Published online 30 April 2009 | Nature | doi:10.1038/news.2009.422


Age-defying dinosaur collagen

Hadrosaur, dead 80 million years, yields oldest protein yet sequenced.

HadrosaurThe ancient protein was found inside a fossilised hadrosaur bone.Wikimedia Commons / Debivort

Scientists have isolated and detailed the sequences of eight fragments of a protein from the fossilized thigh bone of a duck-billed dinosaur. The protein — the bone connective tissue collagen — was isolated from an 80-million-year-old hadrosaur fossil, making it the oldest ever to be sequenced1.

A previous report by Mary Schweitzer of North Carolina State University in Raleigh and colleagues had described finding collagen proteins from a 68-million-year-old Tyrannosaurus rex2. But the results proved controversial, with some questioning whether the samples they had obtained had become contaminated with proteins from modern species.

The new study by the group adds heft to the earlier finding.

"Three years ago, I entered this field as a complete sceptic", says Phillip Lars Manning, a bioengineer who heads palaeontology at the University of Manchester and wasn't involved in the study. "Now I do think the Schweitzer group is onto something."

Better bones

"We didn't realize the level of criticism we would get from the earlier article," says John Asara, whose mass spectrometry lab at Harvard's Beth Israel Deaconess Medical Center in Boston sequenced the dinosaur proteins. "Hopefully, we are moving the field forward by our reporting."

In the earlier T. rex study, Asara reported spectra for seven collagen proteins. Spectra for five of these were criticized3,4 — though the spectra of at least two have withstood the challenges.

Dinosaur bone cellBone cell from the dinosaur thigh bone.M. Schweitzer

This time, Asara had mass spectrometry conducted by two separate labs to confirm the new results. And he is making public immediately the nearly 27,800 spectra from which the collagen proteins were identified. A similar cache of 48,000 spectra wasn't made available for the T. rex specimen until more than a year after the initial publication, prompting criticism from some researchers, including Pavel Pevzner, a proteomics specialist at the University of California at San Diego5.

Schweitzer said the team also set out to locate fossils they believed would be better preserved than the T. rex fossil they found earlier.

In 2006, Schweitzer and her colleagues found a skeleton from the hadrosaur Brachylophosaurus canadensis. The next year, the team returned to the field, removed about seven metres of earth, and extracted a femur still encased in sandstone. That was important, Manning says, because "structural proteins like collagen are locked in the mineral lattice in the bone". This serves like "a strait-jacket for the collagen", he says.

Cleaner and faster

Even researchers who sharply criticized the earlier T. rex study welcome the new report.

"They now are analyzing proteins and reporting the mass spectrometry data the way it should be", says Pevzner. But he remains concerned about the potential for contamination — because of a report earlier this year that the T. rex study included protein from ostrich haemoglobin (see 'Origin of 'T. rex' protein questioned').

"This new report is clearly an advance", says Martin McIntosh, a proteomics researcher whose lab at the University of Washington in Seattle uncovered the haemoglobin data. "The door is closing on alternative explanations, but it is not completely closed yet."

Schweitzer believes the methods her team used to extract and analyse the hadrosaur under sterile conditions may help to usher in a new era of paleontological field work — one where specimen scrutiny uses sterile techniques like a crime scene investigation.


Rapid analysis is also important to limit potential problems with specimen degradation or contamination. It took just three-and-a-half weeks from the time the fossil was removed until microscopy photos were made of vessel-like structures in the bone. Last summer, she said, the team used a new portable laboratory to cut that time to eight hours. "I think this will make a huge difference", she says.

This may be important in the next phase of Cretaceous fossil analysis for other proteins. Collagen is the easiest to find because it is the most common connective tissue in bone. But antibody tests for other proteins, such as laminin and elastin, were also positive.

Asara is cautious. The presence of those proteins has not been confirmed by mass spectrometry sequencing, he says, so those results should still be considered tentative.

"It doesn't mean they aren't there", says Asara, who hopes to describe other protein sequences in a future report. "The mass spectrometry is not sensitive enough to sequence them." 

  • References

    1. Schweitzer, M.H. et al. Science 324, 626-631 (1 May 2009).
    2. Schweitzer, M.H. et al. Science 316, 277-280 (2007). | Article | PubMed | ChemPort |
    3. Kaye, T.G., Gaugler, G. & Sawlowicz, Z. PLoS ONE 3, e2108 (2008). | Article | PubMed | ChemPort |
    4. Buckley, M. et al. Science 319, 33 (2008). | Article | PubMed | ChemPort |
    5. Pevzner, P.A., Kim, S. & Ng, J. Science 321, 1040 (2008). | Article | PubMed | ChemPort |
Commenting is now closed.