National Science Review

RESEARCH ARTICLE 0: 1–8, 2020

doi: 10.1093/nsr/nwz206
Advance access publication 12 January 2020

Downloaded from [Link] by Jamia Millia Islamia University user on 30 April 2020
MOLECULAR BIOLOGY & GENETICS

Evidence of proteins, chromosomes and chemical

markers of DNA in exceptionally preserved
1 Key Laboratory of

Vertebrate Evolution
dinosaur cartilage
and Human Origins,
Institute of Vertebrate Alida M. Bailleul 1,2,∗ , Wenxia Zheng3 , John R. Horner4 , Brian K. Hall5 ,
Paleontology and
Paleoanthropology,
Casey M. Holliday6 and Mary H. Schweitzer3,7,8
Chinese Academy of
Sciences, Beijing
100044, China; 2 CAS ABSTRACT
Center for Excellence A histological ground-section from a duck-billed dinosaur nestling (Hypacrosaurus stebingeri) revealed
in Life and microstructures morphologically consistent with nuclei and chromosomes in cells within calcified cartilage.
Paleoenvironment, We hypothesized that this exceptional cellular preservation extended to the molecular level and had
Beijing 100044, China;
3 Department of
molecular features in common with extant avian cartilage. Histochemical and immunological evidence
supports in situ preservation of extracellular matrix components found in extant cartilage, including
Biological Sciences,
North Carolina State
glycosaminoglycans and collagen type II. Furthermore, isolated Hypacrosaurus chondrocytes react
University, Raleigh, positively with two DNA intercalating stains. Specific DNA staining is only observed inside the isolated
NC 27695, USA; cells, suggesting endogenous nuclear material survived fossilization. Our data support the hypothesis that
4 Honors Program, calcified cartilage is preserved at the molecular level in this Mesozoic material, and suggest that remnants of
Chapman University, once-living chondrocytes, including their DNA, may preserve for millions of years.
Orange, CA 92866,
USA; 5 Department of Keywords: cartilage, dinosaur, nuclei, chromosomes, collagen II, DNA markers
Biology, Dalhousie
University, Halifax,
B3H 4R2, Canada;
INTRODUCTION AND HISTOLOGICAL cellular structures still sharing a single lacuna (i.e., a
6 Department of OBSERVATIONS cell doublet) [4,5] were seen, consistent with chon-
Pathology and
drocytes at the end of mitosis (Fig. 1C, pink arrow;
A nesting ground yielding dozens of disarticulated
Anatomical Sciences, Supplementary Fig. 1). Although many lacunae ap-
nestlings assigned to the herbivorous duck-billed di-
University of Missouri, pear empty (Fig. 1B and C, green arrow), other la-
nosaur Hypacrosaurus stebingeri was discovered in
Columbia, MO 65211, cunae (pink arrow) contain a material distinct from
the 1980s in the Upper Cretaceous (Campanian)
USA; 7 North Carolina the matrix, including a darker material consistent in
Two Medicine Formation of northern Montana
Museum of Natural shape and location with a nucleus (Fig. 1C, white ar-
Sciences, Raleigh, NC
(Museum of the Rockies, MOR 548, locality TM-
rows). This is comparable to features of extant calci-
27601, USA and 066, Supplementary Text 1) [1,2]. Several limb and
fied cartilage [4] observed in ground sections of de-
8 Department of skull elements of these nestlings were subjected to
fleshed, juvenile emu skulls, where some lacunae are
Geology, University of microscopic analyses to answer growth-related ques-
empty, and others retain cells and intracellular con-
Lund, 22362, Sweden tions [1] and to describe different types of cartilage
tents including nuclei (Fig. 1G).
[3]. The calcified cartilage seen within a supraoc-
∗ Corresponding
Near the cell doublet (Fig. 1C), other micro-
cipital in ground-section was especially interesting
scopic structures consist of dark, condensed and
author. E-mail: (Fig. 1B–D). Within the chondro-osseous junction
elongated material, aligned along a plane and slightly
[Link]@[Link] (i.e., the part of the growth plate where bone re-
mirroring each other (Fig. 1D). The cell lacuna sur-
places cartilage) closest to the left exoccipital, the tis-
rounding these structures is visible (Fig. 1D, black
Received 11 sues presented excellent microscopic preservation,
arrow), but is even clearer under a different light
December 2019; such that cartilage could be distinguished from bone
Revised 9 January
setting (with a condenser, Supplementary Fig. 1).
by exhibiting a translucent, amorphous extracellu-
2020; Accepted 10 This dark material shares microstructural features
lar matrix (ECM) and round, hypertrophic chon-
January 2020 with condensed chromatin, more precisely of chro-
drocyte lacunae (Fig. 1B). At higher magnification,
mosomes in metaphase of the cell cycle [6]. Similar

C The Author(s) 2020. Published by Oxford University Press on behalf of China Science Publishing & Media Ltd. This is an Open Access article distributed under the terms of the Creative

Commons Attribution License ([Link] which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original
work is properly cited.
2 Natl Sci Rev, 2020, Vol. 0, No. 0 RESEARCH ARTICLE

Hypacrosaurus

Downloaded from [Link] by Jamia Millia Islamia University user on 30 April 2020
Ground sections ~100 µm
Emu

Figure 1. Ground section of Hypacrosaurus (MOR 548) supraoccipital shows exceptional histological preservation of calcified
cartilage. (A) An isolated supraoccipital (So) of Hypacrosaurus in dorsal view. (B–D) Ground section of another So showing
calcified cartilage with hypertrophic chondrocyte lacunae. (C) Some cell doublets appear empty (green arrow), but others
(pink arrow) present darker, condensed material consistent in shape and location with a nucleus (white arrows). (D) Dark,
condensed, and elongated material with morphological characteristics of metaphase chromosomes. The limit of the cell
lacuna is visible (black arrow). (E) Caudal view of a juvenile emu skull (∼8–10 months old) showing the So and exoccipitals
(Exo) in articulation. (F, G) Ground section (stained with Toluidine blue) of calcified cartilage from this emu skull showing cell
doublets (pink arrows) with remnants of nuclei (white arrows) and others without intracellular content (green arrow).

chromosome-like structures have been observed in a phy. Chondrocytes encased in lacunae secrete ECM
fossil fern from the Jurassic [7], but the present study components including collagen II and glycosamino-
reports this type of exceptional microscopic preser- glycans [4,8]. Conversely, osteoblasts and osteo-
vation, at the sub-cellular level, in a fossil vertebrate cytes secrete collagen type I, minimal amounts
and validates the observations with biochemistry. of glycosaminoglycans [9], and non-collagenous
We hypothesized that this exceptional proteins [10].
morphological preservation would extend to the Hypertrophic chondrocytes of the chondro-
molecular level when methods commonly used to osseous junction have different fates (Supplemen-
identify molecular and chemical markers in extant tary Text 2) but many undergo cell death [4,5,11],
cartilage were applied to these fossil tissues. To eventually resulting in empty lacunae (e.g., Fig. 1G)
test this hypothesis, we investigated molecular followed by replacement with bone [4,12]. Even
preservation of Hypacrosaurus cartilage at the though one fossil cell preserves structures morpho-
extracellular, cellular and intracellular levels in an- logically consistent with mitotic chromosomes in
other supraoccipital from the same nesting ground metaphase (Fig. 1D), we propose, based upon living
(Fig. 1A), similar in size to the one in which these comparisons, that it is not undergoing normal cell
structures were originally observed (Fig. 1B–D). division, but rather presents the characteristics of
This study specimen had not been previously em- a hypertrophic chondrocyte undergoing an early
bedded in resin. We capitalized on distinct chemical stage of cell death called chondroptosis [11,13,14]
differences between cartilage and bone within this (Supplementary Fig. 1; Supplementary Text 2, 3).
second Hypacrosaurus supraoccipital (Fig. 1A), and In this type of abnormal mitosis, specifically iden-
used the supraoccipitals of juvenile emus (Dromaius tified in senescent hypertrophic chondrocytes
novaehollandiae) as extant homologues. of avian growth plates, the DNA condenses and
chromosomes adopt a metaphase-like arrangement
[13–15] (Supplementary Text 2).
RESULTS AND DISCUSSION
Histology Histochemistry
During ontogeny, the supraoccipital arises as a The ECM of extant cartilage and bone can be dif-
primary cartilage which is then replaced by bone ferentiated using Alcian blue, a histochemical stain
via endochondral ossification [3]. Growth occurs that reacts intensely with acidic materials and much
mostly at the chondro-osseous junction, where less so with basic ones. Cartilage incorporates acid
chondrocytes undergo cell division and hypertro- mucins and glycosaminoglycans not found in bone
RESEARCH ARTICLE Bailleul et al. 3

Unstained Alcian blue stain

Cartilage Bone Cartilage Bone

Hypacrosaurus
Paraffin sections - 5 µm

Downloaded from [Link] by Jamia Millia Islamia University user on 30 April 2020
Emu

Figure 2. Alcian blue histochemical stain capitalized on differential presence of glycosaminoglycans in the calcified cartilage
and bone of Hypacrosaurus. Unstained (A, B, E, F) and Alcian-blue stained (C, D, G, H) paraffin sections of Hypacrosaurus and
emu cartilage and bone. A strong, positive blue staining is seen in Hypacrosaurus cartilage (C), comparable to the intense, but
darker stain found in modern emu cartilage (G). This suggests that glycosaminoglycans are still present in the cartilaginous
matrix of this dinosaur. In contrast, the fossil and extant bones show a very light blue stain (D, H). Images are at the same
scale.

[4,16]. Alcian blue was previously employed to not all epitopes would have been present in the living
differentiate tissues in Tyrannosaurus rex [16] and dinosaur.
Yanornis martini [17]; we applied it here to paraf- To account for the possibility of non-specific
fin sections of demineralized Hypacrosaurus carti- binding, we digested the fossil and extant tissues
lage and bone. Fossil (Fig. 2C) and extant (Fig. 2G) with collagenase II, an enzyme specific for colla-
cartilage both demonstrated intense staining when gen II [18]. At the same integration times and pa-
compared to stained demineralized bone from the rameters as in non-digested tissues, antibody reac-
same organisms (Fig. 2D, H), supporting chemical tivity was significantly decreased after digestion of
differentiation between dinosaur tissues similar to both extant (Fig. 3H) and fossil tissues (Fig. 3D);
that seen in extant tissues, and suggesting preserva- providing further support for the presence of col-
tion of the original chemistry in these ancient tissues. lagen II, and confirming the specificity of antibod-
ies used in this study. As an additional specificity
control, fossilized cartilage was exposed to anti-
Immunofluorescence and bodies raised against avian collagen I, the domi-
Immunohistochemistry nant protein in bone [10]. Because extant primary
cartilage does not usually express collagen I [4],
Immunohistochemistry supports the persistence no binding was expected, and none was observed
of protein epitopes consistent with collagen II in either Hypacrosaurus (Fig. 3J) or emu cartilage
(Fig. 3), the most abundant protein comprising (Fig. 3L). To control for non-specific binding of the
the ECM of extant cartilage [4]. Ultra-thin sections secondary antibody or fluorescent label, primary an-
of demineralized Hypacrosaurus cartilage were ex- tibodies were omitted, while keeping all other con-
posed to antibodies raised against avian collagen II. ditions identical; no cross-reactivity was observed
Green fluorescence on tissues reflects the location (Supplementary Fig. 2). The most parsimonious ex-
of antibody-antigen complexes (Fig. 3B). In extant planation for these results is that epitopes of collagen
emu cartilage, binding is present throughout the II are preserved in this 75 million year-old dinosaur.
entire matrix (Fig. 3F). Binding in fossil cartilage is Collagen II is not produced by microbes; positing a
represented by a more globular pattern (Fig. 3B), microbial source is not parsimonious or congruent
suggesting that preserved collagen II epitopes are with the data.
not homogenously distributed in the ECM of this
Hypacrosaurus. Furthemore, immunoreactivity in
Hypacrosaurus cartilage is diminished when com-
pared to emu (Fig. 3F), as illustrated by longer in-
Isolation of chondrocytes and DNA
tegration times and fainter signal. This may indicate assays
that fewer epitopes are preserved in the ancient tis- Although osteocytes have previously been isolated
sues, or alternatively, that there is more phylogenetic from dinosaur bone [19,20], here, we show the first
distance between the dinosaur proteins and chicken isolated dinosaur chondrocytes (Fig. 4). Unlike di-
proteins used to generate the antiserum. If the latter, nosaur osteocytes that often present a reddish hue
4 Natl Sci Rev, 2020, Vol. 0, No. 0 RESEARCH ARTICLE

Anti-Chick Collagen II Ab Anti-Chick Collagen II Ab after Collagenase II Although the cells appeared empty under trans-
mitted light (as do emu chondrocytes) (Fig. 4A
Hypacrosaurus

and B), we tested these microstructures for the

Downloaded from [Link] by Jamia Millia Islamia University user on 30 April 2020
presence of chemical markers consistent with DNA
Ultrathin sections - 200 nm

[20] using two complementary histochemical stains,

propidium iodide (PI) [23] and 4 ,6 -diamidino-2-
phenylindole dihydrochloride (DAPI) [24]. PI in-
Emu

tercalates between every four to five base pairs of

double-stranded DNA, with little or no sequence
Anti-Chick Collagen I Ab
preference, and is detected in red frequencies when
Hypacrosaurus

stimulated by fluorescent light [23]. PI does not

Emu
stain DNA in a living cell, but only in dead cells.
Therefore, positive PI staining cannot arise from
contamination with living (i.e., microbial) cells [25].
Figure 3. Immunohistochemical staining of Hypacrosaurus cartilage. (A, C, E, G, I, K) are DAPI binds preferentially to double-stranded DNA
overlay images showing cartilage and localized binding; (B, D, F, H, J, L) are fluorescent in both living and dead cells. It is sequence depen-
images using FITC label. Hypacrosaurus calcified cartilage (A, B) shows positive, local- dent requiring at least three successive AT base pairs
ized staining when exposed to antibodies (Ab) raised against avian Collagen II, with as a binding site [24].
green fluorescent signal representing antibody-antigen complexes arranged globularly Specific staining of both PI (Fig. 4C) and DAPI
in the extracellular matrix. Immunoreactivity is diminished, as illustrated by longer inte- (Fig. 4D) is observed inside the isolated carti-
gration time and fainter signal, when compared to calcified and hyaline cartilage from lage cells of Hypacrosaurus, following the pattern
an emu (E, F). Antibody reactivity was decreased after collagenase II digestion in Hy- seen in extant cells (Fig. 4F and G), but dimin-
pacrosaurus (C–D) and emu cartilage (G–H), demonstrating that reactivity to Collagen ished in the ancient ones. This not only supports
II is specific for epitopes of that protein. Hypacrosaurus (I, J) and emu cartilage (K, L)
that the compound within these cells is chemically
shows no staining when exposed to antibodies rained against avian Collagen I. Images
are at the same scale.
consistent with DNA, but that material is dou-
ble stranded, and of a minimum length of 6 base
Unstained PI stain DAPI stain pairs [24]. These molecules must be reduced in
concentration relative to the extant ones because
Hypacrosaurus

of the greatly reduced area stained; this pattern

Isolated chondrocytes

was also observed with dinosaur osteocytes (in

Tyrannosaurus rex and Brachylophosaurus canaden-
sis) [20]. Typical taphonomic alteration of DNA
involves, among other processes, backbone break-
age [26], a phenomenon expected to have occurred
Emu

in cells of this antiquity. Additionally, because the

PI stain in the fossil cells localizes to such a small
area, it may indicate that the nuclear material was
in a condensed state at the time of death. DNA
Figure 4. Isolated chondrocytes of Hypacrosaurus and their positive response to two condensation occurs naturally during mitotis, inter-
DNA assays. (A, B, E) Isolated chondrocytes of Hypacrosaurus and emu photographed phase, and all types of cell death (Supplementary
under transmitted light (green arrows). Hypacrosaurus chondrocytes were success- Text 2) [27,28]. Because the isolated cells of Hy-
fully isolated as individual cells (A) and cell doublets (B). Hypacrosaurus (C) and emu
pacrosaurus come from the chondro-osseous junc-
chondrocytes (F) showing positive response to propidium iodide (PI), a DNA interca-
tion, where most chondrocytes eventually die, it is
lating dye, to a small and circular region that locates intracellularly (white arrows).
Hypacrosaurus (D) and emu chondrocytes (G) also show a similar binding when ex- possible that the cells showing PI and DAPI staining
posed to 4 ,6 -diamidino-2-phenylindole dihydrochloride (DAPI), another DNA-specific (Fig. 4C and D) were senescent (e.g., chondroptotic,
stain (black arrows) although in both cases, emu cell staining is significantly greater or apoptotic cells), with condensed nuclear material.
than in the dinosaur cells. Extant senescent chondrocytes can show different
morphologies of condensed nuclear material (with
due to iron inclusions [21], Hypacrosaurus chondro- or without a fragmented nuclear membrane), such
cytes are transparent (Fig. 4A and B), suggesting as condensed chromatin granules, a chromosome-
a different preservation mode. They lack filopodia like arrangement, or multiple apoptotic condensa-
and present the round morphology consistent with tions (Supplementary Text 2, 3).
all vertebrate chondrocytes [22] (including emu, An alternative hypothesis, that the staining arises
Fig. 4E), but distinct from the fusiform shape of os- from microbial contaminant, is not supported (con-
teocytes [20]. A few cell doublets were also isolated tra [25,29]); there is no mechanism for exoge-
in Hypacrosaurus (e.g., Fig. 4B). nous DNA to penetrate an intact membrane and
RESEARCH ARTICLE Bailleul et al. 5

localize to a single point specifically inside the cell, Here, data derived from dinosaurian calcified car-
demonstrating no reactivity in any other region. tilage add to the growing evidence across specimens
We caution that only the sequencing of the mate- and biomolecules that original, endogenous organ-

Downloaded from [Link] by Jamia Millia Islamia University user on 30 April 2020
rial reactive to both PI and DAPI can confirm that ics persist into deep time under exceptional con-
it is dinosaurian in origin, however the combined ditions. Interestingly, all the materials collected at
data at the histological, cellular and molecular lev- this nesting ground were disarticulated, suggesting
els (Figs 1–4) robustly support the hypothesis that that a phenomenon other than rapid burial allowed
the cartilage of Hypacrosaurus has remnants of orig- such exquisite preservation (Supplementary Text
inal chondrocytes, original nuclear material, and en- 1). Other juvenile hadrosaur material from the Two
dogenous compounds chemically consistent with Medicine formation showed exquisite preservation
DNA. In addition, these data support that structures of cartilage [41]. Calcified cartilage may represent
morphologically consistent with nuclei and chromo- a better candidate than bone for biomolecule
somes, first seen in petrographic ground sections preservation, because it exhibits multiple factors
(Fig. 1; Supplementary Fig. 1) are endogenous to that contribute to molecular stabilization [42].
this dinosaur. Although it has been suggested that These include an ECM without vascularization,
similar, cell-like structures recovered in dinosaur making it less porous, with less surface area available
bones could be the result of biofilm infiltration [25], for ground water and microbes that result in biode-
the pattern of reactivity observed when biofilm was terioration; an ECM with a higher mineral:organic
exposed to DAPI and PI staining during a previ- ratio than that of bone [43]; and hypoxia [44].
ous study [20] is inconsistent with the one observed Oxygen levels are lowest specifically in the calcified,
here. It is reasonable and logical to propose that hypertrophic chondrocyte zone, which may serve to
fossil dinosaur bone contains contaminating micro- sequester cells, preventing oxidative damage.
bial communities [29], but the specific case that we
present here, where isolated chondrocytes show in-
tracellular reactivity with PI and DAPI (Fig. 4), does
CONCLUSIONS
not match the staining pattern of ‘cell clusters’ of The identification of chemical markers of DNA in
contaminating biofilms [29]. Hypacrosaurus suggest it may preserve much longer
We have shown that cartilage found in MOR than originally proposed [30,31]. Even though
548 is extremely well preserved, not only at the it is clear that contamination does exist in fossil
histological level with an unreported, sub-cellular material and complicates identifications of original
type of preservation (Fig. 1; Supplementary Fig. 1), organic molecules, it can be accounted for with
but also at the cellular and the molecular levels proper controls. Contamination is not a plausible
(Figs 2–4). This study provides the first clear chem- explanation in this case, and to this date, the possible
ical and molecular demonstration of calcified carti- preservation of original proteins and DNA in deep
lage preservation in Mesozoic skeletal material, and time has not been convincingly eliminated with
suggests that in addition to cartilage-specific colla- data. Although extensive research and sequencing
gen II, DNA, or at least the chemical markers of is required to further understand DNA preservation
DNA (for example, chemically altered base pairs that in Mesozoic material, along with its chemical
can still react to PI and DAPI), may preserve for and molecular alterations, our data suggest the
millions of years. preserved nuclear material in Hypacrosaurus was
The assumption of a temporal limit on molecu- in a condensed state at the time of the death of the
lar longevity has hindered the pursuit of molecular organism, which may have contributed to its stabil-
data from fossils older than ∼1 million years (MA). ity. We propose that DNA condensation may be a
A short temporal range is predicted for informative favorable process to its fossilization. Additionally, as
biomolecules (∼1 MA for proteins, and ∼100,000 was suggested for protein fossilization [20,45,46],
years for DNA; with 700,000 years as the oldest crosslinking may be another mechanism involved in
genome report) [30–32]. However, these assump- the preservation of DNA in deep time.
tions have been challenged by multiple studies on
Mesozoic fossil remains reporting evidence of chem-
METHODS
ical and organic remnants, including extracellular
proteins and pigments (e.g., [33–37]), cytoskeletal All assays using fossil material were conducted in lab
proteins [20], compounds that localize to cell inte- facilities dedicated to fossil analyses, using aseptic
riors that are chemically consistent with DNA [20; techniques, and did not at any time come in contact
and the present study] and peptide sequence data with homologous tissues from emu used as positive
including histone proteins, a protein not found in controls. The latter were analyzed in a separate lab,
bacteria [38–40]. using dedicated instruments and solutions.
6 Natl Sci Rev, 2020, Vol. 0, No. 0 RESEARCH ARTICLE

Fossil material Histochemistry/Alcian blue staining

We used two disarticulated supraoccipitals (bones Fragments taken from the periphery of the supraoc-
of the basicranium) of Hypacrosaurus stebingeri cipital of MOR 548 were chosen to optimize chances

Downloaded from [Link] by Jamia Millia Islamia University user on 30 April 2020
(MOR 548) that were similar in size and preserva- of obtaining mostly cartilage, rather than bone. Emu
tion state. The first was embedded in polyester resin samples also included calcified cartilage from the
and prepared at the MOR in the 1980s (Fig. 1B–D). supraoccipital-exoccipital synchondrosis, as well as
Because resin interferes with some molecular some underlying bone.
methods, a second non-embedded supraoccipital These fragments from extant emu were fixed with
(Fig. 1A) from the same fossil site was chosen to per- 10% NBF overnight, then demineralized in 500 mM
form chemical and molecular analyses (Figs 2–4). EDTA (disodium ethylenediaminetetraacetic acid)
Based on their size, these elements were estimated (pH 8.0) until all mineral was removed, and again
to belong to nestlings with a skull length of about subjected to fixation as above. Because of the frag-
20 cm, and an overall full length of 2 m [1,3]. Adults ile nature of demineralized fossil tissues, MOR
of this species usually reach a full length of 9 m [47]. 548 samples were embedded in 3% agar (Becton
Dickinson Cat# 214530) to stabilize the tissues
prior to sectioning. Extant tissues and agar embed-
Extant material ded tissues from MOR 548 were then subjected
to routine preparation for paraffin histology, em-
We used a total of four juvenile emus (Dromaius bedded in paraffin, cut on a microtome at 5 mi-
novaehollandiae) donated cadaveric to the MOR by crons, and stained with Alcian blue (Supplementary
Montana Emu Ranch, between a few weeks and a Methods).
few months post-hatching (between 8–10 months).
Two were kept frozen, and two were sent to a der-
mestid beetle colony and defleshed for other stud- Immunofluorescence/
ies on emu skulls [48] (in emu skulls, most of the Immunohistochemistry
hyaline, unmineralized cartilage is eaten by bee-
tles, leaving mostly calcified cartilage). Data for the Fossil fragments were first embedded in 3% agar
ground section (Fig. 1E–G) is from the emu skull (Becton Dickinson Cat# 214530), then demineral-
MOR OST 1802 [48]. Histochemical, immuno- ized in EDTA (0.5 M, pH 8.0) for two weeks, washed
logical and DNA staining were performed on the with 1× phosphate buffered saline (PBS) and
other three emu specimens at the supraoccipital- prepared for further embedding and sectioning for
exoccipital junction (homologous to the calcified analyses (Supplementary Methods). In a separate
cartilage region of interest in Hypacrosaurus). laboratory, cartilage tissues were sampled from a dry
emu skull fragment (MOR-OST 1800), fixed for one
hour at room temperature in 10% NBF (Supplemen-
tary Methods).
Ground sections Sections (200 nm) were cut on a Leica EM
One supraoccipital of MOR 548 was embedded UC6 Ultramicrotome and exposed to a Rabbit anti-
whole in polyester resin (according to standard chicken collagen type II (Abcam ab21290) primary
methods) [49]. A relatively thick slice taken using antibody (Supplementary Methods). For antibodies
a Buehler Isomet 1000 precision saw was attached raised against avian collagen I, we used the antibody
to a glass slide with epoxy and ground to desired US Biological C7510–13B. Sections were also incu-
thickness (100 microns) with a Buehler Ecomet bated in antibody dilution buffer only, to which no
Grinder. The slide (Fig. 1B–D) was studied by light primary antibodies were added, to control for spuri-
microscopy under transmitted light with a Nikon ous binding of the secondary antibody (Supplemen-
Optiphot-Pol polarizing microscope. Photographs tary Fig. 2).
were taken with a Nikon DS-Fi1 digital sight camera
and the NIS ELEMENTS BR 4.13 software.
For comparison with extant emu material Isolation of chondrocytes and DAPI/PI
(Fig. 1E–G), fragments including basicranial carti- Staining
laginous joints (synchondroses) from the skull of Fragments were demineralized in EDTA (0.5 M, pH
MOR OST 1802 were extracted with a dremel and 8.0) for at least 2 weeks with daily changes. Result-
diamond blade, fixed in 10% neutral buffered for- ing debris after demineralization were collected from
malin (NBF), and further prepared for embedding the bottom of the wells and centrifuged for 2 min
and stained with Toluidine blue (Supplementary at 400 rcf. After removing the supernatant, the pel-
Methods). lets were re-suspended in 1× PBS to remove the
RESEARCH ARTICLE Bailleul et al. 7

remaining EDTA and the process was repeated three lyzed cellular data; all authors analyzed and interpreted some
times. After final centrifugation, the pellets were in- results; J.R.H. designed original paleohistological study; J.R.H.,
cubated with the iron chelator pyridoxal isonicoti- M.H.S. contributed materials/reagents and laboratory space;

Downloaded from [Link] by Jamia Millia Islamia University user on 30 April 2020
noyl hydrazone (PIH) solution (10 mM PIH in A.M.B, M.H.S wrote the manuscript; A.M.B, M.H.S, W.Z. wrote
the supplementary materials; J.R.H., B.K.H., C.M.H. commented
50 mM NaOH [50]) overnight at room temper-
on the manuscript.
ature, then washed with 1× PBS and centrifuged
to form cell pellets, used for DAPI/PI staining Conflict of interest statement. None declared.
(Supplementary Methods).
Emu cartilage layers were excised from the
supraoccipital-exoccipital synchondrosis of a young REFERENCES
emu (but more cartilage was needed so femoral 1. Horner JR and Currie PJ. Embryonic and neonatal morphology
cartilage was excised as well), incubated in 0.1% and ontogeny of a new species of Hypacrosaurus (Ornithischia,
Sodium Azide for 1.5 hours then cut into 0.5–1 mm Lambeosauridae) from Montana and Alberta. In: Carpenter K,
slices with a sterile razor blade. Slices were then Hirsch KE and Horner JR (eds). Dinosaur Eggs and Babies. New
digested with 3 mg/mL collagenase type II (Wor- York: Cambridge University Press, 1994, 312–36.
thington CLS-2) in Dulbecco’s phosphate-buffered 2. Varricchio DJ and Horner JR. Hadrosaurid and lambeosaurid
saline (D-PBS) pH 7.2 with 0.1% Sodium Azide at bone beds from the Upper Cretaceous Two Medicine Formation
37◦ C overnight or until the majority of the cartilage of Montana: taphonomic and biologic implications. Can J Earth
matrix was removed. The samples were centrifuged Sci 1993; 30: 997–1006.
at 300 rcf for 5 min to pellet cells. Cells were re- 3. Bailleul AM, Hall BK and Horner JR. First Evidence of Di-
suspended in 10% NBF (pH 7.2–7.4) at room tem- nosaurian Secondary Cartilage in the Post-Hatching Skull of
perature for 30 min to fix, then centrifuged as be- Hypacrosaurus stebingeri (Dinosauria, Ornithischia). PLoS One
fore, and washed twice with PBS. Cells were pelleted 2012; 7: e36112.
and subjected to DAPI/PI staining (Supplementary 4. Hall BK. Chapter 3 - Cartilage. In: Bones and Cartilage. San
Methods). Diego: Academic Press, 2005, 33–47.
5. Roach HI, Erenpreisa J and Aigner T. Osteogenic differentiation
of hypertrophic chondrocytes involves asymmetric cell divisions
SUPPLEMENTARY DATA and apoptosis. J Cell Biol 1995; 131: 483–94.
6. Nurse P. A long twentieth century of the cell cycle and beyond.
Supplementary data are available at NSR online.
Cell 2000; 100: 71–8.
7. Bomfleur B, McLoughlin S and Vajda V. Fossilized nuclei and
chromosomes reveal 180 million years of genomic stasis in royal
ACKNOWLEDGEMENTS ferns. Science 2014; 343: 1376–7.
We thank Zhonghe Zhou for fruitful discussions and comments. 8. Roughley PJ and Lee ER. Cartilage proteoglycans: structure and
We thank John Scannella, the MOR, the Gabriel Lab for Cellular potential functions. Microsc Res Tech 1994; 28: 385–97.
and Molecular Paleontology, Ellen-Thérèse Lamm, Karen Chin 9. Wang X, Ford BC and Praul CA et al. Characterization of the non-
and Allison Gentry for laboratory help; Don Collins (Montana
collagenous proteins in avian cortical and medullary bone. Comp
Emu Ranch) for donating cadaveric emus; Elena Schroeter and
Biochem Physiol B Biochem Mol Biol 2005; 140: 665–72.
Dana Rashid for discussions; and Mikhail Kovalenko (University
10. Hall BK. Chapter 2 - Bone. In: Bones and Cartilage. San Diego:
of Missouri) for whole-slide imaging.
Academic Press, 2005, 13–32.
11. Roach H, Aigner T and Kouri J. Chondroptosis: a variant of apop-
FUNDING totic cell death in chondrocytes? Apoptosis 2004; 9: 265–77.
12. Farquharson C and Jefferies D. Chondrocytes and longitudinal
This work was supported by the National Science Foundation
bone growth: the development of tibial dyschondroplasia. Poult
INSPIRE grant (Division of Earth Sciences) (EAR-1344198 to
Sci 2000; 79: 994–1004.
M.H.S. and W.Z); Franklin M. and Susan Packard Orr; Vance and
Gail Mullis (to M.H.S. and W.Z); Gerry Ohrstrom (to A.M.B. and 13. Erenpreisa J and Roach HI. Aberrant death in dark chondrocytes
J.R.H.); the American Association of Anatomists Short-Term Vis- of the avian growth plate. Cell Death Differ 1998; 5: 60–6.
iting Scholarship (to A.M.B.); the Chinese Academy of Sciences- 14. Ērenpreisa J and Zhukotsky A. Interphase genome as the active
President’s International Fellowship Initiative (CAS-PIFI) (to space: chromatin dynamics during chick embryo chondrogene-
A.M.B) and NSF-IOS grant (Division of Integrative Organismal sis. Mech Ageing Dev 1993; 67: 21–32.
Systems) (1457319 to C.M.H and A.M.B). 15. Erenpreisa J and Roach HI. Aberrations of cell cycle and cell
death in normal development of the chick embryo growth plate.
Mech Ageing Dev 1999; 108: 227–38.
AUTHOR CONTRIBUTIONS 16. Schweitzer MH, Zheng W and Zanno L et al. Chemistry supports
A.M.B, M.H.S. designed the study; A.M.B, W.Z. performed ex- the identification of gender-specific reproductive tissue in Tyran-
periments, and collected the data; A.M.B, B.K.H, M.H.S. ana- nosaurus rex. Sci Rep 2016; 6: 23099.
8 Natl Sci Rev, 2020, Vol. 0, No. 0 RESEARCH ARTICLE

17. Bailleul AM, Li Z and O’Connor J et al. Origin of the avian predentary and evi- 35. Lindgren J, Uvdal P and Engdahl A et al. Microspectroscopic evidence of Cre-
dence of a unique form of cranial kinesis in Cretaceous ornithuromorphs. Proc taceous bone proteins. PLoS One 2011; 6: e19445.
Natl Acad Sci USA 2019; 116: 24696–706. 36. Pan Y, Zheng W and Moyer AE et al. Molecular evidence of keratin and

Downloaded from [Link] by Jamia Millia Islamia University user on 30 April 2020
18. Buschmann MD, Gluzband YA and Grodzinsky AJ et al. Mechanical compres- melanosomes in feathers of the Early Cretaceous bird Eoconfuciusornis. Proc
sion modulates matrix biosynthesis in chondrocyte/agarose culture. J Cell Sci Natl Acad Sci USA 2016; 113: E7900–7.
1995; 108: 1497–508. 37. Wiemann J, Yang T-R and Norell MA. Dinosaur egg colour had a single evolu-
19. Schweitzer MH, Wittmeyer JL and Horner JR et al. Soft-tissue vessels and tionary origin. Nature 2018; 563: 555–8.
cellular preservation in Tyrannosaurus rex. Science 2005; 307: 1952–5. 38. Asara JM, Schweitzer MH and Freimark LM et al. Protein sequences from
20. Schweitzer MH, Zheng W and Cleland TP et al. Molecular analyses of dinosaur mastodon and Tyrannosaurus rex revealed by mass spectrometry. Science
osteocytes support the presence of endogenous molecules. Bone 2013; 52: 2007; 316: 280–5.
414–23. 39. Cleland TP, Schroeter ER and Zamdborg L et al. Mass spectrometry and
21. Schweitzer MH, Wittmeyer JL and Horner JR. Soft tissue and cellular preser- antibody-based characterization of blood vessels from Brachylophosaurus
vation in vertebrate skeletal elements from the Cretaceous to the present. Proc canadensis. J Proteome Res 2015; 14: 5252–62.
R Soc B 2007; 274: 183–97. 40. Schroeter ER, DeHart CJ and Cleland TP et al. Expansion for the Brachy-
22. Smith MM and Hall BK. Development and evolutionary origins of vertebrate lophosaurus canadensis collagen I sequence and additional evidence
skeletogenic and odontogenic tissues. Biol Rev 1990; 65: 277–373. of the preservation of Cretaceous protein. J Proteome Res 2017; 16:
23. Yeh C-JG, Hsi B-L and Faulk WP. Propidium iodide as a nuclear marker in 920–32.
immunofluorescence. II. Use with cellular identification and viability studies. 41. Barreto C, Albrecht RM and Bjorling DE et al. Evidence of the growth plate and
J Immunol Methods 1981; 43: 269–75. the growth of long bones in juvenile dinosaurs. Science 1993; 262: 2020–3.
24. Kapuscinski J. DAPI: a DNA-specific fluorescent probe. Biotech Histochem 42. Collins MJ, Nielsen–Marsh CM and Hiller J et al. The survival of organic matter
1995; 70: 220–33. in bone: a review. Archaeometry 2002; 44: 383–94.
25. Kaye TG, Gaugler G and Sawlowicz Z. Dinosaurian soft tissues interpreted as 43. Goret-Nicaise M and Dhem A. Comparison of the calcium content of different
bacterial biofilms. PLoS One 2008; 3: e2808. tissues present in the human mandible. Acta Anat (Basel) 1985; 124: 167–72.
26. Hofreiter M, Serre D and Poinar HN et al. Ancient DNA. Nat Rev Genet 2001; 44. Haselgrove JC, Shapiro IM and Silverton SF. Computer modeling of the oxygen
2: 353–9. supply and demand of cells of the avian growth cartilage. Am J Physiol 1993;
27. Roach H and Clarke N. Physiological cell death of chondrocytes in vivo is 265: C497–506.
not confined to apoptosis: new observations on the mammalian growth plate. 45. Wiemann J, Fabbri M and Yang T-R et al. Fossilization transforms verte-
J Bone Joint Surg Br 2000; 82: 601–13. brate hard tissue proteins into N-heterocyclic polymers. Nat Commun 2018; 9:
28. Swedlow JR and Hirano T. The making of the mitotic chromosome: modern 4741.
insights into classical questions. Mol Cell 2003; 11: 557–69. 46. Schweitzer MH, Zheng W and Cleland TP et al. A role for iron and oxygen chem-
29. Saitta ET, Liang R and Lau MC et al. Cretaceous dinosaur bone contains recent istry in preserving soft tissues, cells and molecules from deep time. Proc R Soc
organic material and provides an environment conducive to microbial commu- B 2014; 281: 20132741.
nities. eLife 2019; 8: e46205. 47. Lull R and Wright N. Hadrosaurian Dinosaurs of North America: Geological So-
30. Lindahl T. Instability and decay of the primary structure of DNA. Nature 1993; ciety of America Special Papers No. 40. 1942.
362: 709–15. 48. Bailleul AM and Horner JR. Comparative histology of some craniofacial sutures
31. Willerslev E and Cooper A. Ancient DNA. Proc R Soc B 2005; 272: 3–16. and skull-base synchondroses in non-avian dinosaurs and their extant phyloge-
32. Orlando L, Ginolhac A and Zhang G et al. Recalibrating Equus evolution using netic bracket. J Anat 2016; 229: 252–85.
the genome sequence of an early Middle Pleistocene horse. Nature 2013; 499: 49. Lamm E-T. Preparation and Sectioning of Specimens. In: Padian K and Lamm
74–8. E-T (eds). Bone Histology of Fossil Tetrapods: Advancing Methods, Analysis,
33. Schweitzer MH, Suo Z and Avci R et al. Analyses of soft tissue from Tyran- and Interpretation. Berkeley and Los Angeles: University of California Press,
nosaurus rex suggest the presence of protein. Science 2007; 316: 277–80. 2013, 55–160.
34. Lindgren J, Kuriyama T and Madsen H et al. Biochemistry and adaptive coloura- 50. Buss JL, Neuzil J and Gellert N et al. Pyridoxal isonicotinoyl hydrazone analogs
tion of an exceptionally preserved juvenile fossil sea turtle. Sci Rep 2017; 7: induce apoptosis in hematopoietic cells due to their iron-chelating properties.
13324. Biochem Pharmacol 2003; 65: 161–72.