Subscriber access provided by UNIV DE GRANADA
Crystal Growth and Design is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Article
Identification of Some Active Proteins in
the Process of Hen Eggshell Formation†
A. Herna#ndez-Herna#ndez, J. Go#mez-Morales, A. B.
Rodri#guez-Navarro, J. Gautron, Y. Nys, and J. M. Garci#a-Ruiz
Cryst. Growth Des.,
2008, 8 (12), 4330-4339 • Publication Date (Web): 04 November 2008
Downloaded from http://pubs.acs.org on December 11, 2008
More About This Article
Additional resources and features associated with this article are available within the HTML version:
•
Supporting Information
•
Access to high resolution figures
•
Links to articles and content related to this article
•
Copyright permission to reproduce figures and/or text from this article
Identification of Some Active Proteins in the Process of Hen
Eggshell Formation†
A. Hernández-Hernández,‡ J. Gómez-Morales,*,‡ A. B. Rodrıguez-Navarro,§ J. Gautron,|
Y. Nys,| and J. M. Garcıa-Ruiz‡
Laboratorio de Estudios Cristalográficos, IACT (CSIC-UGRA), Edificio Instituto López Neyra, AV
da.
Conocimiento, s/n. P.T. Campus Salud, 18100, Armilla (Granada), Spain, Departamento de
Mineralogıa y Petrologıa, UniV
ersidad de Granada, Campus FuentenueV
a, s/n 18002 Granada, Spain,
and INRA, UR83, Recherches AV
icoles, 37380 Nouzilly, France
ReceiV
ed July 20, 2008; ReV
ised Manuscript ReceiV
ed October 2, 2008
ABSTRACT: The hen eggshell is formed in a three-stage sequential process, namely, the initial stage (nucleation of calcite crystals),
the active growth phase (linear deposition), and the terminal stage (inhibition of crystal formation). During these phases, different
proteins are sequentially expressed in the uterine fluid. Some of them are thought to regulate the shell mineralization and particularly
crystal growth. To identify proteins that are actively involved in this process, we have analyzed and compared the effects on CaCO3
precipitation in vitro of some commercial egg white proteins, some noncommercial purified fractions of the eggshell organic matrix
and, also of uterine fluids extracted from the hen uterus during each one of the stages of calcification. Uterine fluids strongly increased
the nucleation density with respect to commercial proteins and purified fractions of the shell. A few identified acidic proteins
(ovocleidin-17, ovocalyxin-32, osteopontin and ovocleidin-116) showed a strong affinity for the CaCO3 surfaces and were selectively
removed from the solution during its precipitation. These proteins may have an active role on CaCO3 growth, aggregation, and
inhibition. Other proteins, with very different isoelectric points, seem to regulate the chemical environment in which the precipitation
takes place, that is, by buffering the pH favoring crystal growth as the couple ovocleidin-17 and ovocalyxin-21.
Introduction
The hen eggshell is a calcified structure which functions as
a protective barrier for the egg content and allows the extrau-
terine development of the chick embryo.1,2 It consists mainly
of a mineral part (>95%) made of calcite crystals and a
pervading organic matrix (1-3.5%), making a composite
material with excellent mechanical properties.2-4 It is deposited
in a very short time (less than 20 h), while the egg passes
through the hen’s uterus (shell gland). The eggshell formation
is a very well regulated spatio-temporal process, resulting in a
material with very well defined compositional and ultrastruc-
tural/microstructural characteristics which are common to all
avian species. It is also a useful model to understand biomin-
eralization processes in other biological systems. Three different
stages during eggshell formation can be differentiated: (a) initial,
(b) active-fast growth, and (c) terminal.
It is normally assumed, based on morphological features, that
the avian eggshell is composed of six layers. However, only
the two innermost non-calcified layers (the inner and outer shell
membranes made of a network of organic fibers) are well
differentiated. In the calcified part, there is continuity across
its thickness as a result of the continuous mode of eggshell
deposition. The inner zone of the calcified shell is made of
irregular cones (mammillary knob layer), the tips of which are
penetrated by the outer membranes fibers. During the initial
stage of eggshell formation (stage I) the deposition of the
mammillary knobs on the outer shell membrane and the sub-
sequent nucleation of calcite spheruliths take place. The growth
continues until adjacent spheruliths fuse together and from there
columnar calcite units arise as the lateral growth of crystals is
constrained. This type of process in a restricted space is
denominated competitive crystal growth as crystals compete for
the limited available space.5 Columnar calcite crystals extend
beyond the bases of the cones during the fast growth stage (stage
II) forming the palisade layer. Columnar crystal growth proceeds
until the arrest of calcification, which is characterized by the
deposition of a thin vertical crystals layer and the cuticle (stage
III). The vertical crystal layer is composed of small crystallites
aligned in a perpendicular form to the shell surface. The cuticle
is a rich organic coating containing hydroxyapatite in its inner
zone6 and pigments.2,7
The mineral part of the shell is made of calcite crystals
precipitating from the uterine fluid, which is an acellular milieu
secreted by the distal part of the oviduct. This fluid contains
calcium and bicarbonate ions greatly in excess of the solubility
product of calcite, as well as other ionic species such as
magnesium and phosphate and the organic precursor components
of the eggshell organic matrix.8 Organic components comprise
proteins, glycoproteins, and proteoglycans. Among the numerous
components identified are egg white proteins as ovalbumin,
lysozyme and ovotransferrin, ubiquitous components such as
osteopontin and clusterin9-12 and organic constituents unique
to the process of shell calcification. This last group comprises
dermatan and keratan proteoglycans,3,13 ovocleidins and
ovocalyxins.9,14,15 A recent proteomic survey of the acid-soluble
organic matrix of the calcified chicken eggshell layer has
allowed the identification of several hundred proteins in the
shell.16 The number of these organic components and their
concentration change in the uterine fluid along the different
stages of eggshell deposition in a well defined way.17 In each
stage specific organic components are expressed at a given
concentration. Also, these organic components are secreted at
specific times and locations in the oviduct and incorporated at
specific substructural regions of the eggshell. For instance, type
X collagen is mostly found at the eggshell membranes, keratan
† Part of the special issue (Vol 8, issue 12) on the 12th International
Conference on the Crystallization of Biological Macromolecules, Cancun,
Mexico, May 6-9, 2008.
* To whom correspondence should be addressed. E-mail: jaime@lec.csic.es.
‡ IACT (CSIC-UGRA).
§ Universidad de Granada.
| INRA.
CRYSTAL
GROWTH
& DESIGN
2008
VOL. 8, NO. 12
4330–4339
10.1021/cg800786s CCC: $40.75
2008 American Chemical Society
Published on Web 11/04/2008
sulfate proteoglycans are the main components of the mam-
millary knobs,18,19 and osteopontin has been found in eggshell
associated with particular crystallographic faces of calcite in
the palisade region.20 On the other hand, in vitro precipitation
tests show that some of these components influence calcium
carbonate precipitation. In particular, they affect the nucleation
flux, the polymorph selection, the crystal size and the morphol-
ogy. For instance, eggshell membrane collagen and osteopontin
are known to inhibit calcite crystal growth, whereas mammillary
keratan sulfate proteoglycan is involved in calcite nucleation.
It is becoming commonly accepted that macromolecules
contribute to the regulation of the eggshell mineralization
process. However, the mechanisms involved in the interaction
of the organic macromolecules with the inorganic phase have
not yet been fully elucidated. It is not only the action of one
macromolecule in a given condition that must be established,
but also the spatio-temporal interactions of a big array of
macromolecules.21 These biological macromolecules differ in
molecular weight, isoelectric point (pI), and concentration. It
has been demonstrated by in vitro experiments that model
globular proteins with different pI and varying concentrations
influence CaCO3 precipitation at different levels, that is, during
nucleation, polymorph selection, crystal growth, and crystal
morphology.22
In order to find evidence of the implication of macromolecules
in eggshell formation and to identify some of the macromol-
ecules actively involved in the process, we decided to analyze
and compare the effects on calcium carbonate precipitation of
macromolecular solutions with different levels of complexity,
that is, (1) egg white proteins which are commercially available,
(2) purified eggshell matrix extracts (results previously pub-
lished),23 and (3) uterine fluids collected at the three main stages
of eggshell formation. In particular, we have analyzed the effects
on the nucleation, crystal growth, polymorphism and morphol-
ogy of the precipitated crystals, and then, we have compared
the morphology of synthetic crystals obtained in vitro with those
present in vivo in various eggshell regions.
Materials and Methods
Organic Macromolecular Solutions. In a first stage, we have used
three pure commercial egg white proteins, that is, hen egg white
lysozyme (Seikagaku Lot E98301, pI 11.35, MW 14313 Da), hen egg
white ovalbumin (Sigma, lot 93H7105, pI 4.5-4.6, MW 45000 Da)
and ovotransferrin (type I chicken egg white Lot 107F8020, pI 6.1,
MW 77000 Da). Stock solutions of the three proteins were prepared
and their concentrations were measured by the Layne method,24 using
a Varian Cary UV-vis spectrometer at a wavelength of 280 nm. These
stock solutions were used to prepare protein bearing experiments. The
concentration of CaCl2 solution was kept constant at 20 mM, meanwhile
the concentration of each one of the three proteins varied (16, 128,
512, 2000, and 5000
μg/mL). In a second stage we have analyzed
previous results of the influence on CaCO3 precipitation of some purified
fractions of the eggshell organic matrix23 and used these results for
their comparison with those of this work. All the purified fractions
were an array of proteins and other macromolecular components except
one of them which was composed mainly of ovocleidin-116 (OC-116).
In a third stage we have used as protein bearing solutions uterine fluids
extracted at the three main stages of the eggshell formation. Uterine
fluids were extracted as already described.17 Laying hens (ISA
BROWN) were housed individually in cages located in a windowless,
air conditioned poultry house. They were subjected to a cycle of 14 h
of light: 10 h of darkness and were fed ad libitum on a layers diet.
Cages were equipped with a computerized system to record the precise
time of daily egg laying (oviposition). Ovulation was considered to
occur 0.5 h after oviposition. Eggs were expelled with an intravenous
injection of 50
μg/hen of prostaglandin (prostaglandin F2R) at 6-9 h
(initial stage, I), 14-18 h (active growth calcification phase, II) and
22-23 h (terminal phase of shell calcification, III) after the preceding
oviposition. Uterine fluid was collected immediately after the egg
expulsion by gravimetry into a plastic test tube placed at the entrance
of the everted vagina. An aliquot of these three fluids was used to
analyze the total calcium concentration by atomic absorption spectros-
copy and the total protein content by the Bradford method25 using
ovalbumin as standard. Uterine fluid proteins were electrophoretically
separated using the sodium dodecyl sulfate polyacrylamide gel elec-
trophoresis (SDS-PAGE) method. Gels were loaded with 40 and 20
μL of each uterine fluid, before and after CaCO3 precipitation. Proteins
were stained with Sypro Ruby Protein Gel Stain. The identification of
stained bands displayed in the electrophoretic profile was carried out
by comparing them to previously published electrophoretic profiles.2,17
Gels were scanned using an infrared imaging system (Odyssey, Licor)
and relative protein quantification was performed as already described.26
Precipitation Experiments. Precipitation experiments were carried
out by CO2 vapor diffusion using the sitting drop crystallization method
on a “crystallization mushroom” at 20 °C and 1 atm total pressure.
Calcium carbonate precipitated in aqueous drops of 40
μL containing
the corresponding protein bearing solution and 20 mM CaCl2, (800
μg/
mL Ca), except in the case of the uterine fluids of stages I, II, and III
whose Ca concentration were higher. In addition to the added Ca (800
μg/mL) these fluids contain also natural Ca (170
μg/mL in fluid I, 370
μg/mL in fluid II and 480
μg/mL in fluid III). The pH of the drops was
raised by the diffusion of NH3 gas released by the underlying NH4HCO3
solution (10 mM) contained in the bottom part of the
mushroom. The
reagent concentrations of the control experiment (20 mM CaCl2 and
10 mM NH4HCO3) were the same as those reported in previous
works.22,23 Experiments were finished at 48 h, the mushrooms were
opened and both solutions and solids were withdrawn from the
microbridges for subsequent analysis.
Analysis. The evolution of the pH along the experiments, the
induction times for nucleation (
ti), the crystalline phase and crystalline
habit of the precipitated particles have been analyzed following the
procedures described in previous papers.22,23 The total number of
crystals (TNC) in each drop has been determined by counting the
crystals with the use of an optical microscope. For this purpose the
drops are divided in four quadrants in top view, and the cross-section
is divided into three regions: surface of the drops, wall and bottom of
the microbridges where drops are deposited.
In some cases, prior to FESEM observations, the crystals were treated
with proteinase K buffer (PKb) to highlight the inorganic material
removing all the proteins, according to the protocol described by Shiao
et al.27 When using the uterine fluids as protein bearing solutions the
remaining solutions after the experiments of CaCO3 precipitation were
also analyzed by the SDS-PAGE method. The residual calcium
concentration of these fluids was analyzed by atomic absorption
spectroscopy.
Results
Compared Precipitation of CaCO3 in Macromolecules
Bearing Solutions. In Table 1 are summarized the main results
of the precipitation experiments using commercial egg white
proteins and uterine fluids collected at the stages I, II, and III.
The results of the control experiment are also included. Figure
2 shows a general view of the drops containing CaCO3
precipitates when using the uterine fluids and the control sample.
In lysozyme, ovalbumin and ovotransferrin bearing experi-
ments,
ti were typically around 1000 min, while TNC ranged
from 50 to 130. There was always a mixture of the three main
CaCO3 polymorphs (calcite, aragonite and vaterite), though with
increasing protein concentration, the percentage of calcite tends
to increase. In contrast, in the uterine fluid bearing experiments,
ti was extremely short (a few seconds), the TNC was at least 2
orders of magnitude higher than in the other protein bearing
experiments using commercial proteins or purified eggshell
fractions (over 10 000 crystals per drop) and the crystalline phase
was always calcite.
pH Evolution in the Control and Uterine Fluid Bearing
Experiments. Figure 3 shows the pH profiles of uterine fluids
bearing experiments. It can be observed that the pH evolution
Active Proteins in Hen Eggshell Formation
Crystal Growth & Design, Vol. 8, No. 12, 2008 4331
in experiments using the uterine fluids of stages I and III
displayed a profile similar to that of the control experiment. In
these experiments the pH increased sharply until reaching the
plateau. However, the pH profile recorded during the experiment
in which we used fluid of the stage II was notably different
during the first 30 min, thus indicating a peculiar behavior of
fluid II during precipitation (Figure 3b). This different behavior
consisted of an initial decrease in the pH and then a slow
increase. Later, the pH increases sharply until reaching the
plateau. The different behavior of uterine fluid II was also
manifested when adding 2
μL of 400 mM CaCl2 to 38
μL of
fluid II during the preparation of the uterine fluid bearing
samples. This addition did not affect significantly the initial pH
of the fluid. However, the addition of the same calcium
concentration to uterine fluids I and III induced a pH decrease
of 0.38 and 1.15 units, respectively. Therefore, we suspect that
fluid II acted as a strong buffer. The buffering capacity of the
uterine fluid of stage II was also shown when using uterine fluids
of the same stage but extracted on two different days (Figure
3b′). The protein concentrations (97 and 479
μg/mL) were very
different on both days, but in both experiments the pH oscillated
to compensate the strong pH increase that the diffusion of NH3
would produce. From a chemical engineering point of view,
the uterine fluid II acted as a pH-stat system favoring the growth
of calcite at this stage.
Electrophoretic Profile (SDS-
PAGE) of the Uterine
Fluids. Figure 4 shows the electrophoretic profile of uterine
fluids, before and after CaCO3 precipitation. In Figure 4a (uterine
fluid I) the more intense stained bands (higher concentration)
correspond to egg white proteins such as ovalbumin (45 kDa)
and ovotransferrin (80 kDa). Additionally, a nonassigned band
at around 240 kDa was observed in one of the samples. At a
much reduced staining intensity, we observed a band at 15 kDa
(assigned to lysozyme). Eggshell matrix proteins specific to the
uterine tissue, such as ovocleidin-17 (OC-17), ovocalyxin-32
(OCX-32), and ovocalyxin-21 (OCX-21), were also observed.
At around 50 kDa, there is a smear of bands that could
correspond to osteopontin. After CaCO3 precipitation the bands
below 45 kDa completely disappeared, with the exception of
the lysozyme band. In contrast, the bands over 45 kDa associated
to ovalbumin, ovotransferrin and the 240 kDa bands remained
with the same staining intensity.
In the electrophoretic profile of fluid II (Figure 4b, left-hand
side) the following bands from top to bottom are observed: a
nonassigned band at 240 kDa, three closely spaced bands
corresponding to ovocleidin-116 (OC-116), ovotransferrin (weakly
Table 1. Precipitation of CaCO3 in Macromolecule Bearing Solutionsa
macromolecular
solution
Cao + Caad
(
μg/mL)
protein
(
μg/mL)
ti (min)
TNC and
polymorphs (%)
crystal
size
L (
μm)
pHi
pHf
control
800
0
≈900
40 (55% C 22% V 23% A)
100-200
5.3
7.3
lysozyme
800
128
52 (84% C 5% V 11% A)
65, 113
5.8
7.2
512
≈700
60 (84% C 6%V 10% A)
68, 113
ovalbumin
800
128
≈900
56 (58% C 26% V 16% A)
120, 180
5.0
7.0
512
1080
133 (61% C 38%V 1% A)
51, 83
5.2
7.1
ovotransferrin
800
128
1000
57 (77% C 21% V 2% A)
90, 110
5.0
7.2
512
1080
98 (88% C 12%V 0% A)
52-74, 134
b
4.9
7.2
Uterine fluid I
970
92
<0.2
.10.000 (100% C)
≈6-20
7.6
9.03
Uterine fluid II
1170
455
<0.2
.10.000 (100% C)
5, 10, 24, 52
7.43
8.25
Uterine fluid III
1280
163
<0.2
.10.000 (100% C)
≈5-22
6.5
8.44
a TNC,
ti, C, A, and V stands for total number of precipitated crystals per drop, induction time, calcite, aragonite, and vaterite, respectively. pHi
(initial pH); pHf (final pH).
b Refers to aggregates of crystals.
Figure 1. FESEM images of the cross-section of a hen eggshell (A) and details of the different structural parts formed during the three consecutive
stages of eggshell formation: (B) mammillary knobs formed at the initial stage or stage I, (C) palisade layer formed at the fast growth stage or stage
II and (D) thin vertical crystal layer covered by the cuticle at the termination stage or stage III.
4332 Crystal Growth & Design, Vol. 8, No. 12, 2008
Hernández-Hernández et al.
stained) and an intensely stained band at around 70 kDa. Below
this band, a smear of bands at around 50 kDa appear that
possibly correspond to osteopontin. Three bands can also be
observed that have been assigned to ovocalyxin-36 (OCX-36),
ovocalyxin-32 (OCX-32), and ovocleidin-17 (OC-17). Below
these bands we found that of lysozyme (weakly stained) and
another band at 13 kDa. After CaCO3 precipitation, the intensely
stained bands of OC-116, the nonassigned band at 240 kDa,
that assigned to osteopontin and the less intense band of OCX-
32 disappeared, thus indicating that these proteins were com-
pletely removed from the fluid during the precipitation process.
The electrophoretic profile of fluid III (Figure 4b, right-hand
side) displays all the bands observed in the previous fluid
samples except those corresponding to OC-116 and lysozyme.
The most intensely stained bands are the following four: two
of them within the interval from 160 to 240 kDa, one at around
70 kDa and another situated around 30 kDa, which seems to
correspond to OCX-32. An additional band was also observed
at very low molecular weight (around 10 kDa). After CaCO3
precipitation most of the bands below 66 kDa including OCX-
32 together with those placed at the top of electrophoretic profile
disappeared.
Effect on Calcite Morphology of Macromolecule Bear-
ing Solutions. Figure 5 compares the morphology of the calcite
crystals found at structurally different eggshell regions, that is,
the mammillary knobs, palisade and vertical crystal layer (Figure
5a-c) with the morphology of calcite crystals obtained from
uterine fluid bearing precipitation experiments. Most representa-
tive morphologies of calcite crystals precipitating in uterine
fluids I, II and III bearing solutions are displayed in Figure 5d-f.
FESEM observation of crystals grown in vitro in the uterine
fluid bearing experiments (Figure 5) shows the presence of
globular agglomerates and modified rhombohedral crystals
(Fluid I) which are similar to those crystals collected from the
mammillary knob at the initial stage of the eggshell formation
(Figure 5d). Aggregates of calcite obtained when using uterine
fluid of the stage II are quasi-spherical or rods made of
microcrystals elongated along the
c-axis (Figure 5e). Those
crystals obtained with Fluid III are elongated along the
c-axis,
compact (they do not show any holes) and are similar to crystals
observed at the final stage of eggshell formation which are also
elongated (Figure 5f). Some of them revealed peanut-like
morphology.
The treatment of crystals and eggshell samples with proteinase
K buffer (PKb, a nonspecific protease), allowed us to observe
the presence of 300-450 nm diameter holes on these crystals.
This finding indicates that most probably the holes contained
aggregates of proteins occluded during eggshell formation or
during the growth of crystals on in vitro experiments. The
occurrence of these holes is very common in crystals obtained
from uterine fluids collected at stage I and II which is also in
agreement with a high frequency of holes found in the
mammillary and palisade layer. In contrast, the holes are neither
Figure 2. Optical microscopy views of drops containing CaCO3 crystals precipitated in the control experiment (a) and in uterine fluids bearing
experiments of either stage I (b: surface of the drop, b′: bottom part of the drop), stage II (c: surface of the drop, c′: bottom part of the drop) and
stage III (d: surface of the drop, d′: bottom part of the drop). The diameter of drop is 6.0 mm. Panels b-d′ correspond to a quadrant of the drop.
Active Proteins in Hen Eggshell Formation
Crystal Growth & Design, Vol. 8, No. 12, 2008 4333
found at the upper part of the eggshell in the vertical layer nor
in the crystals obtained from the uterine fluid at stage III, which
produced compact aggregates of crystals.
The calcite morphologies obtained when using the com-
mercial egg white proteins lysozyme, ovalbumin and ovotrans-
ferrin at two different concentrations, 128 and 512
μg/mL, are
displayed in Figure 6a-c and Figure 6d-f, respectively. Figure
6g shows the morphologies of calcite crystals precipitated in
the presence of a purified fraction of the eggshell organic matrix
which contain mainly OC-116.
Figure 3. pH profiles of (a) uterine fluid bearing experiments of the three main stages of the eggshell formation and of the control experiment. (b)
pH-profile recorded when using the uterine fluid of the stage II and (b′) pH-profile recorded in the experiment using an uterine fluid of the stage
II collected in a different day. Insets in both figures correspond to the pH evolution during the first 30 min.
Figure 4. SDS-PAGE profiles of uterine fluids before and after CaCO3 precipitation. (a) Uterine fluids of stage I and (b) uterine fluids of stages
II and III. In the uterine fluid of the stage I, the stained bands corresponding to OC-17, OCX-21, OCX-25, OCX-32, and OCX-36 disappeared,
meanwhile those of lysozyme, ovalbumin, and ovotransferrin remain after CaCO3 precipitation. In the uterine fluid of the stage II the stained bands
of OC-116, 240 kDa, OCX-25, OCX-32 and OCX-36 disappeared while those of lysozyme, ovotransferrin, OC-17 and OC-21 remained after
CaCO3 precipitation. Finally, in the uterine fluid of the stage III the bands of OC-116, 207 kDa and OCX-32 disappeared, meanwhile those of
ovotransferrin, 10 and 66 kDa remained. There is a smear of bands around 50 kDa that could correspond to osteopontin.
4334 Crystal Growth & Design, Vol. 8, No. 12, 2008
Hernández-Hernández et al.
When using lysozyme and ovalbumin at low protein con-
centration (128
μg/mL), calcite crystals displayed rhombohedral
habits bounded by the {104} face (Figure 6a,b). These crystals
are similar to those found in the control experiment. In contrast,
the use of ovotransferrin at 128
μg/mL modified the rhombo-
hedral habit of calcite by expressing new though not very well
defined faces (Figure 6c,f). The symmetry of the newly
expressed faces in the case of ovotransferrin corresponds to that
of the {110} first-order prismatic form of calcite and to those
of second-order prismatic faces {100} as deducted by compari-
son with calcite crystal model built using SHAPE V 7.1 (Shape
Software). Those new faces were not flat and showed a terraced
morphology with well-defined microsteps advancing parallel to
{104} rhombohedral crystal faces. At higher concentration of
ovotransferrin (512
μg/mL), the edge between rhombohedral
faces were inhibited and new faces with the symmetry of {108}
rhombohedral faces were expressed. The expression of these
new faces reduced the size of the rhombohedral {104} faces of
calcite crystals. In the experiments using lysozyme and oval-
bumin a morphological effect could only be observed in the
case of ovalbumin when its concentration increased to 512
μg/
mL (Figure 6d,e). However, in the later cases the identification
of the specific crystallographic faces could not be possible
because of the large modification of the calcite morphologies.
In Figure 6e, the calcite crystal morphology is almost spherical
and only the remnant of {104} faces are shown. Finally, the
most marked effect on crystal morphology was found when
using a purified fraction of the eggshell organic matrix contain-
ing OC-116, which yielded calcite crystals as agglomerates of
nanocrystals (Figure 6g).
Discussion
To understand the role of organic components during eggshell
formation, we have used macromolecule bearing solutions with
different levels of complexity, from the simplest containing only
one type of macromolecular species to the most complex that
could simulate the natural process. In particular, we have
analyzed the effects on in vitro calcium carbonate precipitation
of (1) commercial egg white proteins also present in eggshell
matrix, and (2) uterine hen fluids extracted at the three main
stages of eggshell formation. In the first group we used three
commercially available proteins which are lysozyme, ovalbumin
and ovotransferrin. Also, we have compared our results with
previous results obtained in our laboratory about the influence
on CaCO3 precipitation of purified extracts of the eggshell
organic matrix.23 This strategy and the resulting information
give us insights into the role and effects of individual organic
components as well as their combined effects and mutual
interactions during the process of eggshell formation.
Our results reveal that individual commercial eggwhite
proteins (lysozyme, ovalbumin, ovotransferrin) at concentrations
comparable to those found in the uterine fluid have a relatively
weak effect on calcite selection, nucleation and growth. In
contrast, the purified eggshell organic matrix fractions and the
uterine fluid bearing solutions were very effective and, for
instance, selected exclusively calcite as precipitating polymorph.
Additionally, some purified fractions of the eggshell organic
Figure 5. FESEM images of (a) crystals of the mammillary layer of the eggshell deposited at the initial stage; (b) eggshell of the palisade layer
deposited during the fast growth phase; (c) eggshell of the terminal phase; (d) crystals grown in the presence of the uterine fluid of the stage I; (e)
crystals grown in the presence of the uterine fluid of the stage II; (f) crystals grown in the presence of the uterine fluid of the stage III.
Active Proteins in Hen Eggshell Formation
Crystal Growth & Design, Vol. 8, No. 12, 2008 4335
matrix studied by Hernandez-Hernandez et al.23 (i.e., fractions
g composed of ovotransferrin, ovomucoid, quiescence-specific
protein and ovocleidin-116, and the fraction
r composed of
quiescence-specific protein, ovocleidin-17, chain A, and some
nonidentified sequences) showed a remarkable ability to inhibit
the nucleation. The fraction g′, which was composed mainly of
OC-116, was very active in modifying the growth mechanism
from a spiral mechanism to an aggregation-agglomeration
growth mechanism. However, uterine fluids with highly complex
organic and inorganic composition (those from which calcite
crystals forming the eggshell precipitate in vivo)2,16 have a much
stronger effect in each one of the precipitation stages (i.e.,
nucleation, crystal growth and morphology modification). The
stronger effect of uterine fluids compared to egg white proteins
or eggshell extracts could be due to the following reasons: (1)
there is not a significative alteration of the natural state of the
original components, (2) the possible cooperative or even
synergistic interactions between the wide array of organic
components present in these fluids, and (3) the specific chemical
and environmental scenario, namely, pH, partial pressure of CO2
and biologically available calcium that only exist in this latter
system.22
It is important to understand the evolution of the chemical
environment (i.e., uterine fluid) in which eggshell forms. During
the different stages, the composition of the uterine fluid changed
in the number and concentration of the diverse organic
components.2,13,17,31 At the same time the concentration of
calcium and the partial pressure of CO2 supplied throughout
the uterus varied in a regulated manner to produce the
mineralization of the shell in a very short time (around 20 h).
There is very limited information regarding the specific role
of the individual organic components involved in eggshell
mineralization.2,32 In this respect, comparison of electrophoretic
profiles SDS-PAGE of the uterine fluids before and after
CaCO3 precipitation (Figures 4a,b) with their respective pH
profiles when the uterine fluids were used as protein bearing
solutions in CaCO3 precipitation experiments is specially
relevant in order to shed light to determine which are the active
proteins regulating eggshell growth and those active in regulating
the chemistry of the system. We propose that those proteins
that have been removed by the precipitate in successive stages
(and do not show up in the electrophoretic profiles after CaCO3
precipitation) have a strong affinity toward calcite surfaces and
should have an active role in controlling calcite growth or its
aggregation behavior.23 Among these proteins are the acidic
proteins OC-17, OCX-32 and osteopontin (see Table 2), which
are selectively removed by the precipitate. These proteins should
display a net negative charge during the precipitation process
(which takes place within the pH range from 7.6 to 9.03)
whereas calcite crystals at this pH range are neutral or positively
charged. Calcite point of zero charge, pHpzc, is around 8.3 and
the concentration of Ca2+ in the uterine fluid bearing solutions
is slightly higher than in the control experiment. Under these
conditions electrostatic interactions can take place between these
Figure 6. FESEM images of (a) calcite grown in the presence of 128
μg/mL lysozyme; (b) calcite grown using 128
μg/mL ovalbumin, (c) calcite
grown using 128
μg/mL ovotranferrin; (d) calcite grown using 512
μg/mL lysozyme; (e) calcite grown using 512
μg/mL ovalbumin; (f) calcite
grown in the presence of 512
μg/mL ovotranferrin; (g) calcite crystals grown in the presence of a purified fraction of the eggshell containing mainly
ovocleidin-116. Most representative calcite faces are indexed.
4336 Crystal Growth & Design, Vol. 8, No. 12, 2008
Hernández-Hernández et al.
proteins and calcite leading to their adsorption to calcite surfaces
during crystal growth. The adsorption of proteins to calcite could
occur through deprotonized negatively charged -COO
-
groups
of the proteins and the positively charged surface complexes,
that is, <CO3Ca
+
. As a result of this process, the calcite crystals
show modified rhombohedral morphologies as was found in our
previous work22 using an acidic model globular protein.
Additionally, calcite crystals treated with proteinase-K buffer
reveal holes that indicate the previous presence of adsorbed and/
or occluded proteins.
In fluid II, OC-116, osteopontin and lysozyme (a minority
component) are removed after CaCO3 precipitation. OC-116,
with pI of 6.62, and osteopontin, with pI 4.5-4.6 (Table 2) are
both negatively charged along the precipitation process (the pH
ranged from 7.43 to 8.25), while the CaCO3 crystals must be
neutral or positively charged due to the charge reversion caused
by the higher concentration of Ca2+ respect to that in the control
experiment.
On the other hand, proteins remaining in solution after CaCO3
precipitation should not be so active in the control of mineral-
ization. However, they could also influence calcium carbonate
precipitation by modifying the chemical environment (i.e., pH,
supersaturation, level of Ca). For instance, in fluid I, ovalbumin,
ovotransferrin and lysozyme, proteins that are not removed by
CaCO3 precipitation, slightly favors calcite nucleation in in vitro
experiments. Moreover, OCX-21 and OC-17, which have very
different isoelectric points (pI 9.3 and pI 4.1 respectively), may
be responsible for the high buffering capacity of fluid II in spite
of being minority components of the complex mixture of
proteins. The fact that OCX-21 and OC-17 are minority
components and have different isoelectric points could explain
the similar buffering capacity of the two fluids of the stage II
tested (those collected in different days, whose concentrations
were 97 and 479
μg/mL, respectively). In the formation of this
buffer other basic proteins like lysozyme could have participated.
However, because of the peculiar pH behavior described above
was only found when using uterine fluid of the stage II as protein
bearing solution, we can discard the presence of this protein in
the buffer. This pH-stat chemical scenario favors the growth of
calcite32 and could explain the fast growth stage during eggshell
formation in which large columnar calcite crystals (palisades)
are deposited. Also, in fluid II bearing experiments, lysozyme
is removed by the precipitate, probably adsorbed because of
the surface charge reversion caused by the previous adsorption
of OC-116 and osteopontin. Interestingly, lysozyme, though its
main putative role is as a bactericidal, has been found in the
palisade layer of eggshell.10 In fluid III bearing experiments,
OCX-32 was selectively removed after CaCO3 precipitation. In
general, OCX-32 shows a high absorption affinity to calcite and
has a high inhibitory effect of the precipitation of CaCO3. This
result is consistent with previously reported work that describes
the inhibition of the precipitation that fluid III has on metastable
solutions formed by calcium chloride and sodium bicarbonate.
The authors believe that this protein could have an important
role in the arrest of eggshell mineralization.9
Regarding the control over the crystal growth processes, we
have found that all organic components tested affected calcium
carbonate precipitation at three fundamental levels: (a) selecting
or at least favoring the precipitation of a particular mineral
phase, (b) controlling the number of crystals or the nucleation
flux, and (c) controlling crystal morphology and the aggregation
behavior.
At the first level, calcite formation is favored over the other
calcium carbonate polymorphs (e.g., aragonite and vaterite). This
study shows that eggshell extracts and uterine fluids were
especially effective in selecting calcite and suppressing totally
the appearance of vaterite and aragonite, despite the fact that
the system is supersaturated with respect to the three main
calcium carbonate polymorphs (e.g., calcite, aragonite and
vaterite) as well as other common calcifying minerals (i.e.,
hydroxyapatite). Only in the terminal stage hydroxyapatite is
found. This superior capacity for selecting a specific polymor-
phic mineral phase is well described in other biological
mineralization processes. For instance, some mollusc shell
proteins are able to switch from calcite to aragonite and
back.33,34
Macromolecules also have an important effect on the
nucleation flux. We have observed that some individual egg
white proteins and purified fractions of the organic matrix
are active in controlling calcite nucleation. Nevertheless, uterine
fluids are much more active and induced a more intense
nucleation. These phenomena occurred irrespectively of the
stage of eggshell formation at which these fluids were extracted.
In our experimental system, we have added calcium to the
uterine fluid to reproduce the control conditions. This added
calcium could explain partly a more intense nucleation. How-
ever, the small difference in calcium concentration in the uterine
fluid bearing solution compared to that of the control solution
could not explain by itself such a large increase in nucleation
density. The uterine fluid bearing solutions produced a total
number of calcite crystals per drop which was at least 2 orders
of magnitude more intense than any other individual commercial
proteins or purified extracts of the eggshell organic matrix tested.
In our opinion, the presence of charged proteins and other
Table 2. Molecular Weight, Isoelectric Point and Localization of Some Selected Proteins Associated to the Process of Eggshell Formation15,28-
30
protein
localization
MW (kDa)
pI (theoretical)
ovalbumin
egg white, uterine fluid of the stage I
45.0
4.5
ovotransferrin
eggshell membranes, uterine fluid of the stage I
77.0
6.05
lysozyme
eggshell membranes, uterine fluid of the three stages
specially at the growth phase
14.3
9.3
ovomucoid
21.2
clusterin
the bases of the mammillary cores, and palisade
52
ovocalyxin-21
21
9.3
ovocalyxin-25
25
8.84
ovocalyxin-32
the external part of the calcified shell (palisade, vertical
crystal layer and cuticle)
32
6.81
ovocalyxin-36
eggshell membranes and basal parts of the mammillary
cores, and palisade matrix
38
5.62
ovocleidin-116
mammillary and palisade layers
77.3
6.62
ovocleidin-17 (chain A)
mammillary and palisade layers
15.7
4.03
cystatin
15.6
8.4
ovoglican
207
osteopontin
in the core of the eggshell membranes, at the bases of the mammilla, in
the outer of and throughout the palisade layer
55
4.5-4.6
Active Proteins in Hen Eggshell Formation
Crystal Growth & Design, Vol. 8, No. 12, 2008 4337
organic compounds facilitate adsorption of calcium ions to their
surface therefore favoring the nucleation of calcite by a
mechanism known as the ionotropic effect.35
Alternatively, some proteins have a strong affinity to specific
calcite surfaces and could favor the nucleation more actively
acting as templates. This template effect is due to a stereo-
chemical matching between the disposition of specific functional
groups (i.e., COOH
-
) in the proteins or their motifs and the
arrangement of carbonate or calcium ions in specific target
calcite faces. The large increase in the nucleation could also be
due to a cooperative interaction of different proteins and/or other
organic components present in the fluids since the same
individual components or eggshell extracts composed of a lower
array of organic components show a much more limited effect
on nucleation. However, we cannot discard the possibility that
the lower effect of individual components on CaCO3 precipita-
tion could be due to an alteration of the natural state of proteins
derived from the extraction and purification procedures. If some
proteins act as a template for calcite nucleation at their natural
state, any alteration of their molecular configuration could
significantly decrease their capacity to induce calcite nucleation
by a template mechanism.
Finally, in vitro CaCO3 precipitation experiments show that
individual commercial egg white proteins, at concentrations
similar to those found in the uterine fluids, have very weak effect
on crystal morphology. In contrast, purified extracts of the
eggshell organic matrix and uterine fluids have a strong and
specific effect on calcite crystal morphologies. For instance, fluid
I induced the overexpression of specific crystal faces of calcite,
that is, the {110} and {100} prismatic forms of calcite. On the
other hand, fluid II and III bearing solutions seem to promote
aggregation of calcite crystals and modified the calcite mor-
phologies on a less specific way.
Modifications in crystal growth morphology can provide
insights regarding protein-mineral surface interactions.35 The
type of modification depends on the specificity of the
protein-mineral interactions. In the case that there is stereo-
chemical compatibility between the protein and a particular
crystal face, the protein will be preferentially absorbed to this
face and all symmetrically equivalent faces. These faces will
be expressed as a consequence of reductions in their normal
growth rate and the morphology of the crystals will be modified
accordingly. However, in other cases a nonspecific absorption
of proteins (to minimize hydrophobic interaction with the
solution) can explain a greater modification of calcite crystals
in an unspecific way resulting in rounded crystals without well-
defined crystal faces.22,36
The above results shed light into understanding what occurs
in vivo. Although these results cannot explain how eggshell
growth is stopped, we have shown how specific eggshell organic
extracts are highly effective in decreasing crystal nucleation.
The sum of all these findings suggest the high level of control
exerted by the organism over the biomineralization of the shell
and, in particular, over the formation of its mineral part, which
is composed of calcite crystals of well-defined orientation, size
and morphologies. In addition to the control of crystal growth
exerted by the organic matrix components,2,3,7 there is a
competitive process during the growth of adjacent crystals
resulting in a columnar microstructure with a preferential crystal-
lographic orientation.5,37
This precise control over composition and the microstructural
characteristics of the eggshell has important consequences in
the mechanical properties, integrity and functionality of this
biomaterial.26,38
Conclusions
In summary, to understand the role of organic components
during eggshell formation, we have used organic bearing
solutions with different levels of complexity, from the simplest
(individual eggshell proteins) to the most complex and realistic
composition (uterine fluids) that could simulate the natural
process. Using this approach, we found that the precise control
of eggshell formation can be due to the combined action of the
components present in the uterine fluid which regulate crystal
growth and the chemistry of the system. However, in vitro
CaCO3 precipitation experiments reveal that some specific acidic
proteins (including OC-17, OCX-32, OC-116 and osteopontin)
inside the uterine fluid and in the purified eggshell extracts have
a strong affinity for the surface of CaCO3 and are selectively
removed from the solution during its precipitation. These
proteins must have an active role in CaCO3 growth, aggregation
and inhibition. Other proteins regulate the chemical environment
in which the precipitation takes place, that is, by buffering the
pH favoring crystal growth as for instance the OC-17 and OCX
21. Individual egg white proteins (lysozyme, ovalbumin,
ovotransferrin), at concentrations similar to those found in
uterine fluids, affected slightly the nucleation density and growth
morphology of calcite in in vitro experiments, but did not select
exclusively calcite as precipitating polymorph.
Acknowledgment. This work was supported by funding
received from the European Community for the project Egg
Defense (QLRT-2001-01606). A.H.H., J.G.M. and J.M.G.R also
acknowledge the Excellence project RNM1344 of the Junta de
Andalucıa and to the “Factorıa de Cristalización” (Consolider
Ingenio 2010).
References
(1) Fink, D. J.
MRS Bull. 1992,
17, 27–31.
(2) Nys, Y.; Gautron, J.; Garcıa-Ruiz, J. M.; Hincke, M. T.
C. R. PaleV
ol.
2004,
3, 549–562.
(3) Arias, J. L.; Fink, D. J.; Xiao, S.; Heuer, A. H.; Caplan, A. I.
Int.
ReV
. Cytol. 1993,
145, 217–250.
(4) Simkiss, K.; Wilbur, K. M.
Biomineralization: Cell Biology and
Mineral Deposition; Academic Press: San Diego, 1989.
(5) Rodriguez-Navarro, A.; Garcia-Ruiz, J. M.
Eur. J. Mineral. 2000,
12,
609–614.
(6) Dennis, J. E.; Xiao, S. Q.; Agarval, M.; Fink, D. J.; Heuer, A. H.;
Caplan, A. I.
J. Morphol. 1996,
228, 287–306.
(7) Nys, Y.; Hincke, M. T.; Arias, J. L.; Garcıa-Ruiz, J. M.; Solomon,
S. E.
Poult. AV
ian Biol. ReV
. 1999,
10, 143–166.
(8) Nys, Y.; Zawadzki, J.; Gautron, J.; Mills, A. D.
Poult. Sci. 1991,
70,
1236–1245.
(9) Gautron, J.; Hincke, M. T.; Mann, K.; Panheleux, M.; Bain, M.;
McKee, M. D.; Solomon, S. E.; Nys, Y.
J. Biol. Chem. 2001,
276,
39243–39252.
(10) Hincke, M. T.; Gautron, J.; Panheleux, M.; Garcia-Ruiz, J.; McKee,
M. D.; Nys, Y.
Matrix Biol. 2000,
19, 443–453.
(11) Mann, K.; Gautron, J.; Nys, Y.; McKee, M. D.; Bajari, T.; Schneider,
W. J.; Hincke, M. T.
Matrix Biol. 2003,
22, 397–407.
(12) Pines, M.; Knopov, V.; Bar, A.
Matrix Biol. 1995,
14, 765–771.
(13) Fernandez, M. S.; Moya, A.; López, L.; Arias, J. L.
Matrix Biol. 2001,
19, 793–803.
(14) Gautron, J.; Murayama, E.; Vignal, A.; Morisson, M.; McKee, M. D.;
Rehault, S.; Labas, V.; Belghazi, M.; Vidal, M. L.; Nys, Y.; Hincke,
M. T.
J. Biol. Chem. 2007,
282, 5273–5286.
(15) Hincke, M. T.
Connect. Tissue Res. 1995,
31, 227–233.
(16) Mann, K.; Macek, B.; Olsen, J. V.
Proteomics 2006,
6, 3801–3810.
(17) Gautron, J.; Hincke, M. T.; Nys, Y.
Connect. Tissue Res. 1997,
36,
195–210.
(18) Arias, J. L.; Fernández, M. S.; Dennis, J. E.; Caplan, A. I.
Connect.
Tissue Res 1991,
26, 37–45.
4338 Crystal Growth & Design, Vol. 8, No. 12, 2008
Hernández-Hernández et al.
(19) Fernandez, M. S.; Arias, J. L.
Eur. Symp. Egg Qual. Egg Prod. 1999,
81–86.
(20) Hincke, M. T.
J. Histochem. Cytochem. 2008,
56, 467–476.
(21) Arias, J. L.; Mann, K.; Nys, Y.; Garcia-Ruiz, J. M.; Fernandez, M.
Handbook of Biomineralization; Bauerlein, E., Ed.; Wiley-VCH Verlag
GmbH and Co: Weinheim, 2007.
(22) Hernández-Hernández, A.; Rodrıguez-Navarro, A.; Gómez-Morales,
J.; Jiménez-Lopez, C.; Nys, Y.; Garcıa-Ruiz, J. M.
Cryst. Growth Des.
2008,
8, 1495–1502.
(23) Hernández-Hernández, A.; Vidal, M. L.; Gómez-Morales, J.; Rod-
rıguez-Navarro, A. B.; Labas, V.; Gautron, J.; Nys, Y.; Garcıa-Ruiz,
J. M.
J. Cryst. Growth 2008,
310, 1754–1759.
(24) Layne, E.
Methods Enzymol. 1957,
3, 447–454.
(25) Bradford, M. M.
Anal. Biochem. 1976,
72, 248–254.
(26) Ahmed, A. M. H.; Rodrıguez-Navarro, A. B.; Vidal, M. L.; Gautron,
J.; Garcıa-Ruiz, J. M.; Nys, Y.
Br. Poult. Sci. 2005,
46, 268–279.
(27) Shiao, J. C.; Tzeng, C. S.; Leu, C. L.; Chen, F. C.
J. Fish Biol. 1999,
54, 302–309.
(28) http://www.worthington-biochem.com/OA/default.html.
(29) Diamond, R.
J. Mol. Biol. 1974,
82, 371–391.
(30) Holen, E.; Elsayed, S.
Int. Arch. Allergy Immol. 1990,
91, 136–141.
(31) Fernandez, M. S.; Passalacqua, K.; Arias, J. I.; Arias, J. L.
J. Struct.
Biol. 2004,
148, 1–10.
(32) Mann, K.; Siedler, F.
Biochim. Biophys. Acta 2004,
1696, 41–50.
(33) Belcher, A. M.; Wu, X. H.; Christensen, R. J.; Hansma, P. K.; Stucky,
G. D.; Morse, D. E.
Nature 1996,
381, 56–58.
(34) Falini, G.; Albeck, S.; Weiner, S.; Adaddi, L.
Science 1996,
271, 67–
69.
(35) Addadi, L.; Weiner, S.
Angew. Chem., Int. Ed. 1992,
31, 153–169.
(36) Jimenez-Lopez, C.; Rodriguez-Navarro, A.; Dominguez-Vera, J. M.;
Garcia-Ruiz, J. M.
Geochim. Cosmochim. Acta 2003,
67, 1667–1676.
(37) Garcia-Ruiz, J. M.
Mat. Sci. Eng. C- BioS. 1995,
3, 95–100.
(38) Rodrıguez-Navarro, A. B.; Yebra, A.; Nys, Y.; Jiménez-López, C.;
García-Ruiz, J. M.
Eur. J. Mineral. 2000,
12, 609–614.
(39) Rodrıguez-Navarro, A.; Kalin, O.; Nys, Y.; Garcıa-Ruiz, J. M.
Brit.
Poult. Sci. 2002,
43, 395–403.
CG800786S
Active Proteins in Hen Eggshell Formation
Crystal Growth & Design, Vol. 8, No. 12, 2008 4339
