Denisovan DNA clue found in Homo erectus tooth proteins

If you’re wondering whether Denisovans passed some of their genes to us, the short answer is yes—and new evidence from tooth proteins adds an independent line of support. Scientists have identified a distinctive variant of a tooth-enamel protein in Homo erectus that also appears in Denisovans and in some living people, reinforcing the case that modern humans inherited DNA from Denisovans during ancient interbreeding events.

What’s novel here isn’t the basic idea of Denisovan ancestry—that’s already well established from ancient DNA. The new twist is the use of palaeoproteomics (the study of ancient proteins) to trace an evolutionary signal much older than most recoverable DNA, pushing our view deeper in time and showing that a specific molecular feature persisted from Homo erectus into Denisovans and then into parts of our species.

Key takeaways

A distinctive amino-acid variant in a tooth-enamel protein found in Homo erectus also occurs in Denisovans and in some modern human populations.
Proteins in tooth enamel can survive far longer than DNA, giving us molecular clues from fossils too old or too degraded for genetic sequencing.
The shared protein signature strengthens the case that Denisovans contributed DNA to the ancestors of certain modern humans, especially in parts of Asia and Oceania.
The finding does not mean modern humans interbred directly with Homo erectus; instead, it suggests that Denisovans (an archaic group in Asia) carried and passed on an ancient variant that also existed in Homo erectus.
As with any single marker, caution is warranted: shared features can arise through common ancestry or, less likely, convergent evolution. But the total body of evidence favors Denisovan introgression into our lineage.

Who this explainer is for

Students and readers curious about human evolution and how scientists reconstruct deep ancestry
Science enthusiasts who’ve heard of Neanderthals and Denisovans but want to understand what’s new
Genealogy hobbyists interpreting results that mention Denisovan ancestry
Educators seeking a plain-language overview of palaeoproteomics and archaic introgression

What changed, in plain terms

Before: We knew from ancient DNA that many modern humans (especially in Melanesia, parts of Southeast Asia, and to a lesser extent East and South Asia) carry Denisovan ancestry. However, ancient DNA typically reaches back hundreds of thousands of years at most in favorable conditions.
Now: Researchers have recovered an enamel protein variant from Homo erectus teeth that matches a form found in Denisovans and shows up in some living people. Because proteins can persist well beyond DNA’s survival window, this offers a deeper historical anchor point for a molecular feature that ultimately entered portions of our species.

Quick background: Who were Homo erectus and Denisovans?

Homo erectus: A widespread early human species first appearing around 2 million years ago, known for long-distance dispersal across Africa and Eurasia, larger brains than earlier hominins, and distinctive stone tools. They persisted in some regions until at least several hundred thousand years ago.
Denisovans: A sister group to Neanderthals identified from DNA in fossils from Siberia and the Tibetan Plateau. They lived across parts of Asia and interbred with both Neanderthals and modern humans. Some modern populations carry notable Denisovan ancestry.

How tooth proteins can reveal ancestry when DNA can’t

Tooth enamel is the hardest tissue in the body. During tooth formation, a handful of specialized proteins (including enamelin and amelogenin, among others) guide enamel growth and structure. Most of these proteins are degraded and removed as the enamel matures, but traces become trapped and mineralized within the enamel’s crystalline lattice.

Why this matters:

Proteins are made of amino acids encoded by genes. Small differences in the DNA can alter specific amino acids in the resulting protein. When those differences are preserved in fossil enamel, they become time-stamped molecular clues.
Unlike DNA, which breaks down relatively quickly outside of permafrost, enamel proteins can survive for hundreds of thousands to millions of years in the right conditions.
Using mass spectrometry, scientists can read partial protein sequences—short peptides—and detect amino-acid variants that map back to genetic changes, even when DNA is long gone.

In short: proteins can function like durable, fragmentary pages of a book that DNA used to write.

What the researchers actually found

A specific amino-acid variant in a tooth-enamel protein was recovered from Homo erectus teeth.
The same variant is present in Denisovan material and shows up in some living people.
It is absent or rare in other archaic hominins or in other modern populations, creating a distinctive pattern.

This pattern is most consistent with the idea that Denisovans carried the ancient variant and passed it into the gene pool of certain modern human groups through interbreeding. The presence of the variant in Homo erectus indicates the feature existed deep in time in Eurasia (or potentially in a shared ancestor), and Denisovans retained it.

Does this prove Homo erectus interbred with us?

No. The more parsimonious reading is that Denisovans—who already are known to have interbred with modern humans—were a conduit for this ancient enamel-protein variant. The variant’s appearance in Homo erectus suggests it is quite old in Eurasian hominins, but that doesn’t imply direct mating between Homo erectus and Homo sapiens. Instead, Denisovans likely inherited it (whether through deep ancestry or earlier hominin gene flow) and later passed it to some of our ancestors.

How introgression works (and where Denisovans fit)

Introgression is the transfer of genetic material from one group to another through interbreeding. In the last 15 years, ancient DNA has revealed that:

All non-African modern humans carry Neanderthal ancestry (typically around 1–2%).
Some populations, particularly in Island Southeast Asia, New Guinea, and Australia, carry additional Denisovan ancestry that can total several percent.
Denisovan-derived DNA has been linked to adaptations, such as high-altitude physiology in Tibetans (the EPAS1 region), although most introgressed fragments are neutral or nearly so.

The new tooth-protein signal doesn’t overturn this picture; it adds a complementary source of evidence, showing that a molecular feature present in a deep Eurasian hominin was still circulating in Denisovans and survives in some living people.

How do scientists authenticate ancient proteins?

Working with million-year-old molecules is tricky. Researchers use a suite of checks to ensure signals are ancient and not contamination:

Biochemical wear patterns: Ancient proteins accumulate characteristic chemical modifications over time (for example, deamidation) that are hard to fake.
Stratigraphic controls: The protein is recovered from securely dated layers and anatomically identified fossils.
Consistency across peptides: Multiple fragments from the same protein tell the same sequence story and fit known enamel protein structures.
Phylogenetic plausibility: The inferred amino-acid variant fits with independent evolutionary information rather than contradicting it.

Even with these checks, scientists state uncertainties and replicate results where possible.

Why an enamel-protein variant matters

Extends timelines: Protein sequences let us peek beyond DNA’s preservation limit, anchoring genetic features deeper in prehistory.
Refines relationships: Shared molecular markers help clarify which lineages carried which variants, indicating whether a feature is ancestral, derived, or introduced via gene flow.
Complements genomes: Where ancient DNA exists, protein data can validate and refine interpretations. Where DNA is absent, proteins keep the story moving.

Could the pattern be explained without Denisovan gene flow?

Alternative explanations always deserve consideration:

Retained ancestral variation: The variant could, in theory, have been widespread in a common ancestor and then lost in most lineages except Denisovans and specific modern populations.
Convergent evolution: Two lineages could independently evolve the same amino-acid change, though exact recurrences at the same protein site are statistically less likely.

Context matters. Because ancient DNA already documents Denisovan introgression into modern humans, the simplest explanation is that Denisovans transmitted this variant to our ancestors. The enamel-protein evidence aligns with, rather than contradicts, that established picture.

What this does not tell us (limits and caveats)

It does not pinpoint when or where interbreeding occurred. Protein data alone can’t date introgression events.
It does not measure how much Denisovan ancestry any living group carries. Whole-genome DNA studies are needed for that.
It does not prove functional impact. A change in an enamel protein might have no visible effect on teeth or health, unless shown otherwise.

Where in today’s world is Denisovan ancestry found?

Patterns vary, but broadly:

Highest proportions: Melanesia (for example, Papua New Guinea) and some Aboriginal Australian groups.
Noticeable traces: Parts of Island Southeast Asia, the Philippines, and mainland East/South Asia.
Minimal or none detectable: Much of Africa (though back-migration complicates the picture) and Western Eurasia outside of limited Asian introgression.

The enamel-protein variant’s frequency in modern populations will be uneven, mirroring the patchy distribution of Denisovan ancestry.

How the method works: a 60‑second tour of palaeoproteomics

Sampling: A tiny amount of tooth enamel is powdered under clean conditions.
Extraction: Ancient proteins locked in the enamel matrix are released using controlled chemical treatments.
Mass spectrometry: The peptides are ionized and measured by their mass-to-charge ratios to reconstruct partial sequences.
Mapping: Detected amino-acid variants are compared against known human, archaic, and outgroup sequences.
Interpretation: Researchers assess evolutionary scenarios that could produce the observed pattern, cross-checking with fossil context and other molecular data.

Because tooth enamel is highly mineralized and shields its embedded proteins, it’s one of the best substrates for very old molecular information.

Why this is exciting for human origins research

Bridges data gaps: Many classic fossils predate reliable DNA recovery. Proteins can pull molecular signals from these specimens, connecting morphology to molecules.
Tests hypotheses: Debates about how different archaic groups relate to each other can be evaluated using independent biochemical markers.
Puts traits on a timeline: If an amino-acid variant shows up in a fossil of known age, it sets a minimum date for the underlying genetic change.

Practical implications beyond prehistory

Forensics and conservation biology: Ancient protein methods are spilling over into other fields, improving species identification from fragmentary remains.
Dental research: While functional effects of the specific variant are unknown, cataloging enamel protein diversity can inform studies on tooth development and resilience.
Museum curation: Protein sampling is minimally destructive compared to older techniques, helping balance science and preservation.

Frequently asked questions

Does this mean we have Homo erectus DNA?
- Not directly. The signal points to Denisovans as the immediate source of introgressed DNA in modern humans. Homo erectus carried the variant much earlier, showing its deep roots, but direct Homo erectus–modern human interbreeding is not supported by this evidence.
How long can proteins last?
- Under favorable conditions, enamel proteins can persist for hundreds of thousands to several million years. Preservation depends on temperature, burial environment, and geochemistry.
Is the enamel-protein variant beneficial?
- Unknown. Many amino-acid changes are neutral. Demonstrating a functional effect would require modern genetic, biochemical, and dental studies.
How does this relate to the well-known Tibetan high-altitude gene?
- That adaptation (in EPAS1) is traced to Denisovan DNA via ancient genomes. The enamel-protein finding is a separate marker that reinforces Denisovan introgression from an independent molecular system (proteins rather than DNA).
Could contamination explain the results?
- Researchers apply multiple authenticity checks and compare results across labs and specimens. While no method is perfect, the field has robust protocols to minimize and detect contamination.

The bottom line

A distinctive tooth-enamel protein variant found in Homo erectus, Denisovans, and some living people provides a durable molecular breadcrumb that aligns with what ancient DNA already taught us: Denisovans contributed genes to our ancestors. The value added here is temporal depth and methodological breadth—proteins open a window into eras where DNA often fails, letting us trace specific features further back in time and tightening the links among fossils, archaic humans, and ourselves.

Source & original reading: https://arstechnica.com/science/2026/05/protein-in-homo-erectus-teeth-suggests-denisovans-gave-us-some-of-their-dna/

Ancient tooth proteins link Denisovans to our DNA: what that really means