissues. Increasing hemoglobin is only one way to improve that chain. Tibetans also show elevated resting ventilation, differences in pulmonary blood flow and high capillary density in skeletal muscle, all of which can help move oxygen through the body.

Nitric oxide may also contribute to this system. Studies have found unusually high levels of nitric oxide-related compounds in Tibetans, which can promote blood-vessel dilation and improve circulation. The result is a physiological strategy that does not depend entirely on packing more hemoglobin into every milliliter of blood. Oxygen delivery can instead be improved by making circulation and tissue-level exchange more efficient.

There is another gene worth mentioning here: EGLN1. It encodes an oxygen-sensing enzyme called PHD2 and is also strongly associated with Tibetan high-altitude adaptation. Variants in EGLN1, like variants around EPAS1, have been associated with lower hemoglobin concentrations. Together, these genes point toward a broader alteration of the hypoxia-inducible factor pathway rather than a single genetic switch controlling Tibetan altitude adaptation.

When did the Denisovan DNA become useful?

One of the most intriguing parts of the story is that the Denisovan-like DNA may have entered the ancestors of Tibetans long before people permanently settled the highest parts of the plateau. A 2021 analysis estimated that the relevant Denisovan introgression occurred roughly 48,700 years ago, while the researchers estimated that strong selection on the EPAS1 haplotype began much later, around 9,000 years ago, although both estimates carry substantial uncertainty.

That timing illustrates an important principle in evolution. A genetic variant does not have to be beneficial immediately after it enters a population. It can remain relatively neutral for thousands of years. If the environment changes, or if a population moves into a new ecological niche, an old piece of genetic variation can suddenly become valuable. Natural selection can then raise its frequency dramatically.

In this case, an inherited piece of archaic DNA appears to have been waiting in the human genetic background until life on the Tibetan Plateau made its particular biological effects advantageous. The DNA was ancient. The selection acting on it was comparatively recent.

Evolution did not build the Tibetan solution from one gene

The Denisovan connection is powerful because it gives researchers an unusually clear example of adaptive introgression in humans. But it should not obscure the larger picture. Tibetan high-altitude adaptation involves many physiological systems and numerous genetic regions, including genes associated with oxygen sensing, blood vessels, metabolism and erythropoiesis. EPAS1 is one of the strongest and most unusual signals, not the entire explanation.

The comparison with other high-altitude populations makes that point even clearer. Andean populations have evolved a different strategy, generally maintaining higher hemoglobin concentrations at altitude. Ethiopian highlanders show yet another physiological pattern. Humans have therefore reached high-altitude adaptation through multiple evolutionary routes, shaped by population history, environment and the particular genetic variation available to each group.

The deeper lesson is not simply that Denisovans “gave Tibetans” a gene for surviving mountains. Human evolution was more complicated than that. Ancient populations met, exchanged DNA and eventually disappeared as recognizable groups, but some of their genetic material remained. Much later, in an environment where oxygen was scarce, one inherited segment became extraordinarily useful. Natural selection amplified it, turning a fragment of archaic genetic history into part of a living human adaptation.

The Tibetan EPAS1 story is therefore a rare window into evolution working with material it did not create itself. A useful mutation did not necessarily have to appear at the moment humans needed it. Sometimes evolution can inherit an old solution, test it against a new environment and preserve it when the circumstances are right. In the thin air of the Tibetan Plateau, a piece of DNA from an extinct human lineage appears to have become part of the answer.