A team at the University of Liverpool says it has detected original collagen remnants inside a 66-million-year-old Edmontosaurus sacrum, a finding that, if confirmed, would force scientists to rethink how long fragile proteins can survive inside fossilized bone. The claim rests on three independent analytical techniques applied to a single specimen and arrives amid an ongoing dispute over whether organic signals in ancient fossils represent genuine dinosaur molecules or later contamination by microbes.
What the Liverpool team found in a single bone
The fossil at the center of the debate is catalogued as UOL GEO.1, a sacrum from the duck-billed dinosaur Edmontosaurus. According to work published in Analytical Chemistry, the Liverpool group used attenuated total reflectance Fourier-transform infrared spectroscopy (ATR-FTIR), cross-polarized light microscopy, and mass spectrometry to probe the bone’s internal chemistry. Each method returned signals consistent with collagen, the structural protein that dominates fresh bone tissue. The team argues that the convergence of three separate lines of evidence makes contamination an unlikely explanation.
In ATR-FTIR, infrared light is directed into the sample and the resulting spectrum reveals characteristic absorption bands from specific molecular bonds. The Liverpool spectra showed amide peaks typically associated with protein backbones, along with patterns that resemble modern bone collagen. Under cross-polarized light, thin sections of the sacrum revealed fibrous textures aligned along the bone’s internal architecture, which the team interprets as remnants of the original collagen scaffold rather than mineral infill. Mass spectrometry then detected peptide fragments that match sequences known from vertebrate collagen, strengthening the case that the signals represent real biological material.
Prof. Steve Taylor, who led the work, said in a University of Liverpool press statement that the discovery “suggests that organic biomolecules can survive in some fossils and that we should revisit existing museum collections with these new techniques.” That statement carries practical weight: natural history museums worldwide hold thousands of dinosaur bones that have never been screened with modern spectroscopy or protein-sequencing tools. If even a fraction of those specimens retain endogenous proteins, they could provide new data on dinosaur physiology, growth, and evolutionary relationships.
A decade of contested soft-tissue claims
The Liverpool results did not emerge in a vacuum. The modern debate over dinosaur proteins began when researchers reported evidence interpreted as preserved collagen I in Tyrannosaurus rex soft-tissue structures. In that work, published in Science, investigators used mass spectrometry to identify peptide sequences from T. rex bone that aligned with bird and crocodile collagen, suggesting deep evolutionary conservation of the molecule. Those data were presented alongside microscopy images of flexible, vessel-like structures recovered from within the fossil.
A companion paper in the same journal extended the approach to mastodon bone, using tandem mass spectrometry to recover additional collagen peptides and build a more complete sequence. That study, accessible through a separate Science report, argued that collagen can persist long enough to bridge gaps in the fossil record and anchor molecular phylogenies where DNA has long since degraded. Together, the T. rex and mastodon results provided the first molecular-level data that appeared to match expectations for ancient collagen.
Those early claims drew intense scrutiny. Critics questioned whether the peptides came from the fossils themselves or from modern sources such as lab reagents, skin cells, or bacterial biofilms. Because the samples were heavily processed before analysis, skeptics argued that each chemical step opened a new pathway for contamination. Others pointed out that collagen is abundant in mammals and humans, making it difficult to exclude the possibility that modern proteins had infiltrated the samples during excavation, storage, or lab work.
Subsequent studies tried to address these concerns by minimizing sample handling and analyzing fossil tissues in place. Work on an Early Jurassic sauropodomorph dinosaur used synchrotron FTIR microspectroscopy to identify collagen-like and protein signatures directly within bone microstructures, bypassing the extraction step that critics had flagged as a major contamination risk. A study on Cretaceous mosasaur bone similarly reported proteinaceous molecules retained inside the fossil matrix, with signals spatially restricted to regions that once housed blood vessels and osteocytes. These results built a cumulative case that protein preservation across tens or even hundreds of millions of years is at least physically possible, though each individual study left room for alternative explanations.
One proposed mechanism centers on iron and oxygen chemistry. A paper in Proceedings of the Royal Society B argued that iron released from hemoglobin during decomposition can cross-link proteins and generate free-radical conditions that stabilize tissues against decay. In this model, iron acts almost like a natural fixative, locking collagen and other biomolecules into a resistant network. If iron-rich microenvironments form at the scale of individual bone canals, they could explain why protein signals appear in some spots within a fossil but not others. The Liverpool team has not yet published a geochemical map of iron distribution across the Edmontosaurus sacrum, so the connection between iron content and collagen preservation in this particular specimen remains untested.
The microbial contamination challenge
The strongest counterargument comes from a 2019 study published in eLife. That research demonstrated that Cretaceous dinosaur bone can contain recent organic material and support active microbial communities. Bacteria colonizing porous fossil bone produce their own proteins and lipids, which can mimic the spectroscopic signatures of ancient collagen. The eLife authors argued that fossils behave as open systems, absorbing groundwater and nutrients over millions of years, and that any organic signal must be rigorously distinguished from these modern biological inputs.
This creates a direct tension with the Liverpool findings. The Analytical Chemistry paper interprets the collagen signals as endogenous, meaning original to the dinosaur. The eLife study, by contrast, shows that organic molecules in Cretaceous bone can be recent arrivals. Neither team has published a side-by-side comparison of their respective FTIR spectra, which would help clarify whether the two sets of signals are chemically distinguishable. Until that comparison exists, the question of ancient versus modern origin remains open.
Another complication is that microbial biofilms can grow along pre-existing bone structures, such as vascular canals and lacunae, reproducing the same spatial patterns that researchers have cited as evidence for preserved soft tissue. Under the microscope, these biofilms can form filamentous or tubular shapes that resemble blood vessels, and their extracellular polymers can yield protein-like signals in spectroscopic analyses. Distinguishing biofilm residues from authentic collagen therefore requires high-resolution imaging, careful controls, and, ideally, peptide sequences that are inconsistent with bacterial proteins.
Separating spectral data from broader interpretation
Readers following this debate should pay attention to what each study actually measured versus what it concluded. ATR-FTIR and cross-polarized light microscopy can confirm the presence of protein-like molecular bonds in a sample. Mass spectrometry can identify specific peptide sequences. But none of these techniques directly timestamps the molecules. A collagen signal tells you that collagen is there; it does not, on its own, reveal whether the protein dates to the Late Cretaceous or to a microbial colonization event thousands of years ago.
To move from detection to interpretation, researchers rely on context. Factors such as the depth of burial within the bone, the match between peptide sequences and expected vertebrate collagen, the absence of known laboratory contaminants, and the geochemical conditions of fossilization all inform whether a signal is judged endogenous. The Liverpool team emphasizes that their collagen-like features occur deep within the mineralized matrix, away from obvious cracks or weathered surfaces, and that their mass spectrometry results align with vertebrate collagen rather than bacterial proteins. Critics will want to see those claims tested with independent replication and more exhaustive contamination controls.
For now, the Edmontosaurus sacrum adds another data point to a growing but contentious literature. If future work confirms that its collagen is genuinely ancient, it would bolster the idea that proteins can survive for tens of millions of years under favorable conditions, opening the door to new molecular studies of dinosaur biology. If, instead, the signals turn out to originate from microbes or modern contaminants, the case for long-term protein preservation will narrow, and researchers will need to refine their methods further.
Either way, the debate is pushing paleontology toward more rigorous chemistry. As teams apply increasingly sensitive instruments to old bones, they are forced to confront not only what can be detected, but also what those detections truly mean. In that sense, the controversy over dinosaur collagen is less about any single fossil and more about how science interprets faint molecular echoes from deep time.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
