Biomacromolecules 2003, 4, 1686-1690
1686
Biospecific Fractionation of Chitosan Chiye Sasaki,† Are Kristiansen,‡ Tamo Fukamizo,† and Kjell M. Vårum*,§ Laboratory of Enzyme System Science, Department of Food and Nutrition, Kinki University, 3327-204 Nakamachi, Nara 631-8505 Japan, FMC BioPolymer, Gaustadalleen 21, N-0349 Oslo, Norway, and NOBIPOL, Department of Biotechnology, Norwegian University of Science and Technology, Sem. Saelands vei 6/8 N-7491 Trondheim, Norway Received April 25, 2003; Revised Manuscript Received July 8, 2003
We have previously reported that, although a fully de-N-acetylated chitosan does not bind to hen egg white lysozyme, chitosans with a low fraction of N-acetylated units (FA) bind biospecifically to lysozyme with an affinity strongly dependent upon pH and ionic strength and without concomitant cleavage of glycosidic linkages. In this study, we report on the fractionation of a low FA chitosan with low molecular weight by biospecific adsorption of the chitosan molecules containing N-acetyl groups to immobilized lysozyme. Hen egg white lysozyme was immobilized to CNBr-activated Sepharose 4B, and a chitosan with a fraction of N-acetylated units of 0.045 and a weight average degree of polymerization (DPW) of 22 was applied to the column at suitable conditions for biospecific binding (pH 5.7, 0.15 M NaCl). The chitosan could be separated into two fractions, one that was not adsorbed to the lysozyme-column and one that was adsorbed and could be eluted by decreasing the pH and the ionic strength (0.08M acetic acid of pH 3.0). The fractions were analyzed and the fraction that was not adsorbed was found to be fully de-N-acetylated chitosan with a DPw of 18, whereas the fraction that was adsorbed was a chitosan with FA of 0.080 and DPW of 24. Experimental data were compared with data from theoretical calculations, which was used to calculate the fraction of chitosan molecules with and without acetyl groups, showing good correlation between experimental and theoretical results. Introduction Chitosan is a family of linear copolymers of 2-acetamido2-deoxy-β-D-glucopyranose (GlcNAc; A unit) and 2-amino2-deoxy-β-D-glucopyranose (GlcN; D unit) connected by (1 f 4) linkages, which can be distinguished from chitin by its solubility in dilute aqueous acid solutions.1 It has previously been found that the sequential arrangement of A and D units in the polymer chain is in accordance with Bernoullian statistics.2,3 Lysozyme (EC 3.2.1.17) may in addition to its natural substrate, the bacterial cell wall polysaccharide composed of alternating residues of (1 f 4) linked GlcNAc and N-acetyl muramic acid, also hydrolyze partially N-acetylated chitosans.4-6 Moreover, previous results have shown that a chitosan molecule with a very low degree of N-acetylation bind to lysozyme without depolymerization of the bound chitosan.7,8 The binding of partially N-acetylated chitosans to lysozyme has also been investigated in quantitative terms, and the dependence of these interactions upon the chemical composition of chitosans has been interpreted with respect to active site cleft properties of lysozyme.8,9 The striking feature of all of these studies is that the acetyl groups of the chitosan * To whom correspondence should be addressed.
[email protected]. † Kinki University. ‡ FMC BioPolymer. § Norwegian University of Science and Technology.
E-mail:
molecule are essential for biospecific binding to lysozyme, with an affinity between ligand and enzyme strongly dependent upon pH and ionic strength.7-9 Dissociation constants for partially N-acetylated chitosans and lysozyme have been determined to 10-5 M (at pH ) 5.5, ionic strength of 0.15 M, and 308 K),9 which should be in the appropriate range for interactions that can be used for successful affinity chromatography. This has encouraged us to investigate the possibility of immobilizing lysozyme in order to separate chitosan molecules containing acetyl groups from those without acetyl groups. We here present results showing for the first time that immobilized lysozyme can successfully be used to biospecifically separate a low molecular weight chitosan with a low fraction of N-acetylated units (FA) into a fully de-N-acetylated fraction and a fraction containing chitosan molecules with acetyl groups. This last fraction will be of interest in specific binding studies between chitosan and proteins where the acetyl group is essential, and as inhibitors for, e.g., lysozymes and chitinases. Moreover, the lysozyme column may be used for separation of partially N-acetylated chitosan oligomers. Experimental Section Preparation of Lysozyme-Column. Hen egg white lysozyme (Sigma L-6876) was coupled to CNBr-activated Sepharose 4B (Pharmacia Biotech) through an amino coupling reaction, and a glycine-NaOH solution was added to the gel to block the excess remaining active groups on the
10.1021/bm034124q CCC: $25.00 © 2003 American Chemical Society Published on Web 08/23/2003
Biospecific Fractionation of Chitosan
gel, using the experimental procedure specified by the manufacturer of the column. A column containing the gel with the immobilized lysozyme was prepared (1.5 × 14 cm). To determine the amount of lysozyme coupled to the gel, the supernatant containing lysozyme that was not immobilized to the gel was recovered, and the lysozyme content was measured by the absorbance at 280 nm. Furthermore, to check if the lysozyme bound to the gel is bio-active, N-acetylglucosamine pentamer [(GlcNAc)5] (Seikagaku Corporation) was applied onto the lysozyme column, and the size distribution of the oligomers (pentamer to monomer) in the eluate was analyzed by high performance gel-filtration using a TSK-GEL G2000PW (Tosoh). Preparation of Highly De-N-Acetylated Chitosan. Chitosan with a low fraction of acetylated units of 0.05 was prepared by homogeneous de-N-acetylation of chitin.10 Depolymerization of the chitosan was performed by nitrous acid with subsequent conventional reduction with sodiumborohydride.11 The chitosan was subsequently dialyzed against 0.2 M sodium chloride followed by dialysis against distilled water to convert the chitosan to the chloride salt. These chitosan salts are fully soluble both in water and in dilute acid. The FA of this chitosan was 0.045 as determined from the 1H NMR spectrum,2 and the weight average degree of polymerization (DPw) was 22 as calculated from the sizeexclusion chromatogram. Fractionation of Chitosan. Chitosan with a FA of 0.045 and a DPw of 22 was applied onto the lysozyme column at suitable conditions for biospecific binding (binding buffer: 0.02 M sodium acetate, pH 5.7, containing 0.15 M NaCl) at a flow rate of 0.4 mL/min, and the adsorbed fraction was eluted by decreasing the pH and the ionic strength (0.08 M acetic acid, pH 3.0). The fractions were collected (1 mL/ tube). The amount of chitosan in the fractions eluted from the lysozyme-column was determined by the ninhydrin method.12 Size-Exclusion Chromatography. Each fraction of chitosan was analyzed by gel filtration on two columns (2.5 × 100 cm) of Superdex 30, prep. grade (Pharmacia Biotech) connected in series as previously described.11 The distribution coefficient (Kav) of an oligomer with a given chain length was calculated from the equation Kav ) (Ve - V0)/(Vt - V0) where Ve is the peak elution volume of the oligomer, V0 is the void volume of the column (estimated from the peak elution volume of a high molecular weight chitosan), and Vt is the total volume of the column (estimated from the highest elution volume of methanol). The fully de-N-acetylated hexamer (GlcN)6 (Seikagaku Corporation) was used as an internal standard which was added to the chitosan samples in order to identify the elution volume of the hexamer. A linear relation between the logarithm of the molecular weight of oligomers with DP < 20 and Kav was obtained and was extended to calculate the molecular weight of the low molecular weight chitosans. The polydispersity index (PI) of the nonfractionated chitosan was determined to 1.3. This rather low PI value (as compared to the PI value of 2 which is characteristic of a linear polymer which has been subjected
Biomacromolecules, Vol. 4, No. 6, 2003 1687
to random depolymerization) can be explained by assuming that the low-molecular weight tail has been removed by the dialysis step included in the preparation of the low-molecular weight chitosan. 1 H NMR Analysis of the Chitosans. All samples were dissolved in D2O and transferred to 5-mm NMR tubes. 1H NMR measurements were performed at 90 °C on a Bruker Avance DPX spectrometer at 300 or 400 MHz. Typically, 30° pulse angles with a repetition time of 6 s were used. Chemical shifts were determined relative to internal sodium3-(trimethylsilyl)propionate-d4 (TMSP). For determination of the FA of chitosan samples with a very low FA (less than 0.01), the integration of three acetyl-protons at 2.04 ppm and the H1/H2 protons of a deacetylated unit was used. In addition, the number of scans when obtaining the spectrum were increased in order to improve the signal-to-noise ratio, allowing a more precise determination of such low FA values. Results and Discussion Immobilization of Lysozyme. Lysozyme was effectively immobilized on CNBr-activated Sepharose gel, as more than 90% of the lysozyme was found to be coupled with the gel, amounting to 0.94 µmol lysozyme/mL of gel. To test if the immobilized lysozyme was bioactive, the fully N-acetylated chito-pentasaccharide (GlcNAc)5 was applied onto the lysozyme column, and the oligosaccharide composition of the eluate was analyzed by HPLC. It was found that (GlcNAc)5 was hydrolyzed to (GlcNAc)4, (GlcNAc)3, (GlcNAc)2, and GlcNAc (data not shown). The distribution of the oligosaccharides in the eluate from the column was typical for the hen egg white lysozyme reaction,13 showing that the initial substrate (GlcNAc)5 was hydrolyzed by immobilized lysozyme and that at least a fraction of the immobilized lysozyme was bioactive. There are eleven arginine residues and six lysine residues in hen egg white lysozyme,14 which are all located on the surface of the native enzyme. These basic amino acids could react with cyanate ester groups in CNBr-activated Sepharose. However, only one arginine residue, Arg114, is located at the lower endmost site (subsite F) in the lysozyme binding cleft. Although immobilization of lysozyme to the Sepharose gel through Arg114 may result in an inactive enzyme, it is likely that the immobilization of lysozyme through any of the other basic amino acids would result in an immobilized enzyme that is bioactive. Thus, it seems likely that most of the immobilized lysozyme molecules are bound to the gel in such a way that the active site is available for binding to substrate molecules. Fractionation of Chitosan on the Lysozyme Column. A low N-acetylated and low molecular weight chitosan (FA ) 0.045, DPw ) 22) was applied onto the lysozyme-column, and the column was eluted with five column volumes of a buffer favoring biospecific binding between lysozyme and chitosan (pH 5.7, ionic strength 0.15 M) and subsequently with a buffer which is unfavorable for biospecific binding, i.e., with a pH of 3 and low ionic strength.7 First, the capacity of the lysozyme column was evaluated by applying increasing amounts of chitosan to the lysozyme column (from 5
1688
Biomacromolecules, Vol. 4, No. 6, 2003
Sasaki et al.
Table 1. Effect of the Amount of Chitosan Applied to the Lysozyme Column on the Amount of Chitosan Eluted in the Nonadsorbed Fraction (Using the Ninhydrin Method12) amount of chitosan (mg)
% eluted in nonadsorbed fraction
5 10 15 20 30
47 51 48 54 86
Figure 1. Elution profile of the chitosan from the lysozyme-coupled Sepharose 4B column.
mg to 30 mg). The amount of chitosan that was eluted in the nonadsorbed and the adsorbed fraction was determined using the ninhydrin method.12 The results are given in Table 1, and it is seen that about equal amounts of chitosan are eluted in the nonadsorbed and the adsorbed fraction when less than 20 mg of chitosan is applied to the column, whereas 86% of the chitosan is eluted in the nonadsorbed fraction when 30 mg of chitosan is applied. This is related to capacity of the column (see later discussion). On the other hand, when a fully de-N-acetylated chitosan (FA < 0.0002) was applied onto the lysozyme column, almost 100% of the chitosan applied was eluted in the nonadsorbed fraction. This indicates that the chitosan binding to the lysozyme column is dependent upon N-acetylated residues in the chitosan, consistent with the binding data reported previously.8 Chemical Composition of the Fractionated Chitosan. An example of a chromatogram from the lysozyme column is shown in Figure 1. In the shown example, 20 mg of the chitosan was applied onto the column. The nonadsorbed (fraction 25-39) and the adsorbed fractions were pooled (fraction 102-120) as indicated in Figure 1. The adsorbed and the nonadsorbed chitosan fractions were characterized by 1H NMR analysis and size-exclusion chromatography. Figure 2 shows the anomer region of the 1 H NMR spectrum of the nonfractionated chitosan, the nonadsorbed, and the adsorbed fraction. From the spectra,
Figure 2. Anomer region of the proton NMR spectra of nonfractionated chitosan (A), nonadsorbed fraction (B), and adsorbed fraction (C).
it can be seen that the H-1 resonance from internal A units (at 4.66 ppm) was absent in the spectrum of the nonadsorbed fraction, whereas the same resonance in the spectrum of the adsorbed fraction was increased relative to the nonfractionated chitosan. Thus, it is evident that the lysozyme column has effectively separated the nonfractionated chitosan according to the specific interactions between A units and lysozyme.8 An experiment (see previous section), where a fully de-Nacetylated chitosan (FA < 0.0002) was applied to the column and was eluted only in the nonadsorbed fraction, excludes the possibility that fully de-N-acetylated chitosan molecules are coeluted with the adsorbed fraction. From the 1H NMR spectra, the FA of nonadsorbed and adsorbed fractions were calculated to be
