on August 9, 2018http://rspb.royalsocietypublishing.org/Downloaded from rspb.royalsocietypublishing.orgResearch
Cite this article: Cleland TP, Schroeter ER,
Schweitzer MH. 2015 Biologically and
diagenetically derived peptide modifications in
moa collagens. Proc. R. Soc. B 282: 20150015.
http://dx.doi.org/10.1098/rspb.2015.0015Received: 4 January 2015
Accepted: 20 April 2015Subject Areas:
biochemistry, palaeontology
Keywords:
moa, collagen, diagenesis,
post-translational modificationsAuthor for correspondence:
Timothy P. Cleland
e-mail: clelat@rpi.eduElectronic supplementary material is available
at http://dx.doi.org/10.1098/rspb.2015.0015 or
via http://rspb.royalsocietypublishing.org.& 2015 The Author(s) Published by the Royal Society. All rights reserved.Biologically and diagenetically derived
peptide modifications in moa collagens
Timothy P. Cleland1, Elena R. Schroeter2 and Mary H. Schweitzer2,3
1Department of Biomedical Engineering, Rensselaer Polytechnic Institute, Troy, NY 12182, USA
2Department of Biological Sciences, North Carolina State University, Raleigh, NC 27695, USA
3North Carolina Museum of Natural Sciences, Raleigh, NC 27601, USA
Themodifications that occur on proteins in natural environments over time are
notwell studied, yet characterizing them is vital to correctly interpret sequence
data recovered from fossils. The recently extinct moa (Dinornithidae) is
an excellent candidate for investigating the preservation of proteins, their
post-translational modifications (PTMs) and diagenetic alterations during
degradation. Moa protein extracts were analysed using mass spectrometry,
and peptides from collagen I, collagen II and collagen V were identified. We
also identified biologically derived PTMs (i.e. methylation, di-methylation,
alkylation, hydroxylation, fucosylation) on amino acids at locations consistent
with extant proteins. In addition to these in vivo modifications, we detected
novel modifications that are probably diagenetically derived. These include
loss of hydroxylation/glutamic semialdehyde, carboxymethyllysine and
peptide backbone cleavage, as well as previously noted deamidation.Moa col-
lagen sequences and modifications provide a baseline by which to evaluate
proteomic studies of other fossils, and a framework for defining the molecular
relationship of moa to other closely related taxa.1. Introduction
The first investigations into biomolecular preservation in fossils focused on pro-
teins [1–5]; however, polymerase chain reaction (PCR) and other technological
advances [6–8] resulted in genomic studies superseding those of proteins.
Now, recent advances in high-resolution mass spectrometry have resulted in a
renewed interest in the utility of proteins preserved in fossils (reviewed in [9]).
Ancient proteomic studies have extended the age for which biomolecules and
phylogenetically informative molecular sequences can be recovered to well
beyond the hypothesized limit for DNA preservation [8,10–17].
Proteomic studies on sequences from fossils directly characterize biologically
and/or diagenetically derived post-translational modifications (PTMs). Because
biologically derived PTMs originate from the organism itself, they inform on
the ultimate protein function, phylogenetic changes or evolutionary adaptations
at the molecular level that cannot be determined from DNA sequences alone [9].
In contrast to phylogenetic or physiological results of in vivo PTMs, diagenetically
derived PTMs are the result of post-mortem decay. These modifications provide
evidence that detected proteins are original to the fossil, and inform how pro-
teins/amino acids degrade over geological time [11]. Historically, few
biologically derived PTMs have been identified for extinct species (e.g. hydroxy-
lation of proline, HYP; carboxylation of glutamic acid [16]) and even fewer
diagenetically derived PTMs have been identified [11,13,14]. Although deamida-
tion is the most common diagenetically derived PTM directly identified on
peptides, non-enzymatic glycation has been hypothesized to be a major factor
affecting preservation of ancient proteins because it results in highly cross-
linked protein structures reducing protein solubility [17]. However, it has not
been directly measured or detected in fossil taxa. Because few PTMs from ancient
remains have been elucidated, information regardingwhich PTMs persist into the
rock record and what types of modifications can occur diagenetically is limited.
Here, we investigate moa bone (MOR OST-255) for the preservation of
protein and PTMs. Moa remains are of special interest because in addition to
rspb.royalsocietypu
2
 on August 9, 2018http://rspb.royalsocietypublishing.org/Downloaded from bone, diverse elements (i.e. feathers, bones, mummified
remains, egg shells) are attributed to them [18–20]. Because of
the availability of exceptionally preserved specimens, DNA
recovered from various moa specimens has been used to differ-
entiate species [20,21] and estimate rates of DNA degradation
[22]. However, few complementary studies of moa proteins
have been attempted [3,23].blishing.org
Proc.R.Soc.B
282:201500152. Material and methods
(a) Moa bone
Cortical bone fragments from an indeterminate moa specimen
(MOR OST-255; 800–1000 years old [24]) were extracted as
described by Cleland et al. [23]. Briefly, approximately 1 g of corti-
cal bone was demineralized in 9 ml 0.6 M HCl for 4 h at room
temperature; the pellet was washed with sterile, deionized water
and fragments were further extracted using 50 mM ammonium
bicarbonate at 658C for 5 h. The HCl fraction was dialyzed against
water (2000 MWCO Slide-A-Lyzer Cassettes; Pierce) for 4 days at
48C, then lyophilized to completion. The ammonium bicarbonate
fraction was dried completely without further dialysis using a
speed vacuum. Resultant powders were stored at2808C until use.
(b) Protein digestion and mass spectrometry
Each powder (1 mg for HCl, 1.5 mg for ammonium bicarbonate)
was resuspended in 500 ml of 50 mM ammonium bicarbonate.
50 ul of each fraction were reduced using 10 mM dithiothreitol for
1 h at 378C followed by alkylation with 20 mM iodoacetamide for
1 h in the dark at room temperature. After reduction and alkylation,
the proteins were digested overnight at 378C with 0.2 mg Promega
modified trypsin. Digestion was subsequently stopped with 1 ul
of 100% formic acid and stored at 2808C until mass spectrometry.
Without further sample processing, 10 ul of each extraction
were injected onto a Waters nanoAcquity UPLC trap column
(180 mm  20 mm) with Symmetry C18 and washed for 5 min at
5 ml min21. Peptides were transferred to a Waters nanoAcquity
UPLC (75 mm  250 mm) BEH130C18 (1.7 mm particle size)
analytical column and eluted at 300 nl min21 on aWaters nanoAc-
quity with the following gradient: 2% B (99.9% acetonitrile, 0.1%
formic acid) to 60% B at 30 min, 90% B at 32 min, 90% B at
35 min, 2% B at 37 min, 2% B at 60 min. Buffer A was 0.1%
formic acid. Eluted peptides were analysed on a ThermoScientific
Orbitrap XLwith a scan range of 375–2000 m/z. The top five peaks
for each precursor were fragmented using collision-induced dis-
sociation, and dynamic exclusion was enabled with a repeat
count of 1, exclusion duration of 10 s and repeat duration of 30 s.
(c) Data analysis
The resulting spectra were analysed using three different search
engines: Mascot 2.3 [25] in Proteome Discoverer 1.2 (ThermoScien-
tific), Sequest HT [26] in Proteome Discoverer 1.4 and PEAKS7
[27,28]. Each Mascot file was sequentially searched against several
databases, namely Uniprot chicken, the common repository of
adventitious proteins (The Global Proteome Machine), Uniprot
osteocalcin, Uniprot bone, Uniprot collagen and Uniprot haemo-
globin; all results were compiled and overall peptide and protein
statistics calculated. The following parameters were used for
searching on all databases unless noted in brackets: 10 ppm precur-
sor tolerance; 0.5 Da fragment tolerance; static modification:
carbamidomethyl cysteine (C); dynamic modifications: deami-
dated asparagine and glutamine (NQ), oxidation methionine (M),
oxidation arginine and lysine (KP) for Uniprot chicken, Uniprot
bone and Uniprot collagen, carboxy glutamic acid (E) for Uniprot
osteocalcin. All peptides were filtered with a 5% FDR based on a
decoy database.Spectra were searched using Sequest HT against a Uniprot
Archosauria þ Testudinidae database and a Uniprot collagen
database with the following parameters: 10 ppm precursor toler-
ance, 0.5 Da fragment tolerance, fixed modifications: none,
variable modifications: carbamidomethyl (C), deamidated (NQ),
oxidation (M), carboxymethyl (K). For the collagen database, oxi-
dation (KP) was also added to variable modifications. This mass
shift represents HYP and lysine but is limited only to collagen
sequences. All peptides were filtered with a 5% FDR based on a
decoy database.
Spectra were searched using PEAKS7 against a Uniprot
Vertebrates database using: 10 ppm precursor tolerance, 0.5 Da
fragment tolerance, fixed modifications: none, variable modifi-
cations: carbamidomethylation (C), deamidated (NQ), oxidation
(M), oxidation or hydroxylation (RYFPNKD) or (G) at C-terminal,
and Carboxymethyl (KW) or (X) at N-terminal. A maximum of
five PTMs were allowed per peptide. Non-specific cleavage was
allowed at both ends of the peptide, as well as a maximum of
three missed cleavages. To find additional, unspecified PTMs and
mutations, PEAKS PTM [28] and SPIDER searches were enabled.
Results were filtered with the following parameters: peptides 210
lgP  15 (FDR 0.6%) and proteins 210 lgP  20 (FDR 0.0%).
For all searches, peptides were exported and those with over-
lapping sequences were culled. All peptides for all identified
collagens were aligned against collagen I and II exemplars
from Uniprot in Seaview 4. Consensus sequences were generated
for each search algorithm using a basic majority rule method
in Seaview.3. Results and discussion
We report the first partial collagen I sequence (figures 1 and 2;
electronic supplementary material, tables S1, S2, S5–S7), the
first partial collagen II sequence (electronic supplementary
material, figure S1 and tables S3, S5–S7; for coverage see elec-
tronic supplementary material, table S8) and several peptides
from collagen V (electronic supplementary material, tables
S4–S7; for coverage see electronic supplementary material,
table S8) for moa. Because the collagen V sequences exhibited
limited coverage, we did not align them; however, future
analyses from this specimen and others will provide additio-
nal sequence information for each of the collagen V chains.
Database searching using three algorithms resulted in
some variation in sequence coverage in collagen Ia1 and a2
(figures 1 and 2) when compared with the mature sequence
of chicken (Col1a1: PEAKS 73.7%, Mascot 77.8%, Sequest
70.9%, all combined 84.1%; Col1a2: PEAKS 63.3%, Mascot
47.0%, Sequest 50.7%, all combined 69.3%). This multi-
algorithm approach facilitates identification of more complete
sequences of other fossil taxa that may be missed when only
one algorithm is applied.
This is an indeterminate specimen (i.e. specieswasunable to
be determined, so only identified to a supraspecific level), but
the collagen I sequences, and to a lesser extent the collagen II
sequence, represent an important baseline for expected collagen
sequences in closely related extinct and extant species, and
provide critical data applicable to other undersampled palaeog-
nath taxa. Additionally, these sequences can be used to refine
phylogenetic hypotheses of other archosaurs (e.g. electronic
supplementarymaterial, figure S7), including extant and extinct
crocodylians and extinct non-avian dinosaurs. The phylogeny
resulting from collagen I sequences places moa in a clade
with other palaeognath taxa as well as Galloansarae taxa
(electronic supplementary material, figure S7). Unfortunately,
Moa_Mascot   -------------------GPXGPPGKNGDDGEAGKPGRPGERGPSGPQGARGLPGTAGLPGMK---GFSGLDGAKGQPGPAGPKGEPGSPGENGAPGQM
Moa_Sequest  -------------------GPXGPPGKNGDDGEAGKPGRPGER---------GLPGTAGLPGMK---GFSGLDGAK---------GEPGSPGENGAPGQM
Moa_PEAKS    -------------------GPAGPPGKNGDDGEAGKPGRPGXRGPXGPQGARGLPGTAGLPGMK---GFSGLDGAKGDTGPAGPKGEPGSPGENGAPGQM
Moa_Mascot   GPR------GAPGINGPAGARGNDGAVGAAGPPGPTGPTGPPGFPGAAGAKGEAGPQGARGSEGPQGARGEPGPPGPAGAAGPAGNPGADGQPGAKGATG
Moa_Sequest  GPR------GRPGAPGPAGARGNDGATGAAGPPGPTGPAGPPGFPGAAGAK---------GSEGPQGARGEPGPPGPAGXAGPSGNPGXDGQPGAKGATG
Moa_PEAKS    GPR------GXPGPSGPAGAR------------------------------GETGPQGARGSEGPQGARGEPGPPGPAGAAGPAGNPGADGQPGAKGATG
Moa_Mascot   APGIAGAPGFPGARGXXGPQGPXGAPGPKGNSGEPGAPGNKGDTGAKGEPGPAGVQGPPGPXGEEGKR---GEPGPAGLPGPAGER------GFPGADGI
Moa_Sequest  APGIAGAPGFPGARGPXGPQGPSGAPGPKGNSGEPGAPGNKGDTGAKGEPGPAGVQGPPGPAGEEGKR---GEPGPXGLPGPAGER------GFPGXDGI
Moa_PEAKS    APGIAGAPGFPGARGAXGPQGPSGAPGPKGNSGEPGAPGNKGDTGAKGEPGPAGVQGPPGPAGEEGKR---GEPGPAGLPGPAGER------GFPGADGX
Moa_Mascot   AGPK---------------------------------GLTGSPGSPGPDGKTGPPGPAGQDGRPGPPGPPGARGQXGVMGFPGPKGAAGEPGKPGERGAP
Moa_Sequest  AGPK---------------GSPGESGRPGEPGLPGAKGLTGSPGSPGPDGK----------------------GQAGVMGFPGPKGAAGEPGKPGERGAP
Moa_PEAKS    AGPK---------------GSPGEAGRPGEAGLPGAKGLTGSPGSPGPDGKTGPPGPAGQDGRPGPPGPPGARGQAGVMGFPGPKGAAGEPGKPGERGAP
Moa_Mascot   GPPGAVGAAGKDGEAGAQGPPGPTGPAGERGEQGPAGAPGFQGLPGPAGPPGEAGKPGEQGVPGNAGAPGPAGAR---------GVQGPPGPQGPRGANG
Moa_Sequest  GPPGAVGAAGKDGEAGAQGPPGPTGPAGERGEQGPSGAPGFQGLPGPAGAPGESGKPGEQGVPGDIGAPGPSGAR---------GVQGPPGPQGPRGANG
Moa_PEAKS    GPPGAVGAAGKDGEAGAQGPPGPTGPAGER------------------------------------------------------GVQGPPGPQGPRGANG
Moa_Mascot   APGNDGAK------------------------GAAGLPGAKGDRGDPGPKGADGAPGKDGLRGLTGPIGPPGPAGAPGDKGEAGPSGPAGPTGARGAPGD
Moa_Sequest  APGNDGAK------------------------------------------------------GLTGPIGPPGPAGAPGDKGEAGPSGPAGPTGARGAPGD
Moa_PEAKS    APGNDGAKGDAGAPGAPGSQGAPGL-------GAAGLPGAKGDRGDXGPKGADGAPGKDGLRGLTGPIGPPGPAGAPGDKGEAGPSGPAGPTGARGAPGD
Moa_Mascot   RGEPGPPGPAGFAGPPGADGQPGAKGETGDAGAK------------------------------GXAGPPGATGFPGAAGRVGPPGPSGNIGLPGPPGPA
Moa_Sequest  RGEPGPPGPAGFAGPPGSDGQPGAKGETGDXGAKGDAGPPGPAGPTGAPGPSGAVGAPGPK---GAAGPPGATGFPGAAGRVGPPGPSGNIGLPGPPGPS
Moa_PEAKS    RGEPGPPGPAGFAGPPGADGQPGAKGETGDAGAK------------------------------GSAGPPGATGFPGAAGRVGPPGPSGNIGLPGPPGPA
Moa_Mascot   GK-------GETGPAGRPGEPGPAGPPGPPGEKGSPGADGPIGAPGTPGPQGIAGQRGVVGLPGQRGERGFPGLPGPSGEPGKQGPSGPSGERGPPGPVG
Moa_Sequest  GK-------------------------------GSPGADGPIGAPGTPGPQGIAGQRGVVGLPGQR---GFPGLPGPSGEPGKQGPSGPSGERGXPGXVG
Moa_PEAKS    GK-------GETGPAGRPGEPGPAGPPGPPGEKGSPGADGPIGAPGTPGPQGIAGQRGVVGLPGQR---GFPGLPGPSGEPGKQGPSGXXGER----PAG
Moa_Mascot   PPGLAGPPGESGREGXPGAEGAPGRDGAXGXKGDRGETGPAGPPGAPGAPGAPGPVGPAGKNGDRGETGPAGPAGPXGPAGARGPAGPQGPRGDKGETGE
Moa_Sequest  XPGLXGXPGESGREGAPGAEGAPGRDGAAGPKGDRGETGPXGPPGAPGAPGAPGPVGPAGKNGDRGETGPAGPAGPXGPAGARGPAGPQGPR--------
Moa_PEAKS    PPGLAGPPGESGREGAPGAEGAPGRDGAAGPKGDRGETGPAGPPGAPGAPGAPGPVGPAGKNGDRGETGPAGPAGPPGPAGARGPAGPQGPRGDKGETGE
Moa_Mascot   QGDR------GFSGLQGPPGPPGAPGEQGPSGASGPAGPRGPPGSAGAAGKDGLNGLPGPIGPPGPR---------------------------------
Moa_Sequest  ----------GFSGLQGPPGPPGAPGEQGPSGASGPAGPRGPPGSAGXAGKDGLNGLPGPIGPPGPR---------------------------------
Moa_PEAKS    QGDRGMK---GFSGLQGPPGPPGSPGEQGPSGASGPAGPRGPPGSAGAAGKDGLNGLPGPIGPPGPR---------------------------------
Figure 1. Collagen Ia1 sequence alignments for PEAKS, Mascot and Sequest peptides.
Figure 2. Collagen Ia2 sequence alignments for PEAKS, Mascot and Sequest peptides.
rspb.royalsocietypublishing.org
Proc.R.Soc.B
282:20150015
3
 on August 9, 2018http://rspb.royalsocietypublishing.org/Downloaded from few collagen I sequences are known from palaeognath taxa,
requiring additional sampling of bone from moa and other
extant palaeognaths (e.g. emu, kiwi) to better elucidate
relationships and lend support to hypothesized DNA-based
phylogenies [29].
(a) In vivo post-translational modifications
We detected various biologically derived PTMs: methylation
(figure 3a), di-methylation (electronic supplementary material,
figure S2), alkylation (figure 3a; electronic supplementary
material, figure S3), fucosylation (figure 3b; electronic sup-
plementary material, figure S5) and hydroxylation; listed in
table 2 and electronic supplementary material, tables S1–S4.
With the exception of hydroxylated proline, few other in vivo
PTMs have been identified from fossil remains. PEAKS PTM
detected more PTMs than have been previously observed on
ancient peptides without a priori knowledge of what may or
may not preserve. This search strategy allowedus to detect enzy-
matic glycosylation for the first time in ancient bone proteins
(figure 3; electronic supplementary material, figure S5).Determining the endogeneity of fossil proteins and bio-
logical PTMs is critical for evaluating their biological function.
To support PTM endogeneity, we compared the positions of
fossil PTMs with those on proteins from extant eukaryotic
taxa. We detected well-known methylation not only on lysine
[30] but also on aspartic acid and glutamic acid (figure 3a;
electronic supplementary material, tables S1–S3), only recently
shown to occur in eukaryotic proteins [30]. These modifications
support endogeneity. Additionally, we identified two peptides
containing fucose onserine residues. Serine is one of several resi-
dues that is fucosylated in extant taxa [31]. Lastly, we identified
one of themost common and potentially important positions of
acetylation in extant proteins [32] on the moa lysine residues
(figure 3a; electronic supplementary material, figure S3).(b) Advanced-glycation end products and diagenetically
modified peptides
In addition to in vivo PTMs, we were able to detect diageneti-
cally derived protein modifications. Consistent with previous
300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100 777. 42
1310. 671027 .58
671. 50
388.17
840 .42
627.83
1282.67970.42
728.42
1183.58
883.58 1126.42584.25497.17370.25 1552.001440.50898.33242.17 525.33 1053.42 1480.58300.08 1336.75470.08
b7+
y4+
y7+
y9+ y12+
b9+
b10+ b13+
b17++
methyl acetyl
300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100 938.42
863.00
946 .92796.33
1115. 50
1568 .92
784.50 1167.58
1082.42744.50
1028.67 1512.921415.83576.33343.08
400.17 1737.00667.42 1245.67 1335.58 1494.83519.33 1755.001593.75319.42
1872.83
y12+
b3+
y11+
y7+
y5+
y4+
y14+ y15+
y16+
y17+
y18+
y18++
b4+
b6+
b7 +
b8+
b11+ b13+
b15+
b20++
Fucose
(a)
(b)
re
la
tiv
e 
ab
u
n
da
nc
e
re
la
tiv
e 
ab
u
n
da
nc
e
m/z
1780.83
m/z
Figure 3. (a) Collagen IIa1 peptide (GDRGDVGEKGPEGAPGK) showing methylation and alkylation. (b) Collagen IIa1 peptide (GERGLPGESGAVGPAGPIGS) showing
fucosylation. (Online version in colour.)
rspb.royalsocietypublishing.org
Proc.R.Soc.B
282:20150015
4
 on August 9, 2018http://rspb.royalsocietypublishing.org/Downloaded from analyses of ancient bone [8,14,15], we observe deamida-
tion for both glutamine and asparagine (figure 4; electronic
supplementary material, figure S4) that is incomplete
(e.g. electronic supplementary material, figure S4) and may
be, in part, derived from sample preparation. We observed
additional modifications we ascribe to diagenesis, including
advanced glycation end-products (AGEs), backbone cleavage
and variable HYP. It has been hypothesized that AGEs lead
to preservation of proteins [17]; however, direct observation
of AGEs on intact peptides has not been previously made
from any ancient source. We observe carboxymethyllysine
(CML) on several peptides, which leads to missed cleavages
by trypsin (figure 4), and a potential backbone cleavage
where CML is present at the C-terminus of the peptide (elec-
tronic supplementary material, figure S6). CML modification
in the C-terminal position may not lead to enhanced preser-
vation because it does not form cross-links [33] but instead
modifies the side chain of lysine, resulting in the potential dis-
ruption of collagen tertiary structure and subsequent loss of
fossil collagen. This may explain the selectivity of peptide pres-
ervation noted by San Antonio et al. [34]. Additionally, we
observe heterogeneity in the presence or absence of CML on
the same peptide (electronic supplementary material, table S1).
Further research is necessary to detect and measure the types,
amounts and effects of AGEs in ancient and fossil bones.Cleavage of the protein backbone has been suggested as
one of the major causes for loss of protein from bone [35].
Because trypsin is specific for digestion at arginine and lysine
residues, we hypothesize that peptides that do not terminate
with arginine, lysine or the end of the protein sequence rep-
resent backbone cleavages. We observe several peptides that
may represent backbone cleavage on the collagen Ia1 chain
(table 1). Two of the peptides (i.e. TGPPGPAGQDGRjGj
PPGPPGAR and VGPPGPSGNIGLjPGPPGPAGK, where
j represents potential backbone cleavage positions) show
breakages at proline residues consistent with backbone
cleavage by oxidation [36].
Finally, we hypothesize that variable HYP represents
diagenetic change. In extant collagen, there is little variation
in HYP percentage [37], but 49% (35 of 71) of moa collagen
Ia1 peptides that contain HYP show variability (i.e. peptides
with the same sequence have different numbers of HYP
independent of hydroxylation position). On collagen Ia2 pep-
tides, we observe 30% variability (12 of 39), and on collagen
IIa1 peptides, we observe 56% variability (5 of 9 peptides).
These variations are much higher than would be expected
from a sample preparation artefact because HYP has been
shown to persist in completely hydrolyzed samples and
used to approximate collagen content [38]. Alternatively,
this variation could represent oxidation of proline to glutamic
1000 1200
m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100 621.83
765.33
528.83 779.08
1137.58
1022.58
850.50 1103.00
1242.50323.25 1357.67557.50
1529.831183.58295.17 1394.58 1646.00 1907.081830.75
y12++
b9+
y13++
y15++
b11+ b12+
y13+
b14+
y15+
y12+
y10+y9+
y8+y4+
b8+
b7+
b6+
y10++
y7++
re
la
tiv
e 
ab
u
n
da
nc
e
OHCMLOH OH DEAM
400 600 800 1400 1600 1800 2000
679.42
823.00
Figure 4. Collagen Ia1 peptide (GPAGPPGKNGDDGEAGKPGRPGQR) showing hydroxylation, carboxymethyllysine and deamidation. (Online version in colour.)
Table 1. Examples of hypothesized protein backbone cleavage detected for
collagen I alpha 1 peptides.
EGAPGAEGAPGR TGPPGPAGQDGR
EGAPGAEGAPGRD TGPPGPAGQDGRPG
EGAPGAEGAPGRDG TGPPGPAGQDGRPGPPGPPGAR
EGAPGAEGAPGRDGAAGPK GAPGDRGEPGPPGPAGFAGPPGADGQPGAK
EGAPGAEGAPGRDGAAGPKGDR GAPGDRGEPGPPGPAGFAGPPGADGQPGAKG
VGPPGPSGNIGL
VGPPGPSGNIGLPGPPGPAGK
Table 2. Protein and peptide modifications detected.
biologically derived diagenetically derived
alkylation backbone cleavage
dimethylation carboxymethylation (advanced glycation
end-product)
fucosylation deamidationa
hydroxylationa dehydroxylation/glutamic semialdehyde
methylation
aPreviously detected.
rspb.royalsocietypublishing.org
Proc.R.Soc.B
282:20150015
5
 on August 9, 2018http://rspb.royalsocietypublishing.org/Downloaded from semialdehyde (i.e. an isobaric mass shift as hydroxyproline
[11]). In either case, we observe large variations in modi-
fied proline residues that may be the result of diagenesis.
The loss of hydroxylation may also explain the detection of
serine/alanine substitutions in multiple peptides (i.e. loss of
oxygen from the side chain of serine results in a transition
of that residue to alanine) detected by database searching in
multiple peptides (e.g. GP[S/A]GPPGKNGDDGEAGKPGR
PG[Q/E]R). This sample peptide also shows a glutamine/
glutamic acid residue difference that may only reflect deami-
dation but not an amino acid substitution. Alternatively, the
serine/alanine transition is potentially an effect of algorithm
differences in detection that need to be further resolved for all
palaeoproteomic studies. These diagenetic changes to amino
acid side groups, consistent with previously observed modi-
fications to ancient DNA (e.g. deamination of cytosine [39]),
may lead to incorrect assignments in phylogenetic analyses,
and need to be considered when protein sequences are
used to evaluate evolutionary relationships.4. Conclusion
Protein sequences obtained from this moa provide a baseline
against which peptides recovered from other fossils may
be searched, and identify modifications that may occur in
other fossils, allowing differentiation between inherent geneticchange and diagenetic change of the original proteins. While
others have reported several biologically and diagenetically
derived PTMs [11,13], we identified four biological PTMs
(table 2) out of five total modifications for the first time from
fossil remains. We also detected three or potentially four novel
diagenetic PTMs (table 2) that have not been previously detected
from fossils. Delineating both in vivo and diagenetically derived
PTMs in these extinct taxa will provide more robust hypotheses
regarding the physiology [40] and/or phylogenies of these
organisms, as well as the mechanisms leading to preservation
or loss of proteins from bones of different ages or localities.Data accessibility. All raw mass spectrometry data are available at Dryad
(http://dx.doi.org/10.5061/dryad.q35s1).
Acknowledgements. We thank J. Horner and J. Scannella for access to and
information on this moa specimen, J. Carlson and S. Baxter for access
to the LTQ Orbitrap XL at David H. R. Murdoch Research Institute,
and N. Kelleher and P. Thomas for access to Peaks7. We also thank
the Willi Hennig Society for providing access to TNT, and three anon-
ymous reviewers for helpful critiques for improving this manuscript.
Funding statement. This research was funded by NSF EAR 0541744 to
M.H.S., NSF DGE-0750733 to T.P.C., the David and Lucile Packard
Foundation to M.H.S. and NSF INSPIRE to M.H.S. and E.R.S.
Authors’ contributions. All authors conceived and designed this study.
T.P.C. acquired the data, T.P.C. and E.R.S. performed bioinformatics
and all authors made interpretations. T.P.C. and E.R.S. wrote the
manuscript and all authors revised it leading to the final version.
6 on August 9, 2018http://rspb.royalsocietypublishing.org/Downloaded from Referencesrspb.royalsocietypublishing.org
Proc.R.Soc.B
282:201500151. Lowenstein JM. 1980 Species-specific proteins in
fossils. Naturwissenschaften 67, 343–346. (doi:10.
1007/bf01106588)
2. Rainey WE, Lowenstein JM, Sarich VM, Magor DM.
1984 Sirenian molecular systematics—including the
extinct Steller’s sea cow (Hydrodamalis gigas).
Naturwissenschaften 71, 586–588. (doi:10.1007/
bf01189187)
3. Huq NL, Rambaud SM, Teh L-C, Davies AD,
McCulloch B, Trotter MM, Chapman GE. 1985
Immunochemical detection and characterisation of
osteocalcin from moa bone. Biochem. Biophys. Res.
Commun. 129, 714–720. (doi:10.1016/0006-291x
(85)91950-3)
4. Wyckoff RWG, McCaughey WF, Doberenz AR. 1964
The amino acid composition of proteins from
pleistocene bones. Biochim. Biophys. Acta 93,
374–377. (doi:10.1016/0304-4165(64)90387-3)
5. Bada JL, Kvenvolden KA, Peterson E. 1973
Racemization of amino acids in bones. Nature 245,
308–310. (doi:10.1038/245308a0)
6. Higuchi R, Bowman B, Freiberger M, Ryder OA,
Wilson AC. 1984 DNA sequences from the quagga,
an extinct member of the horse family. Nature 312,
282–284. (doi:10.1038/312282a0)
7. Green RE et al. 2010 A draft sequence of the
neandertal genome. Science 328, 710–722.
(doi:10.1126/science.1188021)
8. Orlando L et al. 2013 Recalibrating Equus evolution
using the genome sequence of an early Middle
Pleistocene horse. Nature 499, 74–78. (doi:10.
1038/nature12323)
9. Cappellini E, Collins MJ, Gilbert MTP. 2014
Unlocking ancient protein palimpsests. Science 343,
1320–1322. (doi:10.1126/science.1249274)
10. Schweitzer MH et al. 2009 Biomolecular
characterization and protein sequences of the
Campanian hadrosaur B. canadensis. Science 324,
626–631. (doi:10.1126/science.1165069)
11. Cappellini E et al. 2012 Proteomic analysis of a
pleistocene mammoth femur reveals more than one
hundred ancient bone proteins. J. Proteome Res. 11,
917–926. (doi:10.1021/pr200721u)
12. Buckley M. 2013 A molecular phylogeny of
Plesiorycteropus reassigns the extinct mammalian
order ‘Bibymalagasia’. PLoS ONE 8, e59614. (doi:10.
1371/journal.pone.0059614)
13. Wadsworth C, Buckley M. 2014 Proteome
degradation in fossils: investigating the longevity of
protein survival in ancient bone. Rapid Commun.
Mass Spectrom. 28, 605–615. (doi:10.1002/
rcm.6821)
14. van Doorn NL, Wilson J, Hollund H, Soressi M,
Collins MJ. 2012 Site-specific deamidation of
glutamine: a new marker of bone collagen
deterioration. Rapid Commun. Mass Spectrom. 26,
2319–2327. (doi:10.1002/rcm.6351)
15. Wilson J, van Doorn NL, Collins MJ. 2012 Assessing
the extent of bone degradation using glutaminedeamidation in collagen. Anal. Chem. 84,
9041–9048. (doi:10.1021/ac301333t)
16. Nielsen-Marsh CM, Ostrom PH, Gandhi H, Shapiro B,
Cooper A, Hauschka PV, Collins MJ. 2002 Sequence
preservation of osteocalcin protein and
mitochondrial DNA in bison bones older than 55 ka.
Geology 30, 1099–1102. (doi:10.1130/0091-7613
(2002)030,1099:SPOOPA.2.0.CO;2)
17. Nielsen-Marsh CM, Richards MP, Hauschka PV,
Thomas-Oates JE, Trinkaus E, Pettitt PB, Karavanic´ I,
Poinar H, Collins MJ. 2005 Osteocalcin protein
sequences of Neanderthals and modern primates.
Proc. Natl Acad. Sci. USA 102, 4409–4413. (doi:10.
1073/pnas.0500450102)
18. Rawlence NJ, Wood JR, Armstrong KN, Cooper A.
2009 DNA content and distribution in ancient
feathers and potential to reconstruct the plumage
of extinct avian taxa. Proc. R. Soc. B 276,
3395–3402. (doi:10.1098/rspb.2009.0755)
19. Oskam CL et al. 2010 Fossil avian eggshell preserves
ancient DNA. Proc. R. Soc. B 277, 1991–2000.
(doi:10.1098/rspb.2009.2019)
20. Allentoft ME, Rawlence NJ. 2012 Moa’s ark or volant
ghosts of Gondwana? Insights from nineteen years
of ancient DNA research on the extinct moa (Aves:
Dinornithiformes) of New Zealand. Ann. Anat. 194,
36–51. (doi:10.1016/j.aanat.2011.04.002)
21. Baker AJ, Huynen LJ, Haddrath O, Millar CD,
Lambert DM. 2005 Reconstructing the tempo and
mode of evolution in an extinct clade of birds with
ancient DNA: the giant moas of New Zealand. Proc.
Natl Acad. Sci. USA 102, 8257–8262. (doi:10.1073/
pnas.0409435102)
22. Allentoft ME et al. 2012 The half-life of DNA in
bone: measuring decay kinetics in 158 dated fossils.
Proc. R. Soc. B 279, 4724–4733. (doi:10.1098/rspb.
2012.1745)
23. Cleland TP, Voegele K, Schweitzer MH. 2012
Empirical evaluation of bone extraction protocols.
PLoS ONE 7, e31443. (doi:10.1371/journal.pone.
0031443)
24. Schweitzer MH, Wittmeyer JL, Horner JR. 2007 Soft
tissue and cellular preservation in vertebrate
skeletal elements from the Cretaceous to the
present. Proc. R. Soc. B 274, 183–197. (doi:10.
1098/rspb.2006.3705)
25. Perkins DN, Pappin DJC, Creasy DM, Cottrell JS. 1999
Probability-based protein identification by searching
sequence databases using mass spectrometry data.
Electrophoresis 20, 3551–3567. (doi:10.1002/
(SICI)1522-2683(19991201)20:18,3551::AID-
ELPS3551.3.0.CO;2-2)
26. Eng JK, McCormack AL, Yates III JR. 1994 An
approach to correlate tandem mass spectral data of
peptides with amino acid sequences in a protein
database. J. Am. Soc. Mass Spectrom. 5, 976–989.
(doi:10.1016/1044-0305(94)80016-2)
27. Ma B, Zhang K, Hendrie C, Liang C, Li M, Doherty-
Kirby A, Lajoie G. 2003 PEAKS: powerful softwarefor peptide de novo sequencing by tandem mass
spectrometry. Rapid Commun. Mass Spectrom. 17,
2337–2342. (doi:10.1002/rcm.1196)
28. Han X, He L, Xin L, Shan B, Ma B. 2011 PeaksPTM:
mass spectrometry-based identification of peptides
with unspecified modifications. J. Proteome Res. 10,
2930–2936. (doi:10.1021/pr200153k)
29. Mitchell KJ, Llamas B, Soubrier J, Rawlence NJ,
Worthy TH, Wood J, Lee MSY, Cooper A. 2014
Ancient DNA reveals elephant birds and kiwi are
sister taxa and clarifies ratite bird evolution. Science
344, 898–900. (doi:10.1126/science.1251981)
30. Sprung R, Chen Y, Zhang K, Cheng D, Zhang T, Peng
J, Zhao Y. 2008 Identification and validation of
eukaryotic aspartate and glutamate methylation in
proteins. J. Proteome Res. 7, 1001–1006. (doi:10.
1021/pr0705338)
31. Sharon N, Lis H. 2008 Glycoproteins: structure and
function. In Glycosciences: status and perspectives
(eds HJ Gabius, S. Gabius), pp. 133–162. New York,
NY: Wiley-VCH Verlag GmbH.
32. Glozak MA, Sengupta N, Zhang X, Seto E. 2005
Acetylation and deacetylation of non-histone proteins.
Gene 363, 15–23. (doi:10.1016/j.gene.2005.09.010)
33. Shapiro BP, Owan TE, Mohammed SF, Meyer DM,
Mills LD, Schalkwijk CG, Redfield MM. 2008
Advanced glycation end-products accumulate in
vascular smooth muscle and modify vascular but
not ventricular properties in elderly hypertensive
canines. Circulation 118, 1002–1010. (doi:10.1161/
CIRCULATIONAHA.108.777326)
34. San Antonio JD, Schweitzer MH, Jensen ST, Kalluri
R, Buckley M, Orgel JPRO. 2011 Dinosaur peptides
suggest mechanisms of protein survival. PLoS ONE
6, e20381. (doi:10.1371/journal.pone.0020381)
35. Schweitzer MH. 2004 Molecular paleontology: some
current advances and problems. Ann. Pale´ontol. 90,
81–102. (doi:10.1016/j.annpal.2004.02.001)
36. Berlett BS, Stadtman ER. 1997 Protein oxidation in
aging, disease, and oxidative stress. J. Biol. Chem. 272,
20 313–20 316. (doi:10.1074/jbc.272.33.20313)
37. Barnes MJ, Constable BJ, Morton LF, Royce PM.
1974 Age-related variations in hydroxylation of
lysine and proline in collagen. Biochem. J. 139,
461–468.
38. Sroga GE, Karim L, Colo´n W, Vashishth D. 2011
Biochemical characterization of major bone-matrix
proteins using nanoscale-size bone samples and
proteomics methodology. Mol. Cell. Proteomics 10,
M110.006718. (doi:10.1074/mcp.M110.006718)
39. Hofreiter M, Jaenicke V, Serre D, Haeseler AV, Pa¨a¨bo
S. 2001 DNA sequences from multiple amplifications
reveal artifacts induced by cytosine deamination in
ancient DNA. Nucleic Acids Res. 29, 4793–4799.
(doi:10.1093/nar/29.23.4793)
40. Seo J, Lee KJ. 2004 Post-translational modifications
and their biological functions: proteomic analysis
and systematic approaches. J. Biochem. Mol. Biol.
37, 35–44. (doi:10.5483/BMBRep.2004.37.1.035)