Ancient enamel proteins redraw the Hominin family tree
Protein fragments from six Homo erectus teeth have forced a major revision of how ancient human groups moved, mixed, and disappeared across Eurasia. The teeth, recovered from Zhoukoudian, Hexian, and Sunjiadong in China and dated to roughly 400,000 years ago, preserved enough molecular material in enamel to allow direct biological comparison long after DNA had broken down. For a species once treated as an evolutionary endpoint, that detail is extraordinary. It turns a fossil that looked silent into one that still carries molecular history.
The key advance here is methodological as well as evolutionary. Tooth enamel is the hardest tissue in the body, and its proteins can survive when genetic material is gone beyond recovery. By extracting ancient proteins rather than DNA, investigators obtained a workable signal from fossils far older than the limit usually associated with genomic analysis. The result was not just confirmation of species identity. It also opened a route into ancestry, migration, and interbreeding in a period that has long been opaque.
All six specimens shared a previously unknown amino acid variant, a single-letter change in a protein sequence that has not been seen in any other hominin, living or extinct. In practical terms, this molecular signature grouped the east Asian Homo erectus fossils together and strengthened the case that the unusual Hexian material truly belongs inside that species. A fossil classification debate settled not by skull shape alone, but by chemistry locked inside enamel.
One sentence in the paper matters far beyond taxonomy: the same broader protein pattern also links these teeth to Denisovans.
Interbreeding, introgression, and a buried signal in living populations
Denisovans remain one of the most elusive branches of the human family, known mostly from a Siberian cave and a sparse genetic trail. Yet the protein evidence from the Chinese Homo erectus teeth suggests that these lineages were not isolated for long stretches of time. A second variant found in the fossil proteins also appears in Denisovans, and the corresponding genetic form is present today at notable frequencies in some populations, including about 21% in the Philippines and about 1% in India. The geographic pattern is consistent with ancient genetic exchange rather than a random coincidence.
The most plausible reading is straightforward: east Asian Homo erectus populations passed this variant to Denisovans through interbreeding, and Denisovans later passed it onward to the ancestors of some modern people in Southeast Asia and Oceania. In population genetics, this kind of transfer is called introgression. It means that genes can cross species boundaries and persist, even when the source population is no longer visible in the fossil record.
This is not a trivial footnote in human evolution. It reinforces a larger pattern now seen across archaic human genomes. Modern humans outside Africa carry roughly 2% Neanderthal DNA. Papuans and Aboriginal Australians carry an additional 2% to 5% Denisovan ancestry. West African populations also show signatures from an unidentified archaic lineage. The picture is no longer one of clean replacement, with one species wiping the slate clean before the next arrived. It is one of repeated contact, overlap, and exchange across deep time.
The implications reach into the biology of adaptation. Some Denisovan-derived variants in Papuan genomes appear to influence immune function, and similar inherited fragments may have helped populations respond to local environments after long periods of separation. The Homo erectus-derived variant identified in this study has unknown function, so no claim can be made about its physiological effect. Still, the precedent is important: introgressed DNA can persist because it is neutral, but it can also persist because it offered an advantage in survival, reproduction, or environmental fit.
The old evolutionary diagram of a neat tree no longer captures that process. A braided river fits better. Branches split, rejoin, and carry material downstream.
Why the method matters for fossils that never yielded DNA
The deeper significance of this work is the tool it establishes for fossils that have never been molecularly informative. Homo erectus survived in Indonesia until perhaps 100,000 years ago. Homo floresiensis occupied Flores. Homo luzonensis lived in the Philippines. None of these populations have yet produced usable DNA, which has limited direct insight into whether they contributed biologically to later humans. Proteins from enamel may change that.
If ancient proteins can survive for 400,000 years in Homo erectus teeth, then other elusive lineages may also be accessible to the same approach. That does not guarantee a genetic breakthrough for every fossil species. It does, however, shift the boundary of what can be tested. Questions that once ended at morphology now extend into molecular ancestry.
There is also a caution built into the finding. A molecular match does not automatically explain function, behavior, or clinical relevance in any modern sense. The variant detected in the teeth tells us that gene flow occurred; it does not tell us which traits it altered, whether selection favored it, or how many generations carried it forward. Those answers remain out of reach for now.
Even so, the study leaves a concrete conclusion in place: the human past was populated by species that encountered one another often enough to leave durable biological traces. A tooth from Peking Man, or from another ancient East Asian site, can now speak across hundreds of thousands of years. In this case, it speaks of interbreeding, persistence, and a legacy that survived long after the original branch was thought to have vanished.