Groundbreaking Israeli Research Reveals What Mysterious Denisovans Looked Like

DNA methylation data allows Israeli scientists to give illustrator a detailed list of the archaic human's characteristics

This image shows a preliminary portrait of a juvenile female Denisovan, based on a skeletal profile reconstructed from ancient DNA methylation maps.
Maayan Harel

The ability of scientists to discover new information from ancient remnants of DNA has revolutionized our understanding of the evolution of animals in general and humans in particular. Now, thanks to breakthrough research, the anatomical features of an extinct group of archaic humans known as the Denisovans has been reconstructed based only on their DNA.

The research, led by Prof. Liran Carmel and David Gokhman of the Hebrew University, was published Thursday as the cover story in the scientific journal Cell.

The Denisovans lived until a few tens of thousands of years ago in eastern and northeastern Asia. And until now, all that was known about them was based on a few small fragments of bone.

They had different morphological characteristics than the Homo sapiens and Neanderthals who lived at the same time. Homo sapiens lived mostly in Africa and the Neanderthals in Europe and northern Asia — where they met the Denisovans and got to know them very well, even bearing children together.

It seems the Denisovan branch of humankind would never have been discovered without the development of new methods for extracting, sequencing and analyzing ancient DNA.

In 2008, scientists found a single finger bone in the Denisova Cave in Siberia’s Altai Mountains (the species was subsequently named after the cave). It seems to have come from a young woman. Two years later, the scientists — led by Swedish geneticist Svante Pääbo — published the full DNA sequence extracted from the bone and showed that it came from a separate human species significantly different than Homo sapiens and Neanderthals. Denisovan teeth were found later, and earlier this year a lower jawbone, also identified as coming from a Denisovan, was found on the Tibetan Plateau.

A model of a juvenile female Denisovan, based on a skeletal profile reconstructed from ancient DNA methylation maps.
Maayan Harel

Since 2010, studies of modern DNA have found vestiges of Denisovan DNA in the genome of modern humans in a number of places worldwide: In indigenous Australian aboriginals; residents of the Melanesian Islands in the Pacific Ocean; and also in Tibetans — who, recent research shows, inherited their ability to live in the thin mountain air from their ancient Denisovan relatives.

Scientists hypothesize that Denisovan DNA is the result of meetings between the different groups that occurred some 50,000 to 60,000 years ago, at a time when Homo sapiens left Africa and reached Asia.

Scientists now theorize that the branch of humans that led to Neanderthals and Denisovans split off from the branch that led to the development of Homo sapiens about 600,000 years ago. The first branch then split again about 400,000 years ago into separate branches: one offshoot led to the Neanderthals; the other led to the Denisovans.

The genetic research has produced a much more complex picture of human evolution, which includes a great number of common descendants and genetic transfer between the various groups of humans.

But excepting this genetic data and the very few fragments of bone and teeth, researchers were unable to say much about who the Denisovans were, how they lived and what they looked like. Now, though, the scientific race is on to find and identify more vestiges of our ancient relatives and resolve some of the mysteries surrounding them.

A different approach

Gokhman, Carmel and their colleagues, including Prof. Eran Meshorer from Hebrew University, Prof. Yoel Rak from Tel Aviv University and Prof. Tomas Marques-Bonet from the Institute of Evolutionary Biology in Barcelona, have adopted a different approach to solving those mysteries. Previous research managed to produce a complete Denisovan DNA sequence — of a standard that is just as good as any sequence produced by modern labs today. Every letter in the ancient DNA sequence was sampled and sequenced about 30 times on average, which guarantees a high quality result.

Carmel says that scientists nowadays do not have the ability to reconstruct the precise anatomy or appearance of individuals based only on their DNA. To circumvent this problem, the scientists used a method of reconstructing epigenetic patterns in the ancient DNA, which was developed in previous research by Meshorer and Carmel. Epigenetics is the science of changes to organisms and heredity that are found outside the genetic code of the DNA sequence itself, usually caused by changes to gene expression — in other words, which genes are turned “on” or “off” in each and every cell. Epigenetics helps explain why the human body has such a great number of different types of cells constructed from the same “instruction book,” aka the DNA sequence. As Carmel puts it, it’s what produces the enormous anatomic differences between a tadpole and a frog.

One of the types of epigenetic changes being studied today, and which the research team focused on, is DNA methylation. This is the process in which methyl groups — a carbon atom with three hydrogen atoms attached (CH3) — are added to the DNA molecule. The methyl groups usually attach to cytosine molecules, one of the four bases that are the foundations for building all DNA, and this generally serves to repress the gene’s transcription.

Over the past 20 years, the methylation process was revealed to be one of the main mechanisms in determining how genes are expressed (turned on or off). Examining DNA methylation allows scientists to reconstruct which genes are active and which are silenced at the level of a single cell — or the entire organism.

This image shows a portrait of a juvenile female Denisovan, based on a skeletal profile reconstructed from ancient DNA methylation maps.
Maayan Harel

In their present research, the Israeli scientists created DNA methylation maps of the Denisovans, Neanderthals, ancient ancestors of modern humans, human beings living today and chimpanzees. In comparing these maps, the researchers focused on the areas where the DNA methylation differs significantly between species and subspecies of primates.

In the next stage, the researchers consulted a very large data bank of monogenetic disorders — caused by a mutation in a single gene. This data bank also includes detailed information about the connection between the gene and the morphological phenotype (the way in which the specific gene variation is actually reflected in the anatomy of the individual who carries this variation).

With the help of these data, the researchers were able to create a list of genes in which a higher or lower incidence leads to various anatomical changes — for example, a specific gene whose increased activity leads to the development of a broader skull.

They then had the basic anatomical structure — that of modern man — to compare to a list of genes whose change causes an anatomical difference from that same basic model, and methylation models that demonstrate how these genes change between the species and various groups of archaic humans that were studied, said Carmel.

When the scientists applied the new method they had developed to the methylation map of the Neanderthals and chimpanzees, they attained an 85 percent level of accuracy in the comparison between their anatomical model and the known scientific findings.

Carmel said further confirmation of the validity of the model they constructed can be found by comparing it to an ancient skull fragment that was discovered about three years ago in China. Until now, they had been unable to identify which human species it belonged to.

An article analyzing the skull fragment described eight prominent characteristics, which the scientists derived based on the fragment that had been found in China, but it did not mention Denisovans. “We compared the findings to our model and found it had seven of those eight characteristics,” explained Carmel.

Carmel said further support for the model was discovered after the article was submitted for peer review in Cell: The mandible that was identified earlier this year as coming from a Denisovan.

In the article that presented the jaw fragment, four prominent characteristics were described, said Carmel. As a result, eight conclusions could be drawn from them — and for each of the characteristics, we can examine whether the jaw of Denisovan man was more similar to that of Homo sapiens or Neanderthals, he said. Once again, seven out of the eight characteristics that were observed based on the analysis of the jaw also appeared in their model.

So, with the list of differences between Denisovan man and Homo sapiens — which was verified as far as possible based on the existing findings — the researchers turned to scientific illustrator Maayan Harel.

Based on their list, she made a model of Denisovan man — in cooperation with Rak, a physical anthropologist who has been studying archaic humans for decades. Based on this, Harel then created for the first time a scientifically based illustration of the faces of those mysterious ancient humans who went extinct tens of thousands of years ago.