DOI: 10.1126/science.1147046 , 33c (2008); 319Science

et al.Mike Buckley,Spectrometry"

Revealed by MassTyrannosaurus rexand Comment on "Protein Sequences from Mastodon

www.sciencemag.org (this information is current as of February 14, 2008 ):The following resources related to this article are available online at

http://www.sciencemag.org/cgi/content/full/319/5859/33cversion of this article at:

including high-resolution figures, can be found in the onlineUpdated information and services,

http://www.sciencemag.org/cgi/content/full/319/5859/33c/DC1 can be found at: Supporting Online Material

found at: can berelated to this articleA list of selected additional articles on the Science Web sites

http://www.sciencemag.org/cgi/content/full/319/5859/33c#related-content

http://www.sciencemag.org/cgi/content/full/319/5859/33c#otherarticles, 6 of which can be accessed for free: cites 12 articlesThis article

http://www.sciencemag.org/cgi/collection/tech_commentTechnical Comments

http://www.sciencemag.org/cgi/collection/paleoPaleontology

: subject collectionsThis article appears in the following

http://www.sciencemag.org/about/permissions.dtl in whole or in part can be found at: this article

permission to reproduce of this article or about obtaining reprintsInformation about obtaining

registered trademark of AAAS. is aScience2008 by the American Association for the Advancement of Science; all rights reserved. The title

CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by theScience

on

Feb

ruar

y 14

, 200

8 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

Comment on “Protein Sequences fromMastodon and Tyrannosaurus rexRevealed by Mass Spectrometry”Mike Buckley,1 Angela Walker,2 Simon Y. W. Ho,3 Yue Yang,1 Colin Smith,4 Peter Ashton,1Jane Thomas Oates,1 Enrico Cappellini,1 Hannah Koon,1 Kirsty Penkman,1 Ben Elsworth,1Dave Ashford,1 Caroline Solazzo,1 Phillip Andrews,2 John Strahler,2 Beth Shapiro,6Peggy Ostrom,5 Hasand Gandhi,5 Webb Miller,6 Brian Raney,7 Maria Ines Zylber,8M. Thomas P. Gilbert,9 Richard V. Prigodich,10 Michael Ryan,11 Kenneth F. Rijsdijk,12Anwar Janoo,13 Matthew J. Collins1*

We used authentication tests developed for ancient DNA to evaluate claims by Asara et al. (Reports,13 April 2007, p. 280) of collagen peptide sequences recovered from mastodon and Tyrannosaurus rexfossils. Although the mastodon samples pass these tests, absence of amino acid composition data, lackof evidence for peptide deamidation, and association of a1(I) collagen sequences with amphibiansrather than birds suggest that T. rex does not.

Early reports of DNA preservation inmultimillion-year-old bones (i.e., fromdinosaurs) have been largely dismissed

(1, 2) (table S1), but reports of protein recoveryare persistent [see (3) for review]. Most of thesestudies used secondary methods of detection,but Asara et al. (2) recently reported the directidentification of protein sequences, arguably thegold standard for molecular palaeontology,from fossil bones of an extinct mastodon andTyrannosaurus rex. After initial optimism gen-erated by reports of dinosaur DNA, there hasbeen increasing awareness of the problems andpitfalls that bedevil analysis of ancient samples(1), leading to a series of recommendations forfuture analysis (1, 4). As yet, there are no equiv-alent standards for fossil protein, so here weapply the recommended tests for DNA (4) to the

authentication of the reported mastodon andT. rex protein sequences (2) (Table 1).

First, the likelihood of collagen survivalneeds to be considered. The extremely hierar-

chical structure of collagen results in unusual,catastrophic degradation (5) as a consequence offibril collapse. The rate of collagen degradationin bone is slow because the mineral “locks” thecomponents of the matrix together, preventinghelical expansion, which is a prerequisite of fibrilcollapse (6). The packing that stabilizes collagenfibrils (6) also increases the temperature sensitiv-ity of degradation (Ea 173 kJ mol−1) (Fig. 1).Collagen decomposition would be much fasterin the T. rex buried in the then-megathermal(>20°C) (7) environment of the Hell Creekformation [collagen half-life (T½) = ~ 2 thousandyears (ky] than it would have been in themastodon lying within the Doeden Gravel Beds(present-day mean temperature, 7.5°C; collagenT½ = 130 ky) (Fig. 1).

This risk of contamination also needs to beevaluated. Collagen is an ideal molecular targetfor this assessment because the protein has ahighly characteristic motif that is also sufficient-ly variable to enable meaningful comparisonbetween distant taxa if enough sequence is ob-tained (Fig. 2). Compared with ancient DNAamplification, contamination by collagen is in-herently less likely. Furthermore, because thebones sampled in (2) were excavated by the

TECHNICALCOMMENT

1BioArch, Departments of Biology, Archaeology, Chemistryand Technology Facility, University of York, Post Office Box373, York YO10 5YW, UK. 2Department of Biological Chem-istry, University of Michigan Medical School, Ann Arbor, MI48109–0404, USA. 3Evolutionary Biology Group, Departmentof Zoology, University of Oxford, OX1 3PS, UK. 4Departmentof Human Evolution, Max Planck Institute for EvolutionaryAnthropology, Deutscher Platz 6, D-04103, Leipzig, Germany.5Department of Zoology, Michigan State University, EastLansing, MI 48824, USA. 6Department of Biology, Pennsylva-nia State University, University Park, PA 16802, USA. 7Centerfor Biomolecular Science and Engineering, University ofCalifornia–Santa Cruz, CA 95064, USA. 8Department of Par-asitology, Kuvin Center, Hebrew University of Jerusalem,Israel. 9Biological Institute, University of Copenhagen, Univer-sitetsparken 15, 2100 Copenhagen, Denmark. 10ChemistryDepartment, Trinity College, 300 Summit Street, Hartford, CT06106, USA. 11Cleveland Museum of Natural History, 1 WadeOval Drive, University Circle, Cleveland, OH 44106, USA.12National Museum of Natural History “Naturalis,” P.O. Box9517, 2300 RA Leiden, Netherlands. 13National HeritageTrust Fund Mauritius, Mauritius Institute, La Chaussée StreetPort Louis, Mauritius.

*To whom correspondence should be addressed. E-mail:mc80@york.ac.uk

Table 1. Key questions to ask about ancient biomolecular investigations [adapted from (4)].

Test Sample Pass Observation

Do the age, environmentalhistory, and preservationof the sample suggest collagensurvival?

Mastodon, 300 to600 ky old

Yes Collagen T½ at 7.5°C = 130 ky

T. rex, 65 millionyears old

No Collagen T½ at 20°C = 2 ky

Do the biomolecular and/ormacromolecular preservationof the sample, the moleculartarget, the innate nature of thesample, and its handlinghistory suggest thatcontamination is a risk?

Biomolecularpreservation

? Range of evidencepresented (8) but noamino acidcompositional data

Macromolecularpreservation

Yes Macromolecularpreservation is not theequivalent ofbiomolecularpreservation (9)

Molecular target YesHandling history Yes? Large (2.5 g) samples increase

risk of contamination?Do the data suggest that thesequence is authentic, ratherthan the result of damage andcontamination?

Mastodonand T. rex

No Errors in interpretationof spectra [see table S1and (13)]?

Damage-induced errorsin sequence

Do the results make sense, andare there enough data to makethe study useful and/or tosupport the conclusions?

Mastodon Yes Weak affinity to mammals

T. rex No Affinity of a1(I) peptidesto amphibians, not birdsor reptiles

www.sciencemag.org SCIENCE VOL 319 4 JANUARY 2008 33c

on

Feb

ruar

y 14

, 200

8 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

authors, obvious contamination sources such asanimal glue (used in conservation) can be ex-cluded. However, concentrating protein from thelarge amounts of bone used (2.5 g) may haveheightened the risk of extraneous proteinsentering the sample during extraction, althoughthere have been no systematic studies of thisphenomenon. Independent extraction and analy-ses would have strengthened claims for theauthenticity of the origin of the peptides (andpotentially ameliorated the original problems ofdata interpretation) (4).

The remarkable soft-tissue preservation ofthe investigated T. rex specimen (MOR 1125)has been documented (8). However, microscop-ic preservation does not equate with molecularpreservation (9). Immunohistochemistry providessupport for collagen preservation, but Asara et al.(2) presented no data regarding inhibition assayswith collagen from different species or cross-reactivity with likely contaminants [e.g., fungi(10)]. Curiously, no amino acid compositionalanalysis was conducted [see (11)], althoughimmonium ions were identified by time-of-flightsecondary ion mass spectrometry. In our experi-ence, collagen-like amino acid profiles have beenobtained in all bones from which we could obtaincollagen sequence (Fig. 1, inset).

Regarding the proof of sequence authenticity,the spectra reported by Asara et al. (12) areinconsistent with some of the sequence assign-ments (13) (table S1). A common diageneticmodification, deamidation, not considered in (2),may shed light on authenticity. The facilesuccinimide-mediated deamidation (14) of aspar-agine occurred at N229G and N1156G in ostrichpeptides (Ost 4 and Ost5) (see table S1 fornomenclature), presumably during sample prep-aration. Direct hydrolytic deamidation is slower(14), and an expectation of elevated levels ofsuch products is reasonable for old samples. Weagree with the most recent interpretation (13) ofthe spectrum illustrated in Fig. 2B as a1(I)G362SEGPEGVR370, the deamidated (Q→E367)form of the sequence found in most mammals(12). By way of contrast, none of the threeglutamine residues in the reported T. rex peptidesare deamidated (table S1). Only time will tell ifQ→E is a useful marker for authentically oldcollagen, but from the evidence presented, themastodon sequence looks more diageneticallyaltered than T. rex.

The unusual, fragmented nature of the re-ported T. rex sequence does not make it ame-nable to standard, model-based phylogeneticanalysis. Instead, we examined the phylogenetic

signal of the a1(I) frag-ments of mastodon andT. rex using Neighbor-Net analysis and uncor-rected genetic distances.Using the sequencesreported in (13), boththe T. rex and masto-don signal display anaffinity with amphibians(Fig. 2A). Our reinter-pretation of the spectra(12) changes the affinityof mastodon but not ofT. rex (Fig. 2B). Inaddition to the a1(I)peptides used in theNeighbor-Net analysis,Asara et al. reportedtwo other peptides fromT. rex (13); we questionthe interpretation of thea1(II) spectra (identicalto frog) but not the a2(I)spectra (identical tochicken).

We requiremore datato be convinced of theauthenticity of the T. rexcollagen sequences re-ported by Asara et al.Nevertheless, the hand-ful of spectra reportedfor the temperate Pleisto-cene mastodon fail nei-ther phylogenetic nordiagenetic tests, thus

highlighting the potential of protein mass spec-trometry to bridge the present gulf in our un-derstanding between the fate of archaeologicaland fossil proteins. To avoid past mistakes ofancient DNA research (1), we recommend thatfuture fossil protein claims be considered in lightof tests for authenticity such as those presentedhere.

Reference and Notes1. E. Willerslev, A. Cooper, Proc. R. Soc. London. B. Biol.

Sci. 272, 3 (2005).2. J. M. Asara, M. H. Schweitzer, L. M. Freimark, M. Phillips,

L. C. Cantley, Science 316, 280 (2007).3. M. H. Schweitzer, Palaeontologia Electronica

5, editorial 2 (2003); http://palaeo-electronica.org/2002_2/r_and_p.pdf

4. M. T. P. Gilbert, H.-J. Bandelt, M. Hofreiter, I. Barnes,Trends Ecol. Evol. 20, 541 (2005).

5. M. J. Collins, M. S. Riley, A. M. Child, G. Turner-Walker,J. Archaeol. Sci. 22, 175 (1995).

6. C. A. Miles, M. Ghelashvili, Biophys. J. 76, 3243(1999).

7. K. R. Johnson, D. J. Nichols, J. H. Hartman, The Hell CreekFormation and the Cretaceous-Tertiary Boundary in theNorthern Great Plains: Geological Society of AmericaSpecial Paper 361, 503–510 (2002).

8. M. H. Schweitzer et al., Science 316, 277 (2007).9. N. S. Gupta, D. E. G. Briggs, R. D. Pancost, J. Geol. Soc.

London 163, 897 (2006).10. P. Sepulveda et al., Infect. Immun. 63, 2173

(1995).

Age

(un

calib

rate

d R

adio

carb

on y

ears

)

–5.0 0.0 5.0 10.0 15.0 20.0 25.0 30.0

Effective Collagen Degradation Temperature (°C)

predicted survival limit

14C dates onbone collagen

50000

45000

40000

35000

30000

25000

20000

15000

10000

5000

0

ardiagenesis

axis

2 (

23%

)

–20% 20% 40% 60%0%

10%

–10%

–30%

–40%axis 1 (25%)

meat andbone meal

100%

–100%

–100%

100%

Fig. 1. Plot of radiocarbon age versus estimated effective collagen degradation temperature for radiocarbon-dated bones fromlaboratory databases (principally Oxford and Groningen). The line represents the expected calendar age at which 1% of the originalcollagen remains following a zero-order reaction; almost no bone collagen survives beyond this predicted limit. (Inset) The 99%confidence intervals of amino acid compositions by first two principal component analyses (48% of total variance) for bones fromNW Europe aged <11 ky (n= 324), 11 to 110 ky (n= 210), 110 to 130 ky (n= 26), and 130 to 700 ky (n= 31). Pliocene samples arenot plotted, as their composition (n = 8) is highly variable and yields of amino acids are low. The orange line indicates acompositional trend observed when compact bone is heated for 32 days at 95°C, which reduces collagen to 1% of the initialconcentration [each inflection represents a separate analysis; n = 32)]. The composition becomes more similar to mixed tissuesamples (meat and bone meal; n = 32), principally due to the depletion of Gly. An amino acid profile for mammoth is consistent withcollagen, unlike the associated sediment sample [data from (11)].

4 JANUARY 2008 VOL 319 SCIENCE www.sciencemag.org33c

TECHNICAL COMMENT

on

Feb

ruar

y 14

, 200

8 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

11. M. Schweitzer, C. L. Hill, J. M. Asara, W. S. Lane,S. H. Pincus, J. Mol. Evol. 55, 696 (2002).

12. See supporting material on Science Online.13. J. M. Asara et al., Science 317, 1324 (2007).14. N. E. Robinson et al., J. Pept. Res. 63, 426

(2004).15. This work was supported by NSF (EAR-0309467),

National Environment Research Council (NE511148,

GR9/01656), National Center for Research Resources(P41-18627), European Commission (MEST-CT-2004-007909, MEST-CT-2005-020601), Wellcome TrustBioarchaeology Fellowships (K.P. and H.K.) AnalyticalChemistry Trust Fund, the Royal Society of ChemistryAnalytical Division, Engineering and Physical SciencesResearch Council, and the Michigan ProteomeConsortium.

Supporting Online Materialwww.sciencemag.org/cgi/content/full/319/5859/33c/DC1SOM TextTable S1References

26 June 2007; accepted 20 November 200710.1126/science.1147046

Pufferfish (Tetraodon)Pufferfish (Fugu)

Stickleback

SkateOstrich

OpossumCowTree shrew

Galago

TenrecHorse

RabbitGuinea pig

MouseRat

Platypus

TyrannosaurusNewt

Mastodon

BullfrogZebrafish

Trout

Halibut

Medaka

Rhesus

CatDog

Green anole

ChickenHuman, Chimp

Elephant

Xenopus0.01

Pufferfish (Tetraodon)Pufferfish (Fugu)

Stickleback

Skate

Green anole

Ostrich

OpossumMouse

Galago

MastodonTenrec

DogCatHorse

Guinea pig

PlatypusXenopus

Bullfrog

Tyrannosaurus

NewtZebrafish

Trout

Halibut

Medaka

Chimp, HumanRhesus

Cow

Elephant

Rabbit

Tree shrew Rat Chicken

0.01

A

B

Fig. 2. Phylogenetic networks of a1(I) sequences using Neighbor-Net analysis (A) with the most recent Asara et al. assignments (13) and (B) after ourreinterpretation of the mass spectrometric data (12). T. rex does not group with bird/reptile using either set of sequence alignments. More sequence is requiredfor a full, model-based phylogenetic analysis.

www.sciencemag.org SCIENCE VOL 319 4 JANUARY 2008 33c

TECHNICAL COMMENT

on

Feb

ruar

y 14

, 200

8 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

Weighing the mass spectrometric evidence for authenticTyrannosaurus rex collagen

Mike Buckley1, Angela Walker2, Simon Y. W. Ho3, Yue Yang1, Colin Smith4, Peter Ashton1,Jane Thomas Oates1, Enrico Cappellini1, Hannah Koon1, Kirsty Penkman1, Ben Elsworth1,Dave Ashford1, Caroline Solazzo1, Phil Andrews2, John Strahler2, Beth Shapiro3, PeggyOstrom5, Hasand Gandhi5, Webb Miller6, Brian Raney7, Maria Ines Zylber8, M. Thomas P.Gilbert9, Richard V. Prigodich10, Michael Ryan11, Kenneth F. Rijsdijk12, Anwar Janoo13,and Matthew J. Collins1,*

1BioArch, Departments of Biology, Archaeology, Chemistry and Technology Facility, University of York,York, UK. PO Box 373, York YO10 5YW, UK 2Department of Biological Chemistry, University of MichiganMedical School, Ann Arbor, MI 48109-0404, USA 3Evolutionary Biology Group, Department of Zoology,University of Oxford, OX1 3PS, United Kingdom 4Department of Human Evolution, Max Planck Institute forEvolutionary, Anthropology, Deutscher Platz 6, D-04103, Leipzig, Germany 5Department of Zoology,Michigan State University, East Lansing, Michigan 48824, USA 6Center for Comparative Genomics andBioinformatics, Penn State, University Park, PA 16802, USA 7Center for Biomolecular Science andEngineering, University of California Santa Cruz, CA 95064, USA 8Dept. of Parasitology, Kuvin Center, TheHebrew University of Jerusalem, Israel 9Biological Institute, University of Copenhagen, Universitetsparken15, 2100 Copenhagen, Denmark 10Chemistry Department, Trinity College, 300 Summit Street, Hartford, CT06106 11Cleveland Museum of Natural History, 1 Wade Oval Dr., University Circle Cleveland, Ohio 44106,USA 12National Museum of Natural History ‘Naturalis’, P.O. Box 9517, 2300 RA Leiden, the Netherlands13National Heritage Trust Fund Mauritius, Mauritius Institute, La Chaussée Street Port Louis, Mauritius

AbstractWe use authentication tests developed for ancient DNA to evaluate claims by Asara et al. of collagenpeptide sequences recovered from mastodon and Tyrannosaurus rex fossils. Although the mastodonpasses, absence of amino acid composition data, lack of evidence for peptide deamidation, andassociation of the α1(I) peptide sequences with amphibians not birds, suggests that T. rex does not.

CommentEarly reports of DNA preservation in multi-million year old bones (i.e. dinosaurs) have beenlargely dismissed (see 1 and SOM T1, 2) but reports of protein recovery are persistent (see 3for review). Most of these studies used secondary methods of detection, but protein sequence,arguably the gold standard for molecular palaeontology, has now been claimed for the firsttime (2). Following initial optimism generated by reports of dinosaur DNA, there arose agradual awareness of the problems and pitfalls which bedevil analysis of ancient samples (1),leading to a series of recommendations for future analysis (1,4). As yet, there are no equivalentstandards for fossil protein, so here we apply the recommended tests for DNA (4) to theauthentication of the reported protein sequences (2) (Table 1).

*To whom correspondence should be addressed. E-mail: mc80@york.ac.uk

UKPMC Funders GroupAuthor ManuscriptScience. Author manuscript; available in PMC 2009 June 11.

Published in final edited form as:Science. 2008 January 4; 319(5859): 33. doi:10.1126/science.1147046.

UKPM

C Funders G

roup Author Manuscript

UKPM

C Funders G

roup Author Manuscript

Likelihood of collagen survivalThe extremely hierarchical structure of collagen results in unusual, catastrophic degradation(5) as a consequence of fibril collapse. The rate of collagen degradation in bone is slow becausethe mineral ‘locks’ the components of the matrix together, preventing helical expansion whichis a pre-requisite of fibril collapse (6). The packing which stabilises collagen fibrils (6) alsoincreases the temperature sensitivity of degradation (Ea 173 kJ mol-1;Fig. 1). Collagendecomposition would be much faster in the T. rex buried in the then megathermal (>20 °C)(7) environment of the Hell Creek formation (collagen t ½ ~ 2 ka) than it would have been inthe mastodon lying within the Doeden Gravel Beds (present day mean temperature 7.5 °C;collagen t ½ 130 ka; Fig. 1).

Risk of contaminationThe molecular target (collagen) is ideal for this investigation; the protein has a highlycharacteristic motif and yet also is sufficiently variable to enable meaningful comparisonbetween distant taxa if sufficient sequence is obtained (Fig. 2). In comparison with ancientDNA amplification, contamination by collagen is inherently less likely. Furthermore becausethe bones were excavated by the authors, obvious contamination sources such as animal glue(used in conservation) can be excluded. Concentrating protein from the large amounts of boneused (2.5 g) may have heightened the risk of extraneous proteins entering the sample duringextraction, but there have been no systematic studies of this phenomenon. Independentextraction and analyses would have strengthened claims for the authenticity of the origin ofthe peptides (and potentially ameliorated the original problems of data interpretation) (4).

The remarkable soft-tissue preservation of the investigated T. rex specimen (MOR 1125) hasbeen documented (8); however microscopic preservation does not equal molecularpreservation (9). Immunohistochemistry provides support for collagen preservation, althoughno data regarding inhibition assays with collagen from different species or cross-reactivity withlikely contaminants (e.g. fungi 10) are presented. Curiously no amino acid compositionalanalysis was conducted (c.f. 11) although immonium ions were identified (by TOF-SIMS). Inour experience, collagen-like amino acid profiles have been obtained in all bones from whichwe could obtain collagen sequence (Fig. 1, inset).

Proof of sequence authenticityThe spectra (12) (SOM of 12) are inconsistent with many of the original sequence assignments(13,14) (SOM Table 1 and 13). A common diagenetic modification, deamidation, notconsidered in the original publication, may shed light on authenticity. The facile succinimidemediated deamidation of N1156G (14) occurred in peptide Ost 5 (see Table 1 of 13 fornomenclature) from ostrich, presumably during sample preparation. Direct hydrolyticdeamidation is slower (14) and an expectation of elevated levels of such products is reasonablefor old samples. We agree with the new interpretation (13) of the spectrum illustrated in Fig.2b of 3 as α1(I) G362SEGPEGVR370, the deamidated (Q→E367) form of the sequence foundin most mammals (12). By way of contrast, none of the three glutamine residues in “T. rex”peptides are deamidated (Table S1 SOM). Only time will tell if Q→E is a useful marker forauthentically old collagen, but from the evidence presented the mastodon sequence looksmore diagenetically altered than T. rex.

The unusual, fragmented nature of the reported T. rex sequence does not make it amenable tostandard, model-based phylogenetic analysis. Instead, we examined the phylogenetic signal ofthe α1(I) fragments of mastodon and T. rex using Neighbor-Net analysis and uncorrectedgenetic distances. Using the originally reported sequences (2), both the T. rex and mastodonsignal displayed an affinity with amphibians (Fig. 2a) (12). Our re-interpretation of the spectrachanges the affinity of mastodon but not of T. rex (Fig. 2b). Two other peptides are reported

Buckley et al. Page 2

Science. Author manuscript; available in PMC 2009 June 11.

UKPM

C Funders G

roup Author Manuscript

UKPM

C Funders G

roup Author Manuscript

from T. rex (3); we question the interpretation of the (frog) α1(II) spectrum, but not the α2(I)assignment which is identical to chicken.

We require more data to be convinced of the authenticity of the T. rex collagen. Nevertheless,the handful of spectra reported for the temperate Pleistocene mastodon fail neither phylogeneticnor diagenetic tests, highlighting the potential of protein mass spectrometry to bridge thepresent gulf in our understanding between the fate of archaeological and fossil proteins (Fig1.). In order to avoid past mistakes of ancient DNA research(1), we would recommend thatfuture fossil protein claims are considered in the light of tests for authenticity such as those wehave used.

Supplementary MaterialRefer to Web version on PubMed Central for supplementary material.

AcknowledgmentsThis work was supported by NSF (EAR-0309467), NERC (NE511148, GR9/01656), NCRR (P41-18627), EU (MEST-CT-2004-007909, MEST-CT-2005-020601), Wellcome Trust Bioarchaeology Fellowships (KP, HK) AnalyticalChemistry Trust Fund, the RSC Analytical Division, EPSRC, and the Michigan Proteome Consortium.

References1. Willerslev E, Cooper A. Proceedings of the Royal Society of London Series B-Biological Sciences

Jan 7;2005 272:3.2. Asara JM, Schweitzer MH, Freimark LM, Phillips M, Cantley LC. Science April 13;2007 316:280.

[PubMed: 17431180]20073. Schweitzer MH. Palaeontologia Electronica 2002;54. Gilbert MTP, Bandelt H-J, Hofreiter M, Barnes I. Trends in Ecology and Evolution Oct;2005 20:541.

[PubMed: 16701432]20055. Collins MJ, Riley MS, Child AM, Turner-Walker G. Journal of Archaeological Science 1995;22:175.6. Miles CA, Ghelashvili M. Biophysical Journal 1999;76:3243. [PubMed: 10354449]7. Johnson KR, Nichols DJ, Hartman JH. The Hell Creek Formation and the Cretaceous-Tertiary

Boundary in the northern Great Plains: Geological Society of America Special Paper 2002;361:503–510.

8. Schweitzer MH, et al. Science April 13;2007 316:277. [PubMed: 17431179]20079. Gupta NS, Briggs DEG, Pancost RD. Journal of the Geological Society Nov;2006 163:897.10. Sepulveda P, et al. Infect. Immun June 1;1995 63:2173. [PubMed: 7768595]11. Schweitzer M, Hill CL, Asara JM, Lane WS, Pincus SH. Journal of Molecular Evolution Dec 01;2002

55:696. [PubMed: 12486528]200212. Supplementary, Material.13. Asara JM, et al. Science September 7;2007 317:1324. [PubMed: 17823333]14. Robinson NE, et al. Journal of Peptide Research 2004;63:426. [PubMed: 15140160]

Buckley et al. Page 3

Science. Author manuscript; available in PMC 2009 June 11.

UKPM

C Funders G

roup Author Manuscript

UKPM

C Funders G

roup Author Manuscript

Figure 1.Plot of radiocarbon age versus estimated effective collagen degradation temperature forradiocarbon dated bones from laboratory databases (principally Oxford and Groningen). Theline represents the expected calendar age at which 1% of the original collagen remainsfollowing a zero-order reaction; almost no bone collagen survives beyond this predicted limit.Inset. 99% confidence intervals of amino acid compositions by first two principal componentanalyses (48% of total variance) for < 11 ka (n = 324), 11-110 ka (n = 210), 110-130 ka (n =26) and 130-700 ka (n = 31) bones from NW Europe. Pliocene samples are not plotted, as theircomposition (n = 8) is highly variable and yields of amino acids are low. The orange lineindicates a compositional trend observed when compact bone is heated for 32 days at 95 °C,which reduces collagen to 1% of initial concentration, each inflection representing a separateanalysis (n = 32). The composition becomes more similar to mixed tissues samples (meat andbone meal, n = 32) principally due to the depletion of Gly. An amino acid profile for mammoth(m) is consistent with collagen, unlike the associated sediment sample (s) (data from 11).

Buckley et al. Page 4

Science. Author manuscript; available in PMC 2009 June 11.

UKPM

C Funders G

roup Author Manuscript

UKPM

C Funders G

roup Author Manuscript

Figure. 2.Phylogenetic networks of α1(I) sequences using Neighbor-Net analysis (A) with the originalassignments (2) and (B) following reinterpretation of the mass spectrometric data (12). T.rex does not group with bird/reptile using either set of sequence alignments. More sequence isrequired for a full, model-based phylogenetic analysis

Buckley et al. Page 5

Science. Author manuscript; available in PMC 2009 June 11.

UKPM

C Funders G

roup Author Manuscript

UKPM

C Funders G

roup Author Manuscript

UKPM

C Funders G

roup Author Manuscript

UKPM

C Funders G

roup Author Manuscript

Buckley et al. Page 6

Table 1Key questions to ask about ancient biomolecular investigations (taken from 4)

Test Sample Pass Observation

Does the age, environmental history and preservationof the sample suggest collagen survival?

Mastodon 300-600 ka Collagen t½ @ 7.5 °C = 130 ka

T. rex 65 Ma Collagen t½ @ 20 °C = 2 ka

Does the biomolecular and/or macromolecularpreservation of the sample, the molecular target, theinnate nature of the sample and its handling historysuggest that contamination is a risk?

Biomolecular preservation ? Range of evidence presented(8) but no amino acidcompositional data.

Macromolecular preservation ...but macromolecular ≠biomolecular preservation(9).

Molecular target

Handling history ...but large (2.5 g) samplesprocessed

Do the data suggest that the sequence is authentic, ratherthan the result of damage, and contamination?

Errors in interpretation ofspectra (see SOM Table 1 and13)? Damage induced errorsin sequence.

Do the results make sense, and are there enough data tomake the study useful and/or support the conclusions?

Mastodon Weak affinity to mammals

T. rex Affinity of α1(I) peptidesamphibians, not birds orreptiles

Science. Author manuscript; available in PMC 2009 June 11.

www.sciencemag.org/cgi/content/full/319/5859/33c/DC1

Supporting Online Material for

Comment on “Protein Sequences from Mastodon and Tyrannosaurus rex Revealed by Mass Spectrometry”

Mike Buckley, Angela Walker, Simon Y. W. Ho, Yue Yang, Colin Smith, Peter Ashton, Jane Thomas Oates, Enrico Cappellini, Hannah Koon, Kirsty Penkman, Ben Elsworth,

Dave Ashford, Caroline Solazzo, Phil Andrews, John Strahler, Beth Shapiro, Peggy Ostrom, Hasand Gandhi, Webb Miller, Brian Raney, Maria Ines Zylber,

M. Thomas P. Gilbert, Richard V. Prigodich, Michael Ryan, Kenneth F. Rijsdijk, Anwar Janoo, Matthew J. Collins*

*To whom correspondence should be addressed. E-mail: mc80@york.ac.uk

Published 4 January 2008, Science 319, 33c (2008)

DOI: 10.1126/science.1147046

This PDF file includes:

SOM Text References

Other Supporting Online Material for this manuscript includes the following: (available at www.sciencemag.org/cgi/content/full/319/5859/33c/DC1

Table S1 Collagen a1 Sequences

Buckley et al., Technical Comment on Asara et al., (Reports 13 April 2007, p. 280) “Protein Sequences from Mastodon and Tyrannosaurus rex Revealed by Mass Spectrometry” Supporting online material.

Comment on “Protein Sequences from Mastodon and Tyrannosaurus rex Revealed by Mass Spectrometry”

Supporting Online Material Mike Buckley1, Angela Walker2, Simon Y. W. Ho3, Yue Yang 1, Colin Smith4 Peter Ashton1, Jane Thomas Oates1, Enrico Cappellini1, Hannah Koon1, Kirsty Penkman1, Ben Elsworth1, Dave Ashford1, Caroline Solazzo1, Phillip Andrews2, John Strahler2, Beth Shapiro6, Peggy Ostrom5, Hasand Gandhi5, Webb Miller6, Brian Raney7, Maria Ines Zylber8, M. Thomas P. Gilbert9, Richard V. Prigodich10, Michael Ryan11, Kenneth F. Rijsdijk12, Anwar Janoo13, Matthew J. Collins1* 1BioArch, Departments of Biology, Archaeology, Chemistry and Technology Facility, University of York, York, UK. PO Box 373, York YO10 5YW, UK 2Department of Biological Chemistry, University of Michigan Medical School, Ann Arbor, MI 48109-0404, USA 3Evolutionary Biology Group, Department of Zoology, University of Oxford, OX1 3PS, United Kingdom 4Department of Human Evolution, Max Planck Institute for Evolutionary, Anthropology, Deutscher Platz 6, D-04103, Leipzig, Germany 5Department of Zoology, Michigan State University, East Lansing, Michigan 48824, USA 6 Department of Biology, Pennsylvania State University, University Park, PA 16802, USA. 7Center for Biomolecular Science and Engineering, University of California Santa Cruz, CA 95064, USA 8Dept. of Parasitology, Kuvin Center, The Hebrew University of Jerusalem, Israel 9Biological Institute, University of Copenhagen, Universitetsparken 15, 2100 Copenhagen, Denmark 10Chemistry Department, Trinity College, 300 Summit Street, Hartford, CT 06106 11Cleveland Museum of Natural History, 1 Wade Oval Dr., University Circle Cleveland, Ohio 44106, USA 12National Museum of Natural History 'Naturalis', P.O. Box 9517, 2300 RA Leiden, the Netherlands 13National Heritage Trust Fund Mauritius, Mauritius Institute, La Chaussée Street Port Louis, Mauritius *To whom correspondence should be addressed. E-mail: mc80@york.ac.uk

Buckley et al., Technical Comment on Asara et al., (Reports 13 April 2007, p. 280) “Protein Sequences from Mastodon and Tyrannosaurus rex Revealed by Mass Spectrometry” Supporting online material.

Thermal age estimates The effective annual temperature experienced by buried mineralized collagen reflects both the mean annual air temperature (MAT) and seasonal fluctuations. Using climatology data compiled in the years 1987 and 1988, the International Satellite Land Surface Climatology Project (ISLSCP) (1) made global estimates of monthly average soil temperature, at depths of 0 - 49 cms and 49 - 100cms, at a 1° by 1° resolution Estimates were calculated from surface temperature, using the expression T(d)

n = (1-c)T(o) + c(a T(s)n + b T(s)

n-1) Equation 1 Where T(o) denotes the mean annual surface temperature, T(s)

n and T(s)n-1 are the surface

temperature for month n and previous month n-1, a and b are constants defining the temperature phase lag (a = b = 0.5) and c is a constant describing the amplitude damping (c = 0.77). Using this data (and given A = 2.11 x1019, Ea = 173.2kJmol-1 for collagen degradation) we estimated the total reaction over the 24 month cycle to make an estimate of the effective collagen degradation temperature (Teff) at each grid point. The Teff represents the worst-case scenario, shallow buried, poorly thermally buffered bone. An effective collagen degradation temperature could be assigned for each bone in the radiocarbon database for which location data was available. No correction is made for altitude, beyond the resolution of the ISLSCP data-set. This analysis does not take into account palaeoclimatic changes during the burial history of the material but clearly displays the influence of burial temperature on collagen preservation. Based upon the thermal age analysis the mastodon MOR 605 (Doeden Gravel Beds, Eastern Montana, Nr Miles City (mean annual temperature 7.5oC; WS 7266612 MILES CITY, 801m Latitude: 46.41 N, Longitude: 105.84 W.) is predicted (within error) to contain collagen.

Amino acid analysis Amino acid analysis was conducted upon the whole bone, soluble and insoluble fractions of samples from (i) Holocene - Pleistocene bones, (ii) meat and bone meal (supplied by Prosper de Mulder) and (iii) bone heated in water at 95ºC for 32 days. The three fractions were prepared as follows; whole bone samples were hydrolysed in 7 M HCl for 18 h @ 110°C, the soluble fractions from 4 h demineralisation with 0.6 M HCl, and concentrated by evaporation were hydrolyzed with 6 M HCl for 18 h @110ºC; insoluble fractions from the same 4 h demineralisation were washed 5 times with deionised water, lyophilised and hydrolyzed with 6 M HCl for 18 h @110ºC. Samples were analysed for amino acid composition using rHPLC with fluorescence detection (as detailed in 2) Imino acids (Pro, Hyp) were not detected by this method.

Collagen Alignments We added to sequences available from public databases, by extracting and conceptually translating the α1(I) collagen genes from a whole-genome alignment of 30 vertebrate species that will soon be made available at the UCSC Human Genome Browser; these additional sequences are included as a FASTA file. These underwent minor editing to remove small sections in which the sequence lacked the characteristic Gly-Xxx-Yyy motif, but otherwise did not alter the conceptual translations. These sequences were used for creation of the Hypothetical Collagens database and in Neighbor-Net analysis.

Buckley et al., Technical Comment on Asara et al., (Reports 13 April 2007, p. 280) “Protein Sequences from Mastodon and Tyrannosaurus rex Revealed by Mass Spectrometry” Supporting online material.

Interpretation of published spectrum The starting points for manual interpretation of spectra (Supporting Online Material from 3)were the sequences originally reported by Asara et.al. The bias was for the collagen GXY motif and the reassignment of all G* calls. For the isobaric I/L/P*, fragment ion intensity (N-terminal side of P*; C-terminal side of I/L) was the prime determinant. Partial de novo sequencing was done where possible. Delta masses were typically +/- 0.2 Da and permitted distinction between amide vs acid side chains. In the text we have used the numbering system of the α1(I) chain of human type I collagen (accession number P02452). The mature chain of α1(I) collagen is from positions 162 to 1217 (1056 residues). Detailed interpretations of the published spectra are presented in Supplementary Online Materials, Table S1.

Neighbor-Net Analysis The collagen α1(I) sequences were aligned with orthologues from 32 other species. The alignment was investigated using Neighbor-Net analysis (4) as implemented by the software SplitsTree 4 (5). We compared published sequences from (3) with our assignments of the same spectra obtained by manual or partial de novo interpretation (Table S1). In our assignments we used the sequences with the non-deamidated residue if we believed that deamidation arose diagenetically and was not part of the original sequence. Neighbor-Net is a statistically consistent method closely related to split decomposition. It produces a splits graph, which is a representation of the phylogenetic split signals present in an alignment (4). Our analysis, based on uncorrected genetic distances, revealed that the phylogenetic signal was ambiguous for amphibians, reptiles, and birds. However, the T. rex sequence showed a closer affinity to the amphibian sequences, particularly newt.

We did not align the solitary α2(I) or α1(II) peptides reported for T. rex which had identity to bird and amphibian respectively.

References 1. B. W. Meeson et al. (Published on CD NASA

(USA_NASA_GDAAC_ISLSCP_001-USA_NASA_GDDAC_ISLSCP_005). 1995), vol. 1-5.

2. S. A. Parfitt et al., Nature 438, 1008 (2005).

3. J. M. Asara, M. H. Schweitzer, L. M. Freimark, M. Phillips, L. C. Cantley, Science 316, 280 (2007).

4. D. Bryant, Molecular Biology and Evolution 21, 255 (2003).

5. D. H. Huson, D. Bryant, Mol Biol Evol 23, 254 (2006).