Soluble silk-like organic matrix in the nacreous layer of the bivalve Pinctada maxima A new insight in the biomineralization field Lucilia Pereira-Mourie ` s 1 , Maria-Jose ´ Almeida 1,3 , Cristina Ribeiro 3 , Jean Peduzzi 2 , Michel Barthe ´ lemy 2 , Christian Milet 1 and Evelyne Lopez 1 1 Laboratoire de Physiologie Ge ´ ne ´ rale et Compare ´ e, UMR CNRS 8572, Muse ´ um National d’Histoire Naturelle, Paris, France; 2 Laboratoire de Chimie des Substances Naturelles, ESA CNRS 8041, Muse ´ um National d’Histoire Naturelle, Paris, France; 3 INEB – Instituto de Engenharia Biome ´ dica, Rua do Campo Alegre, Porto, Portugal Nacre organic matrix has been conventionally classified as both Ôwater-solubleÕ and Ôwater-insolubleÕ,basedonits solubility in aqueous solutions after decalcification with acid or EDTA. Some characteristics (aspartic acid-rich, silk- fibroin-like content) were specifically attributed to either one or the other. The comparative study on the technique of extraction (extraction with water alone vs. demineralization with EDTA) presented here, seems to reveal that this gen- erally accepted classification may need to be reconsidered. Actually, the nondecalcified soluble organic matrix, extrac- ted in ultra-pure water, displays many of the characteristics of what until now has been called Ôinsoluble matrixÕ.We present the results obtained on this extract and on a conventional EDTA-soluble matrix, with various charac- terization methods: fractionation by size-exclusion and anion-exchange HPLC, amino acid analysis, glycosami- noglycan and calcium quantification, SDS/PAGE and FTIR spectroscopy. We propose that the model for the interlamellar matrix sheets of nacre given by Nakahara [In: Biomineralization and Biological Metal Accumulation, Westbroek, P. & deJong, E.W., eds, (1983) pp. 225–230. Reidel, Dordrecht, Holland] and Weiner and Traub [Phil.Trans.R.Soc.Lond.B(1984) 304, 425–434] may no longer be valid. The most recent model, proposed by Levi-Kalisman et al.[J. Struct. Biol. (2001) 135, 8–17], seemed to be more in accordance with our findings. Keywords: nacre; undecalcified soluble matrix; EDTA- soluble matrix; hydrophobicity; silk-fibroin-like-proteins. In the biomineralization field, the mollusk shell is one of the best studied of all calcium carbonate biominerals. Particular attention has been given to the organic matrix [1–5]. The latter is thought to promote the nucleation of the mineral component, to direct the crystal growth and to act as glue, preventing fracture of the shell [6–9]. The main biopolymers present in the organic matrix are essentially proteins, either glycosylated or not, acidic polysaccharides and chitin. In nacre, they represent 1–5% (w/w) of the structure. From the earliest experiments, it was believed that the biochemical properties of matrix constituents depend of the use of a decalcification procedure for removing the mineral component, which is strongly associated with the organic matrix [1,3]. Therefore, all investigations up until now used either EDTA, acetic acid or hydrochloric acid for this demineralization step and, subsequently, two fractions of the organic matrix were separated, based on their solubility in aqueous solutions. Accordingly, a designation of matrix into two classes, the soluble matrix and the insoluble matrix, has evolved from this extraction [10–14]. This paper presents for the first time the results of a comparative study on the organic matrix extracted from the nacreous layer of the shell from the pearl oyster Pinctada maxima by two very different methods. The first is a nondecalcifying technique obtained by an extraction in ultra-pure water. This unconventional approach arises from previous in vivo and in vitro experiments where we showed that biochemical signals from nacre chips were able to diffuse in the surrounding media and to induce new bone formation [15–22]. In an attempt to identify these signal molecules, we have previously perfected this original method of extraction of the organic matrix, without any acid treatment or demineralization, in order to minimize any possible alteration of the activity of the macromole- cules [20,23]. The second method is one of the widely used extraction techniques which involves a demineralization with EDTA followed by intensive dialysis against distilled water. The content of the respective soluble matrix extracts were very different and seemed to raise important questions about the actual conventional classification of the soluble (known as acidic and aspartic acid-rich) and insoluble (said to be hydrophobic and glycine, alanine-rich) matrices and on the current model of nacre organic matrix organization. Correspondence to E. Lopez, Laboratoire de Physiologie Ge ´ ne ´ rale et Compare ´ e, UMR CNRS 8572, Muse ´ um National d’Histoire Naturelle, 7 rue Cuvier, 75231, Paris Cedex 05, France. Fax: +33 1 40795620, Tel: +33 1 40793622, E-mail: lopez@mnhn.fr Abbreviations:EDTA-IM,EDTA-insolublematrix;EDTA-SM, EDTA-soluble matrix; GAG, glycosaminoglycan; PG, proteoglycan; WIM, water-insoluble matrix; WSM, water-soluble matrix. (Received 22 April 2002, revised 16 August 2002, accepted 23 August 2002) Eur. J. Biochem. 269, 4994–5003 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03203.x MATERIALS AND METHODS Organic matrix extraction The powdered nacre (particle size 50–150 lm), obtained from the inner shell layer of the pearl oyster P. maxima,was extracted by either ultra-pure water or EDTA and then fractionated into soluble and insoluble matrix by centrifu- gation. Demineralization of powdered nacre with EDTA was performed as described by Wheeler et al.[24].Fifty grams of powdered nacre was dissolved in 100 mL 10% EDTA disodium salt dihydrate, pH 8, with continuous stirring for 24 h at room temperature. Then, the suspension was transferred in a dialysis bag (Spectrapor 2, 12–14 kDa molecular weight cut-off) and placed in 2 L of the same EDTA solution (replaced by fresh solution every 12 h), with stirring and at 4 °C, until the powdered shell was completely demineralized (about 3 days). The extract was centrifuged at 27 000 g for 30 min to separate the EDTA-soluble matrix (EDTA-SM) from the EDTA-insoluble matrix (EDTA- IM).ToremoveEDTA,thetwosampleswereextensively dialyzed against ultra-pure water (Milli-Q TM ), at 4 °C, and the EDTA-SM was freeze-dried. The water-soluble matrix (WSM) was obtained as described in Almeida et al.[20]. Fractionation of soluble extracts by HPLC The WSM and the EDTA-SM were separated by size- exclusion high performance liquid chromatography (SE-HPLC), and by anion-exchange HPLC (AE-HPLC), as described elsewhere [20]. For the SE-HPLC, a solution of 30 mg (dry weight)ÆmL )1 was prepared for the two soluble extracts and filtered through a 0.22 lm centrifuge tube filter (Spin- XÒ, Costar) before injection. Aliquots of WSM (500 lL) or EDTA-SM (250 lL) were injected onto the preparative column. Proteins were eluted with ultra-pure water rather than with a salt buffer in order to preserve the integrity of the macromolecules [20,21]. Amino acid analysis Samples of proteins were hydrolyzed at 110 °C under vacuum with 6 M HCl constant boiling (Sigma) for 24 h. Phosphoserine was determined from hydrolysis in 4 M HCl for 6 h [25]. The resulting amino acids were separated on a cation exchange PC6A resin (Pierce) and the o-phthaldial- dehyde derivatives of amino acids were detected with a Waters 420 fluorimeter. Proline, hydroxyproline and hydroxylysine were examined at 254 nm by reverse-phase HPLC of their phenylisothiocarbamate derivatives [26], as reported previously [20]. The serine, threonine and tyrosine contents of the hydrolysates were corrected for destruction during the hydrolysis by extrapolation to zero time hydro- lysis. The amino acid compositions, expressed as a mole percent, represent the average of at least three independent determinations. The amount of protein in each extract was calculated from the amino acids’ molar yields. Glycosaminoglycan analysis Organic matrix samples were dissolved in 5 mL of 0.1 M NaOH [27] at room temperature for 24 h with periodic mixing and maceration, followed by centrifugation at 1000 g for 10 min. Sulfated and nonsulfated glycosamino- glycans (GAGs) from the supernatant were estimated by the Whiteman Alcian blue binding technique [28,29], using chondroitin sulfate as standard. The assay was adapted to the estimation of GAG in more dilute samples by increasing the aliquot size, as proposed by Gold [30]. Calcium analysis Calcium analyses were performed after nitric acid hydrolysis of samples, by atomic absorption spectrophotometry, using a GBC 904AA spectrophotometer. Fourier transform infrared (FTIR) spectroscopy Organic compounds binding groups from soluble matrix samples were detected by FTIR spectrometry. The FTIR spectra were obtained using a Perkin Elmer 200FTIR spectrometer. All the samples were prepared as KBr discs and were run at a spectral resolution of 4 cm )1 .One hundred scans were acquired for each sample. Polyacrylamide gel electrophoresis The proteins from the soluble samples were separated under denaturating conditions (30 lg total protein per well) by SDS/PAGE [31] using 12% polyacrylamide mini-gels (Mini Protean III apparatus, Bio-Rad) of 0.75 mm thickness. After electrophoresis, proteins were visualized by silver- staining, as described by Morrissey [32], without the glutaraldehyde step. The molecular masses were estimated using the Amersham Pharmacia Biotech LMW-SDS marker kit for electrophoresis. RESULTS Extraction of soluble matrix After extraction and lyophilization, the two soluble extracts (WSM and EDTA-SM) were re-suspended in ultra-pure water for protein analysis. The protein recovery was similar in the two cases, with about 0.05% by weight of powdered nacre. Fractionation of soluble extracts by HPLC The size-exclusion chromatographic profiles for WSM and EDTA-SM were very different (Fig. 1). As described in Almeida et al. [20], the soluble matrix recovered by an aqueous extraction (WSM) was separated in four distinct fractions (Fig. 1A). On the contrary, under the same conditions, the soluble matrix obtained after nacre demin- eralization by EDTA (EDTA-SM) consisted of only two different fractions (Fig. 1B). Absorbance values were very high for EDTA-SM, in comparison with those obtained for WSM, whereas the volume of EDTA-SM injected was half the WSM volume. The low molecular weight peak due to the use of EDTA, often mentioned in the literature [24,33], was not observed in the size-exclusion profiles of the EDTA-SM extract (Fig. 1B). The fractionation by anion-exchange HPLC was also different for WSM and EDTA-SM. Separation was better Ó FEBS 2002 Soluble silk-like organic matrix in nacre (Eur. J. Biochem. 269) 4995 resolved for EDTA-SM than for WSM (Fig. 2). The main peak from each separation, indicated by an asterisk in Fig. 2, was collected and submitted to amino acid analysis. Amino acid compositions In accordance with previous studies on nacre EDTA-SM on other mollusk shells [34,35], this extract was aspartate-rich (nearly 40 mole percent) (Table 1) and exhibited a charge to hydrophobic ratio (C/HP; Asx, Glx, His, Arg, Lys/Ala, Pro, Val, Met, Ile, Leu, Phe) of 2.86. The main amino acids found in EDTA-SM were aspartate and glycine (69.2% of total amino acids). Previous studies with regard to soluble organic matrix of mollusk shells indicated that more than 80% of the aspartate and glutamate is in the form of aspartic and glutamic acid, respectively [34]. In order to compare our results with published data [4,33,36], we also determined the global amino acid composition of the EDTA-IM. Here again, the composition was as expected, i.e. very hydrophobic (C/HP ¼ 0.42) and glycine, alanine- rich. Unexpectedly, the soluble matrix obtained by aqueous extraction, WSM, likewise exhibited more than 60% of glycine and alanine residues and a large proportion of hydrophobic amino acids, resulting in a very low C/HP value (0.29). Its content in Asx was moderate. These features are exactly the opposite of that found in EDTA-SM which exhibits a high Asx content (39.6%) and a low content for alanine (6.5%). Thus, the alanine and glycine content of WSM was rather similar to that of EDTA-IM. In spite of the presence of mineral, it was possible to analyze the amino acid composition of the water insoluble matrix (WIM). Again we found a high content in Gly-Ala, and the global composition was very similar to the WSM, and consequently the EDTA-IM. Hydroxyproline, hydroxy- lysine and phosphoserine were not detected in WSM, WIM, EDTA-SM and EDTA-IM. The comparison of the amino acid composition of the main peak obtained from AE-HPLC of WSM and EDTA- SM is given in Table 2. These two peaks were characterized by a high content in glycine-alanine (31–30.6%) and serine. As expected from an anion-exchange separation, the two peaks contained acidic proteins. Nevertheless, they were aspartic-rich in EDTA-SM whilst being quite glutamic acid- rich in WSM. Glycosaminoglycan analysis Glycosaminoglycans are highly negatively charged because of the presence of sulfate ester and/or carboxyl groups. Therefore, they interact with cations and are precipitated by cationic dyes like Alcian blue [37]. GAGs were found in WSM and EDTA-SM (Table 3). Nevertheless, their amount in EDTA-SM was about 15 times as much as that of WSM, suggesting that they are firmly tightened to the mineral or the aspartic acid-rich Fig. 2. Anion exchange-HPLC elution profiles of the water-soluble matrix (A) and the EDTA-soluble matrix (B) of Pinctada maxima nacre. Samples (55 lL) containing 400 lg (WSM) or 200 lg(EDTA-SM) protein mixture in 20 m M Tris/HCl pH 7.8 were loaded on a Mono Q HR 5/5 column equilibrated with the same buffer. Proteins were eluted at a flow rate of 1 mLÆmin )1 with a 25-min linear gradient from 0 to 100% solvent (500 m M NaCl, 20 m M Tris/HClbuffer,pH7.8). Absorbance was monitored at 226 nm. The main peak from each separation is indicated by an asterisk. Fig. 1. Size exclusion-HPLC elution profiles of the water-soluble matrix (A) and the EDTA-soluble matrix (B) of Pinctada maxima nacre. Samples of protein in ultra-pure water (500 lL and 250 lL, respect- ively), were injected onto the preparative column (TSKGel G 3000 SW, 600 · 21.5 mm) and proteins were eluted with ultra-pure water at 2.5 mLÆmin )1 flow rate. Absorbance was monitored at 280 nm. The column was calibrated with alcohol dehydrogenase (150 kDa), bovine serum albumin (66 kDa) and lysozyme (17 kDa). 4996 L. Pereira-Mourie ` s et al. (Eur. J. Biochem. 269) Ó FEBS 2002 matrix. Very few GAGs were found in the pellet (EDTA- IM) obtained after demineralization. Calcium analysis In the EDTA-SM, the amount of nondialyzable calcium after decalcification was 51.2 lgÆmg )1 dry weight (Table 3), and one may suppose that some EDTA remained in the soluble extract, associated with calcium, as suggested by Wheeler et al. [24,35]. The presence of residual EDTA after intensive dialysis of EDTA-SM was also observed in amino acid composition determination as EDTA eluted near to the histidine, precluding from detecting this amino acid. The calcium content of the EDTA-IM was, as expected, much higher; about 310 lgÆmg )1 dry weight. This high value may in part be related to the incomplete demineralization of powdered nacre, during extraction. In contrast, WSM contained only 1.1 lg calcium per mg dry weight, confirm- ing that very small amounts of calcium were dissolved in water and/or linked to the matrix components. Fourier transform infrared (FTIR) spectroscopy Infrared spectroscopy provides means for a characterization at the molecular level of the structure and bonding of surface functional groups and adsorbed species. In this study, we used infrared spectroscopy to identify possible differences in composition between the decalcified nacre soluble matrix (EDTA-SM) and the aqueous, nondecalci- fied, nacre organic matrix (WSM). The FTIR spectrum of the EDTA-SM (Fig. 3A) was characterized by two intense bands, one at 3432 cm )1 (OH and/or NH stretching modes of the organic matrix compo- nents) and another, the most intense, at 1593 cm )1 , possibly corresponding to the COO coordinated asymmetric stretch- ing band. The presence of this band resulted from the EDTA, a potent metal chelator that was used to extract the soluble matrix from the crystalline structure. The EDTA molecule has six potential sites for bonding with a metal ion: four carboxyl groups and the two amino groups. When EDTA is dissolved in water, it behaves like an amino acid such as glycine. From the infrared spectrum of a metal chelated compound of EDTA, it is possible to distinguish the coordinate and free COO stretching band. The union- ized and uncoordinated COO stretching band occurs at Table 2. Amino acid compositions (mole percent) of the main peak from anion-exchange (AE) HPLC of water-soluble matrix (WSM) and EDTA-soluble matrix (EDTA-SM). Cysteine, hydroxylysine, hydroxy- proline, phosphoserine, proline and tryptophan were not determined. Amino acid AE-WSM AE-EDTA-SM Asx 9.7 21.0 Thr 3.3 3.7 Ser 27.6 18.3 Glx 13.4 9.5 Gly 22.4 22.9 Ala 8.6 7.7 Val 2.6 2.7 Met 0 0.5 Ile 1.7 2.1 Leu 3.9 2.9 Tyr 1.5 1.5 Phe 1.9 1.4 His 1.7 2.3 Lys 1.3 2.6 Arg 0.8 0.9 C/HP a 1.44 2.10 a Ratio charged to hydrophobic residues (see Results, section Amino acid composition, for details). Table 1. Amino acid compositions (mole percent) of the water-soluble matrix, water-insoluble matrix, the EDTA-soluble matrix and the EDTA- insoluble matrix of Pinctada maxima nacre. Results are expressed as a mole percent and represent the mean of at least three independent determinations. Cysteine and tryptophan were not determined. Hydroxyproline, hydroxylysine and phosphoserine were not detected in all samples. Amino acid Water-soluble matrix Water-insoluble matrix EDTA-soluble matrix EDTA-insoluble matrix Asx 7.3 10.8 39.6 10.8 Thr 0.9 1.1 1.4 1.1 Ser 3.9 4.0 3.6 4.4 Glx 2.2 2.2 4.3 2.2 Gly 37.6 28.7 29.6 31.5 Ala 30.0 31.8 6.5 30.5 Pro 1.5 1.1 1.2 1.3 Val 2.2 2.1 1.4 1.9 Met 1.2 1.1 1.3 1.0 Ile 1.2 1.3 1.5 1.2 Leu 6.7 6.3 3.4 5.9 Tyr 0.8 1.9 1 1.7 Phe 1.3 1.7 1.4 1.4 His 0.1 0.2 ND a 0 Lys 1.2 1.7 1.7 1.2 Arg 1.9 4.1 2.1 3.9 C/HP b 0.29 0.42 2.86 0.42 a ND, not determined. After intensive dialysis the sample still contained residual EDTA. b Ratio charged to hydrophobic residues (see Results, section Amino acid composition, for details). Ó FEBS 2002 Soluble silk-like organic matrix in nacre (Eur. J. Biochem. 269) 4997 1750–1700 cm )1 whereas the ionized and coordinated COO stretching band appears at 1650–1590 cm )1 [38]. The latter frequency depends on the nature of the metal. Although the EDTA-SM was subjected to intensive dialysis against ultra- pure water to remove EDTA, the fractions isolated from the oyster matrix seem to be, for the most part, EDTA protein complexes. This finding recommends that caution must be taken in interpreting binding data of EDTA-extracted matrix. Another strong band appeared at 1409 cm )1 and corresponds to the COO symmetric stretching band. We could also identify the presence of sulfate groups absorbing at 1284, 1260, 927 and 855 cm )1 . Absorption bands that can be attributed to the amide vibrations, namely 1328 cm )1 (C–N stretching vibration, amide III), 808 cm )1 (amide VII), 708 cm )1 (amide V or VII), 639 and 621 cm )1 (amide IV), were also observed. Several bands were present in the 1000–1150 cm )1 zone, which is the major polysaccharide absorption region. A band at 1180 cm )1 was probably due to in-plane NH 2 rocking. It is also possible that the small bands located at 963, 985 and 1032 cm )1 in the EDTA-SM and the bands 997, 1032 and 1047 cm )1 in the WSM correspond to PO 4 3– vibrations. Phosphate as well as sulfate groups are potential calcium-binding moieties and are reported to be present in mollusk shell soluble fractions [1,39]. The FTIR spectrum of the WSM (Fig. 3B) was very different from that of the EDTA-SM, although some of the bands are common to both samples. These corres- pond to the band at 3431 cm )1 (OH and/or NH stretching modes of the organic matrix components) and those in the 2800–3000 cm )1 region (C–H stretching modes). In the WSM spectrum, a strong band at 1656 cm )1 , also assigned as a shoulder in the EDTA- IM and powdered nacre spectra (data not shown), was present and corresponds to the amide I groups (C¼O stretching vibration in the associated state). The absorp- tion occurred in the high-frequency wing of the amide II band and was sometimes partly merged with it [40]. A band at 1542 cm )1 was also visible and is characteristic of the amide II groups. In the WSM spectrum, other bands were clearly assigned, namely at 1455 cm )1 (CH 2 scissor- ing) and 1384 cm )1 (C–N stretching vibration, amide III). Most of the absorption of the later band came within the region 1390 ± 40 cm )1 , in which the methyl or methy- lene deformations are also active. The WSM sample absorbed less in the region between 1000 and 1150 cm )1 compared to the EDTA-SM sample and therefore appeared to contain a smaller proportion of polysaccha- rides in its composition. That confirms the GAGs analysis results. Polyacrylamide gel electrophoresis Proteins from shell soluble matrices are generally not easy to visualize after SDS/PAGE separation [41]. In the present study, most of the proteins from both EDTA-SM and WSMmigratedinthegelwithnodistinctpattern,leavinga dark continuous smear after silver staining (Fig. 4). No discrete individual bands were observed in the WSM sample. However, the EDTA-SM revealed two distinct proteins around 14 and 20 kDa, still presenting with the dark smear background. Table 3. Glycosaminoglycan analysis and calcium measurements of the water-soluble matrix, the EDTA-soluble matrix and the EDTA-insoluble matrix of Pinctada maxima nacre. Sulfated and nonsulfated glycosaminoglycans from the supernatant were estimated by the Whiteman Alcian blue binding technique [28,29], using chondroitin sulfate as standard. Calcium analyses were performed on samples digested with nitric acid, by atomic absorption spectrophotometry. Results are expressed as lgÆmg )1 organic matrix dry weight (mean value ± standard deviation of three determinations). Water-soluble matrix EDTA-soluble matrix EDTA-insoluble matrix Glycosaminoglycans 1.59 ± 0.41 24.38 ± 1.10 0.18 ± 0.02 Calcium 1.1 ± 0.5 51.2 ± 1.7 312.1 ± 155.7 Fig. 3. FTIR spectra of the EDTA-soluble matrix (A) and the water-soluble matrix (B) of Pinctada maxima nacre. Samples were pre- pared as KBr discs and were run at a spectral resolution of 4 cm )1 . 4998 L. Pereira-Mourie ` s et al. (Eur. J. Biochem. 269) Ó FEBS 2002 DISCUSSION For several years, mollusk shell biomineralization has been studied with demineralized structures, in order to obtain the organic material. In this way, the matrix molecules have been classified conventionally into two types based on their solubility in aqueous solutions after demineralization, the insoluble matrix being characterized by the presence of highly hydrophobic molecules and by a rich content in glycine and alanine residues. It is thought to be largely intercrystalline [42] and to provide a framework where mineralization occurs. The soluble matrix is characterized by the predominance of acidic glycoproteins. It is known as intracrystalline and is considered to play an important role in induction of oriented nucleation, inhibition of crystal growth and control of aragonite-calcite polymorphism [43]. At present, only a few constituents of these organic matrices have been identified. One of the reasons for this is that shell proteins are very difficult to isolate by means of traditional biochemical methods (chromatography, electro- phoresis, enzymatic cleavage) due to self-aggregation of the molecules or an unusual resistance to temperature, chem- icals and enzymes. Essentially, the recent advances in isolation and characterization of matrix molecules have been possible due to the genetic approach and cDNA cloning. In a recent paper, Marin et al.[44]describeda combined technique of preparative electrophoresis and Western blot on individual proteins which enables the purification of different proteins in relative large amounts. In this work, we compare the nacre organic matrix obtained by the traditional demineralizing extraction method, to an original method of studying matrix mole- cules, without previous decalcification. We showed that it is possible to extract and study organic compounds of the biomineral nacre, by bypassing the demineralization step. We think that it may present a new perception of how the different fractions of the organic matrix are organized in the biomineral structure. First, the aqueous method affected neither the yield of organic material extraction in general nor the extraction of proteins. Also, the WSM has a low calcium content, confirming that the molecules extracted are not associated with minute particles of CaCO 3 that had not been removed by centrifugation. What changed significantly was the content of the organic material and, presumably, its original location in the biomineral itself. In fact, the FTIR spectra, the amino acid compositions, the chromatographic and electrophoretic fractionations of EDTA-SM and WSM showed a first sign of this dissimilarity. For the FTIR spectra, the main differences were as follows: first, the presence of sulfate groups and several bands corresponding to polysaccharides in the EDTA-SM, which may be assigned to the high content of GAGs, observed in the quantification assay. Second, the organic matrix extracted solely in water, i.e. WSM, exhibited fewer polysaccharide bands, but showed strong protein peaks (amide I and amide II bands) that were previously observed in the insoluble matrix from decalcified nacre [45] or shell material [46]. The position of the amide I band, the major protein absorption band, depends on the conformation of the polypeptide chain [47]. The presence of this amide I absorption band at 1656 cm )1 suggests that the proteins in WSM are in the a-helix or randomly coiled form, two conformations not distinguishable by IR spectrometry [48]. Until very recently, it was thought that the aspartic acid-rich proteins from the decalcified soluble matrix were in part in the b-sheet conformation [6,43], as well as the decalcified, silk-fibroin- like insoluble matrices [49]. Recently Levi-Kalisman et al. [5] modified these assumptions in suggesting that the ÔsilkÕ would coat the chitin core in a homogeneous and completely disorganized phase. Does the random coiled form of WSM correspond to that disorganized phase? If so, the WSM would be related to the ÔsilkÕ matrix and not to the soluble aspartic acid-rich matrix. The electrophoretic pattern of WSM did not give significant information on its protein composition. The continuous smear suggests the presence of GAGs or other sugars bound to the proteins, despite a low GAG content in this extract. Western blot analysis with a WSM fraction- specific antibody revealed the complexity of this kind of matrix, with several closely separated bands [50]. The EDTA-SM showed a 14-kDa band, probably the N14 protein found by Kono et al. [14], and another one around 20 kDa. Attempts to purify and characterize these proteins are presently in progress. Above all these distinctions, the global amino acid composition showed clearly that the proteins extracted by the two methods are not the same. On the one hand, the soluble (aspartic acid-rich proteins) and insoluble (glycine, alanine-rich, hydrophobic proteins) matrices extracted after demineralization with EDTA are in accordance with similar results in corresponding literature [7,51,52]. On the other hand, the amino acid composition of WSM, obtained by an aqueous extraction, was completely different to that of the EDTA-SM and the so-called soluble matrices in general. This extract was highly hydrophobic with a C/HP value of 0.29 and exhibited more than 65% glycine and alanine residues. To begin with, such a characteristic is strange for a Fig. 4. SDS/PAGE of Pinctada maxima nacre soluble matrices. (1) LMW calibration kit; (2) EDTA-SM (30 lg protein); (3) WSM (50 lg protein). Samples in Laemmli buffer were loaded on a 12% poly- acrylamide mini-gel 0.75-mm thick, and silver stained. Ó FEBS 2002 Soluble silk-like organic matrix in nacre (Eur. J. Biochem. 269) 4999 soluble extract. Again, these are predominantly features of what has been called up until now Ôinsoluble matrixÕ,andof the known silk-fibroin-like molecules. When we looked to the WIM, we found that it was similar in amino acid composition to the WSM and the EDTA-IM extractions. This result means that the silk-like matrix is not completely extracted with WSM, being present in the two phases. The acidic matrix would be still enclosed in the mineral phase of the WIM and would not be accessible for quantification. However, when we placed the extracted particles (the pellet) in pure water once again, no more matrix was released (data not shown). This confirms the hypothesis that the silk is indeed present in two states, soluble and insoluble, the latter being in some way protected from dissolution in water. The peculiarity of the WSM led to the question: how can such a hydrophobic pool of molecules be dissolved in pure water? Recently we hypothesized [50] that sugars might be associated with the apolar proteins of WSM and would be responsible for their solubilization in water. Here we showed that GAGs are present in WSM, but in minor amounts compared with the EDTA-SM. GAGs and, more specifically, proteoglycans (PGs) are supposed to be present in mollusk shell organic matrix. Their presence was indicated by IR spectrometry and Alcian blue staining [48,53], though without direct quantification. Our results confirmed their presence in P. maxima nacre organic matrix and showed that GAGs are mainly released with EDTA and less with water. This is in accordance with the observations of Golberg and Takagi [54] who have also detected a loss of PGs in dentine during EDTA deminer- alization. They concluded that it is necessary to study PGs distribution on material fixed either with cryotechniques, where PGs appear as an amorphous substance in dental organic matrix, or with cationic dyes. The role of GAGs is not really understood, but their presence in the organic matrices of bones, teeth, avian eggshell, otoconia and kidney stones shows their importance in biomineralizing systems. Acidic mucopolysaccharides have been considered as possible candidates for the initiation of the crystal formation [55]. Sulfate, as well as carboxylate groups, may cooperate in the induction of oriented crystal nucleation [56]. These molecules may be responsible for fixation of calcium in the shell [57]. Also, PGs which are GAGs associated to a core protein, may act in cell signaling and metabolic activity [58,59]. To proceed with our argument on the nacre organic matrix organization, we may understand why the WSM obtained by an aqueous extraction does not contain acidic proteins, like the other soluble matrices obtained after demineralization. The acidic proteins are thought to be firmly linked to the mineral, and without decalcification it seems difficult to dissociate them from the whole structure. Thus, in WSM, we theoretically extracted the molecules that were directly accessible to water around the sides of the small mineral particles. In the five-layered model of nacre organic matrix organization [7,60] the surface layers, also called the ÔenvelopeÕ by the authors, are the aspartic acid- rich proteins. However, the surface molecules extracted in WSM correspond to the core of the organic matrix layer of the previous model, the silk-fibroin-like molecules. Thus, we may think of a different structure for nacre organic matrix, where the silk (maybe WSM or at least part of it) would not be so deeply located. From the cryo-transmission electron microscopy studies of the matrix of the bivalve Atrina,anew model for the nacreous layer organic matrix was recently proposed by Levi-Kalisman et al. [5]. In this model, which would be in accordance with our findings, a hydrated silk- gel phase would be located between the mainly composed b-chitin interlamellar sheets (Fig. 5). The aspartic acid-rich glycoproteins would be present both within the silk gel prior to mineralization and also as electron-dense patches at the surface of b-chitin. The presence of chitin combined with the silk-fibroin-like molecules was unfortunately not investi- gated in P. maxima WSM. Still, its presence was demon- strated in P. maxima as a chitin–protein complex, which precipitated after demineralization of the nacreous layer with hydrochloric acid [61]. The characterization of the proteins associated to this complex by amino acid analysis of the precipitate after two deacetylation steps revealed the predominance of two amino acids: alanine (39.2%) and glycine (30%). Interestingly, the proportions are very similar to those found in WSM for alanine and glycine. Curiously, the presence of a gel phase in shells was rarely mentioned in the literature until now, although the forma- tion of a jelly like substance within the shells of cultured Crassostrea gigas is frequently reported by farmers. This phenomenon is usually accompanied by a thickening of the shell, with spaces filled by a nonmineralizing organic matrix (a jelly-like substance) and is due to exogenous factors such as exposure to the tributyltin contained in anti-fouling paints. The few data available on the composition of this kind of gel showed that it can not be compared to a silk- fibroin-like substance [62–66]. Though this jelly-like sub- stance is produced in abnormal situations, it shows that bivalve organisms do possess the biochemical machinery needed to produce gel in the calcification process. In our opinion, the importance of the silk-fibroin-like matrix has been neglected until now, in part because of its inaccessibility. The first insoluble molecules, after decalcifi- cation, to be purified from mollusk shell were MSI 60 and MSI 31, both from Pinctada fucata nacreous and prismatic Fig. 5. Schematic representation of the new model for the nacreous layer organic matrix structure, as proposed by Levi-Kalisman et al.[5].The putative silk gel phase is located between the interlamellar sheets of b-chitin. See details in Discussion (courtesy of Professor Steve Weiner and coworkers). Reprinted from Journal of Structural Biology 135, Levi-Kalisman et al., Structure of the nacreous organic matrix, pp 8–17, 2001, with permission from Elsevier Science. 5000 L. Pereira-Mourie ` s et al. (Eur. J. Biochem. 269) Ó FEBS 2002 layers, respectively [67]. At about the same time, Shen et al. [68] isolated lustrin A, a modular and multifunctional protein from Haliotis rufescens nacre. Lastly, N16 [69] and its homologous soluble protein N14 [14] have been charac- terized in the nacreous layer of P. fucata and P. maxima, respectively, and would constitute a new protein family. Many more of the studies on the organic matrix of biominerals were focused on the attempt to characterize and purify the aspartic acid-rich molecules [70–72]. One of the obvious reasons is the solubility of the aspartic acid-rich matrix under the conditions imposed by the decalcification step. Another reason, related to the first one, may be the accessibility of the acidic molecules. Being easily solubilized, it was possible to use these molecules to perform tests in vitro for their control in the biomineralization process. Some important roles were ascribed to them as the control of nucleation, the growth and the inhibition of crystal forma- tion [8,73]. The other compounds of the nacre organic matrix, present in WSM, also possess important biological activities on cellular mechanisms involved in biominerali- zation. Indeed, we have shown that the WSM is able to induce in vitro the differentiation pathway of osteoblasts from precursor cells like fibroblasts, bone marrow cells and preosteoblasts [20,21,74,75]. As silk-fibroin is insoluble after demineralization, it is difficult to isolate these molecules. The significance of the findings presented here is that, in practice, the silk fraction can now be analyzed for primary and secondary conformations, as well as for other biological and physical properties. With our comparative study on P. maxima nacre, it seems that a classification of the organic matrix into soluble and insoluble, to distinguish the acidic proteins from the hydrophobic glycine-alanine-rich molecules, is no longer valid and may even lead to misunderstandings. Some results support the idea that the amino acid sequence of proteins extracted from soluble and insoluble matrices share com- mon features [4]. Such a characteristic indicates that some of these proteins may in fact belong to the same family. ACKNOWLEDGMENTS We would like to express our thank to Professor Steve Weiner and Dr Lia Addadi for providing the illustration in Fig. 5. We are grateful to Dr Sophie Berland and Sandrine Borzeix (Laboratoire de Physiologie, MNHN, Paris, France) for all their useful comments and iconography, respectively. This study was partly funded by a Siera SA/CNRS Research and Development contract and by grants (PRAXIS XXI/BPD/11811 and PRAXIS XXI/BD/20023) from the Science and Technology Foundation of Portugal. REFERENCES 1. Crenshaw, M.A. (1972) The soluble matrix from Mercenaria mercenaria shell. Biomineralization 6, 6–11. 2. Krampitz, G., Engels, J. & Cazaux, C. (1976) Biochemical studies on water-soluble proteins and related components of gastropod shells. In The Mechanisms of Mineralization in the Invertebrates and Plants (Watabe, N. & Wilbur, K.M., eds), pp. 175–190. The Barush B.W. Library in Marine Science no. 5, The University of South Carolina Press, Columbia. 3. Weiner, S., Traub, W. & Lowenstam, H.A. (1983) Organic matrix in calcified exoskeletons. In Biomineralization and Biological Metal Accumulation (Westbroek, P. & deJong, E.W., eds), pp. 205–224. Reidel, Dordrecht, Holland. 4. Keith, J., Stockwell, D., Ball, D., Remillard, K., Kaplan, D., Thannhauser, T. & Sherwood, R. (1993) Comparative analysis of macromolecules in mollusc shells. Comp. Biochem. Physiol. 105B, 487–496. 5. Levi-Kalisman, Y., Falini, G., Addadi, L. & Weiner, S. (2001) Structure of the nacreous organic matrix of a bivalve mollusk shell examined in the hydrated state using cryo-TEM. J. Struct. Biol. 135, 8–17. 6. Weiner, S. & Hood, L. (1975) Soluble protein of the organic matrix of mollusk shells: a potential template for shell formation. Science 190, 987–989. 7. Weiner, S. & Traub, W. (1984) Macromolecules in mollusc shells and their functions in biomineralization. Phil. Trans. R. Soc. Lond. B 304, 425–434. 8. Belcher, A.M., Wu, X.H., Christensen, R.J., Hansma, P.K., Stucky,G.D.&Morse,D.E.(1996)Controlofcrystalphase switching and orientation by soluble mollusc-shell proteins. Nature 381, 56–58. 9. Kamat, S., Su, X., Ballarini, R. & Heuer, A.H. (2000) Structural basis for the fracture toughness of the shell of the conch Strombus gigas. Nature 405, 1036–1040. 10. Hare, P.E. (1963) Amino acids in the proteins from aragonite and calcite in the shells of Mytilus californianus. Science 139, 216–217. 11. Hare, P.E. & Abelson, P.H. (1965) Amino acid composition of some calcified proteins. Carnegie Inst Washington Year Book 64, 223. 12. Watabe, N. (1984) Shell. In Biology of the Integument,Vol.1 Invertebrates (Bereiterhahn, J., Matolsty, A.G. & Richards, K.S., eds), pp. 448–485. Springer-Verlag, Berlin. 13. Mann, S. (1988) Molecular recognition in biomineralization. Nature 332, 119–124. 14. Kono, M., Hayashi, N. & Samata, T. (2000) Molecular mechan- ism of the nacreous layer formation in Pinctada maxima. Biochem. Biophys. Res. Commun. 269, 213–218. 15. Lopez, E., Berland, S. & Le Faou, A. (1994) Nacre, osteogenic and osteoinductive properties. Bull. Inst Oce ´ an. Monaco 14, 49–57. 16. Lopez, E., Vidal, B., Berland, S., Camprasse, C., Camprasse, G. & Silve, C. (1992) Demonstration of the capacity of nacre to induce bone formation by human osteoblasts maintained in vitro. Tissue Cell 24, 667–679. 17. Atlan, G., Balmain, N., Berland, S., Vidal, B. & Lopez, E. (1997) Reconstruction of human maxillary defects with nacre powder: histological evidence for bone regeneration. C. R. Acad. Sci. Paris, Life Sci. 320, 253–258. 18. Atlan, G., Delattre, O., Berland, S., Le Faou, A., Nabias, G., Cot, D. & Lopez, E. (1999) Interface between bone and nacre implants in sheep. Biomaterials 20, 1017–1022. 19. Lamghari, M., Almeida, M.J., Berland, S., Huet, H., Laurent, A., Milet, C. & Lopez, E. (1999) Stimulation of bone marrow cells and bone formation by nacre: in vivo and in vitro studies. Bone 25S, 91S–94S. 20. Almeida, M.J., Milet, C., Peduzzi, J., Pereira, L., Haigle, J., Barthe ´ lemy, M. & Lopez, E. (2000) The effect of water-soluble matrix fraction extracted from the nacre of Pinctada maxima on the alkaline phosphatase activity of cultured fibroblasts. J. Exp. Zool. (Mol. Dev. Evol.) 288, 327–334. 21. Almeida, M.J., Pereira, L., Milet, C., Haigle, J., Barbosa, M. & Lopez, E. (2001) Comparative effects of nacre water-soluble matrix and dexamethasone on the alkaline phosphatase activity of MRC-5 fibroblasts. J. Biomed. Mater. Res. 57, 306–312. 22. Lamghari,M.,Berland,S.,Laurent,A.,Huet,H.&Lopez,E. (2001) Bone reactions to nacre injected percutaneously into the vertebrae of sheep. Biomaterials 22, 555–562. 23. Lopez,E.,Giraud,M.,LeFaou,A.,Berland,S.&Gutierrez,J. (1995) Proce ´ de ´ de pre ´ paration de substances actives a ` partir de la nacre, produits obtenus, utiles comme me ´ dicaments.PatentCNRS no. FR. 9515650, 28/12/1995, WO 97/24133–10/07/97. Ó FEBS 2002 Soluble silk-like organic matrix in nacre (Eur. J. Biochem. 269) 5001 24. Wheeler, A.P., Rusenko, K.W., George, J.W. & Sikes, S. (1987) Evaluation of calcium binding by molluscan shell organic matrix and its relevance to biomineralization. Comp. Biochem. Physiol. 87B, 953–960. 25. Cohen-Solal, L., Lian, J.B., Kossiva, D. & Glimcher, M.J. (1978) The identification of O-phosphothreonine with soluble non- collagenous phosphoproteins of bone matrix. FEBS Lett. 89, 107–110. 26. Cohen, S.A. & Strydom, D.J. (1988) Amino acid analyses utilizing phenylisothiocyanate derivatives. Anal. Biochem. 174, 1–16. 27. Hewitt, B.R. (1958) Spectrophotometric determination of protein in alkaline solution. Nature 187,246. 28. Whiteman, P. (1973) The quantitative determination of glycos- aminoglycans in urine with Alcian blue 8GX. Biochem. J. 131, 351–357. 29. Whiteman, P. (1973) The quantitative measurement of Alcian blue-glycosaminoglycan complexes. Biochem. J. 131, 343–350. 30. Gold, E.W. (1979) A simple spectrophotometric method for esti- mating glycosaminoglycan concentrations. Anal. Biochem. 99, 183–188. 31. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. 32. Morrissey, J.H. (1981) Silver stain for proteins in polyacrylamide gels: a modified procedure with enhanced uniform sensitivity. Anal. Biochem. 117, 307–310. 33. Wheeler, A.P., Rusenko, K.W. & Sikes, S. (1988) Organic matrix from carbonate biomineral as a regulator of mineralization. In Chemical Aspects of Regulation of Mineralization (Sikes,C.S.& Wheeler, A.P., eds). University of South Alabama Publication Services, Mobile, Alabama. 34. Weiner, S. (1979) Aspartic acid-rich proteins: major components of the soluble organic matrix of mollusk shells. Calcif. Tissue Int. 29, 163–167. 35. Wheeler, A.P., Rusenko, K.W., Swift, D.M. & Sikes, S. (1988) Regulation of in vitro and in vivo CaCO 3 crystallization by frac- tions of oyster shell organic matrix. Mar. Biol. 98, 71–80. 36. Halloran, B.A. & Donachy, J.E. (1995) Characterization of organic matrix macromolecules from the shells of the antarctic scallop Adamussium colbecki. Comp. Biochem. Physiol. 111B, 221–231. 37. Scott, J.E., Quintarelli, G. & Dellovo, M.C. (1964) The chemical and histochemical properties of Alcian blue staining. Histochemie 4, 73–85. 38. Nakamoto, K. (1986) IR and Raman Spectra of Inorganic and Coordination Compounds.J.Wiley&Sons,NewYork. 39. Samata,T.&Krampitz,G.P.(1982)Calciumbindingpolypep- tides in oyster shells. Malacologia 22, 225–233. 40. Roeges, N.P.G. (1998) A Guide to the Complete Interpretation of Infrared Spectra of Organic Structures.J.Wiley&Sons,New York. 41. Cariolou, M. & Morse, D. (1988) Purification and characteriza- tion of calcium binding conchiolin shell peptides from the mollusc, Haliotis rufescens, as a function of development. J. Comp. Physiol. 157B, 717–729. 42. Krampitz, G.P. (1982) Structure of the organic matrix in mollusc shells and avian eggshells. In Biological Mineralization and Demineralization (Nancollas, G.H., ed.), pp. 219–232. Springer- Verlag, Berlin, Heidelberg, New York. 43. Weiner, S. & Addadi, L. (1991) Acidic macromolecules of mineralized tissues: the controllers of crystal formation. Trends Biochem. Sci. 16, 252–256. 44. Marin, F., Pereira, L. & Westbroek, P. (2001) Large-scale frac- tionation of molluscan shell matrix. Protein Expr. Purif. 23, 175–179. 45. Balmain, J., Hannoyer, B. & Lopez, E. (1999) Fourier transform infrared spectroscopy (FTIR) and X-ray diffraction analyses of mineral and organic matrix during heating of mother of pearl (nacre) from the shell of the mollusc Pinctada maxima. J. Biomed. Mater. Res. Appl. Mat. 48, 754. 46.Marxen,J.C.,Hammer,M.,Gehrke,T.&Becker,W.(1998) Carbohydrates of the organic shell matrix and the shell forming tissue of the snail Biomphalaria glabrata (Say). Biol. Bull. 194, 231–240. 47. Miyazawa, T. & Blout, E.R. (1961) The infrared spectra of poly- peptides in various conformations – amide I and II bands. J. Am. Chem. Soc. 83, 712–719. 48. Worms, D. & Weiner, S. (1986) Mollusk shell organic matrix: fourier transform infrared study of the acidic macromolecules. J. Exp. Zool. 237, 11–20. 49. Weiner, S. & Traub, W. (1980) X-ray diffraction study of the insoluble organic matrix of mollusk shells. FEBS Lett. 111,311– 316. 50. Pereira-Mourie ` s, L., Almeida, M.J., Marin, F., Gay, M., Milet, C. & Lopez, E. (2002) Soluble proteins of the nacre from the giant oyster Pinctada maxima with highly hydrophobic amino acids. Biomineralization: Formation, Diversity, Evol. Appl, in press. 51. Nakahara, H., Kakei, M. & Bevelander, G. (1980) Fine structure and amino acid composition of the organic ÔenvelopeÕ in the prismatic layer of some bivalve shells. Venus Jpn. J. Malacol. 39, 167–177. 52. Falini, G., Albeck, S., Weiner, S. & Addadi, L. (1996) Control of aragonite or calcite polymorphism by mollusk shell macro- molecules. Science 271, 67–69. 53. Simkiss, K. (1965) The organic matrix of the oyster shell. Comp. Biochem. Physiol. 16, 427–435. 54. Golberg, M. & Takagi, M. (1993) Dentine proteoglycans: com- position, ultrastructure, and functions. Histochem. J. 25, 781–806. 55. Wilbur, K.M. (1976) Recent studies of invertebrate mineraliza- tion. In The Mechanisms of Mineralization in Invertebrates and Plants (Watabe, N. & Wilbur, K.M., eds), pp. 79–108. The Uni- versity of. South Carolina Press, The Belle W. Baruch Library in Marine Science no. 5, Columbia. 56. Addadi, L., Moradian, J., Shay, E., Maroudas, N.G. & Weiner, S. (1987) A chemical model for the cooperation of sulfates and car- boxylates in calcite crystal nucleation: Revelance to biominer- alization. Proc. Natl Acad. Sci. USA 84, 2732–2736. 57. Marxen,J.C.&Becker,W.(2000)Calciumbindingconstituentsof the organic shell matrix from the freshwater snail Biomphalaria glabrata. Comp. Biochem. Physiol. 127B, 235–242. 58. Tanaka, Y., Adams, D.H. & Shaw, S. (1993) Proteoglycans on endothelial cells present adhesion-inducing cytokines to leuko- cytes. Immunol. Today 14, 111–115. 59. Jackson, R.L., Bush, S.J. & Cardin, A.D. (1991) Glycosami- noglycans: molecular properties, protein interactions, and role in physiological processes. Physiol. Rev. 71, 481–539. 60. Nakahara, H. (1983) Calcification of gastropod nacre. In Bio- mineralization and Biological Metal Accumulation (Westbroek, P. & deJong, E.W., eds), pp. 225–230. Reidel, Dordrecht, Holland. 61. Zentz, F., Be ´ douet,L.,Almeida,M.J.,Milet,C.,Lopez,E.& Giraud, M. (2001) Characterization and quantification of chitosan extracted from nacre of the abalone Haliotis tuberculata and the oyster Pinctada maxima. Mar. Biotechnol. 3, 36–44. 62. Alzieu, C., Thibaud, Y., Heral, M. & Boutier, B. (1980) Evaluation des risques dus a l’emploi des peintures anti-sallissures dans les zones conchyoliques. Rev. Trav. Inst. Pe ˆ ches Marit. 44, 301–348. 63. Alzieu, C., Heral, M., Thibaud, Y., Dardignac, M.J. & Feuillet, M. (1982) Influence des peintures anti-sallissures a ` base d’organostanniques sur la calcification de la coquille de l’huıˆtre Crassostrea gigas. Rev. Trav. Inst. Pe ˆ ches Marit. 45, 101–116. 64. Krampitz, G., Drolshagen, H. & Deltreil, J.P. (1983) Soluble matrix components in malformed oyster shells. Experientia 39, 1105–1106. 65. Alzieu, C. & Heral, M. (1984) Ecotoxicological effects of orga- notin coumpounds on oyster culture. In Ecotoxicological Testing 5002 L. Pereira-Mourie ` s et al. (Eur. J. Biochem. 269) Ó FEBS 2002 for the Marine Environment, 2 (Persoone, G., Jaspers, E. & Claus, C., eds), pp. 187–196. State University Ghent and Inst. Mar. Sci. Res., Bredene, Belgium. 66. Almeida, M.J. (1996) Crescimento em ostras: factores que afectam a formac¸a ˜ o e estrutura da concha. PhD Thesis, Instituto de Cieˆ ncias Biome ´ dicas Abel Salazar, University of Porto, Portugal. 67. Sudo, S., Fujikawa, T., Nagakura, T., Ohkubo, T., Sakaguchi, K., Tanaka, M., Nakashina, K. & Takahashi, T. (1997) Structures of mollusc shell framework proteins. Nature 387, 563–564. 68. Shen, X., Belcher, A.M., Hansma, P.K., Stucky, G.D. & Morse, D.E. (1997) Molecular cloning and characterization of Lustrin A, a matrix protein from shell and pearl nacre of Haliotis rufescens. J. Biol. Chem. 272, 32472–32481. 69. Samata, T., Hayashi, N., Kono, M., Hasegawa, K., Horita, C. & Akera, S. (1999) A new matrix protein family related to the nacreous layer formation of Pinctada fucata. FEBS Lett. 462, 225–229. 70. Mann, S., Webb, J. & Williams, R.J.P. (1989) Biomineralization: Chemical and Biochemical Perspectives, pp. 35–62. VCH, Weinheim. 71. Simkiss, K. & Wilbur, K.M. (1989) Biomineralization – Cell Biology and Mineral Deposition. Academic Press, San Diego. 72. Sarashina, I. & Endo, K. (1998) Primary structure of a soluble matrix protein of scallop shell: implications for calcium carbonate biomineralization. Am. Mineral. 83, 1510–1515. 73. Weiss, I., Kaufmann, S., Mann, K. & Fritz, M. (2000) Purification and characterization of perlucin and perlustrin, two new proteins from the shell of the mollusc Haliotis laevigata. Biochem. Biophys. Res. Commun. 267, 17–21. 74. Pereira-Mourie ` s,L.,Milet,C.,Almeida,M.J.,Rousseau,M., Borzeix, S. & Lopez, E. (2001) Biological effect of the water- soluble matrix extract of the nacre shell layer of Pinctada maxima on vertebrate cell proliferation and differentiation. In Proceedings of the 3rd International Satellite Symposium on the Comparative Endocrinology of Calcium Regulation (Gay, C.V., Dacke, C.G., Dancks, J.A. & Cox, P.A., eds). http://www.sci.port.ac.uk/-cecr 75. Pereira-Mourie ` s, L., Almeida, M.J., Milet, C., Berland, S. & Lopez, E. (2002) Bioactivity of nacre water-soluble organic matrix from the bivalve mollusk Pinctada maxima in three mammalian cell types: fibroblasts, bone marrow stromal cells and osteoblasts. Comp. Biochem. Physiol. 132, 215–217. Ó FEBS 2002 Soluble silk-like organic matrix in nacre (Eur. J. Biochem. 269) 5003 . Soluble silk-like organic matrix in the nacreous layer of the bivalve Pinctada maxima A new insight in the biomineralization field Lucilia Pereira-Mourie ` s 1 ,. water -soluble matrix, water-insoluble matrix, the EDTA -soluble matrix and the EDTA- insoluble matrix of Pinctada maxima nacre. Results are expressed as a mole