Asprich mollusk shell protein: in vitro experiments aimed at elucidating function in CaCO3 crystallization

www.rsc.org/crystengcomm Volume 9 | Number 12 | December 2007 | Pages 1131–1270

CrystEngComm

COVER ARTICLEAddadi et al.Asprich mollusk shell protein: in vitro experiments aimed at elucidating function in CaCO3 crystallization

HOT ARTICLEde Leeuw and RaboneMolecular dynamics simulations of the interaction of citric acid with the hydroxyapatite (0001) and (010) surfaces in an aqueous environment

HOT ARTICLEPokroy and AizenbergCalcite shape modulation through the lattice mismatch between the self-assembled monolayer template and the nucleated crystal face

This paper is published as part of a CrystEngComm themed issue on:

Biomineralisation

Guest edited by Lia Addadi The Weizmann Institute of Science, Israel

Published in issue 12, 2007 of CrystEngComm

Images reproduced by permission of Lia Addadi (outside) and James De Yoreo (inside) Other papers published in this issue include: Molecular dynamics simulations of the interaction of citric acid with the hydroxyapatite (0001) and (01-10) surfaces in an aqueous environment N. H. de Leeuw and J. A. L. Rabone, CrystEngComm, 2007, DOI: 10.1039/b710974a

Synthesis-dependant structural variations in amorphous calcium carbonate Raymond S. K. Lam, John M. Charnock, Alistair Lennie and Fiona C. Meldrum, CrystEngComm, 2007, DOI: 10.1039/b710895h

Calcite shape modulation through the lattice mismatch between the self-assembled monolayer template and the nucleated crystal face Boaz Pokroy and Joanna Aizenberg, CrystEngComm, 2007, DOI: 10.1039/b710294a

Fine structure of nacre revealed by solid state 13C and 1H NMR Christian Jäger and Helmut Cölfen, CrystEngComm, 2007, DOI: 10.1039/b708600h

Visit the CrystEngComm website for cutting-edge crystal engineering research

www.rsc.org/crystengcomm

Asprich mollusk shell protein: in vitro experiments aimed at elucidatingfunction in CaCO3 crystallization

Yael Politi,a Julia Mahamid,a Harvey Goldberg,b Steve Weinera and Lia Addadi*a

Received 27th June 2007, Accepted 11th September 2007

First published as an Advance Article on the web 17th September 2007

DOI: 10.1039/b709749b

Acidic proteins are key components of the organic matrix of many biologically formed minerals

and are therefore thought to play an important role in their formation. Here we study the effect of

one unusually acidic protein of the Asprich family, associated with mollusk shell prismatic layer,

on the precipitation of CaCO3 in vitro. We show that Asprich induces and transiently stabilizes the

deposition of amorphous calcium carbonate (ACC). Asprich also induces the formation of ACC

when adsorbed onto chitin, a major component of the intracrystalline organic matrix of the

prismatic layer. Based on this evidence, combined with previous studies on the forming prisms in

the shell layer, we suggest that the in vivo function of Asprich is inducing and stabilizing ACC

particles and inhibiting their uncontrolled crystallization until they undergo secondary nucleation

on the growing prisms.

Introduction

The formation of crystals from an amorphous precursor phase

is a common phenomenon in biomineralization.1,2 Organisms

from various phyla use this strategy to form different

crystalline materials, including the transformation of a

disordered ferrihydrite phase into magnetite,3 amorphous

calcium phosphate into carbonated apatite,4 and amorphous

calcium carbonate (ACC) into either calcite5–7 or aragonite.8,9

Organisms must invest energy to reversibly stabilize these

intrinsically unstable transient amorphous phases. Possible

advantages of this strategy are the efficient transport of ions to

the mineralization site, providing alternative pathways for the

synthesis of otherwise difficult-to-form phases and for the

production of skeletal materials in complex shapes.2,10

The mechanisms of formation of the mature crystalline

phase involve several distinct processes: transient stabilization

of the amorphous phase, subsequent transformation into a

stable crystalline phase, polymorph selection and crystal

orientation. In the case of ACC, polymorph selection may

occur at the initial stage based on the observation that the

transient amorphous phase already has the nascent order of

the polymorph into which it will transform.11,12 The nucleation

stage may occur by epitaxial nucleation on a preexisting

crystalline seed5,13,14 or be induced by specific nucleating

proteins in the organic matrix.15 Control over these processes

needs to be exerted at every stage. Much of this control is

assumed to be exercised by highly charged glycoproteins, many

of which are rich in aspartic acid.16–20 Here we investigate the

role of a family of such proteins, called Asprich,21 present in

the calcitic prisms of the shells from the mollusk Atrina rigida.

The Asprich proteins are abundant in the calcite prisms, and

are therefore thought to play a significant role in the formation

of the mineral. They are composed of several domains,

including a calcium binding domain, a domain that possibly

binds Mg ions, and long stretches that are enriched in aspartic

acid, including sequences of poly-Asp. The acidic amino acids

constitute more than 50% of the sequence and around 30% are

hydrophobic amino acids.21 Nudelman et al.14 characterized

the structure and organization of the organic matrix in the

mature and forming prismatic shell layer of the mollusk Atrina

rigida. They show that the intracrystalline matrix is composed

of a network of chitin fibers, and that the prisms grow by

depositing spherical mineral particles on the fibrous scaffold.

Using immuno-labeling, they show that Asprich proteins

are associated with the initially deposited mineral particles.

Interestingly, each prism in the A. rigida prismatic layer

diffracts X-rays as a very well ordered single calcite crystal.22

Here we show that one protein from the Asprich family

(Asprich-c) transiently stabilizes ACC during the precipitation

of CaCO3 from solution. The influence of the protein on the

different stages of crystal growth is characterized, as well as its

effect on crystal nucleation when adsorbed on a glass substrate

and chitin fibers. We suggest that these observations are

relevant to the function of the protein in vivo.

Results

Asprich (His-22) cDNA clone c was expressed in competent

E. coli XL-10 cell systems, and a series of in vitro assays were

performed in which the effect of the expressed protein on the

precipitation of CaCO3 was studied. Note that it is not known

whether native Asprich extracted from the shell is post-

translationally modified. The Asprich used in this study was

expressed in prokaryotic cells, and was thus devoid of post-

translational modifications. The effect of a synthetic poly-

aspartic acid peptide (poly-Asp) was analysed for comparison.

The length of the peptide (8.4 KDa) is of the same order of

aDepartment of Structural Biology, Weizmann Institute of Science,Rehovot 76100, IsraelbCIHR Group in Skeletal Development and Remodeling, Schulich Schoolof Medicine & Dentistry, University of Western Ontario, London,Ontario, Canada N6A 5C1

PAPER www.rsc.org/crystengcomm | CrystEngComm

This journal is � The Royal Society of Chemistry 2007 CrystEngComm, 2007, 9, 1171–1177 | 1171

magnitude as the length of the protein (21.8 KDa). All

experiments were performed using equal molar concentrations

of total amino acids for Asprich and poly-Asp, rather than

using protein/peptide molar concentrations.

Slow precipitation experiments with increasing

concentrations of Asprich

Calcite crystals grown in the presence of low Asprich

concentrations (5 and 10 nmol amino acids mL21) exhibit

well developed {104} faces, often terminating with rounded

edges exhibiting fine steps (Fig. 1a and b). Similar observations

were made during the precipitation of CaCO3 in the presence

of the protein Caspartin.20 At similar additive concentrations

the crystals grown in the presence of poly-Asp have large pits

in the center of the faces (Fig. 1c and d). Similar face centered

pits have previously been observed in calcite.23–25

At intermediate concentrations (35 nmol mL21), the effect

on the calcite crystals is enhanced: the rhombohedral

morphology is strongly altered, the surfaces are often not flat,

corners are rounded and non typical faces are displayed

(Fig. 2a). In addition, spherulites having diameters around

150 mm with characteristically smooth surfaces are formed

(Fig. 2b). At the same concentrations of poly-Asp, only

aggregates of large calcite crystals with heavily affected

morphologies are formed (Fig. 2c). At high concentrations

(70 nmol mL21), the Asprich and poly-Asp effects on the

precipitation are similar. Both solutions deposit small (2–5 mm)

and large (100–200 mm) spherulites, together with large

affected crystal aggregates (Fig. 2d–f).

Fast precipitation experiments

In order to study the effect of Asprich on the initial stages of

CaCO3 precipitation we performed experiments with rapid

diffusion of ammonium carbonate in CaCl2 solutions contain-

ing 20 nmol mL21 protein, and monitored the progress of

precipitation under a light microscope. Approximately 10 min

after initiation of ammonium carbonate diffusion, suspended

colloidal particles appear in solutions containing Asprich or

poly-Asp. In the control solution containing no additives,

some calcite crystals were observed in 10 min at the air/water

interface; the solution however remained clear.

The particles from the colloidal suspension were collected

and observed with TEM imaging and diffraction. Amorphous

colloidal particles were present with both Asprich and poly-

Asp additives. Interestingly, the particles grown in the presence

of Asprich or poly-Asp exhibited different stabilities. With the

beam focused on the particles produced in the presence of

Asprich, gradual crystallization started after approximately

20 min. This resulted in the formation of polycrystalline

disordered calcitic particles (Fig. 3). Particles that were stored

on EM grids at ambient temperature without being exposed to

the electron beam remained amorphous after 20 d when they

were examined in the TEM. In contrast, particles formed in the

presence of poly-Asp crystallized within seconds of beam

Fig. 1 Calcite crystals precipitated after one day in the presence of (a)

5 nmol mL21 Asprich, (b) 10 nmol mL21 Asprich, (c) 5 nmol mL21

Poly-Asp. (d) 10 nmol mL21 Poly-Asp. Protein and peptide concentra-

tions are expressed in nmol (amino acid) mL21.

Fig. 2 Precipitates formed after one day in the presence of (a) and (b)

35 nmol(amino acid) mL21 Asprich. (c) 35 nmol(amino acid) mL21

Poly-Asp (d) and (e) 70 nmol(amino acid) mL21 Asprich (f)

70 nmol(amino acid) mL21 poly-Asp. Scale bars = 20 mm.

Fig. 3 TEM images (a) and (b) and electron diffraction patterns (c)

and (d) of 20 nmol(amino acid) mL21 Asprich induced precipitates

before (left panel) and after (right panel) 30 min of exposure to the

electron beam. Scale bar = 100 nm.

1172 | CrystEngComm, 2007, 9, 1171–1177 This journal is � The Royal Society of Chemistry 2007

exposure, yielding crystals with individually identifiable single

crystal electron diffraction patterns (Fig. 4). When unexposed

samples were observed after 10 d, they were all crystalline.

Similar colloidal particle suspensions were collected 30 min

after the initiation of the experiment and examined by FTIR.

The crystallinity of the samples can be established by following

the n4 in-plane bending vibration at 713 cm21, which is

extremely broad in samples of stable ACC. The ratio of the

intensities of the out-of-plane bending and in-plane bending

vibrations In2/In4 is around 3 in fully crystallized geological or

synthetic calcite, and increases with increasing disorder in the

structure.5,26 The spectra of the particles precipitated in the

presence of Asprich indicated a mixture of ACC and calcite

and in one case only ACC was detected (Fig. 5b). The particles

formed in the presence of poly-Asp under the same conditions

were all calcitic and had In2/In4 ratios around 2.6, implying that

they were fully crystalline (Fig. 5a).

The precipitates that accumulated at the bottom of the wells

containing suspensions formed in the presence of Asprich,

consist of 0.05–5 mm spherulites and calcite rhombohedra. The

smallest particles (Fig. 6a, inset) are presumably colloidal

particles that adhered to the glass, while the larger particles

result from fusion of the former. Interestingly, the {104} faces

of the calcite crystals show relatively high steps covering part

of the face. The steps have very flat surfaces and rough edges.

These steps seem to advance by addition and fusion of

spherulites (colloidal particles) to their edges (Fig. 6b, c).

Spherulites and crystal aggregates of different sizes also preci-

pitate in the presence of poly-Asp (Fig. 7), however, the above

mentioned unique form of crystallization was not observed.

Nucleation assays

Asprich protein and poly-Asp peptide were adsorbed on

silicon coated glass slides by incubation of the slides with a

drop of 50 nmol mL21 protein/peptide solution. The solution

was removed, the slides were floated on top of a CaCl2solution and precipitation was induced. After approximately

1 h the slides were collected and the precipitates were observed

Fig. 4 TEM image (a) and electron diffraction pattern of

20 nmol(amino acid) mL21 poly-Asp induced precipitates before (b)

and after (c) exposure of a few seconds to the electron beam.

Fig. 5 Normalized FTIR spectra of suspended particles grown in the

presence of 20 nmol(amino acid) mL21 poly-Asp (a) and Asprich (b).

Fig. 6 Precipitates accumulated on the glass at the bottom of the well after 30 min of fast precipitation, in the presence of 20 nmol(amino acid)

mL21 Asprich. (a) Large spherulites. Inset: small spherulite similar to spherulites in Fig. 3. scale bar = 50 nm; (b) a crystal displaying the typical

rhombohedral morphology, with colloidal particles accumulating on the steps. (c) Higher magnification of (b), arrows point to colloidal particles.

This journal is � The Royal Society of Chemistry 2007 CrystEngComm, 2007, 9, 1171–1177 | 1173

using a light microscope. The influence of the protein or

peptide on crystal nucleation was evaluated by examining

only the crystals nucleated on the slides, and avoiding crystals

that were homogeneously nucleated in the bulk solution.

Moreover, because the only protein present is adsorbed on the

glass substrate, there should be no effect of the protein on

crystal growth.

In the control containing no additive, calcite crystals

nucleated over the whole area of the glass (Fig. 8c). In

contrast, nucleation of calcite crystals was almost totally

inhibited by the presence of Asprich adsorbed on the glass

slide. This was inferred from the fact that the area where the

protein-containing drop was adsorbed contained relatively

few crystals (Fig. 8a). Nucleation was also inhibited by the

presence of poly-Asp adsorbed on the glass, but to a much

lesser extent (Fig. 8b). However without knowing the amount

of the respective protein adsorption to the surface, we cannot

determine inhibitory potencies. It is possible that the difference

in the degree of inhibition between poly-Asp and Asprich is

related to the difference in the extent to which the poly-Asp

adsorbs compared to the Asprich.

Protein adsorbed on chitin

Chitin and acidic macromolecules are the two major compo-

nents of the intracrystalline organic matrix assembly in the

prismatic layer.14 There are also indications that Asprich is

associated with chitin at the early stage of mineralization.14

In order to test for the combined effect of the two components

on calcium carbonate deposition, we performed precipitation

experiments on a chitin substrate to which Asprich was

adsorbed, in a manner similar to the method described by

Falini et al.27 The results are compared with poly-Asp

adsorbed on chitin, and a chitin only substrate.

In the absence of adsorbed macromolecules calcite crystals

were only present on the chitin surface and no precipitation

occurred within the chitin sheaths. The presence of either

poly-Asp or Asprich adsorbed on the chitin fibers resulted in

precipitation occurring inside the chitin, but the resultant

crystals were significantly different. In the presence of poly-

Asp most of the crystals were affected rhombohedra (Fig. 9c)

Fig. 7 Precipitates accumulated on the glass at the bottom of the well

after 30 min of fast precipitation, in the presence of 20 nmol(amino

acid) mL21 poly-Asp. (a) Large spherulites; (b) stepped calcite crystals

delimited by regular {104} faces. Scale bar = 10 mm.

Fig. 8 Calcite crystals nucleated on siliconized glass slides. The slides were pre-incubated with a drop of (a) Asprich protein, (b) poly-Asp peptide

and (c) no additive, and placed on top of a CaCl2 solution. The dashed line marks the borders where the drop was incubated.

Fig. 9 CaCO3 Precipitation on chitin substrate with Asprich and poly-Asp adsorbed. (a) Spherulite precipitated on chitin substrate incubated with

Asprich. (b) A magnification of the interior of the spherulite showing the mineral spheres and an area with extensive fiber mineralization. (c) A

calcite crystal precipitated on chitin substrate incubated with poly-Asp.

1174 | CrystEngComm, 2007, 9, 1171–1177 This journal is � The Royal Society of Chemistry 2007

and some vaterite spherulites. In contrast, a large number of

spherulites formed in the presence of Asprich protein. A close

examination of the interior of the spherulites revealed that the

crystals grew around the chitin fibers and that the chitin is

decorated with small spherical mineral particles (Fig. 9b). In

the presence of both Asprich and poly-Asp we found many

places in which the chitin fibers are mineralized. Similar

observations were made by Kato and Amamiya28 using chitin

and polyanions, e.g., polyAsp, polyGlu and poly(acrylic acid).

Experimental

Cloning and expression of Asprich

Asprich (His-22) cDNA obtained from the Atrina cDNA

library was cloned into the pET28 expression vector

(Novagen). The vector added an N-terminal 6xHis tag and a

TEV cleavage site to the Asprich cDNA sequence. The amino

acid sequence of the expressed protein is as described (protein

sequence ‘‘c’’). The recombinant DNA procedures were carried

out using methods described by Sambrok 1989.29 The coding

sequence was confirmed by DNA sequencing.

Plasmids were first transfected into competent E. coli XL-10

cells. Colonies were picked and the sequence confirmed. E. coli

BL21 (DE3) were transformed with the expression plasmids,

and grown in phosphate-buffered Super Broth (SB) supple-

mented with 15 mg mL21 kanamycin and 0.4% glucose to an

absorbance of 0.6–0.9. After induction with 2 mM isopropyl-

b-D-thiogalactopyranoside, cultures were grown for a further

4 h, cells recovered by centrifugation, resuspended and then

sonicated in denaturing binding buffer (5 mM imidazole, 0.5 M

NaCl, 0.02 M Tris/HCl, 6 M urea, pH 7.9) as described.14 The

extract was applied to a His-bind column (Novagen), and

Asprich eluted with 0.5 M imidazole-containing elution buffer.

Fractions containing Asprich were pooled and subjected to

purification by fast protein liquid chromatography (FPLC)

using established protocols. Proteins were purified by chro-

matography on a Q-Sepharose Fast Flow column followed by

gel filtration on a Superdex 200PG column (GE Biosciences).

Purification of Asprich was monitored by electrophoresis on

12.5% SDS polyacrylamide gels, stained with Stains-all and

silver nitrate as described. Fractions containing Asprich were

dialyzed against 0.01 M ammonium bicarbonate buffer,

lyophilized in aliquots, and protein content determined by

amino acid analysis. Yield of intact Asprich (y95% purity)

was approximately 400 mg L21 of initial culture media.

Poly-aspartic-acid peptide, mean MW(vis) = 8400 Da, was

purchased from Sigma.

Precipitation experiments

Slow precipitation. Synthetic crystals were grown in Nunc

multi-well dishes (24 wells; 1.5 cm diameter) by diffusion of

ammonium carbonate vapor into calcium chloride solutions.

Glass cover slips were placed on the bottom of each well. A

total volume of 0.5 mL of 10 mM CaCl2 (Merck, A grade)

solution, was introduced into each well. Poly-Asp or Asprich

were added to reach different final concentrations: 5, 10, 35

and 70 nmol (amino acids) mL21, corresponding to 0.55, 1.1,

3.86 and 7.7 mg mL21. Each well was sealed separately with

aluminium foil, pierced with a 25 G needle and sealed with

Parafilm. Control experiments were performed in parallel with

10 mM CaCl2 and no additives. The multi-well dish was placed

in a closed desiccator containing ammonium carbonate vapor

for one day at room temperature.

Fast precipitation. These experiments were performed in

a Nunc multi-well dish, where the central wells contained

the crystallization solutions with 10 mM CaCl2 and

20 nmol (amino acids) mL21 poly-Asp or Asprich additives,

and a CaCl2 control. Four vials containing ammonium

carbonate powder were placed at the corners of the plate.

The vials were covered with aluminium foil and sealed with

Parafilm. One hole was punctured in each of the vials with a

25 G needle. The plates were covered, sealed with Parafilm and

placed under a light microscope to follow the precipitation.

FTIR spectrometry. For the fast precipitation procedure, the

reaction was stopped when the solution became cloudy,

normally after approximately 30 min. The suspension was

transferred with a glass pipette into an Eppendorf tube, and

centrifuged for 3 min at 14000 rpm. The pellet was re-

suspended in ethanol and placed in an agate mortar. After the

ethanol evaporated, the precipitate was lightly ground and

KBr pellets were prepared and analysed by FTIR spectrometry

(Nicolet 380).

Transmission electron microscopy (TEM). (FEI, Philips,

T12). The suspension was collected 10 min after diffusion

was initiated, centrifuged as above and the pellet was re-

suspended in ethanol. A drop of the ethanol suspension was

evaporated on a TEM grid.

Precipitations on a chitin substrate. b-Chitin was obtained

from the pen of the squid Loligo sp. (Mediterranean Sea) and

purified by reflux in 1 M NaOH solution, following Darmon

et al.30 It was then washed extensively with water and stored

dry. Pieces of chitin (0.5 x 0.5 cm) were incubated overnight

with 50 nmol (amino acids) mL21 Asprich or poly-Asp in

10 mM CaCl2 solution. The control chitin was incubated

with CaCl2 solution only. The chitin was then washed with

deionized water and placed in wells containing 0.5 mL 10 mM

CaCl2 solution. The crystallization was performed in the same

way as in the fast crystallization described above, but with no

additives in the solution.

Scanning electron microscopy (SEM). (Philips, XL30

FESEM FEG). The glass slides from each well were rinsed

in purified water and mounted on an aluminum stub with

double-sided carbon tape and Au/Pd sputtered.

Nucleation assay. A drop of 10 mM CaCl2 solution with or

without 50 nmol (amino acids) mL21 Asprich or poly-Asp

was placed in the center of a 12 mm siliconized glass slide

(Hampton Research). The incubation was performed in a

closed box containing open water vials, for 4–6 h at room

temperature. The slide was then rinsed with water, placed

face down on a 10 mM CaCl2 solution in a multi-well dish

and crystallization was initiated as described above for fast

This journal is � The Royal Society of Chemistry 2007 CrystEngComm, 2007, 9, 1171–1177 | 1175

precipitation. After 1 h the slides were collected, rinsed with

water and viewed with a light microscope (Nikon, Eclipse

E600 Pol).

Discussion

We show that Asprich protein inhibits calcite nucleation and

growth. It also induces and stabilizes an otherwise unstable

amorphous calcium carbonate phase. The stabilization is

transient and the mineral eventually transforms into calcite.

The effects of both Asprich and poly-Asp on calcite crystal

growth are not specific for certain crystallographic planes.

Nevertheless, there are significant differences in the manner

Asprich and poly-Asp interact with the growing crystals. The

rounded crystal edges observed during slow growth of calcite

in the presence of low concentrations of Asprich in solution

may result from reversible adsorption of Asprich on high-

energy steps and kink sites on the growing crystal and their

subsequent accumulation at the crystal edges. Alternatively,

they may also result from the addition of colloidal ACC

particles to the growing steps, as observed during the fast

growth experiments. In contrast, poly-Asp induces the forma-

tion of deep pits in the center of the {104} crystal faces, the

so-called ‘hopper crystals’. This morphology may arise from

adsorption of inhibitors on sites of emerging screw dislocations

or from Mullins–Sekerka instabilities, as observed for calcite

grown in the presence of simple alcohols.25 At higher additive

concentrations, the respective effects of poly-Asp and Asprich

become more pronounced, until deposition of spherulites,

characteristic of crystal formation through transient ACC

precursors, is observed for both additives. This occurs however

with Asprich at concentrations lower than with poly-Asp.

During fast precipitation experiments, both Asprich and

poly-Asp promote the formation of a colloidal suspension in

solution before the onset of crystallization in the control

experiment. The ACC precipitates formed in the presence of

Asprich, when stored in air, were stable for at least 20 d,

whereas ACC formed in the presence of poly-Asp crystallized

after 4 to 10 d. After 30 min in water, the former particles

were amorphous, while the latter were crystalline, based on

infrared analysis. Interestingly, colloidal particles formed in

the presence of Asprich, were observed to incorporate onto

layers of growing calcite single crystals. Although we do not

have direct in situ evidence that the particles adding to the

crystal are still amorphous, it is reasonable to assume that once

the colloids come in contact with the crystal they diffuse along

the surface to steps, where they undergo secondary nucleation

to become part of the growing single crystal.

Polyanions, such as poly-aspartate, poly-acrylate and poly-

carboxylate in general are well known for their ability to inhibit

calcite nucleation and induce the formation of amorphous

phases.31–33 Nevertheless, the potency of the Asprich protein is

found to be much higher at lower concentrations compared to

poly-Asp.

ACC may form in the presence of Asprich by inhibition of

nucleation and subsequent increase of the supersaturation

levels in the solution. This has been proposed for other

poly-anions,34 acidic macromolecules extracted from other

biominerals,35–37 and occurs during fast precipitation even

without additives.33 Moreover, Asprich, being negatively

charged, recruits calcium ions and thus locally increases the

supersaturation level such that ACC may precipitate. Asprich

is composed of y60% acidic residues, and y30% hydrophobic

residues. The poly-Asp peptide used here thus has approxi-

mately twice the negative charge/amino-acid concentration

compared to Asprich. When used at the same concentration

(in moles amino acids) mL21 as Asprich, poly-Asp’s ability

to inhibit calcite nucleation is lower. We deduce from this

that the potency of Asprich protein is not solely related to its

polyanionic nature, and that its composition and/or sequence

and structure play a significant role in ACC induction and

stabilization.

A mechanistic interpretation of the results presented must

provide an explanation for four independent effects observed

here in the presence of Asprich: inhibition of crystal growth,

inhibition of de novo crystal nucleation, induction of ACC and

stabilization of ACC particles. Inhibition of crystal growth as

well as concentration of calcium carbonate leading to its pre-

cipitation can both be explained by the poly-anionic character

of the protein that concentrates calcium (and subsequently

carbonate) ions and is adsorbed on crystal steps because of

its multiple charge.38 Inhibition of de novo nucleation by

mixed hydrophilic–hydrophobic peptides and proteins has

been suggested to arise from the hydrophobic domain acting as

a diffusion barrier for lattice ion adsorption to the incipient

nuclei.39,40 This would, however, require sufficient protein

coverage of the crystallization nuclei that could not be easily

achieved at the protein concentration used, unless there is

protein aggregation. Besides, Asprich adsorbed on hydro-

phobic glass also inhibits calcite nucleation much more than

the hydrophobic glass itself. This excludes this explanation in

the present case.

Some interesting information also related to the nucleation-

inhibiting mechanism can be deduced from the Asprich effect

on ACC stabilization. Interestingly, the IR spectra of the

colloidal particles show a broad absorption around the n4

absorption frequency. This is characteristic of both synthetic

and biogenic stable ACC minerals, whereas in transient ACC

this peak has reduced intensity, but remains relatively sharp.

Stable ACC contains structural water in proportions of

one molecule of water per CaCO3, while transient ACC is

practically anhydrous.10 We suggest that, in vitro, the Asprich-

induced ACC phase contains bound water, and that this

structural water plays a role in the stabilization of the phase.

Practically, the particles cannot crystallize unless water is

expelled. The hydrophobic part of the protein may then

conceivably contribute to stabilization of ACC by causing

water ordering. This would be similar to the role suggested in

the inhibition of ice formation by antifreeze proteins.41

Considering the above information on the effect of Asprich

on calcium carbonate deposition in vitro, what can be

concluded concerning the function(s) of Asprich in mollusk

shell prismatic layer formation? Nudelman et al.14 showed that

the deposition cycle leading to formation of the large calcite

crystals in growing A. rigida prismatic layer starts with

secretion of a chitin meshwork, which is then decorated with

20–30 nm particles of calcium carbonate containing Asprich.

While the particles become denser, they fuse with the existing

1176 | CrystEngComm, 2007, 9, 1171–1177 This journal is � The Royal Society of Chemistry 2007

calcite crystal. The final product is a single crystal of calcite.

We suggest that the particles are Asprich-stabilized ACC

particles introduced by the mantle cells into the mineralization

compartment, where they undergo epitaxial nucleation on

the preexisting calcite prism. The in vivo function of Asprich

would be to induce and stabilize ACC particles and to inhibit

uncontrolled calcite nucleation, until the mineral particles

come into contact with the preexisting calcite prism. The

crystallization process would then be similar to the sequence

observed during in vitro crystallization (Fig. 6), where initial

formation of transient ACC particles, followed by their fusion

to growing calcite crystals, occurs. In addition, the in vitro

experiments performed in the presence of chitin support a

possible function of the interaction between Asprich and chitin

in the formation of the biogenic mineral.

As yet there is no direct evidence for the use of a transient

ACC precursor phase in adult bivalve shell biomineralization,

although this has been suggested.14,15,42 It has been shown to

occur in the formation of the larval mollusk shell.8,9 In vitro

studies such as the one described here and by Gotliv et al.43

show that the proteins involved in the mineralization process

of the prismatic and nacreous layers induce the formation of

ACC phase that subsequently crystallizes into either calcite or

aragonite. It is therefore likely that adult bivalves such as

Atrina rigida, also use this apparently widespread strategy of

biomineralization. We note that this strategy can also be

exploited for the design of new synthetic crystalline materials

with complex shapes and superior materials properties.

Acknowledgements

We are especially grateful to Hong H. Chen for preparing the

recombinant Asprich protein. We thank Irit Nudelman for

help in experiments, Dr Talmon Arad for his help in the

electron microscopy work and Fabio Nudelman for useful

discussions. We thank the Israeli Ministry of Science and

the Minerva Foundation for financial support. L.A. is the

incumbent of the Dorothy and Patrick Gorman Professorial

Chair of Biological Ultrastructure, and S.W. is the incumbent

of the Dr Trude Burchardt Professorial Chair of Structural

Biology.

References

1 H. A. Lowenstam and S. Weiner, On Biomineralization, OxfordUniversity Press, New York, 1989.

2 S. Weiner, I. Sagi and L. Addadi, Science, 2005, 309, 1027.3 K. M. Towe and H. A. Lowenstam, J. Ultrastruct. Res., 1967, 17, 1.4 H. A. Lowenstam and S. Weiner, Science, 1985, 227, 51.5 E. Beniash, J. Aizenberg, L. Addadi and S. Weiner, Proc. R. Soc.

London, Ser. B, 1997, 264, 461.6 Y. Politi, T. Arad, E. Klein, S. Weiner and L. Addadi, Science,

2004, 306, 1161.7 R. Dillaman, S. Hequembourg and M. Gay, J. Morph., 2005, 263,

356.

8 J. C. Marxen, W. Becker, D. Finke, B. Hasse and M. Epple,J. Molluscan Studies, 2003, 69, 113.

9 I. M. Weiss, N. Tuross, L. Addadi and S. Weiner, J. Exp. Zool.,2002, 293, 478.

10 L. Addadi, S. Raz and S. Weiner, Adv. Mater., 2003, 15, 959.11 B. Hasse, H. Ehrenberg, J. Marxen, W. Becker and M. Epple,

Chem.–Eur. J., 2000, 6, 3679.12 Y. Politi, Y. Levi-Kalisman, S. Raz, F. Wilt, L. Addadi, S. Weiner

and I. Sagi, Adv. Funct. Mater., 2006, 16, 1289.13 H. Theel, Nova Acta Res. Soc. Sci. Upsala, 1892, 15, 1.14 F. Nudelman, H. H. Chen, H. A. Goldberg, S. Weiner and

L. Addadi, Faraday Discuss., 2007, 136, 9.15 L. Addadi, D. Joester, F. Nudelman and S. Weiner, Chem.–Eur. J.,

2006, 12, 980; F. Nudelman, B. A. Gotliv, L. Addadi and S. Weiner,J. Struct. Biol., 2006, 153, 176.

16 S. Weiner, Calcif. Tissue Int., 1979, 29, 163.17 H. A. Lowenstam, Bull. Marine Sci., 1989, 452, 243.18 D. Tsukamoto, I. Sarashina and K. Endo, Biochem. Biophys. Res.

Commun., 2004, 320, 1175.19 A. Veis and A. Perry, Biochemistry, 1967, 6, 2409.20 F. Marin, R. Amons, N. Guichard, M. Stiger, A. Hecker,

G. Luquet, C. Layrolle, G. Alcaraz, C. Riondet andP. Westbroek, J. Biol. Chem., 2005, 280, 33895.

21 B. Gotliv, N. Kessler, J. L. Sumerel, D. E. Morse, N. Tuross,L. Addadi and S. Weiner, ChemBioChem, 2005, 6, 304.

22 A. Berman, J. Hanson, L. Leiserowitz, T. F. Koetzle, S. Weinerand L. Addadi, Science, 1993, 259, 776.

23 J. Lahiri, G. Xu, D. M. Dabbs, N. Yao, A. Aksay and J. T. Groves,J. Am. Chem. Soc., 1997, 119, 5449.

24 L. A. Gower and D. A. Tirrell, J. Cryst. Growth, 1998, 191, 153.25 S. R. Dickinson and K. M. McGrath, J. Mater. Chem., 2003, 13,

928.26 R. Gueta, A. Natan, L. Addadi, S. Weiner, K. Refson and

L. Kronik, Ang. Chem., 2006, 119, 295.27 G. Falini, S. Albeck, S. Weiner and L. Addadi, Science, 1996,

271, 67.28 T. Kato and T. Amamiya, Chem. Lett., 1999, 3, 199.29 J. Sambrok, F. E. Fritsch and T. Maniatis, Molecular cloning:

A Laboratory Manual, Cold Spring Harbor Laboratory Press,Cold Spring Harbor, NY, 1989.

30 S. E. Darmon and K. M. Rudall, Discuss. Faraday Soc., 1950, 9,251.

31 L. B. Gower and D. J. Odum, J. Cryst. Growth, 2000, 210, 719.32 G. F. Xu, N. Yao, I. A. Aksay and J. T. Groves, J. Am. Chem.

Soc., 1998, 120, 11977.33 J. Rieger, T. Frechen, G. Cox, W. Heckmann, C. Schmidt and

J. Thieme, Faraday Discuss., 2007, 136, 265.34 H. Colfen and S. Mann, Angew. Chem., 2003, 42, 2350.35 J. Aizenberg, G. Lambert, L. Addadi and S. Weiner, Adv. Mater.,

1996, 8, 222.36 S. Raz, P. Hamilton, F. Wilt, S. Weiner and L. Addadi, Adv. Funct.

Mater., 2003, 13, 480.37 B. Pokroy, E. Zolotoyabko and N. Adir, Biomacromolecules, 2006,

7, 550.38 S. Elhadj, J. J. De Yoreo, J. R. Hoyer and P. M. Dove, Proc. Natl.

Acad. Sci. U. S. A., 2006, 103, 19237.39 D. I. Hay, E. C. Moreno and D. H. Schlesinger, Inorg. Persp. Biol.

Med., 1979, 2, 271.40 C. S. Sikes and A. P. Wheeler, Chemical Aspects of Regulation

of Mineralization, University of South Alabama PublicationServices, 1988.

41 Z. Jia and P. L. Davis, TRENDS Biochem., 2002, 27, 101.42 N. Nassif, N. Pinna, N. Gehrke, M. Antonietti, C. Jager and

H. Colfen, Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 12563.43 B. A. Gotliv, L. Addadi and S. Weiner, ChemBioChem, 2003, 4,

522.

This journal is � The Royal Society of Chemistry 2007 CrystEngComm, 2007, 9, 1171–1177 | 1177