Tyrannosaur morsels

Blogging on Peer-Reviewed Research

This story is in the news again, so I’ve reposted my description of the paper from 3½ years ago. This is an account of the discovery of soft organic tissue within a fossilized dinosaur bone; the thought at the time was that this could actually be preserved scraps of Tyrannosaurus flesh. There is now a good alternative explanation: this is an example of bacterial contamination producing a biofilm that has the appearance of animal connective tissue.

Read GrrlScientist’s explanation and Greg Laden’s commentary and Tara Smith’s summary of the recent PLoS paper that tests the idea that it is a biofilm.

Look! A scrap of soft tissue extracted from dinosaur bone:

Demineralized fragments of endosteally derived tissues lining the marrow cavity of the T. rex femur. The demineralized fragment is flexible and resilient and, when stretched (arrow), returns to its original shape.

It has been reported in Science this week that well-preserved soft tissues have been found deep within the bones of a T. rex, and also within some hadrosaur fossils. This is amazing stuff; fine structure has been known to be preserved to this level of detail before, but these specimens also show signs of retaining at least some of their organic composition. What the authors have done is to carefully dissolve away the mineral matrix of the bone, exposing delicate and still flexible scraps of tissue inside.

Here, for example, is a piece of endothelial tissue, or the tubelike epithelia that line blood vessels and form capillaries. It is compared to a similarly prepared piece from fresh ostrich bone; you can tell the T. rex fragment has undergone some changes, but it’s comparable in size and organization to the bird sample.

(I) T. rex vessel fragment showing detail of branching pattern and structures morphologically consistent with endothelial cell nuclei (arrows) in vessel wall. (J) Ostrich blood vessel liberated from demineralized bone after treatment with collagenase shows branching pattern and clearly visible endothelial nuclei.

Looking more closely with a scanning electron microscope, here’s a similar piece of T. rex blood vessel that has ruptured, spilling out its contents. Maybe those cells don’t look perfectly preserved, but they’re darned close.

Exploded T. rex vessel showing small round microstructures partially embedded in internal vessel walls.

And lastly, here’s a closeup of the surface of that epithelia, compared with an ostrich epithelium. The cells here are very, very flat, and the nuclei are the thickest part, bulging up and giving the surface a pebbled appearance. The T. rex epithelium has a similar pebbly look, suggesting that just maybe there is even some subcellular structure preserved.

(E) Higher magnification of a portion of T. rex vessel wall, showing hypothesized endothelial nuclei (EN). (F) Similar structures visible on fixed ostrich vessel. Striations are seen in both (E) and (F) that may represent endothelial cell junctions or alternatively may be artifacts of the fixation/dehydration process.

How could this be? Here’s the authors’ explanation.

…we demonstrate the retention of pliable soft-tissue blood vessels with contents that are capable of being liberated from the bone matrix, while still retaining their flexibility, resilience, original hollow nature, and three-dimensionality. Additionally, we can isolate three-dimensional osteocytes with internal cellular contents and intact, supple filipodia that float freely in solution. This T. rex also contains flexible and fibrillar bone matrices that retain elasticity. The unusual preservation of the originally organic matrix may be due in part to the dense mineralization of dinosaur bone, because a certain portion of the organic matrix within extant bone is intracrystalline and therefore extremely resistant to degradation. These factors, combined with as yet undetermined geochemical and environmental factors, presumably also contribute to the preservation of soft-tissue vessels. Because they have not been embedded or subjected to other chemical treatments, the cells and vessels are capable of being analyzed further for the persistence of molecular or other chemical information.

So, basically, these cells were entombed in a thick mineral sarcophagus, protected from bacteria and other external insults. There have to have been other factors at play—cells are full of enzymes that trigger a very thorough self-destruct sequence at death—so I’m definitely looking forward to the molecular analysis. Even if their form was preserved, I expect these cells to be denatured monomer soup on the inside.

Schweitzer MH, Wittmeyer JL, Horner JR, Toporski JK (2005) Soft-Tissue Vessels and Cellular Preservation in Tyrannosaurus rex. Science 307(5717):1952-1955.


  1. jparenti says

    I’m glad that they’re still working on this. I was shocked myself when this first came to light, and also a little anxious. Could there be DNA? (Of course not, but a nice sequence would have helped evolutionary theory along so much, I kept irrational hope alive.)

  2. Mark Borok says

    Actually, I expect some creationists will claim this as “proof” that the fossil is much younger than claimed.

  3. Tom says

    Damn, you beat me to it, GunOfSod. My comment was going to be: “Does the specimen have a footprint from the contemporaneous human who saddled up and rode this T. Rex?”

  4. says

    There’s nothing really new about preserved proteins in dinosaur bones. In fact, there was hope of finding DNA in some such bones, though it seems that DNA is unlikely to last that long. Here’s one source for the knowledge of preserved proteins, prior to the more spectacular find of preserved and flexible soft dinosaur tissues:

    Two different immunological assays were used to identify the remains of a bone matrix protein, osteocalcin (OC), in the bones of dinosaurs and other fossil vertebrates. Antibodies raised against OC from modern vertebrates showed strong immunological cross-reactivity with modern and relatively young fossil samples and significant reactions with some of the dinosaur bone extracts. The presence of OC was confirmed by the detection of a peptide-bound, uniquely vertebrate amino acid, γcarboxyglutamic acid (Gla). Preservation of OC in fossil bones appears to be strongly dependent on the burial history and not simply on age. These results extend the range of protein preservation in the geologic record and provide a first step toward a molecular phylogeny of the dinosaurs.


    In fact, the creationists were whining about the supposed (and never substantiated) impossibility of proteins in dinosaur bones before the tissues in question. But there seems to be no reason to think that proteins can’t last for millions of years–it’s just not very likely in any one instance.

    Glen D

  5. says

    I love watching the process, as scientists check and counterbalance eachother’s thoughts, weighing evidence and proposing alternate possibilities in their search for the truth.

    Compare and contrast with another “process” in which the priests who held ownership of the sacred T-rex bone with its scrape of tissue, condemned as “intolerant” the apostasy of the other priests who offered the biofilm theory.

    Whenever some creotard tries to make it sound like science and religion are even slightly similar, I replace the frustration I feel at wondering “is it T.Rex or isn’t it?” with respect for the process and wariness for its alternative.

  6. Deepsix says

    “#7 Actually, I expect some creationists will claim this as “proof” that the fossil is much younger than claimed.”

    They already have. They claim that if there is soft tissue, then dinos can’t be nearly as old as evilutionists claim. They think this type of find is “proof” for their claims.

  7. JB says

    But, but, but…

    Schweitzer et al extracted protein from the structures. The protein was analysed by mass spectrometry and was found to be most closely similar to that of chicken (actually ostrich IIRC, but based on mass-spectra rather than sequence, whcih wasn’t available for ostrich. Don’t have the paper to hand, this is all from memory). The next closest matches were for some amphibian (frog?). Not E. coli, streptowotsit or achaeobacterithingy.

    Chicken. Chicken collagen at that. Don’t be put off by the statement that collagen-like proteins have been found in bacteria – why is the alledged contaminant from Rhodococcus trying to look exactly like what we expect from vascular structures? (But check the Rhodococcus genome against the results from Schweitzer someone please…)

    Not completely conclusive, given the relative poor abundance of the analyte. However, they also did the same analysis on the same structures from fossilised mastadon. If these were artefacts, one might expect the same chicken-like artefact. It was most closely similar to modern elephant.

    The same experiment needs to be done on these biofilm samples. If they show the same signature then we may be able to scream artefact. If they come up with something bacterial, then it’s different. As I recall (again) the Schweitzer bone was exceptional from the size point of view and it presumably took some commitment to cut it open and dig around inside. Maybe it’s just different? Maybe the biofilm is contaminated with Tyrannosaurus protein? But from my viewpoint as a protein chemist with a small knolwedge of mass spectrometry it looks as though we still have T. rex protein sequence.

    Shame about the blood though…

  8. says

    This is a wonderful opportunity for science and religion to complement each other. We know that fossils are artifacts of the Devil to confuse the message of the One True God, thus we shouldn’t expect to find soft tissue, DNA, or other sins of the flesh among the bones.

    The Devil however, being a fallen creation of God, must be made of the kind of stuff the rest of us are (sans soul, of course), though it will likely be imbued with some sort of Satanic Substance™. (I shall describe likely sources of Satanic Substance in an upcoming treatment, but for now the most likely source is the innocent stem cells from fœti corrupted by exposure to the skin of evil that surrounds abortionists).

    Thus, it is clear to any thinking man of God (women of God may feel free to have their husbands, brothers, or fathers explain it to them once the shopping is done) that this mysterious organic substance is in fact Devil DNA (not to be confused with Devilled DNA, a dish eaten by atheists at their anthropophagic orgies of godless debauchery).

    As good Christian science requires hypotheses to make predictions which can then be tested (within reason; I shouldn’t have to remind you of God’s feelings towards being tested), this organic tissue, being in and of the Devil, should be highly resistant to heat. If it is not, it is only because our Loving Lord has rendered the Satanic Substance inert so that it may safely be examined by His children. Further, when exposed to Holy Water, this organic tissue will likely become soggy and hydrated; this is because the Glory of God inverts the properties of unholy tissue–dry tissue shall become wet when exposed to Holy Water, whole tissue would lose its integrity and form punctures and tears when pierced with a crucifix.

    So carry on scientists, for you do the work of God though you do not yet know it, and through the fruits of your examination of this relic of Hell shall yourselves others be brought to life in Christ.

  9. uncle frogy says

    I too enjoy the adventure of science. How the first impression is never just left alone and deeper exploration of the data is carried out leading to even more deeper understanding.
    There is more going on than meets the eye. No priest’s fable can get the details of reality nor should it.The conceit of religion is truly amazing.
    I am humbled and aw struck by “nature”. To realize that I am a small part of it and an expression of all of the interacting forces and processes leaves me speachless.

  10. Ast says

    Never mind the tyrannosaur – can we clone a couple of hadrosaurs and finally build that Large Hadrosaur Collider we’ve always needed?

  11. JB says

    Just BLAST searched the entire NCBI microbial database looking for the collagen-like proteins. There certainly are several GXXGXXGXXetc collagen-like sequences. However, a BLAST search of the same database with one of the the presumed T. rex sequences (GLPGESGAVGPAGPIGSR) finds nothing. You will of course now wonder what happens if you BLAST the chicken database with the same sequence.

    Clue: the answer has a ‘100%’ in it.

  12. S. Fisher says

    Damn scientists…always trying to get to the truth of things by worshiping at the Tabernacle of Empirical Knowledge. Scripture is so much easier and has the additional bonus of multiple interpretations.

  13. JB says

    #19 continued – pressed post too quickly…

    #19 is somewhat related to #14 – in case you were thinking WTF!

    Bottom line: probably loads of biofilms all over the place. But Schweitzer’s T. rex bones still had collagen in them that looks a lot more like chicken (or ostrich) than anything bacterial.

  14. Sigmund says

    JB, you aren’t BLASTing properly. The sequence is just a little too short for it to work in that program. Add a couple of amino acids to it and try again.

  15. Timothy Wood says

    @#14 wtf? Satanic Substance is a trade mark? Probably the vatican… or disney.

  16. says

    @#14 wtf? Satanic Substance is a trade mark? Probably the vatican… or disney.

    Indeed. One must fight hellfire with holyfire, and Hell has a lot of lawyers.

  17. Sven DiMilo says

    Finding preserved proteins, especially extracellular proteins like collagen, is not the same thing as finding preserved tissue (which would mean cell structure). The biofilm explanation is an alternative to preserved cells and tissues, and does not conflict with finding actual dinosaur protein molecules. I think.

  18. JB says

    #22 – I am not BLASTing as it is intended to be done, but I am (t)BLAST(n)ing as optimally as is possible. There aren’t any more AAs. Should have asked the T. rex to have died somewhere colder =)

    Works for chicken though – making the point.

  19. t-1000 says

    While we’re on the topic of dinosaurs, I heard that we can’t clone dinosaurs becuase the DNA from the sample(amber or otherwise) is too damaged to glean any information from. So I was wondering if one could use the DNA repair system of Deinococcus radiodurans(a radiation resistant bacteria) to repair this damaged DNA and clone a dinosaur. Much in the same way Craig Venter is using it to assemble synthetic chromosomes.

  20. Confuseddave says

    Because they have not been embedded or subjected to other chemical treatments, the cells and vessels are capable of being analyzed further for the persistence of molecular or other chemical information.

    How in hell did they get electron micrographs of them then?

  21. Phillip Allen says

    @ 18

    I’d pay money to see hadrosaurs colliding! Monster trucks would have nothing on it.

  22. says

    @#29: I think that the DNA damage is a bit beyond that, unfortunately. The results of blindly attempting to patch it up would hopefully be something like The Fly 2, but in all likelihood boringly ineffectual.

  23. Sili says

    I’m sure every cloud has a silver lining for the Fundies.

    I can hear them now:

    “Biofilms! Biofilms!! See!?! The Shroud of Turin is too the Messiah’s!!”

  24. says

    The Cambrian Explosion – Darwin’s Worst Nightmare

    It’s so painful to see Cambrian animals fit into the same nested hierarchies as all of the rest of life. Uh huh.

    You really don’t have any brains at all, do you?

    By the way, bornretarded77, there is no evident break in the genetic record at all that corresponds to the Cambrian “explosion.” It’s just same processes of evolution going back to single-celled eukaryotes. The evidently different evolution occurred before there were the eukaryotes we know, at a time when gene-swapping occurred rather more promiscuously than it does now.

    Glen D

  25. JD says

    bornretarded77 wrote,

    The Cambrian Explosion – Darwin’s Worst Nightmare

    As is always the case, truth and evidence are the creo-tards worst nightmare.

    Whether you realize it or not, the so-called explosion took place over tens of millions of years. Moreover, the information presented in the video is way out-of-date (not surprising given the source and the tendency of creo-tards to rely on ancient texts).

    Please take a moment and find someone (outside the trailer park of course) who can read and explain the following to you:

    How real was the explosion?

    The events that occurred during the Cambrian do not threaten the theory of evolution in any way.

  26. articulett says

    Don’t the real miracles beat the religious miracles (magic crackers, etc.) any day?!

    I mean I could understand being impressed by weird things back in the day when we knew nothing– but look at the totally cool stuff being brought to us by science every day!

    Awesome. How vile that some religions would keep people from learning the astounding facts that no magical scripture ever mentioned.

  27. says

    Large hadrosaurs colliding! I squirted coffee out my nose, and coffee break was hours ago.

    However, I noticed a comment regarding the film “The Fly” somewhat below that. Makes it sound as though somone’s going to use matter transmission chambers to yank a couple of hadrosaurs from the past and slam them together until they become a single organism.

    Any creationist driving by is likely to get some deuced funny … well, funnier … ideas about what goes on in paleontology labs.

  28. Longtime Lurker says

    Whoa, sorry about the bad post… what I MEANT was:

    T. rex or T. licious ?

    Gotta learn to use the preview option.

  29. pipsqueak says

    “Large Hadrosaur Collider”

    *wipes down keyboard*
    When the two hadrosaurs crash, will we all get sucked into a black hole or the dark ages?

  30. andrew says

    “I expect these cells to be denatured monomer soup on the inside.”

    If the endothelial cells are preserving their macroscopic appearance – along with that of the tissue – they will also be in a fair intracellular state of preservation.
    Just how far that goes is the interesting question.

  31. Snoof says

    #46 reminds me of a conversation I had a while back about what biologists would be like if they had to use the same techniques particle physicists use for investigating the internal structure of things.

    We concluded that it would be very, very messy.

  32. says

    JB, you’re using the wrong tool in BLASTing those fragments against the NCBI database. The original Asara and Schweitzer article (2007) claimed their 7 fragments of “T. rex” were closer to chicken than to reptile. Their analysis was based on a mass spectrometry profile, which is a series of peaks representing the masses of tiny protein fragments. These spectrograms have to be searched quite differently from BLAST, and I regret to report that Asara and Schweitzer didn’t know what they were doing. They didn’t include bacterial mass spec data in their database (apparently), and if they had, they’d have found that their fragments matched bacteria better than chicken.
    They wrote a letter to Science in Sept 2007 admitting that one of their 7 fragments was wrongly identified. Then another group published a Commentary in Science in January showing that Asara and Schweitzer’s T. rex DNA didn’t pass the proper (and standard) tests for ancient DNA. (Their mastodon samples did pass, however.)
    So Asara and Schweitzer’s work is just bad science. I just wrote a blogpost myself on this topic. It’s a classic case of bad science, unfortunately: take a plausible and likely true hypothesis – that T. rex is closer to birds than to reptiles – and then look for evidence proving that hypothesis. Here, the “looking” was done by mass spec, but the scientists were guided more by wishful thinking than by objectivity and skepticism. So even though the conclusion that “T. rex tasted like chicken” might be correct, the data is bogus. Too bad.

  33. Adam Cuerden says

    Ah, well. *raises a glass in toast to a cool, but unfortunately failed idea* Farewell, dinosaur proteins. Despite our desire that you might be true, science marches on, and you are the theory stomped under those marching feet.

    On the upside, though, snake fang evolution revealed through evo-devo? Probably right, and equally cool

  34. mikelist says

    i didn’t get enough info from the article to determine, but wouldn’t the biofilm take the shape of the surfaces it developed on? that it is not literally dinosaur soft tissue would only invalidate it for dna analysis, it might still have a great deal of value in determining how systems worked and how they looked.

  35. mikelist says

    sorry, i didn’t rta, having just read one that basically opined that the soft tissue idea was worthless because it wasn’t really soft tissue. i see that the above article covers the questions i had. i wish i would have read it before posting.

  36. John Moore says

    I seem to recall that the Science paper had a peptide sequence that was darn close to the sequence of chicken collagen, but was different enough to be T. rex protein. Presumably, a microbial biofilm would not yield a collagen-like sequence. However, the issue of contamination would have to be resolved by using two different GC-MS machines – one to read only the dino sequences and one to read the chicken/bird controls. I haven’t read the biofilm paper, but if I had to guess, it would set limits on which bones might yield actual results and which would yield microbial artifacts.

    Science 13 April 2007:
    Vol. 316. no. 5822, pp. 280 – 285
    DOI: 10.1126/science.1137614

    Evidence to the contrary:
    Science 4 January 2008:
    Vol. 319. no. 5859, p. 33
    DOI: 10.1126/science.1147046

    Mastodon sequences are okay, but T. rex “protein” is more like amphibian.

    Rebuttal of the naysayers:
    Science 4 January 2008:
    Vol. 319. no. 5859, p. 33
    DOI: 10.1126/science.1147364

    The biofilm argument should mean that neither the mastodon nor the dinosaur bones should have any soft tissue survive, but there are no absolutes in biology. The tougher the molecule, the more likely it might survive should the bones be isolated by some environmental fluke from microbial scavenging. DNA will not, but collagen might just survive. This debate will go on until they know which bones are impervious to microbial colonization.

  37. John Moore says

    According to this paper (http://www.jbc.org/cgi/content/abstract/278/34/32313), bacteria and phage have collagen like proteins. The discerning factor is that bacteria can not make hydroxyproline. If your GC-MS machine shows the existence of hydroxyproline, chances are it came from the animal, not the bacterial slime assuming there is no contamination. However, it is up to the original group to prove their results and prove that these remnants aren’t artifactual. The biofilm paper seems to be good science. A debate like this is good for the field.