Biomacromolecules 2003, 4, 1380-1385
1380
Characterization of Shrimp Shell Deproteinization Aline Percot, Christophe Viton, and Alain Domard* Laboratoire des Mate´ riaux Polyme` res et des Biomate´ riaux, UMR-CNRS 5627, Baˆ timent ISTIL, Domaine Scientifique de la Doua, 15 Bd. Andre´ Latarjet, 69622 Villeurbanne Cedex, France
Downloaded via UNIV OF EDINBURGH on January 29, 2019 at 09:13:53 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
Received April 15, 2003; Revised Manuscript Received June 13, 2003
The aim of this paper was to contribute to the interpretation of the mechanism of shrimp shell deproteinization. We used amino acid analysis to quantify the amount of proteins remaining in chitin. NaOH 1 M was added to a demineralized shrimp shell powder with a solution-to-solid ratio of 15 mL/g at ambient temperature. Because of the limited precision of the technique, after 24 h the protein content measured by elemental analysis had to be considered as negligible. However, with the use of amino acid analysis, it was still possible to determine with precision this content down to 0.25%. We also showed that among the peptides remaining linked to chitin after deproteinization, acidic amino acids were always in proportion higher than alkaline ones, but the balance between the two kinds of residues increased in favor of the latter with time. The kinetic study of the deproteinization clearly revealed a three-step mechanism with very different rate constants. The variation of these constants with temperature was used to calculate the energies of activation and the frequency factors of collision, thus allowing us to propose a new interpretation of the mechanism of deproteinization. Introduction The polysaccharide chitin corresponds to the structure polymer of the cuticles of all of the crustaceans and insects but also of most of the fungal cell walls. Chitin and chitosan, the deacetylated counterpart of chitin, have excellent biological properties: biocompatibility, biodegradability, nontoxicity, sorption properties, etc. They also have been identified as elicitors of biological responses in both mammals and plants.1-5 These specific properties involve the development of a new range of biotechnological applications, which require products with very high purity levels and, in some cases, relatively high molecular weights.5-7 In the native material, chitin is closely associated with proteins and minerals. The usual extraction procedure consists of a demineralization process followed by a deproteinization step. Classical demineralizations are carried out in acidic media, and the residual content in minerals can be detected by very sensitive methods such as atomic absorption spectrophotometry.8,9 The presence of minerals can be thus accurately evaluated down to the parts per million. The deproteinization process is more complex because we do not know the exact structure of the proteins, as well as the nature of their interactions with chitin. Moreover, no simple and precise technique exists for the detection of the remaining proteins. Classical deproteinization processes correspond to the hydrolysis of proteins by NaOH 1 M at various temperatures. In most studies, the amount of residual proteins is determined using the Kjeldahl method and elemental analysis involving many assumptions associated with large experimental errors. * To whom correspondence
[email protected].
should
be
addressed.
E-mail:
A few authors used amino acid analysis to determine the amount and the identity of the residual peptides or amino acids,10-14 and the results were used to compare several kinds of raw material or treatments. This technique involves the complete hydrolysis of the samples, and after reaction with ninhydrine, a specific reactive to free amine groups, both amino acids and glucosamine residues can be detected accurately. In this paper, we describe the use of amino acid analysis to quantify the amount of residual proteins and to characterize the composition of the proteins remaining associated with chitin as a function of the deproteinization time. The kinetics of deproteinization are also studied as a function of temperature, and the energies of activation, as well as the frequency factors of collision, are calculated and then discussed. Materials and Methods Raw Material and Preparation. Salted cephalothoraxes of Parapenaeopsis stylifera shrimps were obtained from France-Chitine. The manufacturer used salting to preserve the shells for a long time, especially during their transportation. Then, prior to use, the shrimp shells were washed thoroughly in distilled water until the conductivity of the rinsing water reached the value of pure water. The shells were freeze-dried and then cryoground under liquid nitrogen. The powder thus obtained was sieved, and the fraction below 80 µm was used hereafter. The shells were demineralized in HCl 1 M at ambient temperature with a solution-to-solid ratio of 10 mL/g for 24 h.15 The reaction media was filtered and the demineralized shrimp shell powder was washed to neutrality and freezedried. This demineralized shrimp shell powder was used as the starting material for the deproteinization.
10.1021/bm034115h CCC: $25.00 © 2003 American Chemical Society Published on Web 07/24/2003
Biomacromolecules, Vol. 4, No. 5, 2003 1381
Characterization of Shrimp Shell Deproteinization
Kinetics of Deproteinization. For the deproteinization step, NaOH 1 M was added to the demineralized powder with a solution-to-solid ratio of 15 mL/g. The deproteinization kinetics was monitored by several methods. First, it was simply followed by measuring the amount of released proteins. The absorbance at 280 nm, characteristic for tryptophan residues present in protein structures, was representative of the peptide concentration in the supernatant. By means of a calibration curve previously obtained through the use of the extracted proteins as a standard, we could calculate the exact protein concentration.15 Samples of 200 µL were collected as a function of time and then filtered and analyzed for the evaluation of the protein concentration in the supernatant. In the second part, the amount of proteins remaining in chitin was estimated by elemental and amino acid analyses. A representative sample was subtracted from the deproteinization reactor at 5, 20, 60, 180, 300, and 1440 min. Then, the partly deproteinized shrimp shell powder was removed in each sample by filtration, washed to neutrality, freezedried, and analyzed for the protein content. For elemental analysis, the percentage of proteins was calculated using the following equation.6,8,12,16 %P ) (%N - 6.9) × 6.25
(1)
where %P represents the percentage of proteins remaining in the obtained powder, %N represents the percentage of nitrogen measured by elemental analysis, 6.9 corresponds to the theoretical percentage of nitrogen in fully acetylated chitin (this value was adjusted as a function of DA, the degree of acetylation), and 6.25 corresponds to the theoretical percentage of nitrogen in proteins. The protein content in chitin was also measured by analyzing the amino acid content. Thus, 2 mg of sample was hydrolyzed with HCl 6 M in the presence of trifluoroacetic acid and thioglycolic acid (2 vol/1 vol/5%) for 45 min at 150 °C under vacuum. The samples were then solubilized in a buffer (Beckman), and an aliquot was used for analysis on Beckman equipment 6300 amino acid analyzer (column 120 mm × 4 mm). The amino acids were thus revealed by ninhydrine at 570 and 440 nm (for proline and hydroxyproline, respectively).10,12-14 The percentage of proteins was calculated from the total amino acid weight. To determine the limit of detection of this technique, we prepared standard mixtures with pure N-acetylglucosamine and bovine serum albumin (BSA) mixed in known proportions varying from 0% to 1% (w/w) of BSA. To obtain these mixtures, stock solutions of N-acetylglucosamine and BSA were prepared in water. Appropriate volumes of the two solutions were then mixed to achieve the desired composition, and these aqueous solutions were freeze-dried. The resulting powder was analyzed for the amino acid content. It has been observed that the hydrolysis of N-acetylglucosamine also generated residues that react with ninhydrine. These parasitic peaks appear on the chromatograms among the amino acid peaks. For both chitin and the model mixtures, the peak corresponding to glucosamine is superimposed with that of tyrosine; as a consequence, the tyrosine content is systematically considered as negligible (tyrosine represents only 4%
in the pure protein extract). No analysis for cysteine was obtained because of its destruction during the preparative hydrolysis. Results and Discussion One major problem for the utilization of natural polymers and then of chitin and chitosan in various industrial fields, especially those related to a contact with biological media, remains the possible presence of protein traces in the processed materials. The existing measuring methods for the determination of the nitrogen content in chitin samples generally call for the use of the Kjeldahl method and elemental analysis. For both methods, we assume that proteins contain 6.25% nitrogen, and we also take into account the degree of acetylation (DA) of the obtained chitin, which is measured with a nonnegligible error. All of these calculations imply an experimental error over 1% and thus a too high level of uncertainty for biomedical applications. In some cases, the prepared chitins are hydrolyzed with NaOH 2 M at 121 °C for 15 min.17 These conditions are supposed to be efficient enough to extract the residual proteins, which are further detected by the Lowry method using BSA as a standard. Once more, there is an important error in the obtained result because we can never be sure that we have removed all of the proteins from chitin and also because BSA is not a perfect reference for the shrimp proteins. As a consequence, we preferred to use the amino acid analysis to quantify both the amount of proteins remaining in chitin and their amino acid composition. In Figure 1, we can see an amino acid chromatogram characteristic for a nondeproteinized shrimp shell sample containing 15% (w/w) amino acids, initially demineralized in conditions previously described.15 As expected, chitin hydrolysis results in the production of several residues that react with ninhydrine and that are detected on the chromatogram. The most important peak corresponds to glucosamine and is superimposed with the tyrosine peak precluding the analysis of the latter. The large peak located at about 52 min corresponds to ammonia produced during hydrolysis.18 All of the peaks due to amino acids are well defined and can be easily integrated. In the first step, we validated the method by studying model mixtures made with known proportions of N-acetylglucosamine and protein solutions. BSA was used as a reference for the proteins. The first stage consisted in the complete hydrolysis of the mixture by acidic vapors. The same kind of chromatogram was obtained as that with the model mixture of BSA and N-acetylglucosamine, with the apparition of new parasite peaks in the second part of the chromatogram at times over 30 min corresponding to the hydrolysis products of Nacetylglucosamine. For this reason, only the first part of the chromatogram (including all of the amino acids eluted until alanine and corresponding to 50% (w/w) of the amino acids in pure BSA) was used in the calculation. To estimate the total weight of amino acids, the amount of amino acids detected before 30 min was multiplied by 2. The amino acid content was calculated for the model mixtures, and the results
1382
Biomacromolecules, Vol. 4, No. 5, 2003
Percot et al.
Figure 1. Amino acid analysis chromatogram of demineralized shrimp shell powder. The elution of the amino acids is represented as a function of the elution time for a detection at 570 nm.
Figure 2. Determination of the amino acid content in a model mixture of N-acetylglucosamine and BSA by amino acid analysis as a function of the theoretical ratios. The black line represents the theoretical results. The bars of error correspond to the standard deviation between the experimental results and the theoretical values.
are reported in Figure 2. Each experiment was carried out five times. We notice a good agreement between the measured and theoretical values. We were then able to determine a relatively precise protein content down to 0.25% ( 0.02% (w/w). For protein contents below this value, the presence of amino acids was still detected but the precision of the results was not better than 20%. This method was used to characterize our deproteinization kinetics. Thus, NaOH 1 M was added to a demineralized shrimp shell powder with a solution-to-solid ratio of 15 mL/g at ambient temperature. Representative samples of the mixture were collected at different deproteinization times. These partly deproteinized chitin samples were then washed and freeze-dried. The protein content of each sample was analyzed by both elemental and amino acid analyses. Figure 3 shows the decrease of the protein content as a function of time. All experiments were repeated three times. From a general point of view, the two detection methods gave consistent results. After 24 h, the protein content measured by elemental analysis was about 0.0% ( 1.0% and thus had to be considered as negligible. Therefore, the low precision
Figure 3. Kinetics of deproteinization in NaOH 1 M at ambient temperature. Variation of the protein content in the obtained chitin measured by elemental analysis (0) and amino acid analysis ([).
of this method did not allow a precise detection in the case of a low amount of residual proteins. However, the amino acid analysis revealed that 0.6% of proteins remained in the sample after 24 h of deproteinization. Moreover, the complete amino acid composition could be known as a function of time. This composition is shown in Table 1. In the native demineralized shrimp shell, glutamic acid, aspartic acid, and alanine are prevalent (more than 10%). Proline, glycine, arginine, valine, and serine represent between 6.9% and 8.1% of the overall composition,11,16,19 suggesting that proline, glycine, and alanine can play a role in determining and restricting the conformation. Both glycine and alanine have small side chains promoting the packing interaction of β-pleated sheets. The literature is relatively poor in information about the general organization of the chitin-protein lattices. In the mollusk shell nacreous layer,20 it has been observed that the core was a highly ordered β-chitin-protein complex, of which the proteins are rich in glycine and alanine and have a β-sheet structure. The surfaces of this core are coated with a layer of proteins rich in aspartic acid. The nucleation of the aragonite crystals takes place at a specific site on the surface, which is known to have unique calcium-binding properties. We can imagine a similar arrangement in the case of shrimp shells, the high ratio of acidic and basic amino acids playing crucial roles in the
Biomacromolecules, Vol. 4, No. 5, 2003 1383
Characterization of Shrimp Shell Deproteinization Table 1. Amino Acid Composition of the Partly Deproteinized Chitin as a Function of the Deproteinization Time Given as a Percentage (w/w) time of deproteinization (min) aspartic acid glutamic acid lysine histidine arginine threonine serine cysteine methionine glycine alanine valine leucine isoleucine proline tyrosine phenylalanine tryptophan
0
5
20
60
180
300
1440
12.3 17.1 3.1 2.8 7.7 4.3 6.9 b 1.3 8.0 10.0 7.0 3.2 2.6 8.1 b 5.9 b
11.0 16.6 3.3 2.6 7.0 4.7 7.0 b 1.6 8.8 10.7 7.3 2.9 2.5 8.2 b 5.8 b
10.7 17.2 3.1 2.6 8.02 4.6 6.7 b 1.5 8.3 11.0 7.5 2.4 2.5 8.3 b 5.7 b
10.7 17.2 3.1 2.9 8.3 4.8 6.5 b 1.7 8.5 11.3 8.7 1.8 2.4 7.0 b 5.1 b
9.7 16.5 3.1 3.3 8.8 4.3 5.8 b 0.8 9.2 11.2 7.5 2.6 2.3 10.3 b 3.9 b
9.3 15.7 3.8 1.16 9.3 4 5.7 b 0.9 9.7 10.8 7.8 2.2 2.2 10.1 b 7.4 b
9.1 14.0 5.2 3.7 11.51 3.4 5.3 b 1.2 10.0 10.1 7.8 0 1.8 9.4 b 7.3 b
a The values are the average of two experiments with an experimental error below 8% for each amino acid. b Not determined.
Figure 4. Variation of the polarity of the proteins embedded in the chitinous matrix before and after 24 h of deproteinization in NaOH 1 M at ambient temperature. Acidic amino acids include aspartic and glutamic acids; basic amino acids include lysine, histidine, and arginine; nonpolar amino acids include proline, alanine, valine, isoleucine, and leucine.
interaction with calcium carbonate and the chitinous matrix, respectively. Because we can estimate the amino acid composition for each deproteinization time, we can easily control the composition of the proteins remaining more strongly entrapped in the chitin structure. In Figure 4, we have represented the variation with time of the proportion of remaining amino acids, acidic amino acids corresponding to aspartic and glutamic acids, basic amino acids represented by lysine, histidine, and arginine, and nonpolar amino acids especially proline, alanine, valine, isoleucine, and leucine. All experiments were repeated two times. After 24 h of extraction with NaOH 1 M, although acidic amino acids were still in a higher amount compared to the others, their relative proportion had decreased. The sequence of the rate of peptide extraction, corresponding to acidic > hydrophobic > basic, allows us to propose two possibly connected propositions. (I) This sequence is directly related to a physical freedom of the proteins in the cuticles of shrimps (after demineral-
ization), increasing from acidic to neutral and basic proteins. (II) Basic amino acids are involved in a particular process of interaction with chitin chains limiting their extraction, and acidic and neutral amino acids are free or involved in lowenergy or reversible interactions (at alkaline pH) with chitin chains. We must also keep in mind that after the demineralization step, acidic proteins are more vulnerable in relation to the destruction of their interaction with CaCO3. If we may easily assume the presence of several kinds of proteins, the second proposition seems to play a major role. The presence of basic amino acids closely associated with polysaccharides has already been proposed in the case of marine organisms called diatoms.21 These unicellular algae are able to produce an external skeleton of hydrated silica called frustule surrounded by an organic coverage made of proteins and polysaccharides (including chitin). The role of this organic casing in biosilication is still unknown, but polycationic peptides have been detected. Their amino acid composition is particularly rich in lysine and arginine. The same arrangement between polysaccharides and basic peptides is observed in the case of highly deproteinized chitin, suggesting an important role of the free amine groups of these amino acids in the interaction with chitin. The possibility of a covalent linkage between chitin and proteins has been considered for a long time, but there is no convincing proof in favor of this interaction. Gottschalk22 suggested the formation of a covalent linkage corresponding to a glycosidic ester between a N-acetylated hexosamine entity and either the γ-carboxyl group of a glutamyl residue or the β-carboxyl function of an aspartyl unit. This kind of covalent bond has already been observed in globulin and ovalbumin.22,23 However, these glycosidic esters should be subjected to hydrolysis in mild alkali, and their presence could only partially explain the increase in the amount of the residual alkaline amino acids with time. Other types of stable linkages can exist, for example, via a carboxyl to the amino group of glucosamine. No more additional information could be obtained from the amino acid analysis on the nature of the link between chitin and proteins. Nevertheless, instead of covalent bonds, we can also propose the formation of numerous hydrogen-bonding and hydrophobic interactions involving, on one hand, the amino groups of alkaline amino acids and, on the other hand, the hydrophobic part of nonpolar amino acids. Indeed, these interactions can participate, alone or in association with other entities, in the protection of the concerned protein sequences regarding their alkaline hydrolysis. Kinetics of Deproteinization. The concentration of NaOH and temperature play an important role in deproteinization. However, in the range studied in our work, the influence of the solution-to-solid ratio was insignificant, provided that a minimal value of 5 mL/g was achieved to maintain a sufficient fluidity during deproteinization. In this paper, the NaOH concentration and the solution-to-solid ratio were fixed to 1 M and 15 mL/g, respectively, for all of the experiments, and only the temperature was varied from 16 to 70 °C. The amount of released peptides was deduced from the UV absorbance at 280 nm, and the kinetics was followed for each temperature during 8 days. In Figure 5, as expected,
1384
Biomacromolecules, Vol. 4, No. 5, 2003
Percot et al. Table 2. Variation of the Three Rate Constants and Their Coefficients of Regression of the Deproteinization Process as a Function of Temperature temp k1 (°C) (10-2 min-1) 16 27 50 70
0.68 ( 0.01 1.01 ( 0.02 2.10 ( 0.04 2.68 ( 0.05
r
k2 (10-3 min-1)
r
k3 (10-4 min-1)
r
0.99 0.98 0.99 0.93
1.87 ( 0.09 2.44 ( 0.12 3.12 ( 0.16 1.52 ( 0.08
0.99 0.99 0.99 0.98
1.38 ( 0.28 1.28 ( 0.26 0.48 ( 0.10 1.35 ( 0.27
0.84 0.95 0.61 0.94
Table 3. Energies of Activation (Ea) and Frequency Factors of Collision (A) Deduced from the Variation of the Rate Constants as a Function of Temperature in the Three Stages Defined in Figure 6 Figure 5. Kinetics of deproteinization in NaOH 1 M at various temperatures: variation of the protein concentration in the supernatant as a function of time at 16 (b), 27 (9), 50 (0), and 70 °C (O).
Figure 6. Logarithmic variation of the protein content in chitin as a function of the deproteinization time performed in NaOH 1 M at 70 °C.
we notice an increase of the protein concentration in the supernatants as a function of time and a decrease of the peptide release at decreasing temperatures. Whatever the temperature, we can conclude that after about 24 h, a plateau is reached (between 13 and 17 mg/mL). In Figure 6, the logarithmic variation of the percentage of proteins remaining in chitin is plotted as a function of the deproteinization time at 70 °C in NaOH 1 M. Whatever the temperature (figures not shown for the other temperatures), the deproteinization from shrimp shells appears to obey a first-order reaction kinetics in three stages. Changes in the reaction rates occur within 25-70 min and then 500-700 min. Therefore, in each domain, the deproteinization mechanism can be described by a first-order equation dP/dt ) -kP, where P represents the protein content remaining in the shrimp shell powder, t the treatment time, and k the reaction rate constant. In Table 2, the three rate constants are reported as a function of the deproteinization temperature. We can observe that they strongly decrease with time for all of the temperatures with the following sequence: k1 > k2 > k3. In the first part of the reaction, depending on temperature, the rate constants vary between 0.68 × 10-2 and 2.68 × 10-2 min-1. This result is in the same order as the result obtained by Chang and Tsai8 with shells of shrimp Solenocera melantho. After the first hour, the constant rates decrease considerably to achieve (1.52-3.12) × 10-3 min-1 in the second range
stages
Ea (kJ‚mol-1) A (min-1)
1
2
3
21 ( 2 55 ( 12
11 ( 2 2(2
0.00 ( 0.40 1(2
and (0.48-1.38) × 10-4 min-1 in the third one. We can imagine that this drastic decline in the deproteinization rate suggests an important change of mechanism during the deproteinization process. The rate can also be influenced by several parameters such as the location of the proteins in the chitinous matrix (on the surface of the particles, in the amorphous regions, or in the crystalline regions), the kind of interaction between chitin and proteins (hydrogen bonding, ionic interactions, hydrophobic interactions, or covalent bonding), and the composition of the proteins involved. Several studies on chitin/protein complexes of insect cuticles24 reveal the presence of several kinds of proteins associated with chitin. The changes that we observed in the amino acid ratios as a function of time confirm this result. Nevertheless, nobody knows the exact nature of the links between the carbohydrates and proteins. Gottschalk22 suggested that in the attachment of a carbohydrate to a peptide, almost any of the functional groups of the latter may be involved, resulting in a variety of modes of linkages and in the fact that these links can also be protected from hydrolysis by being embedded in a chitinous matrix just as the acetyl groups are relatively protected in the crystalline chitin. As proposed above, we may attribute an important role to the low-energy interactions between proteins and the polysaccharide chains possibly arranged in a complex manner with other substrates. The logarithmic plot representing the variations of the three rate constants as a function of temperature (not shown) was used to determine the energy of activation (Ea) and the frequency factor of collision for each range of time defined in Figure 6. Then, the energy of activation calculated for the first stage was found to equal 21 ( 2 kJ‚mol-1. For the second stage, the value was divided by almost a factor of 2, and this energy was then considered as near zero in the third stage (Table 3). Thus, the energies of activation decrease considerably from the first to the third stage. We could have expected the contrary because the protein extraction seems easier in the first period and concerns loosely bound proteins. This is probably due to the higher freedom of the concerned peptides and then to a lower stress on the peptide bonds in the
Characterization of Shrimp Shell Deproteinization
concerned proteins. In relation to the results of Table 1, we could attribute this first stage either to the peptide sequences essentially involved in electrostatic interactions with the few cationic sites present on the structure of chitin, which are destroyed in alkaline media,25 or to ester bonds between the amino acids and chitin also broken in this kind of media. The second step could be attributed to peptides involved in very low-energy interactions such as hydrophobic interactions with chitin and the last range to hydrogen bondings. In the last two cases, the decrease of the energy of activation has to be related to the increase of the stress induced by the formation of such interactions on the geometry of the peptides involved. We also calculated the frequency factors of collision of the rate constants (Table 3). We notice the same behavior as that for the energy of activation with an important decrease of these factors in relation to the succession of the three steps. As a consequence, the deproteinization is favored in the two last cases by an increase of the stresses on the bonds concerned by the hydrolysis, but the accessibility to these sites tends toward zero. Thus, the last peptide structures retained in the chitin matrix must be regarded as weakly accessible to water and NaOH and therefore highly protected from hydrolysis. These structures are certainly well protected against hydrolysis thanks to their hydrophobic domains preventing the diffusion of both water and sodium hydroxide. These structures should also be in ordered conformations in relation to the low values of the energies of activation. It is thus possible to propose lipoproteins as the last entities remaining in the system. Conclusion Amino acid analysis is a powerful technique to determine with precision the protein content in the presence of chitin down to 0.25%. It is still possible to detect proteins under this percentage but without any quantification. Amino acid analysis can be successfully used to monitor the deproteinization and to measure accurately the amount of peptides remaining even after 24 h of extraction. We observed that the amino acid composition of the proteins remaining in the chitin varies as a function of the deproteinization time. We could detect high initial percentages of basic and acidic amino acids, which probably play key roles in the interaction with chitin and minerals. After 24 h of treatment, the ratio of basic amino acids increases, suggesting their involvement in a typical situation. The study of the deproteinization kinetics reveals three levels of first-order reaction kinetics corresponding to three mechanisms. The first stage is attributed to loosely associated proteins in alkaline media, but the measured energy of activation is higher than that for the second stage attributed
Biomacromolecules, Vol. 4, No. 5, 2003 1385
to proteins more strongly attached or embedded in lessaccessible or crystalline domains of the chitinous matrix. In the third stage, after 10 h, the protein extraction carries on very slowly, but both the energy of activation and the frequency factor of collision become close to zero. This is attributed to an important decrease in the freedom and the accessibility of the chain conformation, which is probably due to the formation of a dense network of hydrogen bonding in structures of ordered lipoproteins. It still remains to really prove a possible direct chemical interaction between chitin and proteins, especially in the case of covalent bonds. Acknowledgment. This work belongs to the CARAPAX project financially supported by the EC through the 5th PCRD. We also acknowledge the fruitful help of Dr. Daniel Herbage (IBCP, Lyon, France). References and Notes (1) Usami, Y.; Okamoto, Y.; Takayama, T.; Shigemasa, Y.; Minami, S. J. Biomed. Mater. Res. 1998, 42 517. (2) Lu, J. X.; Prudhommeaux, F.; Meunier, A.; Sedel, L.; Guillemin, G. Biomaterials 1999, 20 (20), 1937. (3) Cho, Y. W.; Cho, Y. N.; Chung, S. H.; Yoo, G.; Ko, S. W. Biomaterials 1999, 20 (22), 2139. (4) Tucci, M. G.; Riccoti, G.; Mattioni-Belmonte, M.; Gabbanelli, F.; Lucarni, G.; Orlando, F.; Viticchi, C.; Bigi, A.; Panzavolta, S.; Roveri, N.; Morganti, G.; Muzzareli, R. A. A. J. Bioact. Compat. Polym. 2001, 16 (2), 145. (5) Vander, P.; Vårum, K. M.; Domard, A.; El Gueddari, N. E.; Moerschbacher, B. M. Plant Physiol. 1998, 118 (4), 1353. (6) Kumar, M. N. V. R. React. Funct. Polym. 2000, 46 (1), 1. (7) Hon, D. N.-S. In Polysaccharides in Medicinal applications; Dumitriu, S., Ed.; Marcel Dekker: New York, 1996; p 631. (8) Chang, K. L. B.; Tsai, G. J. Agric. Food Chem. 1997, 45, 1900. (9) Chen, R. H.; Yang, M. H. J. Fish. Soc. Taiwan 1994, 21 (3), 293. (10) Gildberg, A.; Stenberg, E. Process Biochem. 2001, 36 (8-9), 809. (11) Hunt, S.; Nixon, M. Comp. Biochem. Physiol. 1981, 68B, 535. (12) Cremades, O.; Ponce, E.; Corpas, R.; Gutierrez, J. F.; Jover, M.; Alvarez-Ossorio, M. C.; Parrado, J.; Bautista, J. J. Agric. Food Chem. 2001, 49 (11), 5468. (13) Brine, C. J.; Austin, P. R. Comp. Biochem. Physiol. 1981, 70B, 173. (14) Shahidi, F.; Synowiecki, J. J. Agric. Food Chem. 1991, 39 (8), 1527. (15) Percot, A.; Viton, C.; Domard, A. Biomacromolecules 2003, 4 (1), 12. (16) No, H. K.; Meyers, S. P.; Lee, K. S. J. Agric. Food Chem. 1989, 37 (3), 575. (17) Ito, M.; Hidaka, Y.; Nakajima, M.; Yagasaki, H.; Kafrawy, A. H. J. Biomed. Mater. Res. 1999, 45 (3), 204. (18) Rosset, R.; Caude, M.; Jardy, A. Chromatographie en phase liquide et supercritique; Masson Ed.: Paris, Milan, Barcelone, Bonn, 1991. (19) Welinder, B. S. Comp. Biochem. Physiol. 1974, 47 (2), 779. (20) Addadi, L.; Weiner, S. Angew. Chem., Int. Ed. Engl. 1992, 31, 153. (21) Perry, C. C.; Keeling-Tucker, T. J. Biol. Inorg. Chem. 2000, 5 (5), 537. (22) Gottschalk, A.; Murphy, W. H.; Graham, E. R. B. Nature 1962, 194, 1051. (23) Gottschalk, A. Nature 1960, 186, 949. (24) Rudall, K. M. AdV. Insect Physiol. 1963, 1, 257. (25) Sorlier, P.; Denuzie`re, A.; Viton, C.; Domard, A. Biomacromolecules 2001, 2 (3), 765.
BM034115H
