Abstract
The role of proteins in biomineralization and the mechanism of eggshell formation are not well understood. We have isolated and purified the major protein, ansocalcin from goose eggshell matrix. The amino acid sequence study indicates that ansocalcin is homologous to the chicken eggshell protein, ovocleidin 17, and C-type lectins. Ansocalcin nucleates polycrystalline aggregates of calcite crystals in in vitro mineralization experiments. The polycrystalline aggregates obtained at higher concentration of ansocalcin appears to be similar to the crystals observed at the mamillary layer of the eggshell.
Biomineralization is commonly referred to as the formation of biocomposites consisting of layered assembly of biomacromolecules such as proteins with well ordered calcium-rich inorganic phase (1–3). This highly complex process involves a series of molecular events such as selective recognition and deposition of calcium salts by proteins followed by the formation of a mineral phase comprising crystallites of uniform size, crystallographic orientation, and morphology (4, 5). In most cases, the minerals are formed over a biomolecular scaffold or mold resulting in composite materials (6, 7). The role of such biologically programmed composites varies from skeletal tissues to protective shells against the predators (8, 9). These highly complex, hierarchically ordered and multifunctional composites are formed under mild conditions with size ranges from nano- to centimeter scale, exhibiting unusual mechanical properties that outperform synthetic materials (10–15). Such structures appeared to originate from organized assembly of biomacromolecules such as proteins, polysaccharides or proteoglycans, and the inorganic salt (16). The organic macromolecules act as templates through self-assembly to facilitate interaction with the insoluble matrix and to induce the desired stereochemistry for the construction of organized structures (17–19). Biomacromolecules are known to be involved in controlling the nucleation, growth, size, and shape of the mineral phases (20). Often these macromolecules are functionalized with acidic groups such as carboxylic acid, sulfonate, and phosphate moieties, which allow them to be an effective metal ion chelator to interact with the inorganic matrix (21–23).
Avian eggshells present a unique and interesting model for exploring the process of biomineralization, in which CaCO3 layers are created by the selective nucleation and deposition of calcite crystals by proteins. Moreover, the active sites on the protein recognize calcium ions and induce nucleation of specific polymorph of CaCO3 and control the morphology of mineral phase. In the case of chicken eggshells, ≈5 g of CaCO3 is deposited within 22 h in an acellular medium as the egg passes through the oviduct, which makes avian eggshells one of the fastest mineralized hard tissues in biological systems (24). The structure of an avian eggshell consists of two principal external layers, namely a membrane layer that surrounds the albumin and a calcified shell. The calcified shell exhibits various morphologically different domains (Fig. 1; ref. 25). The inner portion of the calcified shell, known as mamillary layer, is partially embedded and attached to the outer surface of the membrane fibers. Next to the mamillary layer is the pallisade layer (Fig. 1 A and B), which forms 65–70% of the shell thickness. The demarcation between the mamillary and pallisade layer involves a gradual alteration in the crystal packing. It is believed that the calcite crystal nucleation and eggshell calcification start from the mamillary caps on the mamillary layer. To reveal the mamillary caps, we removed the membrane layer by perchlorate bleaching by using a procedure reported in the literature (26). Fig. 1C shows mamillary caps, which contain tiny aggregates of calcite crystal buds. Because growth of the biomineralized structure takes place in a spatially controlled manner, it has been hypothesized that matrix vesicles might fulfill such a role in the eggshell calcification (27).
Figure 1.
Schematic representation of the structure of the eggshell (B) with SEM micrographs showing the cross section (A), and the mamillary caps (see C Inset for detailed picture).
The biomacromolecules also exert exquisite control over the polymorphism of a particular mineral phase. CaCO3 has three known polymorphs: calcite, aragonite, and vaterite phases. The aragonite and calcite are stable polymorphs and observed in nature, whereas the vaterite is a metastable polymorph and rarely seen in biological systems (9). Falini et al. (28, 29) showed that macromolecules extracted from aragonite layer of a natural hard tissue induced nucleation of aragonite crystals under appropriate microenvironment. Moreover, Belcher et al. (30) demonstrated that a soluble polyanionic protein (16 kDa) is efficient to allow the transformation of calcite to aragonite phase. However, the exact mechanism by which the macromolecules control the polymorphism in both biotic and abiotic systems is not yet well understood. Our interest is focused on understanding the molecular mechanism of biomineralization process and designing synthetic composites that resembles their biological counterparts. Here we report the purification and characterization of various matrix proteins from goose (Anser anser) eggshells and identify their role in in vitro mineralization of calcium salts. We believe that the nature and organization of functional groups at the surface of the proteins are crucial to achieving the desired selectivity in polymorph nucleation, as well as controlling the crystalline nature and morphology of the inorganic phase.
Experimental Procedures
Extraction of the Eggshell Matrix Proteins.
Commercially available fresh goose eggshells were powdered and thoroughly washed with Millipore deionized water. They were decalcified with 1 N HCl for 30 min, filtered, and centrifuged. The supernatant solution was microconcentrated using Amicon micro concentrator (YCO5 membrane, 500 MW cutoff) at 4°C. The turbid solution was centrifuged again at 4,000 rpm for 15 min and the supernatant liquid was taken for purification using reversed-phase (RP)-HPLC.
Protein Purification.
The proteins were fractionated on a Jupiter C18 reverse-phase column (5 μm, 250 mm × 10 mm) by using a Vision Workstation (Perkin–Elmer Biosystems). The column was equilibrated with 0.1% trifluoroacetic acid (TFA) and a linear gradient of acetonitrile was used for elution. The microconcentrated sample (≈15 mg protein) was injected onto the column and eluted at a flow rate of 2 ml/min. The elution of the proteins was monitored both at 215 and 280 nm.
Electrospray Ionization Mass Spectrometry.
Precise masses of the proteins were determined by electrospray ionization mass spectrometry using a Perkin–Elmer Sciex API 300 triple quadrupole instrument equipped with an ionspray interface. The ionspray 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 used as the nebulizer gas. The sample was injected into the mass spectrometer at a flow rate of 20 ml/min and scanned from mass to charge (m/z) 500 to 3,000. The multiple charged spectrum was deconvoluted into the mass scale by using the BIOSPEC RECONSTRUCT software supplied with the instrument data system.
Analysis of the Amino Acid Composition.
Amino acid composition was obtained using the Shimadzu HPLC model LC-10 A/C-R7A amino acid analysis system. To about 200 μg of the pure proteins in 300 μl of 6 N hydrochloric acid was added and hydrolyzed in vacuo at 112°C for 22 h. The hydrolyzed products, after evaporation of HCl, were separated on an ion exchange column, post column derivatized with o-phthalaldehyde and detected using a fluorescent detector.
Amino Terminal Sequencing.
Amino terminal sequencing of the purified proteins/peptides were performed by automated Edman degradation using a Perkin–Elmer Applied Biosystems 494 pulsed-liquid phase protein sequencer (Procise) with an on-line 785A PTH-amino acid analyzer.
Size Exclusion Chromatography (SEC).
The SEC was performed on a 1.5 × 57-cm Sepharose 6B-CL resin (Sigma) with CaCl2 (10 mM) as the mobile phase. The ovalbumin, BSA, and chicken egg lysozyme were used as the standards. All of the standards and the extracted proteins were dissolved in CaCl2 solution and eluted on the column.
Crystal Growth Experiments.
CaCO3 crystals were grown on glass coverslips placed inside the CaCl2 solution kept in a Nunc dish (4 × 6; ref. 16). Typically, 1 ml of 7.5 mM CaCl2 solution was introduced into the wells containing the coverslips and the whole set-up was covered with aluminum foil with a few pinholes on the top. Crystals were grown inside a closed desiccator for 2 days by the slow diffusion of gases released by the decomposition of ammonium carbonate placed inside the desiccator. To study the role of ansocalcin in the CaCO3 crystallization, aliquots of protein (0.01–500 μg/ml) dissolved in 7.5 mM CaCl2 solution were introduced into each well containing the coverslips. After 2 days, the slides were carefully lifted from the crystallization vessels, rinsed gently with Millipore water, air dried at room temperature, glued to copper stubs, and used for further analysis.
Scanning Electron Microscopy (SEM).
SEM studies on the CaCO3 crystals were carried out using JEOL 2200 scanning electron microscope at 15/20 kV. The crystals were sputter coated with gold to increase the conductivity before characterization using SEM.
Results
Structural Characterization of a Goose Eggshell Protein, Ansocalcin.
The eggshell mineral layer incorporates 3–5% biomacromolecules by its weight, which includes proteins, glycoproteins, and proteoglycans. A number of such biomacromolecules have been extracted, purified, and identified from the chicken eggshells (31–38). These biomacromolecules, especially the matrix proteins, are believed to be active in controlling the morphology and selective nucleation of polymorphs of the CaCO3 layer of the eggshell (39). We focused our attention on understanding the molecular mechanism of eggshell calcification through extraction, purification, and characterization of the matrix proteins (40). Recently, a number of proteins and peptides from the goose eggshell extract were isolated, purified, and characterized. The supernatant liquid obtained after decalcification and microconcentration of the eggshell extract was fractionated by reversed-phase (RP)-HPLC. A 15-kDa protein was identified from the positive ion electrospray ionization mass spectra and named ansocalcin. Its role in the mineralization using in vitro CaCO3 crystallization was evaluated.
The amino acid composition (Table 1, Fig. 2) shows that ansocalcin is rich in acidic Glx and Asx residues and basic Lys, Arg, and His amino acids. However, ansocalcin appeared to be more acidic than OC-17. Unlike other soluble, partially purified matrix protein fractions, ansocalcin and OC-17 have a considerable amount of basic amino acid residues (1, 20, 41).
Table 1.
A comparison of amino acid composition of ansocalcin extracted from the goose eggshell and ovocleidin 17 extracted from chicken eggshell
| Amino acid | Ansocalcin composition, mol % | Amino acid proportion* | Ovocleidin 17 composition,† mol % |
|---|---|---|---|
| Asx | 10.73 | 14.37 | 7.51 |
| Thr | 2.14 | 2.87 | 3.9 |
| Ser | 8.08 | 10.83 | 6.81 |
| Glx | 11.92 | 15.97 | 7.65 |
| Pro | 3.32 | 4.45 | 6.27 |
| Gly | 8.64 | 10.91 | 10.14 |
| Ala | 12.02 | 16.11 | 17.85 |
| Val | 4.88 | 6.54 | 3.47 |
| Ile | 3.58 | 4.8 | 1.27 |
| Leu | 6.06 | 8.12 | 8.1 |
| Tyr | 3.07 | 4.11 | — |
| Phe | 4.94 | 6.62 | 4.76 |
| His | 4.94 | 6.62 | 2.6 |
| Lys | 7.42 | 9.94 | 2.25 |
| Arg | 7.31 | 9.8 | 18.56 |
The amino acid content is expressed in mol %.
Number of residues per protein of mass 15,342.14. The data are presented as percentage molar composition of the amino acid and number of amino acid residues per protein taking into consideration that the average molecular weight of the amino acid is 114.62 and the protein does not have any posttranslational modifications and/or glycosylated.
Data taken from ref. 32.
Figure 2.
Amino acid sequence homology of ansocalcin with other related proteins. Ovocleidin isolated from chicken eggshell (35), c-type lectin found from snake venom (36). The identical amino acid residues are shaded in black; asterisks indicate the pair of acid residues observed in ansocalcin.
The amino acid sequence of ansocalcin was determined by automated Edman degradation of native and pyridylethylated protein, as well as that of peptides obtained by chemical and enzymatic digests of the protein. Ansocalcin has 132-aa residues comprising hydrophilic and hydrophobic domains (Fig. 2). Both ansocalcin and ovocleidin (OC-17) (35) proteins belong to the C-type Lectin (CTL) family of proteins. Ansocalcin showed significant homology to ovocleidin 17 extracted from chicken eggshell (50%) and CTL (49%). All cysteines of ansocalcin are conserved in OC-17 (except Cys-38), and CTL has an additional Cys at position 133. Ansocalcin has many acid residues (e.g., Asp, Glu) arranged in pairs, which are known to be potential calcium binding moieties. We identified six such pairs in the primary structure (marked with asterisk) of ansocalcin (Fig. 2). It will be interesting to examine the role of these six pairs of acidic residues in the in vitro mineralization of CaCO3.
Mineralization Properties of Ansocalcin.
Fig. 3 shows some of the salient features of calcite crystals formed in the presence of ansocalcin. As expected, the rhombohedral calcite crystals were observed in the absence of protein (Fig. 3A). The aggregation of calcite crystals was strongly influenced by the concentration of the ansocalcin protein in the crystallization medium. At lower levels of proteins (below 50 μg/ml), calcite crystals that exhibit screw dislocations in the form of “hopper” crystals with rounded pits on the {10.4} faces were formed (Fig. 3B). This may be due to the interaction of protein with the crystal faces of growing calcite crystals. An abrupt transition from single crystals to polycrystalline aggregates took place above 50 μg/ml of added ansocalcin protein. The crystal aggregates incorporated more and more rhombohedral calcite crystallites as the protein concentration increased to 100 μg/ml (Fig. 3C). The overall shape of the crystal aggregates gradually changed from spherical to ellipsoidal as the concentration of the ansocalcin was further increased to 500 μg/ml (Fig. 3D).
Figure 3.
Morphological changes in the calcite crystals grown in the presence of ansocalcin at various levels. (A) Control; (B) 10 μg/ml; (C) 100 μg/ml; (D) 500 μg/ml. See Inset for an enlarged crystal. (Scale bar, 50 μm.)
The surface of the crystals grown at a high concentration of ansocalcin revealed a porous texture (Fig. 3D Inset). The SEM investigation of the crystals collected under various conditions also showed that the number, size, and distribution of crystal aggregates changed significantly as the concentration of ansocalcin was increased. The aggregates formed at the higher concentration of ansocalcin indicated resemblance to the morphology of the mamillary caps (Figs. 1C and 3D). The observed concentration-dependant changes in morphology of calcite crystals may imply that the active nucleation sites and the growth of the crystals are influenced by the change in concentration of the proteins. These observations directed our attention to investigating possible aggregation properties of ansocalcin via self-assembly at higher concentration. To study the aggregation properties, we examined the elution properties of ansocalcin solution at various concentrations by using a Sepharose 6B-CL resin column (Sigma, 1.5 × 57 cm) with CaCl2 solution as the mobile phase. The commercially available proteins such as ovalbumin, BSA, and chicken egg lysozyme dissolved in CaCl2 solution were used as the standards. As shown in Fig. 4, the ansocalcin protein exists as homodimers at a concentration of 0.1 mg/ml and homotrimers at 0.5 mg/ml.
Figure 4.
Size exclusion chromatogram (SEC) of ansocalcin in CaCl2 solution at various concentrations, 0.1 mg/ml (1) and 0.5 mg/ml (2). The dotted lines represent the position of the standards: (a) BSA (66 kDa), (b) egg white albumin (44 kDa), and (c) chicken egg lysozyme (14 kDa).
Discussion
The detailed understanding of the structure and functions of eggshell proteins and their contribution to biomineralization will enhance our capability to develop functional biomaterials and synthesize advanced ceramics. As a first step, we have purified various eggshell matrix proteins to homogeneity from the goose eggshell and examined the contribution of ansocalcin, the most abundant protein on CaCO3 crystallization. Amino acid analysis showed that ansocalcin has high amount of acidic amino acid residues typical of a eggshell matrix protein (Fig. 2). In vitro crystal growth experiments showed that individual crystals transformed into polycrystalline aggregates as the concentration of ansocalcin was varied (Fig. 3). The formation of spiral pits at lower concentration of ansocalcin (10 μg/ml) indicates strong interaction between the acidic groups (Asp or Glu) present in the protein and the growing crystal nuclei. Ansocalcin at higher concentrations (above 50 μg/ml) induces the formation of crystal aggregates. This observation may be due to the aggregation of ansocalcin and thus accumulation of nucleating sites. More and more calcite crystallites were incorporated in the aggregates as the concentration of ansocalcin increases. Ansocalcin exists as monomer at lower concentration and homodimer or homotrimer at higher concentrations (Fig. 4). Such aggregation is anticipated because of the presence of high charge on the molecule.
By correlating the morphological changes with the aggregation features of the ansocalcin, it is clear that at a low concentration level, the protein tend to behave like individual proteins, which get adsorbed onto the growing calcite crystal faces and thereby creating screw dislocation features. However, as the concentration increases, the aggregation of proteins facilitates the aggregation of crystallites by providing large surface area and increased number of surface functional groups. Other research groups also reported similar mechanisms. For example, Banfield et al. (42) proposed a self-assembly based mechanism for the naturally occurring iron oxyhydroxide biomineral formation. The enamel extracellular matrix protein, amelogenin, produced clusters of calcium hydroxyapatite crystals by forming tightly associated spheres of 15–20 nm dimensions (43). Therefore, we believe that eggshell calcification begins with the self-assembly of matrix macromolecule(s), which then direct the growth of calcite crystals.
Conclusions
We discuss an attempt to explore the molecular mechanism of eggshell formation by using a biochemical model system of goose eggshell matrix. Ansocalcin was extracted from the goose eggshell and fully characterized. The in vitro calcite crystal nucleation was studied to understand the role of this protein in the eggshell mineralization. The highly charged ansocalcin protein appears to exist in aggregate form at higher concentration and influence the nucleation and aggregation of calcite crystals in solution. The results indicate significant structural and shape similarity between the crystals observed at the mamillary layer of the goose eggshell and the crystals obtained from in vitro mineralization using ansocalcin. We believe that such approaches focusing on understanding the role of various functional groups and the recognition processes at the surface of protein aggregates are important to unraveling the role of proteins and other biomacromolecules toward the development of hard tissues in nature.
Acknowledgments
We wish to acknowledge the financial support of the National University of Singapore through research grants and support from the technical staff at the Department of Chemistry and Biological Sciences.
Abbreviation
- SEM
scanning electron microscopy
Footnotes
This paper was submitted directly (Track II) to the PNAS office.
Data deposition: The sequence reported in this paper has been deposited in the Swiss-Prot database (accession no. P83300).
References
- 1.Weiner S. Biochemistry. 1983;22:4139–4145. [Google Scholar]
- 2.Berman A, Addadi L, Kvick A, Leiserowitz L, Nelson M, Weiner S. Science. 1990;250:664. doi: 10.1126/science.250.4981.664. [DOI] [PubMed] [Google Scholar]
- 3.Mann S. Struct Bonding (Berlin) 1983;54:125–174. [Google Scholar]
- 4.Lowenstam H A. Science. 1981;211:1126–1131. doi: 10.1126/science.7008198. [DOI] [PubMed] [Google Scholar]
- 5.Soten I, Ozin G A. Curr Opin Colloid Interface Sci. 1999;4:325–337. [Google Scholar]
- 6.Berman A, Addadi L, Weiner S. Nature (London) 1988;331:546–548. [Google Scholar]
- 7.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, et al. Science. 1992;255:1098–1105. doi: 10.1126/science.1546311. [DOI] [PubMed] [Google Scholar]
- 8.Addadi L, Weiner S. Angew Chem Int Ed Engl. 1992;31:153–169. [Google Scholar]
- 9.Weiner S, Addadi L. J Mater Chem. 1997;7:689–702. [Google Scholar]
- 10.Krampitz G, Graser G. Angew Chem Int Ed Engl. 1988;27:1145–1156. [Google Scholar]
- 11.Jackson A, Vincent J E, Turner R M. Proc R Soc London Ser B. 1988;234:415–440. [Google Scholar]
- 12.Currey J D. Proc R Soc London Ser B. 1977;196:443–463. [Google Scholar]
- 13.Mann S. Nature (London) 1993;365:499–505. [Google Scholar]
- 14.Currey J D. J Mater Educ. 1987;9:118–296. [Google Scholar]
- 15.Simkiss K. In: Biomineralization. Allemand D, editor. Monaco: Mussee Oceanographique; 1993. p. 49. [Google Scholar]
- 16.Albeck S, Aizenberg J, Addadi L, Weiner S. J Am Chem Soc. 1993;115:11691–11697. [Google Scholar]
- 17.Mann S. Nature (London) 1988;332:119–124. [Google Scholar]
- 18.Knight C A, Cheng C C, Devries A L. Biophys J. 1991;59:409–418. doi: 10.1016/S0006-3495(91)82234-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Addadi L, Weiner S. Proc Natl Acad Sci USA. 1985;82:4110–4114. doi: 10.1073/pnas.82.12.4110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Greenfield G, Wilson D C, Crenshaw M A. Am Zool. 1984;24:925–932. [Google Scholar]
- 21.Veis A. In: Biomineralization: Chemical and Biochemical Perspectives. Mann S, Webb J, Williams R J P, editors. Weinhein: VCH; 1989. p. 189. [Google Scholar]
- 22.Addadi L, Weiner S, Geva M. Z Kardiol. 2001;3,Suppl.:92–98. doi: 10.1007/s003920170049. [DOI] [PubMed] [Google Scholar]
- 23.Albeck S, Weiner S, Addadi L. Chem Eur J. 1996;2:278–284. [Google Scholar]
- 24.Arias J L, Fink D J, Xiao S Q, Heuer A H, Caplan A I. Int Rev Cytol. 1993;145:217–250. doi: 10.1016/s0074-7696(08)60428-3. [DOI] [PubMed] [Google Scholar]
- 25.Watabe N, Dunkelberger D G. Scanning Electron Microsc. 1979;II:403–416. [Google Scholar]
- 26.Hincke M T, Bernard A M, Lee E R, Tsang C P W, Narbaitz R. Br Poultry Sci. 1992;33:505–516. doi: 10.1080/00071669208417489. [DOI] [PubMed] [Google Scholar]
- 27.Fraser A C, Bain M M, Solomon S E. Br Poultry Sci. 1998;39:225–228. doi: 10.1080/00071669889169. [DOI] [PubMed] [Google Scholar]
- 28.Falini G, Albeck S, Weiner S, Addadi L. Science. 1996;271:67–69. [Google Scholar]
- 29.Levi Y, Albeck S, Brack A, Weiner S, Addadi L. Chem Eur J. 1998;4:389–396. [Google Scholar]
- 30.Belcher A M, Wu X H, Christensen R J, Hansma P K, Stucky G D, Morse D E. Nature (London) 1996;381:56–58. [Google Scholar]
- 31.Hincke M T, Gautron J, Tsang C P W, McKee M D, Nys Y. J Biol Chem. 1999;274:32915–32923. doi: 10.1074/jbc.274.46.32915. [DOI] [PubMed] [Google Scholar]
- 32.Hincke M T, Tsang C P W, Courtney M, Hill V, Narbaitz R. Calcif Tissue Int. 1995;56:578–583. doi: 10.1007/BF00298593. [DOI] [PubMed] [Google Scholar]
- 33.Hincke M T. Connect Tissue Res. 1995;31:227–233. doi: 10.3109/03008209509010814. [DOI] [PubMed] [Google Scholar]
- 34.Gautron J, Hincke M T, Garcia-Ruiz J M, Dominguez-Veva J M, Nys Y. In: Eggs and Egg Products Quality. Kijowski J, Pikul J, editors. Pozhaus, Poland: Proceedings of the VII European Symposium; 1997. pp. 66–75. [Google Scholar]
- 35.Mann K, Siedler F. Biochem Mol Biol Int. 1999;47:997–1007. doi: 10.1080/15216549900202123. [DOI] [PubMed] [Google Scholar]
- 36.Hirabayashi J, Kusunoki T, Kasai K. J Biol Chem. 1991;266:2320–2326. [PubMed] [Google Scholar]
- 37.Mann K. FEBS Lett. 1999;463:12–14. doi: 10.1016/s0014-5793(99)01586-0. [DOI] [PubMed] [Google Scholar]
- 38.Gautron J, Hincke M T, Mann K, Panheleux M, Bain M, McKee M D, Solomon S E, Nys Y. J Biol Chem. 2001;276:39243–39252. doi: 10.1074/jbc.M104543200. [DOI] [PubMed] [Google Scholar]
- 39.Panheleux M, Bain M, Fernandez M S, Morales I, Gautron J, Arias J L, Solomon S E, Hincke M T, Nys Y. Br Poultry Sci. 1999;40:240–252. doi: 10.1080/00071669987665. [DOI] [PubMed] [Google Scholar]
- 40.Solomon S E. Br Poultry Sci. 1999;40:5–11. doi: 10.1080/00071669987764. [DOI] [PubMed] [Google Scholar]
- 41.Halloran B A, Donachy J E. Comp Biochem Physiol. 1995;111:221–231. doi: 10.1016/0305-0491(94)00245-p. [DOI] [PubMed] [Google Scholar]
- 42.Banfield J F, Welch S A, Zhang H, Ebert T T, Lee Penn R. Science. 2000;289:751–754. doi: 10.1126/science.289.5480.751. [DOI] [PubMed] [Google Scholar]
- 43.Moradian-Oldak J, Tan J, Fincham A G. Biopolymers. 1998;46:225–228. doi: 10.1002/(SICI)1097-0282(19981005)46:4<225::AID-BIP4>3.0.CO;2-R. [DOI] [PubMed] [Google Scholar]




