FTIR, XPol, and LC-MS confirm endogenous collagen in Edmontosaurus fossil bone

doi:10.21203/rs.3.rs-3576191/v1

Download PDF

Research Article

FTIR, XPol, and LC-MS confirm endogenous collagen in Edmontosaurus fossil bone

https://doi.org/10.21203/rs.3.rs-3576191/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this older preprint version

Read the latest preprint version →

Reports of proteins in fossilized bones have ignited an on-going debate because previously it was generally assumed that fossilization results in destruction of all organic components. We combine Attenuated Total Reflection (ATR) Fourier Transform InfraRed (FTIR) spectroscopy, cross-polarized light microscopy (XPol), and two variations of Liquid Chromatography Mass Spectrometry (LC-MS), to detect and verify endogenous collagen protein in fossilized bone. Data from an Edmontosaurus dinosaur excavated from Upper Cretaceous strata of the South Dakota Hell Creek Formation were compared with that from thermally decayed modern turkey (Meleagris) bone. The four methods show independent evidence for collagen presence in the fossilized bone. ATR-FTIR spectra show a signal for carbonyl (C = O) groups, a repeat moiety in protein amino acid sequences. XPol shows the birefringence of collagen within Meleagris, as well as Edmontosaurus. LC-MS bottom-up proteomics showed m/z agreement between multiple ions from trypsin-digested samples and identifies peptide sequences thus far only identified in a hadrosaur. Tandem LC-MS unambiguously identified hydroxyproline, a unique collagen-indicator amino acid, in acid-digested samples from the same.

Collagen

Edmontosaurus

FTIR

XPol

Tandem LC-MS

The bones of vertebrate animals contain proteins including collagen [1]. After death these molecules decay over time as a consequence of bio- and environmentally-induced degradation [2]. However, due to the initially high bone collagen concentration [3] as well as the availability of increasingly sensitive techniques to assess proteins, it is possible to identify and quantify relatively small amounts of the remnants of this protein in ancient bone samples.

Bone stability and diagenesis (and with it the temporal decay of organic molecules including collagen) is of interest to palaeontology [4, 5] and archaeology [6], but also forensics [7]. Theoretically the state of decay could provide information about a bone fossil such as burial information (depth, temperature, aerobic/anaerobic conditions, etc.) [8], disease status [9], and an estimate of the age of the fossil in question [10–15]. However, before this can be realized, many variables need definition and formulation.

Detection of the components of soft tissues (e.g., proteins—particularly collagen—and in some cases DNA) in animal (fossil) bones of various taphonomic stages is a growing field of study, with more discoveries reported over time. This study contributes to the expanding list of such findings.

Among the earliest of such reports (1966) was that of collagenous material in an Upper Cretaceous dinosaur from the Gobi desert, by means of electron microscopy [16].

Later, T. rex (Hell Creek Formation, Montana, Museum of the Rockies MOR 555) collagen fibres were revealed by transmission electron microscopy (TEM) in 1999 [17]. In this case, some of the preserving conditions were attributed to rapid burial. Microstructures similar in shape to red blood cells were also observed. Attempts to identify residual haemoglobin and heme were inconclusive [18].

The same group used using scanning electron microscopy (SEM) in the examination of another T. rex bone (MOR 1125) from the Hell Creek Formation [19]. The flexibility of the tissue found in this theropod sample was unanticipated. Schweitzer et al. propose a modelling of processes that could explain the preservation of soft-tissues in taphonomy. Secondary ion mass spectrometry (SIMS) was used two years later to confirm the endogeneity of the reported T. rex proteins [20].

In 2008, using SEM, more than 25 layers of collagenous fibres were reported in Psittacosaurus skin from China [21].

Five years later, a study of sauropodomorph embryos from the Lower Jurassic of China was published [22]. The FTIR analysis showed, besides the apatite mineral, organic presence in the form of amide groups, potentially of protein or peptide origin.

A 2017 study used four independent techniques including Raman and FTIR spectroscopy, and mass spectrometry-based protein identification, to confirm the presence of collagen in samples of modern, medieval, and ice-age bones from a variety of ages, settings, and taxa [23]. The authors proposed that Second Harmonic Generation (SHG) imaging may non-destructively estimate collagen content in partly mineralised bone.

In 2019 SEM imaging of Edmontosaurus annectens bone from the Hell Creek Formation showed what appeared to be “fibrous tissue consistent with collagenous bone matrix” [4]. The survival of this material was attributed to rapid burial in a low-oxygen, high-iron environment. The authors comment that more work is needed on different taxa from varying locations to further evaluate cellular preservation in fossils.

Another hadrosaur study (Hypacrosaurus stebingeri, 2020) reports collagen II using immunohistochemistry. As microbes are incapable of producing collagen, such contamination could be eliminated as the protein source [24]. In addition, histochemical evidence for DNA was found using two stains that specifically bind to double-stranded DNA. Propidium iodide, a widely used red-fluorescent nuclear and chromosome counterstain, is membrane-impermeable and only works on dead cells and any living cells are excluded. DAPI, a blue-fluorescent DNA counterstain, is membrane-permeable and binds to three sequential AT base pairs in DNA from both living and dead cells. Both types of staining were observed in a manner similar as they do in extant samples.

The above summarizes key findings and studies relating to the subject of proteinaceous remains in fossilised bone. Previous studies claiming original collagen from dinosaurs are listed in Table 1. Findings of protein remnants and chromosome structures in paleontological bones in the last two decades have been controversial and are oft ascribed to contamination. However, it seems that these protein discoveries are not rare, with reports (over 80, as of 2019) from different strata and geographies [25]. Some have questioned whether the process of protein degradation (particularly collagen) is fully understood [11]. Others maintain that all organic materials should long ago have been replaced by the process of permineralization and could no longer be found in-situ [26–29]. When considering these arguments, the physico-chemical fact that collagen is virtually insoluble in water should not be ignored [30]: this insolubility would be expected to delay degradation processes, whether they be biologically- or environmentally-catalysed.

ATR-FTIR [31, 32] offers an independent and accessible technique and uses near-infrared light to pass through the sample several times before being detected. Molecular vibrations absorb energy at wavenumbers specific to the type of stretching, which in turn depends on the molecular configuration. FTIR requires little sample preparation and is ideal for acquiring initial specimen information prior to more elaborate experimental methodology, such as mass spectrometry.

Cross-polarized light microscopy (XPol) has long been used to identify bone collagen and remains in use today on fossil specimens [33]. When a bone sample containing collagen is rotated approximately 90 degrees under cross polars and a first order red filter, it undergoes a characteristic colour shift from gold to blue or vice versa.

Bottom-up proteomic analysis employs enzymatic digestion (almost always with trypsin) to produce a mixture of peptides, followed by LC-MS/MS and MS/MS to produce molecular weight measurements and partial amino acid sequences. These are then used in database searching to identify proteins [34]. On the other hand, amino acid analysis uses acid digestion to break all amide bonds, followed by identification and quantification of the individual amino acids. This can be performed by conventional colorimetric methods or can be done with high sensitivity and specificity by LC-MS/MS. We use these mass spectrometric approaches to add independent, quantitative perspectives, and to corroborate ATR-FTIR and cross-polarized light microscopy. The experiments were benchmarked against a sample of modern turkey bone (Meleagris gallopavo).

Table 1

Publications of original collagen claims in dinosaur fossils. Some are briefly summarised in the introduction above (*) and below (**).
Taxon	Associated tests for collagen	GSA system	Assigned age (Ma)	Formation, geography	Year	Ref.
phalange bone of a dinosaur	Electron microscope	Cretaceous	80		1966	[16]*
dinosaur bone scraps	Analysis by gas chromatography	Cretaceous	100	V6841: Kaiparowits formation, Utah V6842: Belly River formation, Alberta, Canada V6870: Hell Creek beds, Montana V67145: Mesa Verde formation, Wyoming V67147: Lance formation, Montana	1976	[35]
bones and teeth of reptiles	GC-MS	Cretaceous	72–74	Judith River Formation, Alberta	1990	[36]
Tyrannosaurus rex (MOR 1125)	LC-MS/MS protein sequencing	Cretaceous	68	Hell Creek Formation, eastern Montana	2007	[37]**
Tyrannosaurus rex (MOR 1125)	AFM, TEM, ELISA (enzyme linked immunosorbent assay), TOF-SIMS	Cretaceous	68	Hell Creek Formation, eastern Montana	2007	[20]*
Edmontosaur sp. (MRF-03)	FTIR	Cretaceous	66	Hell Creek Formation, North Dakota	2009	[38]**
Brachylophosaurus canadensis (MOR 2598)	TOF-SIMS, LTQ-Orbitrap XL mass spec	Cretaceous, Campanian	80	Judith River Formation, eastern Montana	2009	[39]**
Ornithischia and Saurischia	SEM, TEM, ToF-SIMS	Cretaceous	75	Dinosaur Park Formation and Lance Formation	2015	[40]
Lufengosaurus (CXPM Z4644)	Synchroton Radiation FTIR	Jurassic, Sinemurian	195	Dawa, Lufeng county, Yunnan province, China	2017	[41]**
Brachylophosaurus canadensis (MOR 2598)	FT-ICR (ion cyclotron resonance) mass spec	Cretaceous, Campanian	80	Judith River Formation, eastern Montana	2017	[42]**
Edmontosaurus annectens (SRHS-DU-###)	SEM	Cretaceous	66	(Standing Rock Hadrosaur Site) Hell Creek Formation, South Dakota	2019	[4]*
Hypacrosaurus stebingeri (MOR 548)	Immunohistochemistry (collagen II)	Cretaceous, Campanian	75	Two Medicine Formation, northern Montana	2020	[24]*
Triceratops horridus (HCTC-31)	Light microscopy	Cretaceous	65	Hell Creek Formation, Montana	2021	[43]

Table 1 lists publications claiming original collagen in dinosaur bones. Entries marked ** use similar techniques as described in this paper and are briefly summarised below.

In 2007, Asara et al. sequenced sub-femtomole peptides from a Tyrannosaurus rex using LC-MS/MS, showing alpha chains from multiple organisms, including chicken, newt, and frog [37]. No dinosaur peptides were identified, possibly due to the absence of such sequences in then-existing databases.

In 2009, a study of Edmontosaurus (sp.) using FTIR spectroscopy suggested the presence of amide-containing compounds by absorption around 1,650 cm⁻¹, and pyrolysis GC-MS confirmed endogenous organics [38].

Published in the same year, Time-of-flight (ToF)-SIMS was used in the study of Brachylophosaurus canadensis fossils [39]. Hydroxyproline (Hyp, C₅H₉NO₃) was identified, a relatively rare amino acid but abundant in collagen. This result is consistent with an ancient source and not contamination, as microbes cannot produce this. Further study on the same bone confirmed earlier findings: in 2009, collagen I peptides were identified on specimen MOR 2598, and eight years later using high-resolution mass spectrometry, a further eight collagen I peptides were sequenced [42]. The authors suggested the next step is amino-acid sequencing and validation.

In 2017, Lee et al. published their evidence of preserved collagen in a Jurassic sauropod Lufengosaurus by means of synchrotron FTIR, which consisted of characteristic amide I band (carbonyl group) absorption of collagen type I [41]. No absorption was seen in the bone matrix, suggesting original organic material.

To elucidate the specific question of collagen endogeneity, different techniques were used on a mature Edmontosaurus sp. (Hadrosauridae) fossil bone (UOL GEO.1) excavated from Cretaceous sediments.

Bone samples

Herbivorous Edmontosaurus sp. (Hadrosauridae) sacrum bone fossils were excavated from the Upper Cretaceous zone of the Hell Creek Formation in Harding County, South Dakota, USA. A 22 kg sample from this duck-billed dinosaur fossil, together with samples of the sediment in which the fossil was found, was donated to and accessioned at the repository of the Victoria Gallery & Museum of the University of Liverpool under UOL GEO.1.

The modern bone (control) was from a common turkey (Meleagris gallopavo). Small bone segments were dried in an oven at 60°C for several hours in preparation for crushing (powderisation), and analysis following the same protocols as used for the Edmontosaurus samples.

Motion Photogrammetry

Motion photogrammetry procedures were used to capture a digital 3D model of the Edmontosaurus sp. bone fossils prior to sampling.

The model was reconstructed using 1061 images in Agisoft Metashape software version 1.7.2. Sony alpha ILCE-6400 cameras were used with Sony 50 mm Macro F2.8 FE Lens. Even diffuse lighting was used, and for the smaller bones light tents were used. Images were saved in both RAW and fine JPG format, and JPGs were used for the reconstruction of the models. Due to the level of granular detail in the objects, and the controlled conditions, high resolution (100% of image resolution) was used for the camera alignment process. The surfaces of the specimens were highly matte. This permitted the generation of high-resolution depth maps without concern for the potentially erroneous influence of specular highlights. Textures were also generated at high resolution (8k JPGs) using the software mosaic algorithm. As the specimens were not overly complex (i.e., there were no appendages and few occluding surfaces), only a single JPG texture file was generated for each skeletal fragment assigned to this specimen. The final models varied between 2–5 million faces and have been retained for future detailed measurements. The models were down-sampled to 64 thousand faces for online loading speeds while retaining adequate accuracy for general inspection (see Supplementary Table S1).

The samples were ground bone shards (cross-sections of 1–3 mm width, Supplementary Figure S1) prepared using a mortar and pestle. These were cleaned with powdered bicarbonate and hot-tap water (~ 50°C) before final rinsing with de-ionised water. A 50-micron stackable zooplankton sieve was used to filter the particles onto a freshly cut piece of aluminium foil for transfer into new vials, ready for use with FTIR or further sample preparation for LC-MS.

FTIR

FTIR was performed using an Attenuated Total Reflectance accessory (ATR) with a germanium window on a Bruker Vertex 70© equipped with Deuterated Lanthanum α Alanine doped TriGlycine Sulphate (DLaTGS) detector. Each spectrum combined an average of 32 scans, with a resolution of between 2–4 cm⁻¹ in the range 4,500 to 650 cm⁻¹. Spectra were collected and analysed with OPUS software and compared with authentic Ca₃PO₄ from the library (©Nicodom, 2014). Peak intensities correspond to the moiety abundance in the sample absorbing the energy at a certain frequency.

For this measurement samples were ground to powder with particle sizes of no more than 50 µm [44]. This provides a sufficient contact area between the sample and the crystal.

XPol

Thin sections of UOL GEO.1 were prepared and imaged using a Motic Polarizing Microscope BA310 with a Sony ILCE-7RM4 detector. Images from several focal planes were collected then stacked using Photoshop 24.5.

LC-MS bottom-up proteomics

Initial analyses were completed in the Materials Innovation Factory (MIF) laboratories, University of Liverpool, using high resolution LC-MS. The preliminary results led to further analyses at the Centre for Proteome Research, also at the University of Liverpool, which are reported herein.

Bone samples were ground by hand in a mortar & pestle. Twenty milligrams each of Edmontosaurus bone, turkey bone, and pure collagen (Bovine tendon collagen, lyophilized fibrous powder, Sigma-Aldrich product #5162) was dispensed into separate polypropylene microcentrifuge tubes. Each sample was treated with aqueous ammonium bicarbonate (AmBic, 80 µl, 25 mM) and RapiGest SF Surfactant solution (1% RapiGest solution in AmBic, 5 µl, Waters) with continuous gentle shaking (450 rpm, 80˚C, 10 min.). Cysteine reduction was then performed by the addition of dithiothreitol (DTT, 11.1 mg/ml in 25 mM AmBic, 5 µl). After mixing and incubation (60˚C, 10 min.) alkylation of free thiols was performed using iodoacetamide (46.6 mg/ml in 25 mM Ambic, 5 µl, 30 min. in the dark). Excess iodoacetamide was quenched with DTT (4.7 µl as above), and samples were acidified (neat trifluoroacetic acid, 2 µl) to a pH of 2 or less (checked with pH indicator paper). Digestion was carried out with trypsin (Promega sequencing grade, 0.2 µg/µl in 50 mM aqueous acetic acid) with incubation (37˚C, 16 hr). Following centrifugation (13000g, 15 min., 4˚C) the supernatants were transferred to clean microcentrifuge tubes and stored frozen until analysis.

Samples were analysed using nanobore reversed-phase chromatography (Ultimate™ 3000 RSLC, Thermo Scientific™, Hemel Hempstead) coupled to hybrid linear quadrupole/orbitrap mass spectrometry (Q Exactive™ HF Quadrupole-Orbitrap™, Thermo Scientific™) equipped with a nanospray ionization source. Samples (2 µl) were loaded onto the trapping column (Thermo Scientific, PepMap100, C18, 300 µm X 5 mm) equilibrated in aqueous formic acid (0.1%, v/v) using partial loop injection over seven minutes at a flow rate of 12 µl/min. After the direction of eluent flow was reversed components were transferred to and resolved on an analytical column (Easy-Spray C18 75 µm x 500 mm, 2 µm particle size) equilibrated in 96.2% eluent A (water/formic acid, 100/0.1, v/v) and 3.8% eluent B (acetonitrile/water/formic acid, 79.95/19.95/0.1, v/v/v) and eluted (0.3 µl/min.) with a linear increasing concentration of eluant B (min./% B; 0/3.8, 30/50). The mass spectrometer was operated in a data-dependant positive ion mode (FWHM 60,000 orbitrap full-scan, automatic gain control (AGC) set to 3e⁶ ions, maximum fill time (MFT) of 100 ms). The seven most abundant peaks per full scan were selected for high energy collisional dissociation (HCD, 30,000 FWHM resolution, AGC 1e⁵, MFT 300 ms) with an ion selection window of 2 m/z and a normalised collision energy of 30%. Ion selection excluded singularly charged ions and ions with ≥ + 6 charge state. A 60 s dynamic exclusion window was used to avoid repeated selection of the same ion for fragmentation.

Survey analyses of each sample were first used to determine the amount of each sample, calculated by extrapolation, needed to give a full orbitrap scan base peak intensity (BPI) of 1–2e⁹. These analyses were performed with a compacted 15 min. gradient (Supplementary Table S2). Based on these BPI results, 2 µl of neat Edmontosaurus and Hadrosaurus samples were used for the full analysis. The modern turkey sample was diluted 1:100 and Bovine collagen sample was diluted 1:1000 in water/acetonitrile/trifluoroacetic acid (97/3/0.1, v/v/v).

Particular attention was given to the order of sample injections to avoid carry-over from analysis to analysis. Four 30 min. blank analyses (injection solvent only) were performed after the turkey and bovine samples to minimize carry over, then the three fossilized samples were analysed on the 1 hr program.

Database searches

Analysis of the LC-MS/MS data was performed using MASCOT PD using the Swissprot, UniCow, UniTurkey and UniChick databases. The search parameters included cysteine carbamidomethylation, variable methionine oxidation, a precursor mass tolerance of 10 ppm, a product mass tolerance of 0.01 Da, and a maximum of one missed cleavage.

The data files were also imported onto Peaks (Peaks 11 Software) for further database searching. In this case the search parameters included cysteine carbamidomethylation, methionine oxidation, variable lysine and proline oxidation, a precursor mass tolerance of 10 ppm, a product mass tolerance of 0.01 Da, and a maximum of one missed cleavage. This software permits database searching for multiple post translation modifications (PTMs) as opposed to MASCOT PD that cannot be searched with more than two PTMs. The ‘SPIDER’ feature uses an algorithm that is specially designed to detect peptide mutations and performs cross-species homology search. The Edmontosaurus sample was searched against the Swissprot database, bovine collagen (96%) was searched against UniCow, and the modern turkey sample was searched against both UniChick and UniTurkey databases. The contaminants (cRAP) database was also included in each search.

LC-MS/MS of hydroxyproline

After being frozen with liquid nitrogen, both fossilized and turkey bone samples were manually crushed to a fine powder with a mortar and pestle. One-gram portions of the powdered bone or sediment were dispensed into polypropylene microcentrifuge tubes, suspended in water (1 ml), mixed vigorously, sonicated in a bath sonicator (30 min.), centrifuged (2000g, 15 min.), and the supernatants transferred to new tubes. The extraction procedure was repeated on the pellet by adding methanol (1 ml) and the samples were mixed, sonicated, and centrifuged as above. The supernatants were pooled and reserved for future bottom-up proteomics. The pellets were then treated with HCl (2 ml, 6 N) and incubated (2 hrs, 60°C) before the samples were dried in a vacuum centrifuge. The HCl treatment was repeated until the samples ceased effervescing after HCl addition, each time with drying in a vacuum centrifuge between acid treatments. Generally, it took two or three such treatments before effervescing ceased. The final dried samples were treated again with HCl (500 µl, 6 N) and incubated (12 hrs, 120°C) to effect protein hydrolysis. The samples were dried overnight in a vacuum centrifuge and then treated with n-butanolic HCl (300 µl, 3 N), incubated (2 hrs, 60°C) to make the butyl esters, and dried again in a vacuum centrifuge. Lastly, the samples were reconstituted in water (200 µl), mixed vigorously and centrifuged (5 min., 16,000g, room temperature). The supernatants were transferred to HPLC vials and aliquots (typically 10 µl), injected onto a reversed-phase HPLC column (Phenomenex Kinetex, 2.6 µm Polar C18, 100 Å, 100 x 2.1 mm), equilibrated in eluant A, and eluted (100 µl/min) with a step-wise linearly increasing concentration of eluant C (acetonitrile/formic acid, 100/0.1, v/v; min/%C, 0/1, 5/1, 20/25, 22/1, 60/1). The effluent from the column was passed through an electrospray ionization source connected to a hybrid linear ion trap/orbitrap mass spectrometer (Thermo Scientific Orbitrap LTQ XL) scanning either in the positive linear ion trap (for quantitative measurements) or the positive orbitrap (for accurate m/z measurements) modes.

For quantitative measurements, the mass spectrometer was set to fragment preselected parent ions under standard MS/MS fragmentation conditions for the butyl esters of hydroxyproline (Hyp_be, MH⁺ at m/z 188). For the accurate m/z measurements the orbitrap was scanned (FWHM of 100,000 at m/z 400) immediately after calibration with LTQ ESI Positive Ion calibration solution mixture. Data were collected and interrogated with instrument manufacturer-supplied software (Xcalibur 2.05).

Several control samples were included with each batch of bone samples. These were negative control samples devoid of added bone extracts (in triplicate), sediment in which the fossil was sitting at the excavation site (20 mg/sample, in triplicate), pure collagen (Sigma, 20 mg/sample, in triplicate), and authentic Hyp standard in a range of amounts (typically 0, 2, 10, 20 and 50 nmoles/sample, in duplicate). These samples were prepared and processed with each batch of bone samples.

The order in which samples were analysed was carefully arranged. Injections of water (solvent blanks) were used at the start of the analysis of each batch of samples to check that there were no peaks for Hyp_be resulting from carry-over from the analysis of previous sample batches. After verification that the LC/MS system was clean a typical order of sample analysis was: negative control samples 1–3; water blank #1; sediment samples 1–3, water blank #2, Edmontosaurus fossilized bone samples 1–3, water blanks #3 and #4, turkey bone samples 1–3, water blanks #5 and #6, collagen samples 1–3, water blanks #7 and #8, Hyp_be standards 1–10, water blanks #9–#11.

The data from the standard Hyp_be samples were processed by plotting the known amount of Hyp per sample against the measured chromatographic peak areas corresponding to the Hyp_be peak. The trendline equation was then used to interpolate or extrapolate the amount of Hyp in each sample.

The fossilized Edmontosaurus sp. sacrum (UOL GE0.1) was excavated in 2019. After initial cleaning, bone material weighed in total approximately 20 kg and the main fragment was intact (Fig. 1), needing little stabilising. The 3D model (Supplementary Table S1) allows detailed inspection of the surface topography of the bone aiding surface identification of post-depositional breaks and geometric measurements.

FTIR

The inorganic component of the assemblage of bone (bioapatite) is largely composed of calcium phosphate (Ca₃PO₄) [32]. Antisymmetric stretching of PO₄ occurs between approximate wavenumbers 1,000–1,100 cm‾¹, depending on the dipole change of the moiety [45, 46]. As expected, the FTIR spectra recorded for Edmontosaurus fossilized bone, turkey bone, and inorganic calcium phosphate all show a strong absorption around wavenumber 1,050 cm‾¹ (Fig. 2). The organic component of fresh bones comprises mostly type I collagen. An FTIR spectrum of collagen will show a peak for amide I group (carbonyl, C = O) absorption around 1,650 cm‾¹ [33]. The peaks visible at 1,652 cm⁻¹ in the bone spectra (Fig. 2) indicated carbonyl groups, likely from protein or peptide remnants [45]. This peak was not evident in the spectrum obtained from calcium phosphate. The peak ratio for carbonyl over phosphate (indicated herein by “CO/P”) is used as a proxy for collagen presence and abundance [47, 48].

CO/P was calculated to be 0.065 (0.011/0.168) for UOL GE0.1, and 0.455 for the turkey sample, indicating a reduced organic component relative to bioapatite in the dinosaur sample.

XPol

XPol revealed permineralization within microscopic regions of UOL GEO.1 bone tissue (Fig. 3). One such inclusion, likely calcite, appears in the lower right corner of Fig. 3C as a hard-edged angular green shape without lacunae when the sample was viewed with a 1st order red filter. In contrast, bone tissue has lacunae, striations, irregular outlines, and more diffused edges than precipitated minerals. It appears mostly dark red or almost black under crossed polars (Fig. 3B). With a first order red filter, XPol revealed gold-coloured regions (Fig. 3C inset and Fig. 3D) that turned blue-green when rotated 90⁰. This birefringence characterizes collagen that pervades fresh bone but occurs only in patches within the fossil. We interpret the regions within bone that have no birefringence as regions where collagen has completely decayed, given its age.

To evaluate this interpretation, similar birefringence was observed in artificially decayed turkey bone under XPol with a first order red filter (Fig. 4). No permineralization was observed, as expected. In this case almost-black areas remain dark after rotating the sample at an approximate right angle (105⁰). However, still-collagenated micro-regions traded gold for blue and vice versa upon rotation, similar to the fossil. These results are consistent with the presence of collagen remnants in the dinosaur bone and suggest that permineralization does not result in complete decomposition of bone collagen, at least within the time frame of UOL GEO.1 preservation.

LC-MS bottom-up proteomics

Despite the use of compacted 15-minute survey analyses to determine the amount of each sample required to obtain comparable signal intensities, the base peak intensity (BPI, and Total ion intensity (data not shown)) 30-minuted chromatograms of the dinosaur and turkey bone samples were relatively sparsely populated with peaks (Fig. 5A and Fig. 5B). For reasons that remain unexplained this was in stark contrast to what was expected and what is usually obtained from bottom-up proteomic analyses of tissue samples, and what was obtained from authentic bovine collagen (Fig. 5C).

Nevertheless, careful analysis of the data sets revealed six collagen-derived peptides in the Edmontosaurus sample, with errors ranging from 1.3–3.6 ppm, and oxidized at a minimum of one position (Table 2). All these sequences correspond to peptides reported in the Swissprot database for Brachylophosaurus canadensis, another duck-billed dinosaur that together with Edmontosaurus is classed in the Hadrosauridae family. Five of the sequences represent the amino acids from the collagen alpha-1(I) chain, in total accounting for covering 73.45% of the entire sequence (positions 19–33 and 64–78 absent). The remaining detected peptide (m/z 805.38074) accounts for 50% coverage of the collagen alpha-2(I) chain, with positions 19–36 unaccounted.

Table 2

Edmontosaurus (UOL GEO.1) peptides detected from alpha-1 and alpha-2 Type I collagen. Masses were calculated using ExPASy PeptideMass and configured as follows: Cysteine residues treated with iodoacetamide, methionine residues oxidized, monoisotopic mass, allowance for a maximum of one missed tryptic cleavage, mass between 500–unlimited Dalton, z equal to 1 + or 2 + selected.
Collagen ions	Data search match	ppm error	Protein	Assigned sequence	Residue assignment
786.8897	786.8918	-2.7	P86289\|CO1A1_BRACN	GATGAPGIAGAPGFPGAR	1–18
766.87427	766.8762	-2.5	P86289\|CO1A1_BRACN	GETGPAGPAGPPGPAGAR	96–113
795.90826	795.9097	-1.8	P86289\|CO1A1_BRACN	GLTGPIGPPGPAGAPGDK	46–63
730.34717	730.3498	-3.6	P86289\|CO1A1_BRACN	GSAGPPGATGFPGAAGR	79–95
581.80048	581.8018	-2.3	P86289\|CO1A1_BRACN	GVQGPPGPQGPR	34–45
805.38074	805.3818	-1.3	P86290\|CO1A2_BRACN	GSNGEPGSAGPPGPAGLR	1–18

A total of 41 discovered collagen sequences from UOL GEO.1 (Supplementary table S3) matched a mixture of extant and extinct fauna found in the Swissprot database, and the most abundant of these matched peptides were from domestic chicken (Gallus gallus), followed by American mastodon (Mammut americanum).

In addition, peptides from other proteins were detected in the Edmontosaurus sample. Of the most abundant 17, 13 matched keratin and of these a dozen matched Homo sapiens keratin. These were likely laboratory contaminants as they routinely appear in analyses of this type conducted in a standard laboratory without use of a laminar-flow hood during sample preparation. Over a hundred actin peptides were found. Furthermore, 61 hemoglobin, 158 histone, and 92 tubulin peptides were also detected.

The turkey bone data checked against the UniChick database (Supplementary table S5) gave 55.33% coverage of collagen alpha-1(I) chain and 74.69% coverage of the alpha-2(I) chain. Surprisingly, the UniTurkey database provided no matches for type I collagen (Supplementary table S6).

No Brachylophosaurus peptides were found in the turkey bone data set (compared against UniChick and UniTurkey databases), but six were discovered in a femur fossil of another Hadrosaur (GDFM 04.001) that matched the Swissprot database, with a coverage of 58.41% of the collagen alpha-1(I) chain (Supplementary table S4).

185 of the 188 peptides found in bovine tendon (96% pure) collagen matched that of Bos taurus, accounting for 98.40% of the amino acid sequence. The remaining three were marked as contamination (Supplementary table S7).

LC-MS/MS of hydroxyproline

Under ESI conditions, Hyp_be (butyl ester of hydroxyproline) produces an intense parent ion at m/z 188 (MH⁺), which upon fragmentation under the specified MS/MS conditions, yields fragment ions at m/z 132 and 86. The former likely results from loss of the butyl side chain (C₄H₈) with a proton transfer to the CO group: thus, (C₅H₉NO₃⁺H)⁺ would be the corresponding elemental composition. Further neutral loss of 46 u (H₂CO₂) through proton transfer and cleavage of the acid group would create the m/z 86 fragment (C₄H₈NO⁺).

Following injection of the derivatized bone samples, reconstructed ion traces for m/z 188–132 and 188–86 transitions consistently showed significant peaks at or near the retention of authentic Hyp_be (Fig. 6, top two traces). Despite all precautions taken, the negative controls always revealed a peak for Hyp. However, the intensity of this peak was significantly less than what was obtained for the bone samples (Table 3).

Table 3

Average concentration of hydroxyproline in 7 independent experiments (nmoles/sample).
Experiment	Negative control	Geological sediment	Hadrosaurus bone	Edmontosaurus bone	Turkey bone	Bovine collagen
1	0.12	-	2.50	2.71	537.90	151.81
2	1.41	-	6.83	9.25	197.07	66.38
3	1.18	1.52	9.02	6.18	145.52	87.89
4	0.08	2.13	4.35	0.03	434.03	519.14
5	0.95	0.12	-	1.99	232.28	87.89
6	0.06	-	-	0.30	-	-
7	0.03	-	-	0.14	-	-
Mean	0.55	1.26	5.67	2.94	309.36	182.62
Stdev	0.56	1.03	2.85	3.53	168.20	190.83
n	7	3	4	7	5	5

A co-chromatography experiment was performed because the Hyp analyses were marked by a greater than usual variation in retention times. After analysing the individual samples, the dinosaur bone and authentic Hyp_be samples were mixed, and the result from the mixed sample revealed a single sharp peak. This co-chromatography experiment showed the material in dinosaur bone was chromatographically indistinguishable from Hyp_be.

Furthermore, the ratios of the peak areas for the two transitions (188–132, 188–86) from authentic Hyp_be and the peak from derivatized bone extract were the same (Table 4).

Table 4

Ratio of the peak areas for the two transitions (188–132/188–86).
	Authentic Hyp_be (n = 10)	Assigned Hyp_be peak from bone samples (n = 6)
Mean	1.20	1.21
Standard Deviation	0.02	0.06

Further testing of the validity of the assertion that the chromatographic peak assigned as Hyp_be in the fossilized samples came from accurate mass measurements completed with the orbitrap as the chromatographic detector. In this investigation, the measured mass of the parent ion from the fossilized bone sample was recorded at m/z 188.12802 that is within 0.5 ppm of the theoretical calculated value (calculated for C₉H₁₇NO₃⁺H⁺, 188.12812), and this lies within the tolerance of the instrument.

Finally, the Hyp_be concentration in the fossilized bone samples was estimated by interpolating peak areas for the 188–132 transition using a standard curve constructed from a series of samples of decreasing Hyp_be concentration. Six separate 1 mg samples of bone taken from the same bone specimen had Hyp_be concentrations ranging from 6.7 to 41.7 nmoles of Hyp_be/gram of bone (Table 5), a result reflecting an uneven distribution of collagen within the fossilized sample. As noted below, XPol imaging also revealed an uneven distribution of bone collagen within microscopic fossil bone regions.

Table 5

Hydroxyproline (Hyp_be) concentration in dinosaur bone calculated based on transition 188–132.
Sample ID	nanomoles Hyp/gram of bone
1	41.7
2	19.1
3	6.7
4	8.9
5	15.4
6	8.6

Virtually incontrovertible data is presented from numerous experiments attesting to the presence of collagen in Edmontosaurus sp. fossil bone. The bottom-up proteomics results also revealed other proteins in the sample, although this point is not addressed in detail here. The experiments set a scheme that may be followed by others for similar investigation of other fossilized bones. Such investigations will lay essential groundwork for using the data to ascertain the state of bone decay that may ultimately provide burial information, disease status, and an estimate of the age of the fossil(s) in question.

Bone and Photogrammetry

Decomposition of bone organics is expected to relate to time elapsed since mortality, the temperature history, the oxidizing or reducing strength of the surrounding milieu, and other degradative factors such as radiation, acidity/alkalinity, salinity and corrosive environment, and especially microbial activity. The older the bone sample, the less intact will be the organic molecules [49, 50]. The mineral portion of bone is largely comprised of calcium phosphate. Bone tissue integrity overall fails rapidly once the collagen matrix decays and exposes the remaining hydroxyapatite to the environment [51]. However, the photogrammetric analysis showed that UOL GEO.1 had residual integrity.

Two diagenetic modes were therefore considered. In one, some organic components were reinforced by permineralization, whereby exogenous minerals precipitated within the bone pore spaces. In the other, original bone biomineral was replaced (mineralization) to retain only the original morphology, but little of the original biochemistry. ATR-FTIR was used as an initial test between these two diagenetic modes, with the expectation that a carbonyl peak would contraindicate the second model; that of complete replacement via mineralization [23].

FTIR

The organic component (C = O; herein “CO”) is found around wavenumber 1,650 cm‾¹, and the phosphate vibrations representing apatite are found between 960–1,100 cm‾¹ [33, 45, 48, 52]. A peak intensity ratio of CO/P, with wavenumbers around ^{1,650 cm⁻¹}/_{1,035 cm⁻¹}, was chosen to counter for any baseline shift; numerator and moderator are similarly affected and baseline shifts would be largely annulled. In completely mineralised bone, the CO peak would be indistinguishable from the baseline absorbance. Here, the CO peak above baseline with a CO/P ratio of 0.168 suggested residual organic material.

XPol

The data revealed microscopic regions within UOL GEO.1 that retain the characteristic birefringence of collagen. These regions were surrounded by dark and purple areas where collagen has apparently decayed. Other brightly coloured spots within pores or cracks are assigned as permineralized calcite crystals. Like those from FTIR, these results are consistent with the first mode of diagenesis.

For comparison, XPol images were also collected from artificially decayed Meleagris. The results revealed a similar birefringence in microscopic regions surrounded by apparently non-collagenated dark purple areas. No permineralization was evident. This was expected since our artificial decay did not include exposure of the bones to sediments or hydrothermal simulation. The similar birefringent patterns between artificially and actually decayed bone are consistent with endogenous collagen in UOL GEO.1.

LC-MS bottom-up proteomics

Since the 1990s, mass spectrometry is the preferred method for protein identification [54]. Providing unparalleled sensitivity and specificity, not only is this useful for the diagenesis of proteins, but the data can also ultimately be used to develop bone regeneration models and potentially might elucidate disease processes [51]. Theropod and hadrosaur collagen sequences have been published and offer comparisons for new species [39, 42].

Our results indicate that past conditions allowed retention of some endogenous collagen. Collagen fragments (41) have been discovered from the UOL GEO.1 Edmontosaurus sample (Supplementary table S3).

The focus of the results in this paper is on the peptides assigned to Brachylophosaurus alpha-1 and alpha-2 helices. As these peptides, thus far, are unique to this hadrosaur, they can be regarded as original tissue instead of contamination from any extant creature. It is probable that similar taxa (Brachylophosaurus and Edmontosaurus) had many proteins in common.

LC-MS/MS of hydroxyproline

Hydroxyproline (Hyp) is found—but uncommon—in few proteins other than collagen [55], but comprises somewhere between 4–10% of collagen residues in extant vertebrates. Because collagen is by far the most abundant protein in bone tissue, Hyp was targeted as an indicator of collagen. Our results unequivocally verify the presence of Hyp in acid-digested samples.

Half a century of reports that show organics including proteins in fossils suggest that they harbour molecular clues for phylogenetic and taphonomic reconstructions. However, controversy remains over the endogeneity of Mesozoic fossil organics. The aim of the present research was to examine a not-previously studied upper Cretaceous hadrosaurid Edmontosaurus sp. bone (UOL GEO.1) using spectroscopic, microscopic, and proteomic analyses as independent techniques to evaluate such claims.

Previous studies using FTIR of Jurassic [41] and Cretaceous [38] dinosaurs have shown the organic amide I group around the absorption band 1,650 cm⁻¹. The vibration of the carbonyl moiety indicates collagen presence in the sample. Here we confirm these results by showing the same in UOL GEO.1.

XPol revealed the birefringence expected from collagenated bone remnants in both fossil and artificially decayed bone tissues. The presence of protein distributed throughout the sample is an indication of endogeneity. This result suggests XPol microscopy as a direct means to detect endogenous bone collagen remnants, as opposed to indirect chemical methods. It also offers the potential to describe microscopic collagen decay patterns within both artificially decayed and actually old bone tissues.

In the past, mass spectrometry has indicated the presence of organic materials in T. rex [37] and twice in B. canadensis [39, 42]. The T. rex results yielded alpha chains, which could not at the time be referenced against a database with dinosaurian peptides. The Brachylophosaurus data also yielded peptide chains, eight in 2009 and a further eight in 2017. The 2009 study had already shown the presence of Hyp. This paper confirms these previous conclusions by sharing similar results in Edmontosaurus. Our bottom-up proteomic analyses revealed a total of 41 collagen polypeptides. Five of those assigned to Brachylophosaurus canadensis collagen alpha-1 (I) helix; one belongs to the collagen alpha-2 (I) helix of the same, totalling six sequences that have previously only been found in one other hadrosaurid.

For the first time we quantify the amount of collagen by singling out authentic Hyp, an abundant collagen amino acid in samples of known provenance. This otherwise unusual amino acid constitutes 9.6, 7.8, and 4.0 residue percent of collagen from rat, bovine, and codfish, respectively [56].

Overall, results from four independent techniques, including the first quantification of hydroxyproline in Mesozoic fossil bone, support the idea that protein remnants in dinosaur bones are original (endogenous) to the fossils. Further research using controlled experiments is encouraged to elucidate the diagenetic processes to which vertebrate skeletal structures are subjected.

Data availability statement

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Acknowledgments

Dr Ardern Hulme-Beaman for (3D) imaging.

The Centre for Proteome Research (specifically Dr Katie Kennedy) in assisting with the bottom-up proteomics and sequencing.

This work made use of shared equipment located at the Materials Innovation Factory; created as part of the UK Research Partnership Innovation Fund (Research England) and co-funded by the Sir Henry Royce Institute.

Author contributions

LT, BT and ST designed the research. ST had the Edmontosaurus sample accessioned. LT, BT, SR, KP, and GE performed the research. LT, BT, SR, KP, GE, KF, and ST analysed the data. LT, BT, GE, KF, and ST wrote the paper.

Additional Information

Supplementary Information: The online version contains supplementary material available at [DOI here]

The authors declare no competing interests.

Huizen, N. A., Ijzermans, J. N. M., Burgers, P. C. & Luider, T. M. Collagen analysis with mass spectrometry. Mass Spectrometry Reviews 39, 309 (2020).
Wolfenden, R. & Snider, M. J. 938 (ACS AMERICAN CHEMICAL SOCIETY, United States, 2001).
Shoulders, M. D. & Raines, R. T. Collagen structure and stability. Annu Rev Biochem 78, 929-958, doi:10.1146/annurev.biochem.77.032207.120833 (2009).
Ullmann, P. V., Pandya, S. H. & Nellermoe, R. Patterns of soft tissue and cellular preservation in relation to fossil bone tissue structure and overburden depth at the Standing Rock Hadrosaur Site, Maastrichtian Hell Creek Formation, South Dakota, USA. Cretaceous Research 99, 1-13, doi:10.1016/j.cretres.2019.02.012 (2019).
Kontopoulos, I. et al. Petrous bone diagenesis; a multi-analytical approach. Palaeogeography, Palaeoclimatology, Palaeoecology 518, 143-154, doi:10.1016/j.palaeo.2019.01.005 (2019).
Madden, O., Chan, D. M. W., Dundon, M. & France, C. A. M. Quantifying collagen quality in archaeological bone: Improving data accuracy with benchtop and handheld Raman spectrometers. Journal of Archaeological Science: Reports 18, 596-605, doi:10.1016/j.jasrep.2017.11.034 (2018).
Delannoy, Y. et al. Characterization of bone diagenesis by histology in forensic contexts: a human taphonomic study. International Journal of Legal Medicine 132, 219-227, doi:10.1007/s00414-017-1699-y (2018).
Matthew et al. The half-life of DNA in bone: measuring decay kinetics in 158 dated fossils. Proceedings of the Royal Society B: Biological Sciences 279, 4724-4733 (2012).
Sawafuji, R. et al. Proteomic profiling of archaeological human bone. Royal Society Open Science 4, 161004, doi:doi:10.1098/rsos.161004 (2017).
Sinex, F. M. Aging and the Lability of Irreplaceable Molecules—II. The Amide Groups of Collagen. Journal of Gerontology 15, 15-18, doi:10.1093/geronj/15.1.15 (1960).
Collins, M. J., Riley, M. S., Child, A. M. & Turner-Walker, G. A Basic Mathematical Simulation of the Chemical Degradation of Ancient Collagen. Journal of Archaeological Science 22, 175-183, doi:10.1006/jasc.1995.0019 (1995).
Naito, Y. I., Yamane, M. & Kitagawa, H. Vol. 34 (2020).
Herrando-Pérez, S. Bone need not remain an elephant in the room for radiocarbon dating. Royal Society Open Science 8, 201351, doi:10.1098/rsos.201351.
Tomassetti, M., Marini, F., Bucci, R. & Campanella, L. A survey on innovative dating methods in archaeometry with focus on fossil bones. TrAC Trends in Analytical Chemistry 79, 371-379, doi:https://doi.org/10.1016/j.trac.2015.11.017 (2016).
Cersoy, S., Daheur, G., Zazzo, A., Zirah, S. & Sablier, M. Pyrolysis comprehensive gas chromatography and mass spectrometry: A new tool to assess the purity of ancient collagen prior to radiocarbon dating. Analytica Chimica Acta 1041, 131-145, doi:10.1016/j.aca.2018.07.048 (2018).
Pawlicki, R., Korbel, A. & Kubiak, H. Cells, Collagen Fibrils and Vessels in Dinosaur Bone. Nature 211, 655-657, doi:10.1038/211655a0 (1966).
Schweitzer, M. H. & Horner, J. R. Intravascular microstructures in trabecular bone tissues of Tyrannosaurus rex. Annales de Paleontologie 85, 179-192, doi:10.1016/S0753-3969(99)80013-5 (1999).
Schweitzer, M. H. BLOOD FROM STONE. Scientific American 303, 62-69, doi:10.1038/scientificamerican1210-62 (2010).
Schweitzer, M. H., Wittmeyer, J. L., Horner, J. R. & Toporski, J. K. Soft-Tissue Vessels and Cellular Preservation in Tyrannosaurus rex. Science 307, 1952-1955, doi:10.1126/science.1108397 (2005).
Mary Higby, S. et al. Analyses of Soft Tissue from Tyrannosaurus rex Suggest the Presence of Protein. Science 316, 277, doi:10.1126/science.1138709 (2007).
Lingham-Soliar, T. A unique cross section through the skin of the dinosaur Psittacosaurus from China showing a complex fibre architecture. Proc Biol Sci 275, 775-780, doi:10.1098/rspb.2007.1342 (2008).
Reisz, R. R. et al. Embryology of Early Jurassic dinosaur from China with evidence of preserved organic remains. Nature (London) 496, 210-214, doi:10.1038/nature11978 (2013).
Thomas, B. et al. Second-harmonic generation imaging of collagen in ancient bone. Bone Reports 7, 137-144, doi:https://doi.org/10.1016/j.bonr.2017.10.005 (2017).
Bailleul, A. M. et al. Evidence of proteins, chromosomes and chemical markers of DNA in exceptionally preserved dinosaur cartilage. National Science Review, doi:10.1093/nsr/nwz206 (2020).
Thomas, B. & Taylor, S. Proteomes of the past: the pursuit of proteins in paleontology. Expert Review of Proteomics 16, 881-895, doi:10.1080/14789450.2019.1700114 (2019).
van der Valk, T. et al. Million-year-old DNA sheds light on the genomic history of mammoths. Nature 591, 265-269, doi:10.1038/s41586-021-03224-9 (2021).
Wadsworth, C. & Buckley, M. 605 (John Wiley & Sons, Ltd, Great Britain, 2014).
Buckley, M. et al. Comment on 'Protein sequences from mastodon and Tyrannosaurus rex revealed by mass spectrometry'. (2008).
Shapiro, B. & Hofreiter, M. Ancient DNA. methods and protocols. (Humana Press, 2012).
Sionkowska, A., Skrzyński, S., Śmiechowski, K. & Kołodziejczak, A. The review of versatile application of collagen. Polymers for Advanced Technologies 28, 4-9, doi:https://doi.org/10.1002/pat.3842 (2017).
Lebon, M., Gallet, X., Reiche, I., Bellot-Gurlet, L. & Zazzo, A. Rapid quantification of bone collagen content by ATR-FTIR spectroscopy. Radiocarbon 58, 131-145, doi:10.1017/RDC.2015.11 (2016).
Aufort, J. et al. Macroscopic electrostatic effects in ATR-FTIR spectra of modern and archeological bones. American Mineralogist 103, 326-329, doi:10.2138/am-2018-6320CCBYNCND (2018).
de Campos Vidal, B. & Mello, M. L. S. Collagen type I amide I band infrared spectroscopy. Micron 42, 283-289, doi:https://doi.org/10.1016/j.micron.2010.09.010 (2011).
Saitta, E. T. et al. NON-AVIAN DINOSAUR EGGSHELL CALCITE CONTAINS ANCIENT, ENDOGENOUS AMINO ACIDS. bioRxiv, 2020.2006.2002.129999, doi:10.1101/2020.06.02.129999 (2020).
Wyckoff, R. W. G. & Davidson, F. D. Pleistocene and dinosaur gelatins. Comparative Biochemistry and Physiology Part B: Comparative Biochemistry 55, 95-97, doi:https://doi.org/10.1016/0305-0491(76)90179-6 (1976).
Ostrom, P. H., Macko, S. A., Engel, M. H., Silfer, J. A. & Russell, D. Geochemical characterization of high molecular weight material isolated from late cretaceous fossils. Organic Geochemistry 16, 1139-1144, doi:https://doi.org/10.1016/0146-6380(90)90149-T (1990).
John, M. A., Mary, H. S., Lisa, M. F., Matthew, P. & Lewis, C. C. Protein Sequences from Mastodon and Tyrannosaurus Rex Revealed by Mass Spectrometry. Science 316, 280, doi:10.1126/science.1137614 (2007).
Manning, P. L. et al. Mineralized soft-tissue structure and chemistry in a mummified hadrosaur from the Hell Creek Formation, North Dakota (USA). Proceedings of the Royal Society B: Biological Sciences 276, 3429-3437, doi:10.1098/rspb.2009.0812 (2009).
Schweitzer, M., H. et al. Biomolecular Characterization and Protein Sequences of the Campanian Hadrosaur B. canadensis. Science 324, 626, doi:10.1126/science.1165069 (2009).
Bertazzo, S. et al. Fibres and cellular structures preserved in 75-million–year-old dinosaur specimens. Nature Communications 6, 7352, doi:10.1038/ncomms8352 (2015).
Lee, Y.-C. et al. Evidence of preserved collagen in an Early Jurassic sauropodomorph dinosaur revealed by synchrotron FTIR microspectroscopy. Nature Communications 8, 14220, doi:10.1038/ncomms14220 (2017).
Schroeter, E. R. et al. Expansion for the Brachylophosaurus canadensis Collagen I Sequence and Additional Evidence of the Preservation of Cretaceous Protein. Journal of Proteome Research 16, 920-932 (2017).
Armitage, M. H. First Report of Peripheral Nerves in Bone from Triceratops horridus Occipital Condyle. Microscopy Today 29, 20-25 (2021).
Kontopoulos, I., Presslee, S., Penkman, K. & Collins, M. J. Preparation of bone powder for FTIR-ATR analysis: The particle size effect. Vibrational Spectroscopy 99, 167-177, doi:https://doi.org/10.1016/j.vibspec.2018.09.004 (2018).
Thompson, T. J. U., Islam, M. & Bonniere, M. A new statistical approach for determining the crystallinity of heat-altered bone mineral from FTIR spectra. Journal of Archaeological Science 40, 416-422, doi:10.1016/j.jas.2012.07.008 (2013).
Moore, E. Fourier Transform Infrared Spectroscopy (FTIR) : Methods, Analysis, and Research Insights. (Nova Science Publishers, Inc, 2016).
Thompson, T. J. U., Gauthier, M. & Islam, M. The application of a new method of Fourier Transform Infrared Spectroscopy to the analysis of burned bone. Journal of Archaeological Science 36, 910-914, doi:https://doi.org/10.1016/j.jas.2008.11.013 (2009).
Paschalis, E. P., Mendelsohn, R. & Boskey, A. L. Infrared Assessment of Bone Quality A Review : Bone Quality: Forum Bench to Bedside. Clinical orthopaedics and related research 469, 2170-2178 (2011).
Keenan, S. W. From bone to fossil: A review of the diagenesis of bioapatite. American Mineralogist 101, 1943-1951, doi:10.2138/am-2016-5737 (2016).
Tahoun, M., Engeser, M., Namasivayam, V., Sander, P. M. & Müller, C. E. Chemistry and Analysis of Organic Compounds in Dinosaurs. Biology 11, doi:10.3390/biology11050670 (2022).
San Antonio, J. D. et al. Dinosaur Peptides Suggest Mechanisms of Protein Survival. PLOS ONE 6, e20381, doi:10.1371/journal.pone.0020381 (2011).
Pestle, W. J., Ahmad, F., Vesper, B. J., Cordell, G. A. & Colvard, M. D. Ancient bone collagen assessment by hand-held vibrational spectroscopy. Journal of Archaeological Science 42, 381-389, doi:10.1016/j.jas.2013.11.014 (2014).
Simon, H. J. et al. Vol. 141 157-165 (Royal Society of Chemistry, Great Britain, 2016).
Steen, H. & Mann, M. The abc's (and xyz's) of peptide sequencing. Nature Reviews Molecular Cell Biology 5, 699-711, doi:10.1038/nrm1468 (2004).
Gorres, K. L. & Raines, R. T. Prolyl 4-hydroxylase. Critical Reviews in Biochemistry and Molecular Biology 45, 106-124, doi:10.3109/10409231003627991 (2010).
Carvalho, A. M., Marques, A. P., Silva, T. H. & Reis, R. L. Evaluation of the Potential of Collagen from Codfish Skin as a Biomaterial for Biomedical Applications. Marine drugs 16, 495, doi:10.3390/md16120495 (2018).

FTIRXPolandLCMSconfirmendogenouscollageninEdmontosaurusfossilboneSupplementaryInformation.docx

Download PDF

Version 1

posted

You are reading this older preprint version

Read the latest preprint version →

FTIR, XPol, and LC-MS confirm endogenous collagen in Edmontosaurus fossil bone

Status:

Version 1

Abstract

Figures

Introduction

Methods

FTIR

Results

FTIR

Discussion

FTIR

Conclusion

Declarations

References

Supplementary Files

Status:

Version 1