The Denisovans: the third human species revealed by its genome

In 2010, a team of palaeogeneticists at the Max Planck Institute for Evolutionary Anthropology in Leipzig announced an unprecedented discovery: a tiny piece of finger bone found in a cave in the Altai Mountains of Siberia had yielded the DNADNAThe molecule carrying genetic information, used to reconstruct kinship between species.→ of an entirely unknown human species. Not Homo sapiensHomo sapiensThe present-day human species, which emerged in Africa around 300,000 years ago, the only surviving human lineage after the extinction of Neanderthals and Denisovans.→. Not Neanderthal. A third lineage of hominins that had shared the Earth with us and our Neanderthal cousins for hundreds of thousands of years, in complete historical obscurity. The Denisovans had just entered the catalogue of humanity -- and they would overturn everything we thought we knew about our evolutionary past.¹

Denisova Cave and the enigmatic phalanx

Denisova Cave is a limestone cavern in the foothills of the Altai Mountains, in the Altai region of Russia, approximately 150 kilometres south of the city of Barnaul. It takes its name from an eighteenth-century hermit, Denis, who reportedly lived there. Known to Russian archaeologists since the 1970s, the cave has revealed a remarkable stratigraphic sequence spanning several hundred thousand years, with lithic tools and faunal remains indicating ancient and continuous occupation by hominins.

It was in 2008 that a fragment of finger phalanx -- specifically the distal portion of the fifth finger -- was unearthed in layer 11 of the cave by a Russian-German team. The piece, no larger than a grain of rice, belonged to an individual whose size, sex, and physical appearance remained completely unknown. What was known, however, was that it had survived exceptionally well in the cold, stable conditions of the cave: its DNA was partially intact.

Svante Paabo, director of the department of palaeogeneticsPalaeogeneticsThe study of ancient DNA extracted from remains (bones, teeth, sediments, walls) to reconstruct the past of populations.→ at the Max Planck Institute in Leipzig, and his team specialised in extracting and sequencingSequencingReading the order of the bases (A, T, G, C) of a DNA molecule; high-throughput sequencing reads millions of fragments in parallel.→ ancient DNAAncient DNAGenetic material preserved in old remains, often degraded, sequenced with cutting-edge techniques.→ from fossils. They had already decoded several Neanderthal genomes and were tackling prehistoric human diversity. When the Denisova phalanx arrived in their laboratory, sequencing of the mitochondrial DNA -- the DNA of mitochondria, transmitted only through the mother and present in large numbers of copies per cell, facilitating extraction -- produced a stunning result: the sequence was twice as divergent from those of NeanderthalsNeanderthalsA fossil humanity of Eurasia, robust and cold-adapted, extinct around 40,000 years before present.→ or modern humans as these two are from each other. It was an entirely distinct lineage.

Genome sequencing: a third human species revealed

Definitive confirmation came in 2010 with the publication in Nature of the complete nuclear genome sequence of the individual. The nuclear genome, contained in the nucleus of every cell, is far richer in information than mitochondrial DNA, but also far harder to extract and reconstruct from a fossil. Paabo and his colleagues achieved this feat using ancient DNA purification and amplification techniques they had been developing for years.

The result was unambiguous: the individual from the phalanx belonged to a lineage that had separated from the common ancestor of Neanderthals and modern humans approximately 800,000 years ago, and from the Neanderthal lineage between 400,000 and 500,000 years ago. In other words, Denisovans and Neanderthals are more closely related to each other than either is to Homo sapiens -- but they form two distinct branches that followed separate evolutionary trajectories for hundreds of thousands of years.²

This result overturned the then-dominant scenario of PleistocenePleistoceneThe geological epoch of the great ice ages (c. 2.6 Ma–11,700 BP), spanning most of human prehistory.→ human evolution: a Europe and Asia occupied by Neanderthals and Homo erectus, then invaded by modern Homo sapiens from AfricaAfricaThe cradle of humankind: the continent where the first hominins appeared, then Homo sapiens around 300,000 years ago, before the expansion to the rest of the world.→. Now, a third human population in Asia had to be counted -- a population about which virtually nothing was known physically, but whose genome would yield extraordinary information.

A "ghost" in our genomes: DenisovanDenisovanAn extinct human population, cousin of the Neanderthals, identified in 2010 from the DNA of remains in Denisova Cave (Siberia).→ introgressionIntrogressionThe lasting transfer of DNA segments from one population or species into another through repeated interbreeding, detectable in genomes long afterwards.→

The most spectacular discovery did not concern the Denisovans themselves, but their legacy in living human populations. By comparing the Denisovan genome with genomes of modern human populations from around the world, Paabo's team detected something astonishing: Melanesians -- the populations of Papua New Guinea, the Solomon Islands, Vanuatu, Fiji, and Australia -- carry in their genome between 3 and 6% of sequences of Denisovan origin. Indigenous peoples of Australia and the Philippines, as well as certain indigenous groups of the Americas, also carry significant traces of this interbreedingInterbreedingGenetic mixing between human populations or species; between Neanderthals and Homo sapiens it left 1 to 2% of Neanderthal DNA in non-Africans.→.

The conclusion was inescapable: before or during their dispersal into Southeast Asia and Oceania, the ancestors of Melanesian populations had met and had children with Denisovans. This interbreeding, called "introgression", most likely took place between 50,000 and 60,000 years ago, somewhere in Southeast Asia or the islands of Oceania. The introgresssed Denisovan genes were not neutral residues: some of them had been positively selected, meaning they conferred an adaptive advantage on their carriers and therefore increased in frequency in populations across generations.

The most remarkable example of this phenomenon concerns the EPAS1 geneEPAS1 geneA gene regulating the response to low oxygen; a variant inherited from Denisovans helps Tibetans live at high altitude.→, which codes for a protein regulating the body's response to hypoxia -- oxygen deprivation. Modern Tibetans carry a variant of this gene that allows them to live and work efficiently at altitudes above 4,000 metres, where most other human populations suffer from altitude sickness. This variant, absent from all other human groups except Han Chinese (at low frequency), proved to be of Denisovan origin: it had been introgresssed from Denisovans into the ancestors of Tibetans approximately 40,000 years ago, conferring a crucial advantage for colonising the Tibetan plateau.³

Denny, the impossible hybrid: when species interbred

In 2018, the analysis of a new bone from Denisova Cave -- a phalanx labelled Denisova 11 but nicknamed "Denny" by researchers -- produced a revelation that astonished even the most experienced palaeogeneticists: Denny was a first-generation hybrid between a Neanderthal and a Denisovan. Her mother was Neanderthal, her father was Denisovan. A female whose genome was exactly half from each of the two human species.⁴

This discovery had several major implications. First, it confirmed that Denisovans and Neanderthals shared the same habitat at least at certain times and in certain regions, and that they were reproductively fertile together -- evidence that they were still, biologically, close enough to produce viable offspring. Second, by examining the Neanderthal genome of Denny's mother, researchers determined that she belonged to a western Neanderthal population, probably from Western Europe, genetically distant from the Siberian Neanderthals that might have been expected. This suggested that Neanderthals had greater long-distance mobility or connectivity than previously thought.

Finally, the probability of encountering a first-generation hybrid in the meagre sample of known Denisovan remains suggests that these crossings were relatively frequent at and around Denisova Cave -- far more frequent than random chance would predict if hybrids were exceptional events. Denisova Cave appears to have been a meeting place where Denisovans and Neanderthals regularly mingled.

The Denisova bracelet: unsuspected symbolic intelligence

The question of the cognitive complexity of Denisovans remained open as long as only tiny bone fragments were available. The lithic tools found in the cave layers corresponding to Denisovans (attributed to the "Denisova Complex") were relatively unremarkable and did not allow their makers to be distinguished from other Middle PalaeolithicMiddle PalaeolithicA Palaeolithic period (c. 300,000 to 40,000 years ago) associated mainly with Neanderthals and early Homo sapiens, marked by Levallois tools.→ hominins. Then an exceptional object came to change the picture.

In the deepest layers of Denisova Cave, archaeologists unearthed a fragment of bracelet carved from green chlorite schist -- a metamorphic rock -- of remarkable workmanship. The bracelet had been drilled with a perfect hole, probably to attach a pendant or to fix it to an object. The inner surface had been polished to an exceptional sheen. It had been decorated with a pattern of grooves. Analyses indicate a date of approximately 50,000 years ago and a Denisovan attribution on stratigraphic grounds.

This bracelet is extraordinary for several reasons. The chlorite schist used does not occur in the immediate vicinity of the cave: it comes from a deposit located approximately 200 kilometres away, implying either deliberate transport of the raw material or exchange with other groups. The drilling technique used -- probably a form of rotary drill -- is not attested in Homo sapiens until the Upper Palaeolithic, considerably later. Its existence among Denisovans forces us to recognise a technical and symbolic capacity comparable to that of our ancestors.

The Xiahe mandible: Denisovans at the heart of Tibet

Denisova Cave, despite its paramount importance, is not the only site where Denisovan remains have been identified. In 2019, an international team published the analysis of a nearly complete mandible discovered in 1980 in Baishiya KarstKarstA limestone landscape shaped by rock dissolution, rich in caves and passages; its sediments can preserve bone and DNA over long timespans.→ Cave, in Xiahe, in China's Gansu province, on the north-eastern edge of the Tibetan Plateau at an altitude of 3,280 metres.⁵

The mandible had a very robust morphology, with enormous teeth whose size is not compatible with modern Homo sapiens or a typical Neanderthal. In the absence of extractable DNA -- the specimen had been too exposed to preserve intact DNA -- researchers used palaeoproteomics, the analysis of ancient proteins preserved in the dental collagen. The protein profile obtained was clearly different from those of Homo sapiens and Neanderthals, and consistent with what would be expected from a Denisovan. The mandible was dated to at least 160,000 years ago -- significantly older than the remains from Denisova Cave itself.

This discovery was fundamental: it proved that Denisovans had reached the Tibetan Plateau -- one of the most inhospitable regions on the planet, where altitude reduces oxygen content to approximately 60% of sea-level values -- at least 160,000 years ago, long before Homo sapiens settled there. The Denisovan EPAS1 gene acquired by modern Tibetans takes on its full meaning here: Denisovans may themselves have evolved an altitude adaptation that was later transferred to Homo sapiens through interbreeding.

Geographic range, chronology, and multiple populations

With Denisova Cave in Siberia, the Xiahe mandible in Tibet, and genetic traces in the populations of Melanesia, Australia, the Philippines, and potentially Southeast Asia, the Denisovans prove to have occupied a colossal geographic range -- probably the largest ever achieved by any homininHomininMember of the subtribe Hominina, comprising the human lineage (Homo, Australopithecus, Paranthropus…) but excluding orangutans and gibbons. The term progressively replaces "hominid" in its narrow sense.→ population of the Pleistocene. From Siberia to the islands of the Pacific, from the Himalayas to the tropics of Southeast Asia: a distribution stretching nearly 7,000 kilometres from north to south, and as much from east to west.

Analysis of the Denisovan genomes from the cave (several individuals have now been sequenced) reveals considerable genetic diversity, suggesting either a long evolutionary history in this region or the coexistence of several distinct populations. The traces of introgression in living human populations are themselves heterogeneous: Papuans and Australians carry a Denisovan signature very different from that detected in certain groups of mainland Southeast Asia, suggesting at least two, perhaps three, distinct episodes of interbreeding between different Denisovan populations and successive waves of Homo sapiens.

The chronology of the Denisovans is equally extended: the Xiahe fossils attest their presence at 160,000 years ago, the Denisova Cave remains range from approximately 200,000 to 50,000 years ago, and the introgression events with Homo sapiens are situated between 50,000 and 30,000 years ago. This means that Denisovans survived for at least 150,000 years in their known form, co-existing successively with Neanderthals and then with the first dispersing Homo sapiens in Asia.

What did Denisovans look like?

The question of the physical appearance of Denisovans is one of the most fascinating -- and most difficult to resolve -- in contemporary human palaeontology. With so few bone remains (one finger fragment, a few teeth, the Xiahe mandible), it is impossible to reconstruct their anatomy by traditional methods. Two genomic approaches have, however, been explored.

The first, published in 2019 by an Israeli team led by Liran Carmel, consists of analysing methylation markers in the Denisovan DNA -- chemical modifications that regulate gene expression -- and comparing them to corresponding markers in Neanderthals and modern humans. These methylation markers influence which genes are expressed in which tissues, and their comparison allows us to infer which genes were active (or not) in the bones, cartilage, muscles, and other tissues of Denisovans. The researchers reconstructed a probabilistic anatomical portrait: Denisovans would have had a wide, flattened skull (like Neanderthals), a broad face, and massive dentition -- consistent with the Xiahe mandible. They would also have had prominent brow ridges, a robust build, and possibly dark skin pigmentation adapted to the low sunlight of high altitudes.

The second approach uses genetic variants specific to Denisovans to identify the genes involved in their distinctive morphology -- skin colour, skeletal form, cranial capacity -- compared to Homo sapiens. These analyses are still ongoing and the results are partial, but they suggest that Denisovans were morphologically closer to Neanderthals than to modern humans, while displaying distinctive features reflecting their independent evolutionary history in Asia.

Svante Paabo and the 2022 Nobel Prize in Medicine

Svante Paabo's work on Denisovans and Neanderthals was crowned in 2022 with the Nobel Prize in Physiology or Medicine, which the Swedish Academy awarded to him "for his discoveries concerning the genomes of extinct hominins and human evolution". Paabo is the founder of palaeogenetics, an entirely new discipline that combines molecular biology and palaeontology to extract and analyse DNA from fossils.⁶

The prize celebrated not only the discoveries about Denisovans, but an entire methodological revolution: Paabo and his team had developed techniques for decontaminating fossil samples of modern human DNA (contamination being the number one problem in palaeogenetics), amplifying infinitesimal quantities of fragmented DNA preserved in bones, and reassembling them into coherent genomic sequences. These techniques have since been adopted by laboratories worldwide and now allow the analysis of DNA from specimens hundreds of thousands of years old.

The discovery of the Denisovans remains perhaps the most spectacular revelation of this palaeogenetic revolution: the identification of an entire human species from a single small bone, without any prior anatomical clues, solely through reading its genome, is to date without equivalent in the history of the natural sciences. It illustrates in a striking way the power of molecular biology applied to archaeology and human evolution.

Perspectives: what we still do not know

Despite the spectacular progress of the last decade, our knowledge of Denisovans remains full of gaps. We do not know with certainty what relationship the Denisovans had with other hominin populations in Asia, such as Homo erectus -- present in Asia for more than a million years -- or the mysterious populations of Southeast Asia whose genetic traces in living populations cannot be explained by known Denisovans, Neanderthals, or anatomically modern Homo sapiens.

Homo luzonensis, a species recently discovered on the island of Luzon in the Philippines and dated to 50,000 to 67,000 years ago, displays very distinctive morphological features that have not yet been explained. Could it be a population of insular Denisovans that evolved independently? The question remains open. Similarly, the "Negrito" populations of the Philippines carry the highest proportion of Denisovan genes of any non-Oceanian group -- suggesting that the islands of Southeast Asia were a particularly intense meeting point between the two humanities.

Palaeoproteomics -- the analysis of ancient proteins that survive where DNA has degraded -- offers a promising future perspective: proteins can be extracted from fossils more than a million years old, where DNA rarely survives beyond 500,000 years. Fossil Asian hominin specimens whose classification was uncertain might, through this technique, reveal whether they belong to the Denisovan lineage or another.

Finally, numerous "archaicArchaicRefers to an ancient, now-extinct human population or form (Neanderthals, Denisovans, ghost lineages), as opposed to anatomically modern humans.→" hominin fossils from East Asia -- the Dali, Jinniushan, and Harbin skulls in China -- remain unclassified within the standard framework of human evolution. The Harbin skull (the "Dragon Man"), described in 2021 as potentially the closest known relative of modern humans, has been proposed by some researchers as belonging to the Denisovan lineage. If this attribution were confirmed, Denisovans would have developed, in East Asia, a brain of comparable volume to that of modern humans -- a crucial piece of information for understanding their intelligence and culture.

The Denisovans remind us forcefully that the history of humanity is far more complex, far richer in branchings and encounters, than the linear schemas we have long tried to impose upon it. We are not the product of a straight march towards modernity, but the heirs of an interwoven tapestry of humanities, some of which vanished leaving only their genes in our bodies.