Purification and Characterization of a Vaterite-Inducing Peptide

Biomacromolecules 2005, 6, 1429-1437

1429

Purification and Characterization of a Vaterite-Inducing Peptide, Pelovaterin, from the Eggshells of Pelodiscus sinensis (Chinese Soft-Shelled Turtle) Rajamani Lakshminarayanan,† Emma Ooi Chi-Jin,† Xian Jun Loh,† R. Manjunatha Kini,‡ and Suresh Valiyaveettil*,† Departments of Chemistry and Biological Sciences, National University of Singapore, 3 Science Drive 3, Singapore 117 543 Received November 15, 2004; Revised Manuscript Received February 8, 2005

Proteins play a crucial role in the biomineralization of hard tissues such as eggshells. We report here the purification, characterization, and in vitro mineralization studies of a peptide, pelovaterin, extracted from eggshells of a soft-shelled turtle. It is a glycine-rich peptide with 42 amino acid residues and three disulfide bonds. When tested in vitro, the peptide induced the formation of a metastable vaterite phase. The floretshaped morphology formed at a lower concentration (∼1 µM) was transformed into spherical particles at higher concentrations (>500 µM). The solution properties of the peptide are investigated by circular dichroism (CD), fluorescence emission spectroscopy, and dynamic light scattering (DLS) experiments. The conformation of pelovaterin changed from an unordered state at a low concentration to a β-sheet structure at high concentrations. Fluorescence emission studies indicated that the quantum yield is significantly decreased at higher concentrations, accompanied by a blue shift in the emission maximum. At higher concentrations a red-edge excitation shift was observed, indicating the restricted mobility of the peptide. On the basis of these observations, we discuss the presence of a peptide concentration-dependent monomer-multimer equilibrium in solution and its role in controlling the nucleation, growth, and morphology of CaCO3 crystals. This is the first peptide known to induce the nucleation and stabilization of the vaterite phase in solution. Introduction Calcium carbonate is an important biological mineral utilized by organisms for myriad purposes.1,2 All three stable crystalline polymorphs of calcium carbonate (calcite, aragonite, and vaterite) and the hydrated amorphous calcium carbonate are known to exist in organisms under specific conditions.3-6 Calcite is thermodynamically the most stable polymorph under standard conditions: however, altering the kinetics of crystallization by means of added divalent ions such as Sr2+ and Ba2+ induces the formation of aragonite or vaterite phases.7,8 Biomacromolecules such as proteins control the nucleation, polymorph selection, and final morphology of mineral tissues such as bones, teeth, and shells.9-13 Since formation of mineral phase take place in the presence of a complex mixture of proteins and other macromolecules, a number of attempts have been made to purify and characterize the active proteins to understand the mechanism of biomineralization. In the past few years, purification and amino acid sequences of several proteins and peptides isolated from various aragonitic shells have been reported.14-24 In vitro studies showed that the purified proteins induce or inhibit the nucleation of calcite phase, and no aragonite nucleation was observed unless the crystal* To whom correspondence should be addressed: tel +65 6874 4327; fax +65 6779 1691; e-mail [email protected]. † Department of Chemistry. ‡ Department of Biological Sciences.

lization was performed in the presence of magnesium ions.14-24 Eggshells are the fastest forming hard acellular biocomposites, which protect the developing embryo. For example, in the case of chicken eggshell, ∼5 g of the mineral phase is produced within 24 h inside the oviduct.25 The calcified layer of the eggshell consists of ∼95% mineral and ∼5% organic molecules such as proteins.26 The mineral phase acts as a mechanical support as well as allowing the diffusion of gases, water, and ions; therefore it is essential for the survival of the embryo. So far, several matrix proteins from chicken eggshells have been purified and characterized.27-29 These proteins are subdivided into three groups: noncollagenous bone proteins (osteopontin), eggshell-specific proteins (ovocleidins and ovocallyxins), and egg white proteins (ovalbumin, ovotransferrin, and lysozyme). Hincke et al.30 showed that egg white lysozyme and ovotransferrin influence the morphology of CaCO3 crystals. Dermatan sulfate, chondroitin sulfate, and hyaluronic acid were also identified in the chicken eggshell matrix.31-33 We have recently reported the extraction and characterization of a goose eggshell matrix protein, ansocalcin.34,35 Eggshells of soft-shelled turtles are mineralized with aragonite polymorph.36 Figure 1 shows the ultrastructure of the mineral phase present in the chicken (calcitic) and turtle (aragonitic) eggshells. The mineral phases, in both cases, are formed over the shell membrane on regiospecific sites and organized into various domains. In the case of turtle shell, the crystals grow in fan-shapped

10.1021/bm049276f CCC: $30.25 © 2005 American Chemical Society Published on Web 03/26/2005

1430

Biomacromolecules, Vol. 6, No. 3, 2005

Figure 1. Cross section of the eggshells: electron micrographs of (A) chicken and (B) sea turtle eggshells. The various layers are indicated by the letters M, mamillary; P, palisade; C, cuticle; and SM, shell membrane. Scale bar ) 50 µm.

units encompassing aragonite needles. However, the calcite crystals acquire a conelike shape in the chicken eggshells. This project was initiated to investigate the identity, similarity, and differences in the composition of soluble organic matrixes (proteins) associated with the eggshell matrix of soft-shelled turtle, Pelodiscus sinensis. Here, we report the extraction, purification, determination of the amino acid sequence, secondary structure, and in vitro mineralization studies of the major matrix peptide purified from turtle eggshell matrix. Materials and Methods Taxonomic Information. The taxonomic position of the animal, according to Wiegmann (1834), is detailed below: Kingdom Phylum Subphylum Class Order Family Genus Species

Animalia Chordata Vertebrata Reptilia Testudines Trionychidae Pelodiscus sinensis

Extraction of the Soluble Organic Matrix. Fresh turtle eggshells were purchased from Sin Yaw Enterprise (M) Sdn Bhd, Malaysia. The eggshells were treated with hypochlorite bleach solution (5%) for about 15 min. The bleach treatment removed the shell membrane and other proteins/peptide on the surface of the eggshell, thus allowing the identification of intracrystalline components. The clear white crystals of the bleached eggshells were powdered, decalcified by slow addition of 2 N hydrochloric acid overnight at 4 °C, and centrifuged for 30 min at 3000g. A significant amount of insoluble components were present in the extract. The centrifugate was dialyzed against Millipore water at 4 °C for 2 days, and the desalted solution was lyophilized and used for further purification.

Lakshminarayanan et al.

Purification of the Major Matrix Component from SOM. The SOM was fractionated on a Jupiter C18 reversedphase column (5 µm, 250 mm × 10 mm) by use of a Vision Workstation (Perkin-Elmer PerSeptive Biosystems). The extract (∼20 mg) was injected onto a column that was equilibrated with 0.1% trifluoroacetic acid, and the bound proteins were eluted with a linear gradient of acetonitrile at a flow rate of 2 mL/min. The elution of the peptide was monitored at 215 and 280 nm. Electrospray Ionization Mass Spectrometry. The mass of the purified peptide was determined by ESI-MS on a Perkin-Elmer Sciex API 300 triple-quadrupole instrument equipped with an ion spray interface. The ion spray voltage was set at 4.6 kV and the orifice voltage at 30 V. Nitrogen was used as a curtain gas with a flow rate of 0.6 L/min, while compressed air was utilized as the nebulizer gas. The sample was injected into the mass spectrometer at a flow rate of 50 µL/min and scanned from mass to charge ratio (m/z) 500 to 2500. The multiply charged spectrum was deconvoluted into the mass scale with the Biospec Reconstruct program and the homogeneity of the fractions was confirmed. Reduction and Alkylation of Native Peptide. The lyophilized peptide (∼3 mg) was dissolved in 1000 µL of 130 mM Tris-HCl, 1 mM ethylendiaminetetraacetic acid (EDTA), and 6 M guanidine hydrochloride (pH 7.5). 2-Mercaptoethanol (20 µL/mg of protein) was added and the mixture was incubated under nitrogen atmosphere for 2 h at 37 °C. The alkylating agent, 4-vinylpyridine (200 µL/mg of protein) was subsequently added and the mixture was incubated under nitrogen for another 2 h at room temperature. The S-pyridylethylated peptide was separated from the reaction mixture by reversed-phase high-performance liquid chromatography (HPLC) on a Jupiter C18 (5 µm, 250 mm × 10 mm) column with a linear gradient of acetonitrile. MALDI-TOF Mass Spectrometry. Matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectra of the pyridylethylated peptide were obtained on a Voyager-DE Biospectrometry workstation equipped with a 337-nm nitrogen laser. To obtain a good signal-to-noise ratio, 150-200 single-shot spectra were collected. Saturated sinapinic acid in 50% acetonitrile was used as the matrix. The fractionated pyridylethylated peptide (0.5 µL) was mixed with 0.5 µL of the matrix and dried prior to the analysis. Amino-Terminal Sequencing. Amino-terminal sequencing of the native and S-pyridylethylated peptides was performed by automated Edman degradation method on a Perkin-Elmer Applied Biosystems 494 pulsed-liquid-phase protein sequencer (Procise) with an online 785A phenylthiohydantoin- (PTH-) amino acid analyzer. Crystal Growth Experiments. The supersaturated calcium bicarbonate solution was prepared by passing purified CO2 through a stirred suspension of CaCO3 in Milli-Q water (2 g/L) for 90 min. The suspension was filtered through a 0.45 µm nylon filter, and CO2 was passed through the clear solution for 30 min to dissolve the residual nuclei. The Ca2+ concentration was found to be 7.2 ( 0.12 mM and the pH was 5.9-6.1. CaCO3 crystals were grown on 12 mm glass coverslips placed inside the supersaturated Ca(HCO3)2 solu-

Vaterite-Inducing Peptide from Eggshell Matrix

tion kept inside a 12 cm Petri dish. Typically, 1 mL of the Ca(HCO3)2 solution along with aliquots of peptide (5 µg/ mL to 5 mg/mL) dissolved in 7.5 mM CaCl2 solution were introduced onto the coverslips. The whole setup was covered with aluminum foil with a few pinholes on the top. To study the role of proteins in the CaCO3 crystallization, the slides were carefully lifted from the crystallization wells after 24 h intervals, rinsed gently with Millipore water, air-dried at room temperature, glued to copper stubs, and used for analysis. The morphology of the CaCO3 crystals nucleated was examined on a Phillips XL 30 FEG scanning electron microscope at 15/20 kV after sputter coating with gold to increase the conductivity. The mineral phase was identified from micro-Raman spectra acquired at room temperature in the backscattering geometry with a single-grating Raman spectrometer (Model Spex T64000, Jobin Yvon, France). Circular Dichroic Spectroscopy. The secondary structure of the protein was investigated on a Jasco J-810 spectropolarimeter. The CD spectra of the protein at a concentration range of 0.1-5 mg/mL in water were collected in 0.1 cm sample cells. To study the effect of Ca2+ ions on the protein structure, spectra were also recorded in 7.5 mM CaCl2 solution. The instrument optics were flushed with 30 L/min nitrogen gas. A total of three scans were recorded and averaged for each spectrum and baseline-subtracted. Fluorescence Spectroscopy. Steady-state fluorescence emission spectrum was recorded on a Shimadzu RF-5301PC fluorescence spectrophotometer, with the emission and excitation band-passes set at 5 nm. The excitation wavelength was set at 295 nm and the spectra were recorded from 300 to 430 nm. Spectrum of the peptide were recorded either in 7.5 mM CaCl2 solution or in water. The concentration of the peptide was varied from 0.5 to 5 mg/mL. For the rededge excitation shift (REES) studies, the excitation wavelength was changed from 290 to 305 nm. Dynamic Light Scattering. About 120 µL of the peptide (2.5 and 5 mg/mL) in 7.5 mM CaCl2 solution or in water was used to perform dynamic light scattering experiments on a PDDLS/batch (Precision Detectors, Franklin, MA) instrument. Intensity data from each sample were collected in five replicates and analyzed by use of the PRECISION DECONVOLVE program, which yielded size-versus-fraction distribution plots. Results Purification and Characterization of Soluble Organic Matrix. Figure 2A shows the elution profile of proteins present in the SOM of the turtle eggshell. The HPLC profile showed a diffuse band around 25-35% B and a sharp major fraction eluting at around 50% B. The yield of this major fraction was about 200 µg per gram of the eggshell. ESIMS of this fraction is homogeneous with a mass of 4189.92 ( 0.89 Da (Figure 2B). Because of its ability to nucleate the vaterite phase (see below), this peptide is named as pelovaterin (Pelodiscus sinensis vaterite inducing protein). Amino acid sequence of the native peptide yielded 38 residues (DDTPSSRXGS GGWGPXLPIV DLLXIVHVTV GXSGGFGX) with nonidentifiable residues at positions 8,

Biomacromolecules, Vol. 6, No. 3, 2005 1431

Figure 2. Purification and characterization of pelovaterin. (A) Reversed-phase high-performance liquid chromatography of the SOM of bleach-treated turtle eggshells. The dotted line indicates the percentage gradient of buffer B (0.1% TFA in 80% ACN). The fractions indicated by the horizontal line contain the peptide pelovaterin. (B) ESI-MS of pelovaterin. The inset depicts the reconstructed mass spectrum. (C) MALDI-TOF mass spectrum of the pyridylethylated pelovaterin.

16, 24, 32, and 38. To identify these residues and complete the sequence, the peptide was reduced and pyridylethylated. The molecular mass of the pyridylethylated peptide determined by MALDI-TOF was 4827.8 Da (Figure 2C). The increase in mass accounted for the incorporation of six pyridylethyl groups confirming that pelovaterin contains six cysteine residues. Amino-terminal sequencing of the pyridylethylated pelovaterin unambiguously yielded 42 amino acid residues as DDTPSSRCGS GGWGPCLPIV DLLCIVHVTV GCSGGFGCCR IG. The calculated mass of this sequence is 4189.85 with the existence of cysteine bridges. This value is in good agreement with the observed mass of 4189.92 ( 0.89 Da obtained from ESI-MS.

1432



Table 1. Comparison of Amino Acid Sequence of Pelovaterin with Other Peptides Purified from Aragonitic Shells

a

This report. b NCBI accession no. AAK00635 (ref 22). c NCBI accession no. P82595 (ref 19). d NCBI accession no. AAQ08227 (ref 23).

When the sequence was appended into the databases such as FASTA and BLAST search, no proteins with significant similarity were found. On the basis of the amino acid sequence, pelovaterin is rich in glycine (21.4%), cysteine (14.3%), serine (9.5%), and valine (9.5%). The N-terminus contains six hydrophilic residues while the rest of the sequence is characterized by predominantly hydrophobic residues. Table 1 gives the comparison of the amino acid sequence of pelovaterin with other proteins of low molecular mass extracted from aragonite shells whose sequence have been previously determined or deduced. Perlustrin19 and pelovaterin contain a high degree of disulfide bonds compared to the other proteins, and no correlation in structure or property exists among this group of proteins. Therefore, pelovaterin is a unique peptide associated with the turtle eggshell matrix. Interaction of Pelovaterin with CaCO3 Crystals. To test the role of pelovaterin in biomineralization, we carried out in vitro calcium carbonate crystallization experiments. The crystal growth was performed by the slow evaporation of supersaturated calcium bicarbonate solution prepared by the procedure of Kitano et al.7 The peptide concentration was varied from 5 µg/mL to 5 mg/mL, and the crystals collected after 24 h were characterized. The representative electron micrographs of the crystals grown in the presence of pelovaterin are shown in Figure 3. Peptide used at a concentration range of 5-100 µg/mL induced the formation of floret-shaped crystals, characteristic of vaterite phase, whereas at higher concentrations (g0.5 mg/mL) spherical particles (25-30 µm) of vaterite were observed exclusively. With increasing concentration of the peptide, a significant decrease in the size with no changes in morphology of the spherules was observed. Some of the spherules were intertwined to form larger aggregates. In the absence of pelovaterin only rhombohedral calcite crystals were formed, suggesting that the vaterite nucleation was mediated by the peptide (Figure 3). Figure 4 shows the Raman scattering data of the spherules formed at various concentrations of pelovaterin. At all concentrations of the peptide, the spectrum exhibited a strong peak at 1090 cm-1 and a weak absorption around 1075 cm-1 in the ν1 region. The observed strong peak around 118 cm-1 and weak absorptions around 750, 303, and 270 cm-1 are characteristic of the vaterite phase.37,38 Recently, in vitro studies demonstrated that some of the proteins associated with the aragonitic shells induce calcite nucleation.14-24 In the presence of pelovaterin, predominantly the metastable vaterite phase was observed even at a very low peptide concentration (∼50 nM) and no transition to aragonite phase was observed within 2 weeks. The crystal growth experiments indicate that pelovaterin induces the formation of

Figure 3. Nucleation of CaCO3 by pelovaterin: representative electron micrographs of the CaCO3 crystals grown at various concentration of pelovaterin. The concentration of the peptide is indicated on the images.

vaterite polymorph and alters the morphology in a concentration-dependent manner. Solution Properties of Pelovaterin. To understand the substantial increase in the amount of vaterite particles at higher concentrations of the peptide, solution properties of the peptide were investigated by circular dichroism, tryptophan fluorescence, and dynamic light scattering studies. Concentration-Dependent Unordered h β-Sheet Transition in Pelovaterin. The conformation of the peptide was examined by far UV-CD at 23 ( 1 °C in 7.5 mM CaCl2 solution and in water by diluting a concentrated stock solution of the peptide. The CD spectrum of the peptide exhibits remarkable concentration dependence, suggesting that the peptide at higher concentrations undergoes intermolecular interactions assisted via self-organization (Figure 5A). At 5 mg/mL, the spectrum is characterized by a negative dichroic minimum at 217 nm and a positive maximum at 205 nm, suggesting that the peptide adopts a high degree of β-sheet conformation in addition to a small amount of type



Figure 5. Secondary structure of pelovaterin: circular dichroic spectra of the protein recorded in (A) CaCl2 solution and (B) water. The concentrations of the peptide used were (1) 0.1, (2) 0.25, (3) 0.5, (4) 1, (5) 1.5, (6) 2.5, and (7) 5 mg/mL. The inset shows the CD in millidegrees at 217 nm, which is a measure of β-sheet content, as a function of pelovaterin concentration. For a better understanding, positive values of the CD are plotted in the graph.

Figure 4. Characterization of the mineral phase formed in the presence of pelovaterin. Raman spectra of the CaCO3 crystals were grown at various concentrations of pelovaterin: (a) 0.5, (b) 1, (c) 2.5, and (d) 5 mg/mL. The characteristic vaterite absorption peaks are indicated in the figure. The spectra of calcite (e) and aragonite crystals (f) are given for comparison.

II β-turn. The appearance of a positive band in the region 200-210 nm indicates the contribution from aromatic side

chains.39 A closer examination of Figure 5A reveals a considerable increase in noise for the most concentrated solutions at lower wavelength regions, which indicate enhanced scattering of light. Upon lowering the concentration (2.5 mg/mL), the intensity was increased and the negative minimum was shifted to 208 nm with an additional small minimum at 196 nm, suggesting that the conformation is dominated by antiparallel β-sheet folding and a very small or negligible amount of R-helix structure.40 At a peptide concentration of 1.5 mg/mL, the negative minimum value is shifted to 203 nm without much change in the intensity and a shoulder peak appears around 208 nm, presumably due to the presence of a mixture of antiparallel β-sheet and unordered conformations of the proteins. At 1 mg/mL, the protein exhibited fine structure around 197-205 nm with a negative dichroic minimum at 201 nm. The negative minimum was further blue-shifted to 199 nm as the concentration was lowered from 0.5 mg/mL onward, suggesting that at high dilutions (i.e., low concentrations) the peptide exists in an unordered state. At high concentrations, the secondary structure is dominated by a highly ordered

1434


β-sheet structure. Thus the data demonstrate reversible concentration-dependent transitions from unordered to β-sheet transition with changes in concentration. The CD spectra in water are given in Figure 5B. A few important differences compared to the CD studies in CaCl2 solution are interesting to mention here: (i) at high concentration, significantly more scattering was observed in water than in calcium chloride solution; (ii) the positive maximum observed at high concentrations of the protein was red-shifted slightly in water (211 nm); (iii) CD absorption at 217 nm indicates an increase in β-sheet content with increasing concentration of the protein in CaCl2 solution (Figure 5 insets); and (iv) the transition to ordered structure is more rapid in water than in CaCl2 solution. It is conceivable that the solution structure was independent of Ca2+ ions, although subtle differences were observed at high concentration. These conformational changes are reminiscent of amyloidforming proteins that also form β-sheet structure, rendering them susceptible to oligomerization.41 It has been observed that peptides that undergo transition from an unordered state to an ordered state exhibit an isodichroic point (i.e., the point where CD intensities are equal), which is characteristic of a two-state transition.42,43 No single isodichroic point was observed in the CD spectra at various concentrations of pelovaterin, suggesting that the transition from monomers to multimers is not a simple two-stage transition but proceeds through a series of intermediates. A number of β-sheet peptides have been shown to spontaneously self-assemble and form a macroscopic matrix in a concentration-dependent manner in the presence of monovalent salts or by variation of temperature or pH.44-46 The results presented here gave evidence for the self-assembly of a 42-mer peptide to form rigid β-sheet structure in solution at higher concentration and influence the nucleation of calcium carbonate crystals in solution. We also observed that when a concentrated solution of pelovaterin was left undisturbed, protofibrils were precipitated after 1 week, suggesting the formation of large peptide aggregates (data not shown). Intrinsic Tryptophan Fluorescence of Pelovaterin. Further insight into the structure and dynamic changes in the peptide is obtained from fluorescence spectroscopy investigations. Since pelovaterin contains one tryptophan (W13) residue, the intrinsic tryptophan fluorescence emission was recorded at various concentrations in 7.5 mM CaCl2 solution and in water at an excitation wavelength of 295 nm. The emission spectrum changed significantly with the concentration of the peptide (Figure 6A). At low concentration (0.5 mg/mL), the emission spectrum of pelovaterin is not a smooth Lorentzian curve as expected for a protein containing single, uncomplexed tryptophan; instead the peptide exhibited a plateau from 332 to 350 nm. With increasing concentration of the protein, a smooth Lorentzian curve with a maximum centered at 350 nm was observed. At high peptide concentrations (2.5 and 5.0 mg/mL) the quantum yield dropped significantly (indicated by the area under the peak), followed by a slight blue shift in the maximum (347 nm, Figure 6A inset). The appearance of a plateau in the region 332-350 nm at a low concentration of the peptide could be due to overlapping of spectra as a result of the existence of various microstates of the residue W13


in the rapidly altering unordered conformation.47 The emission maximum at 350 nm with increasing concentration suggests the exposed nature of the fluorophore to the polar environment. The decrease in the emission intensity and blue shift of the emission maximum at higher concentrations is consistent with the self-aggregation of pelovaterin.48 Thus the results from CD and fluorescent emission spectral analysis indicated the existence of monomers and multimers of pelovaterin in solution. The emission spectrum of a fluorophore is independent of excitation wavelength and solvent viscosity.49 However, for polar fluorophores in a moderately polar and viscous solvent, the emission maximum is observed to shift to the longer wavelength.50 A shift in the fluorescence emission maximum toward the red edge of the absorption spectrum is termed as red-edge excitation shift (REES). REES is observed for the polar fluorophores in viscous solution or condensed phases.51,52 Red-edge effects were first observed in the steady-state spectra of various dyes in condensed phase and are more pronounced for proteins containing a single tryptophan residue.53 The insets of Figure 6B-F show the plots of emission maxima obtained with change in excitation wavelength for the given concentration of pelovaterin in 7.5 mM CaCl2 solution. At lower concentrations, there is no change in the emission maximum, indicating that the motion of the dipoles surrounding the tryptophan is fast compared to the lifetime of the excited state.54 However, at higher concentrations, a red-edge effect was observed with shifts of ∼5 nm at 2.5 mg/mL and ∼11 nm at 5 mg/mL. These observed shifts in the fluorescence emission maximum indicated a delayed mobility of the dipole groups in the microenvironment of the tryptophan residue. Similar changes were observed in water, indicating that the calcium ions do not influence the conformation or spectral properties of the peptide. Dynamic Light Scattering Studies. To obtain better insight into the aggregation properties of pelovatarin in solution, dynamic light scattering measurements were performed in 7.5 nM CaCl2 solution. The mean diffusion coefficient (〈D〉), which is related to the motion of the macromolecules in solution, and the mean hydrodynamic radius (〈Rh〉) obtained by these experiments are summarized in Table 2.55 The observed changes in 〈D〉 and 〈Rh〉 values imply a concentration-dependent change in the size of the peptide aggregates. In CaCl2 (5 mg/mL) solution, particles with a mean hydrodynamic radius of ca. 20 nm were observed. Dilution of this solution to 2.5 mg/mL yielded a particle size of ca. 9 nm. When the DLS experiments were performed in water, similar concentration-dependent changes were observed. This suggests that the peptide aggregation is independent of Ca2+ ions, although some differences were observed in the values of 〈D〉 and 〈Rh〉. These results, in combination with other spectroscopic data, confirm the formation of aggregated pelovaterin particles in solution and a concentration-dependent aggregation process. Discussion Calcification is an important process in many living beings. Though a number of biomacromolecules have been postu-



Figure 6. (A) Emission spectra of pelovaterin in 7.5 mM CaCl2 at various concentrations: (1) 0.5, (2) 1, (3) 1.5, (4) 2.5, and (5) 5 mg/mL. The concentration-dependent changes in the maximum emission intensity (Fmax) are shown in the inset. (B-F) Red-edge excitation shift (REES) of pelovaterin at various concentrations: (B) 0.5, (C) 1, (D) 1.5, (E) 2.5, and (F) 5 mg/mL. Insets show the plots of excitation wavelength (λex) versus maximum emission wavelength (λmax).

1436



Table 2. Aggregation Parameters Obtained from Dynamic Light Scattering Measurements of Pelovaterin in Water and in CaCl2 〈Rh〉, nm

〈D〉, cm2/s

peptide concn (mg/mL)

CaCl2

water

CaCl2

water

5 2.5

20.2 9.4

18.1 14.3

2.85 × 10-7 6.4 × 10-7

3.2 × 10-7 4.1 × 10-7

lated to play an important role in the skeletal construction, understanding the molecular mechanism of calcification is still an interesting area of research. We have purified and characterized a 4 kDa peptide (pelovaterin) from the turtle eggshell and determined its amino acid sequence, solution structure, and ability to nucleate CaCO3 crystals. Pelovaterin shows no homology with any known proteins. When tested in vitro, the new peptide induced nucleation of vaterite phase, which is rarely seen with other proteins. When 50-500 ng/ mL pelovaterin was introduced into the crystallization medium, a mixture of calcite and vaterite crystals was observed. Increasing the concentration to 5 µg/mL resulted in the nucleation of exclusively floret-shaped vaterite phase (Figure 2), suggesting specific interactions between the peptide and the mineral phase.56 Presence of a very small amount of pelovaterin (5-100 µg/mL) induced the formation of floret-shaped vaterite crystals, whereas spherical vaterite particles were observed at high concentrations (500-5000 µg/mL). The formation of spherulitic vaterite particles indicates that the peptide forms an aggregated structure with the exposed carboxylate groups for calcium binding.57 As the concentration of the peptide was increased in the crystallization medium, the size of the vaterite spherules was markedly reduced with a concomitant increase in the number of crystallites. These results demonstrate the ability of pelovaterin to alter the polymorphism, stabilize the thermodynamically less stable phase, and accelerate the mineral growth. Davey et al.58 suggested two possible transformation pathways from unstable to stable phase. The first is the solution-mediated transformation wherein the dissolution of less stable polymorph is followed by the nucleation of stable polymorph. In water, the vaterite phase is known to irreversibly transform to the more stable calcite phase within 80 h.59 When the crystallization experiment was performed at a low concentration of the peptide, no such transition was observed even after 2 weeks, indicating that the vaterite phase is kinetically stabilized by the pelovaterin. It is likely that pelovaterin forms strong Ca-O bonds and prevents the dissolution of vaterite particles. The second mechanism is the solid-state transformation where the internal rearrangement of the lattice occurs. Wilbur and Watabe60 observed the formation of vaterite phase during the repair of defective shells in the mollusk ViViparus intertextus and postulated this to be a precursor to the aragonite phase. The absence of any of these transformations in our assay suggests that the conditions used in our laboratory did not mimic the in vivo process. Although stable biogenic vaterite phase is known to exist in tropical ascidians, Herdmania momus, and two fish species, its transformation to stable aragonite or calcite phase has not been observed.61 Recently, partial biomimetic reconstitution of the avian eggshell has been demonstrated

by use of the cooperative interaction between the shell membrane, carbonic anhydrase, and proteoglycans.32 Further analysis of SOMs and their interaction with the shell membrane may probably uncover the mechanism of aragonite nucleation. The observed increase in the number of vaterite particles with increasing peptide concentration suggests that the nucleation of the crystals may be due to the aggregation of the peptide. On the basis of CD, emission spectra, and DLS studies, it is clear that pelovaterin exists predominantly in monomeric (at low concentration) and aggregated (at higher concentrations) forms in solution. The observation of REES and decrease in the diffusion coefficient suggest that the motion of the peptide chain in solution is hindered in the aggregated state. Such restricted movement provides a scaffold for the nucleation of vaterite particles as it contains numerous Ca2+ binding motifs. The concentration dependence of the CD spectra suggests that the transition from monomer to multimeric state is not a simple one- or twostep transition as no single isodichroic point was observed. Particle size analysis further revealed a reduction in size when the concentration of the protein was halved, confirming the concentration-dependent aggregation of pelovaterin. In conclusion, a novel peptide, pelovaterin, has been extracted from the turtle eggshell matrix, purified, and fully characterized. Vaterite crystals with floret and spherical morphologies were nucleated by pelovaterin at various concentrations. The solution studies with dynamic light scattering, CD, and fluorescence spectroscopy showed an aggregated structure for the peptide at high concentration. The observed formation and stabilization of vaterite phase by a protein extracted from turtle or other eggshell matrix is interesting owing to the metastable nature of the vaterite and stabilization effect of the peptide, without which other polymorphs were formed. Acknowledgment. R.L. thanks the Singapore Millennium Foundation for the award of a fellowship. S.V. acknowledges financial support from the National University of Singapore through academic research funds (ARF) and technical support from the laboratories in the departments of chemistry and biological sciences. Supporting Information Available. Crystallization of CaCO3 in the presence of pelovaterin and size distribution of pelovaterin in 7.5 mM calcium chloride solution obtained by DLS. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Weiner, S.; Addadi, L. J. Mater. Chem. 1997, 7, 689-702. (2) Weiner, S.; Addadi, L.; Wagner, H. D. Mater. Sci. Eng., C 2000, 11, 1-8. (3) Lowenstam, H. A. Science 1981, 211, 1126-1131. (4) Krampitz, G.; Graser, G. Angew. Chem., Int. Ed. 1988, 27, 11451156. (5) Heuer, A. H.; Fink, D. J.; Laraia, V. J.; Arias, J. L.; Calvert, P. D.; Kendall, K.; Messing, G. L.; Blackwell, J.; Rieke, P. C.; Thompson, D. H.; Wheeler, A. P.; Veis, A.; Caplan, A. I. Science 1992, 255, 1098-1105. (6) Addadi, L.; Raz, S.; Weiner, S. AdV. Mater. 2003, 15, 959-970. (7) Kitano, Y.; Park, K.; Hood, D. W. J. Geophys. Res. 1962, 67, 48734874.

Vaterite-Inducing Peptide from Eggshell Matrix (8) Easton, D.; Claugher, D. Mineral. Magn. 1986, 50, 332-336. (9) Belcher, A. M.; Wu, X. H.; Christensen, R. J.; Hansma, P. K.; Stucky, G. D.; Morse, D. E. Nature 1996, 381, 56-58. (10) Levi, Y.; Albeck, S.; Brack, A.; Weiner, S.; Addadi, L. Chem.s Eur. J. 1998, 4, 389-396. (11) Falini, G.; Albeck, S.; Weiner, S.; Addadi, L. Science 1996, 271, 67-69. (12) Gotliv, B. A.; Addadi, L.; Weiner, S. ChemBioChem 2003, 4, 522529. (13) Hasse, B.; Ehrenberg, H.; Marxen, J. C.; Becker, W.; Epple, M. Chem.sEur. J. 2000, 6, 3679-3685. (14) Miyamoto, H.; Miyashita, T.; Okushima, M.; Nakano, S.; Morita, T.; Matsushiro, A. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 96579660. (15) Sarashina, I.; Endo, K. Am. Mineral. 1998, 83, 1510-1515. (16) Bedouet, L.; Schuller, M. J.; Marin, F.; Milet, C.; Lopez, E.; Giraud, M. Comp. Biochem. Physiol B: Biochem. Mol. Biol. 2001, 128, 389400. (17) Kono, M.; Hayashi, N.; Samata, T. Biochem. Biophys. Res. Commun. 2000, 269, 213-218. (18) Samata, T.; Hayashi, N.; Kono, M.; Hasegawa, K.; Horita, C.; Akera, S. FEBS Lett. 1999, 462, 225-229. (19) Weiss, I. M.; Gohring, W.; Fritz, M.; Mann, K. Biochem. Biophys. Res. Commun. 2001, 285, 244-249. (20) Mann, K.; Weiss, I. M.; Andre, S.; Gabius, H. J.; Fritz, M. Eur. J. Biochem. 2000, 267, 5257-5264. (21) Weiss, I. M.; Kaufmann, S.; Mann, K.; Fritz, M. Biochem. Biophys. Res. Commun. 2000, 267, 17-21. (22) Michenfelder, M.; Fu, G.; Lawrence, C.; Weaver, J. C.; Wustman, B. A.; Taranto, L.; Evans, J. S.; Morse, D. E. Biopolymers 2003, 70, 522-533. (23) Zhang, Y.; Xie, L. P.; Meng; Q. X., Jiang; T. M.; Pu, R. L.; Chen, L.; Zhang, R. Q. Comp. Biochem. Phys. A 2003, 135, 565-573. (24) Blank, S.; Arnoldi, M.; Khoshnavaz, S.; Treccani, L.; Kuntz, M.; Mann, K.; Grathwohl, G.; Fritz, M. J. Microsc. (Oxford) 2003, 212, 280-291. (25) Arias, J. L.; Fink, D. J.; Xiao, S. Q.; Heuer, A. H.; Caplan, A. I. Int. ReV. Cytol. 1993, 145, 217-250. (26) Simkiss, K. Biol. ReV. Cambridge Philos. Soc. 1961, 36, 321-367. (27) Hincke, M. T.; Gautron, J.; Tsang, C. P. W.; McKee, M. D.; Nys, Y. J. Biol. Chem. 1999, 274, 32915-32923. (28) Gautron, J.; Hincke, M. T.; Mann, K.; Panhe´leux, M.; Bain, M.; McKee, M. D.; Solomon, S. E.; Nys, Y. J. Biol. Chem. 2001, 276, 39243-39252. (29) Mann, K.; Siedler, F. Biochem. Mol. Biol. Int. 1999, 47, 997-1007. (30) Hincke, M. T.; Gautron, J.; Panhe´leux, M.; Garcia-Ruiz, J. M.; McKee, M. D.; Nys, Y. Matrix Biol. 2000, 19, 443-453. (31) Carrino, D. A.; Rodriguez, J. P.; Caplan, A. I. Connect. Tissue Res. 1997, 36, 175-193. (32) Fernandez, M. S.; Passalacqua, K.; Arias, J. I.; Arias, J. L. J. Struct. Biol. 2004, 148, 1-10. (33) Arias, J. L.; Neira-Carrillo, A.; Arias, J. I.; Escobar, C.; Bodero, M.; David, M.; Fernandez, M. S. J. Mater. Chem. 2004, 14, 2154-2160.

Biomacromolecules, Vol. 6, No. 3, 2005 1437 (34) Lakshminarayanan, R.; Valiyaveettil, S.; Rao, V. S.; Kini, R. M. J. Biol. Chem. 2003, 278, 2928-2936. (35) Lakshminarayanan, R.; Kini, R. M.; Valiyaveettil, S. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 5155-5159. (36) Packard, M. J.; Packard, G. C. J. Morphol. 1979, 159, 131-144 (37) Gauldie, R. W.; Sharma, S. K.; Volk, E. Comp. Biochem. Phys. A 1997, 118, 753-757. (38) Kontoyannis, C. G.; Vagenos, N. V. Analyst 2000, 125, 251-255. (39) Tetin, S. Y.; Prendergast, F. G.; Venyaminov, S. Y. Anal. Biochem. 2003, 321, 183-187. (40) Dufton, M. J.; Hider, R. C. CRC Crit. ReV. Biochem. 1983, 14, 113171. (41) Kelly, J. W. Curr. Opin. Struct. Biol. 1998, 8, 101-106. (42) Terzi, E.; Gunter, H.; Seelig, J. Biochemistry 1994, 33, 1345-1350. (43) Aggeli, A.; Bell, M.; Keen, J. N.; McLeish, T. C. B.; Nyrkova, I.; Radford, S. E.; Semenov, A. J. Mater. Chem. 1997, 7, 1135-1145. (44) Stefani, M.; Dobson, C. M. J. Mol. Med. 2003, 81, 678-699. (45) Zhang, S. G.; Rich, A. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 2328. (46) de la Paz, M. L.; Goldie, K.; Zurdo, J.; Lacroix, E.; Dobson, C. M.; Hoenger, A.; Serano, L. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 16052-16057. (47) Gopalan, V.; Golbik, R.; Schreiber, G.; Fersht, A. R.; Altman, S. J. Mol. Biol. 1997, 267, 765-769. (48) Ruano, M. L. F.; Garcia-Verdugo, I.; Miguel, E.; Perez-Gil, J.; Casals, C. Biochemistry 2000, 39, 6259-6537. (49) Lakowicz, J. R.; Keating-Nakamoto, S. Biochemistry 1984, 23, 30133021. (50) Macgregor, R. B.; Weber, C. Ann. N.Y. Acad. Sci. 1981, 336, 140154 (51) Demchenko, A. P. Eur. Biophys. J. 1988, 16, 121-129. (52) Demchenko, A. P.; Ladokhin, A. S. Eur. Biophys. J. 1988, 15, 369379. (53) Galley, W. C.; Purkey, R. M. Proc. Natl. Acad. Sci. U.S.A. 1970, 67, 1116-1121. (54) Demchenko, A. P. Essays Biochem. 1986, 22, 120-157. (55) Lafont, S.; Veesler, S.; Astier, J. P.; Boistelle, R. J. Cryst. Growth 1994, 143, 249-255. (56) Mann, S.; Heywood, B. R.; Rajam, S.; Walker, J. B. A. ACS Symp. Ser. 1991, 444, 28-41. (57) Naka, K.; Tanaka, Y.; Chujo, Y. Langmuir 2002, 18, 3655-3658. (58) Davey, R. J.; Cardew, P. T.; McEwan, D.; Sadler, D. E. J. Cryst. Growth 1986, 79, 648-653. (59) Nassrallah-Aboukais, N.; Boughriet, A.; Laureyns, J.; Aboukais, A.; Fischer, J. C.; Langelin, H. R.; Wartel, M. Chem. Mater. 1998, 10, 238-243. (60) Wilbur, K. M.; Watabe, N. Ann. N.Y. Acad. Sci. 1963, 109, 82112. (61) Lowenstam, H. A.; Abbott, D. P. Science 1975, 188, 363-365.

BM049276F

Purification and Characterization of a Vaterite-Inducing Peptide

Recommend Documents