Galactose-Substituted Alginate: Preliminary Characterization and

Abstract. Coupling of alginate with 1-amino-1-deoxygalactose in the presence of .... polymannuronate, 0, 1, 0, 0, 1, 0 ...... Biomacromolecules 2005 6...
0 downloads 0 Views 114KB Size
Download PDF
Biomacromolecules 2003, 4, 624-631

624

Galactose-Substituted Alginate: Preliminary Characterization and Study of Gelling Properties Ivan Donati,* Amedeo Vetere, Amelia Gamini, Gudmund Skja˚k-Bræk,† Anna Coslovi, Cristiana Campa, and Sergio Paoletti Department of Biochemistry, Biophysics and Macromolecular Chemistry, University of Trieste, Via L. Giorgieri 1, I-34127 Trieste, Italy Received October 23, 2002

Coupling of alginate with 1-amino-1-deoxygalactose in the presence of 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide results in a substituted polymer containing galactose side linked via an amide bond. To clarify the degree and pattern of substitution, a 1H NMR study on the anomeric region of modified alginate, polymannuronate, alginate enriched in guluronic acid (G-enriched alginate), and polyalternating MG, was carried out (G, R-L-guluronic acid; M, β-D-mannuronic acid). From the resonance of the proton at position 1 of galactosylamine, it was possible to determine the amount of galactose linked to mannuronic and to guluronic residues, respectively. Furthermore, 1H NMR spectroscopy revealed a higher reactivity of guluronic residues for low degrees of conversion. Modified alginates with 7% and 19% of substitution are both able to form stable beads in the presence of calcium ions. The effect of galactose substitution on the dimensions, swelling, and stability of the beads has been studied and the cytotoxicity of the modified polymer evaluated in preliminary biological tests. Introduction Alginates are a family of polysaccharides produced by brown algae1 and bacteria.2,3 They are copolymers of 1f4 linked β-D-mannuronic acid (M) and R-L-guluronic acid (G) in which the fraction and the sequence of the two monomers varies over a wide range according to the species or the tissue they are isolated from. The rapid gel formation in the presence of calcium ions, at room temperature and neutral pH, allows alginate to be a very versatile material for biotechnological and medical applications,4-6 in particular in the field of cell and tissue encapsulation.7 In recent years many researchers have shown that polymeric biomaterials with covalent coupling of cell-specific ligands or extracellular signaling molecules (the synthetic Extracellular Matrix (ECM)) exhibit peculiar advantages, e.g., both by enhancing the embedding of cells within the biomaterial and by controlling their growth, differentiation, and behavior in culture.8,9 As an example, Mooney and coworkers introduced an RGD-containing cell adhesion ligand on alginate. This modified material has been demonstrated to provide for the adhesion, proliferation, and expression of differentiated phenotype of skeletal muscle cells.10 It therefore seemed very challenging to introduce such cellspecific ligands on alginate in order to use it as a threedimensional synthetic ECM for immobilization and encapsulation of cells in tissue engineering.11,12 It is nevertheless important to notice that modifications of the polymer chain usually affect some important functional properties of alginate gel beads, such as cooperative binding of cations, * Corresponding author. Tel: +39 040 558 3692. Fax: +39 040 558 3691. E-mail address: [email protected]. † On leave of absence from: Department of Biotechnology, Norwegian University of Science and Technology, Sem Sælands vei 6-8, N-7491 Trondheim, Norway.

osmotic swelling, chemical stability, transparency, and biological activity.13-15 In the present paper, alginate from brown algae has been modified, exploiting the carbodiimide chemistry, introducing 1-amino-1-deoxy-β-D-galactose via an amide bond with the carboxylic group on the polymer. A receptor recognized by an asialoglycoprotein is localized on the hepatocyte cell membrane,16 and it was shown to bind and internalize glycoproteins with terminal β-galactosyl residue. ASGPR (asialoglycoprotein receptor) was the first lectin to be identified in animals. In the normal liver, each hepatocyte performs hundreds of metabolic activities. However, an isolated hepatocyte outside the organism rapidly loses its cellular functions and has only a limited viability. Specific requirements to enhance the viability of hepatocytes have already been addressed. In the first place, hepatocytes need to anchor and adhere to a matrix surface. Furthermore, encapsulation of hepatocytes17 into alginate capsules could produce a very high density cell culture system with mechanical support as well as immunoprotection in the event of implantation.18 An alginate bearing galactose residues and with suitable gel-forming properties could then be proposed as an engineered biomaterial to improve encapsulation and adhesion of hepatocytes.19,20 The main structural features of the galactose-substituted alginate were studied, its calcium bead-forming ability, dimensions, stability, swelling of the beads investigated, and cytotoxicity evaluated in a preliminary biological test. Materials and Methods A commercial sample of sodium alginate isolated from Laminaria hyperborea stipe LF 10/60 was provided by Protan A/S (Norway). A homopolymeric mannuronan was produced from an epimerase negative mutant of Pseudomans

10.1021/bm020114y CCC: $25.00 © 2003 American Chemical Society Published on Web 02/26/2003

Biomacromolecules, Vol. 4, No. 3, 2003 625

Galactose-Substituted Alginate Table 1. Composition of the Polymers Used As Determined by 1H NMRa polymer

FG

FM

FGG

FGM+MG

FMM

NG>121

alginate polymannuronate G-enriched alginate polyalternating MG

0.69 0 0.92 0.47

0.31 1 0.08 0.53

0.56 0 0.86 0

0.26 0 0.12 0.94

0.18 1 0.02 0.06

14 0 ∼33 0

a F denotes the proportion of alginate that consists of guluronic acid. G FGG indicates the proportion of alginate consisting of guluronic acid in blocks of dimers. FMM indicates the proportion of alginate consisting of mannuronic acid in blocks of dimers. FGM+MG indicates the proportion of alginate consisting of mixed sequences of guluronic acid and mannuronic acid. NG>1 indicates the average length of guluronic acid blocks.

aeruginosa as described by Ertesva˚g and Skja˚k-Bræk.22 An alginate enriched in guluronic acid (G-enriched alginate) was prepared from Laminaria hyperborea outer cortex after fractionation,23 while the polyalternating MG was produced by in vitro epimerization of mannuronan with the recombinant mannuronan C-5 epimerase AlgE4.24 The main compositional features of the used samples are reported in Table 1. 1-Ethyl-3-[3-(dimethylamino)propyl]carbodiimide hydrochloride (EDC), calcium chloride, and 3-(4,5-dimethylthiazol-2-yl)-2,5-tetrazolium bromide (MTT) were purchased from Aldrich Chemical Co. (Milwaukee, WI). N-Hydroxysuccinimide (NHS) and 2-[N-morpholino]ethanesulfonic acid were purchased from Sigma Chemical Co. (St. Louis, MO). The pig endothelial cell line PK15 was from ATCC (Rockville, MA). Dulbecco’s modified Eagle’s medium (DMEM) and streptomycin/ampicillin solution and fetal calf serum were from GIBCO-BRL (Grand Island, NY). Multiwell plates (96 wells) for cell culture were from Corning (Corning, NY). Synthesis of Galactosylamine. 1-Amino-1-deoxy-β-Dgalactose was synthesized as reported elsewhere.25,26 Synthesis of Galactose-Substituted Polymers. 1-Amino1-deoxy-β-D-galactose was added to a stirred solution of the sodium form of the polymer (300 mg) in 0.2 M MES buffer (pH 4.5, 80 mL) containing NHS and EDC (molar ratio EDC/ COOH ) 2.5; molar ratio NHS/EDC ) 0.2). The solution was stirred overnight at room temperature, the pH was adjusted to 8.5, the polymer was extensively dialyzed against deionized water and freeze-dried. Table 2 reports the amount of galactosylamine used for each polymer, with the list of products (samples 1 to 12). Elemental Analysis. Determination of C, H, and N content was performed on the Na+-form of samples 3 and 8. A derivatization of 24% for sample 2 and 39% for sample 8, with respect to the monomer unit of alginate, was found. 1 H NMR Spectroscopy. Samples were prepared as described by Grasdalen et al.27 The 1H NMR spectra were recorded at 90 °C with Bruker WM 300 and WM 400 instruments. The chemical shifts are expressed in parts per million downfield from the signal for 3-(trimethylsilyl)propanesulfonate. Preparation of Calcium Beads. The alginate beads were prepared by allowing droplets of aqueous 2% solution of the sodium form of samples 1, 2, or 3, respectively, to fall into 50 mM calcium chloride solution. The size of the droplets was controlled by applying either a coaxial air

Table 2. Reaction between 1-Amino-1-deoxy-β-D-galactose and Alginate, Polymannuronate, Polyguluronate, and Polyalternating MG sample

polymer

1 2 3 4 5 6 7 8 9 10 11 12

alginate alginate alginate alginate alginate alginate polymannuronate polymannuronate G-enriched alginate G-enriched alginate polyalternating MG polyalternating MG

% reaction Gal-1-NH2a derivatizationb conditionsc 0 0.2 0.5 1.5 1 2 0 6 0 6 0 6

0 7 19 36 9 0 0 42 0 25 0 70

/ s, pH 4.5 s, pH 4.5 s, pH 4.5 b,d pH 4.5 s, pH 6.5 / s, pH 4.5 / s, pH 4.5 / s, pH 4.5

a Amount of 1-amino-1-deoxy-galactose expressed as equivalents used per repeating unit. b Total derivatization calculated from 1H NMR. c Key: s, reaction carried out on polymer in solution; b, reaction carried out on polymer beads. d Size of the droplet not controlled by coaxial air stream. Mean diameter of the beads is 2 mm.

stream28 or an electrostatic bead generator,29 as previously described. The beads were kept in 50 mM calcium chloride for 30 min before being used. The diameters (798, 838, and 880 µm for samples 1, 2, and 3, respectively) were measured (n ) 12) by means of a light microscope (Leitz). Swelling in Water. Freshly prepared calcium alginate beads of samples 1, 2, and 3, respectively, were sequentially washed with water and then with solutions of increasing concentration of ethanol (20%, 40%, 60%, 80%, and 100%) and air-dried overnight. The diameters (348, 298, and 240 µm for samples 1, 2, and 3, respectively) were measured (n ) 12) by means of a light microscope (Leitz). The dry beads were then suspended in distilled water and their increase in diameter was monitored against time. The experiment was repeated in triplicate for each sample. Measurement of Dimensions. The dimensional stability of calcium alginate beads of samples 1, 2, and 3, respectively, was measured when 1/2 mL of gel beads was added to 3 mL of 0.9% NaCl solution (saline). The sample was stirred for 1 h. The saline solution was changed several times, and the diameter of the capsules was determined (n ) 12) before each change. Cytotoxicity. PK15 cell culture was performed until confluence in DMEM supplemented with 10% fetal calf serum, 2 mM L-glutamine, 100 U/mL of penicillin, and 100 µL/mL of streptomicyn at 37 °C in a 5% CO2 atmosphere. This cell line was used in cytotoxicity test of samples 1, 2, and 3. Cells were seeded at a density of 2.0 × 104 cells/well into 96-well microtiter plates and incubated for 24 h. The media were removed and 100 µL portions of 0.2 µm filtered solutions of samples 1, 2, and 3 at a concentration of 200 µg/mL were added. Untreated cells in media were used as positive reference. The MTT 30 assay was used to determine cell viability. Cells were incubated with the solutions of the polymers for 24 h before the addition of MTT (20 µg/well) at a concentration of 5 mg/mL in phosphate-buffered saline (PBS). After a further incubation period of 4 h, MTT was removed, the cells were washed with PBS, and DMSO (100 µL) was added. After 1 h the absorbance was read at 540 nm using a Dynatech MR5000 plate reader.

626

Biomacromolecules, Vol. 4, No. 3, 2003

Donati et al.

Figure 1. FT-IR on KBr disk of sample 2.

Results and Discussion a. Preparation and Characterization. a1. Derivatization of Polymers with 1-Amino-1-deoxygalactose. The coupling between polymers and 1-amino-1-deoxy-D-galactose, obtained as pure β anomer,26 was carried out in solution at pH 4.5 for samples 2, 3, 4, 8, 10, and 12 as reported in Table 2. Although galactosylamine tends to hydrolyze giving rise to galactose,25,26 the high rate of the reaction between the EDCNHS activated carboxylic group and the amino moiety leads to a partial derivatization. The low value of pH (4.5) was selected in order to prevent, as much as possible, the hydrolysis of galactosylamine.25 In the case of alginate it can be emphasized that by increasing the equivalents of galactosylamine from 0.2 to 1.5, the degree of derivatization increased from 7% to 36% (samples 2 and 4). The rate of the process is dramatically decreased at higher pH values, as evidenced for sample 6 where no derivatization occurred. The lack of conjugation could stem from the higher rate of hydrolysis of 1-amino-1-deoxy-galactose at neutral conditions. The reaction can also be performed on alginate beads (sample 5) at pH 4.5, but a low degree of derivatization (9%) is achieved with 1 equiv of galactosylamine. The absence of ester bonds (∼1700 cm-1) in the dialyzed polymers was checked by FT-IR spectroscopy on sample 2 (Figure 1). a2. 1H NMR Spectroscopy. We attempted to determine both extent and location of galactose-amide substitution by analyzing the anomeric region of the 1H NMR spectra of the alginate derivatives reported in Figure 2, making reference to the reported 1H NMR study on native alginate.31 On one hand the spectra of Figure 2 indicate that the derivatization does not alter the proton resonances of the alginate backbone. Indeed, the abundance values of both G and GG fractions, easily calculated by established procedures,31 remain unaffected in samples 1-4. On the other hand, spectra of Figure 2 show that on increasing the amount of galacto-

sylamine used for the derivatization, an increasing number of peaks (namely, A, B, and C) becomes detectable. To assign the proton resonances corresponding to signals A, B, and C, 1H spectra of polymannuronate, G-enriched alginate, and polyalternating MG, as well as their galactose-amide derivatives, were recorded (Figure 3a-c). In this respect pure (or very nearly such) homo- and copolymers were used as reference polymers, taken as mimicking separately the MM, GG, and MG-GM sequences, randomly occurring within native alginate chains. With respect to the underivatized polymer (sample 7), the spectrum of the galactose-modified polymannuronate (sample 8) displays two additional weak peaks located at around 4.75 and 4.80 ppm respectively, arising from the H-1 of a galactosylamine linked to a MM sequences (Figure 3a). It is important to underline that the sum of the integrated intensities of those two peaks (42%) is in very good agreement with the total degree of substitution as determined by elemental analysis (39%). The spectrum (Figure 3b) of the substituted polyalternating MG (sample 12), which is different from that of the underivatized sample (sample 11), discloses the presence of both signal A and C, already found in Figure 2 (i.e., signal A located at ∼5.15 ppm and signal C at ∼4.9 ppm). Surprisingly enough, the spectrum (Figure 3c) of the galactose-modified G-enriched alginate (sample 10) showed also the additional peak B at ∼4.6 ppm, beside peak A at ∼5.15 ppm. By comparison of all the spectra of Figure 3, it can be deduced that peak A should correspond to the H-1 of galactosylamine linked to G residues. Being present whenever G residues are present, it should also give the total amount of galactosylamine linked to G residues. Conversely, peak C must correspond to the H-1 of galactosylamine linked to an M residue neighboring a G residue. The area of peak C should then monitor the total amount of galactose-amide linked to an M residue in an alternating sequence. The origin of signal B appears more uncertain; it

Biomacromolecules, Vol. 4, No. 3, 2003 627

Galactose-Substituted Alginate

Figure 2. 300 MHz 1H NMR spectra (anomeric region) of native (sample 1) and modified alginates (samples 2, 3, and 4) with degrees of derivatization of 7%, 19%, and 36%, respectively. Table 3. Extent and Pattern of Derivatization for Modified Alginate, Polymannuronate, Polyguluronate, and Polyalternating MG Determined by 1H NMR

sample

total derivatization (%)

derivatization on G (%)

derivatization on M (%)

Gal-G-(M) (%)a

Gal-G-(G) (%)b

Gal-M-(G) (%)c

Gal-M-(M) (%)d

2 3 4 5 8 10 12

7 19 36 9 42 25 70

7 19 29 9 0 25 37

n.d.e n.d. 7 n.d. 42 n.d. 33

7 7.2 7.3 5.3 0 8 37

n.d. 11.8 21.7 3.7 0 17 0

n.d. n.d. 7 n.d. 0 n.d. 33

n.d. n.d. n.d. n.d. 42 n.d. n.d.

a Derivatization on a G residue neighboring an M residue. b Derivatization on a G residue neighboring a G residue. c Derivatization on an M residue neighboring a G residue. d Derivatization on an M residue neighboring an M residue. e n.d. ) not determined (below the detection limit).

is present only when GG diads are present. Tentatively, this signal might likely arise from a proton, different from the anomeric one, that experiences a peculiar surrounding environment, as can be expected for hydrogen atoms of a galactose moiety linked to a diaxial GG diad. Although the proton resonance remains unidentified, the area of peak B likely monitors the amount of galactose residues linked to a G residue neighboring another G residue. The difference between peak A and peak B would then account for the substitution occurred on G residue of an alternating sequence.

The quantification of derivatized M residue close to another M in alginate (but also, though to a less extent, in polyalternating MG) is almost impossible in view of the broadening and overlapping of signals in the range from 4.60 to 4.80 ppm. On the basis of the above deductions, the analysis of spectra of Figure 2 was performed and the resulting data are reported in Table 3. It must be noticed that the total degree of derivatization reported for sample 3 (i.e., 19%) is in reasonable agreement with the value of 24% obtained by

628

Biomacromolecules, Vol. 4, No. 3, 2003

Donati et al.

Figure 4. Effect of galactose on diameter of freshly prepared (columns A, dimensions of the beads controlled by coaxial air stream) and dry beads (columns B) of native and modified alginates: light gray for sample 1; dark gray for sample 2; white for sample 3. Results are presented as mean ( standard deviation.

Figure 3. (a) 400 MHz 1H NMR spectra (anomeric region) of unmodified (sample 7) and modified polymannuronate (sample 8) with degree of substitution of 42%. (b) 300 MHz 1H NMR spectra (anomeric region) of unmodified (sample 11) and modified polyalternating MG (sample 12) with degree of substitution of 70%. (c) 300 MHz 1H NMR spectra (anomeric region) of unmodified (sample 9) and modified (sample 10) G-enriched alginate with degree of substitution of 25%.

elemental analysis, giving confidence to the interpretation. Although this work is not a mechanistic study of the reactivity of alginate, from the data reported in Table 3 for samples 2, 3, and 4, some qualitative conclusions can be drawn:

(i) The pattern of derivatization discloses a selectivity of the reaction toward G residues for low degrees of conversion. In the case of samples 2 and 3 the substitution occurs exclusively on G while M residues are significantly affected by the derivatization reaction only at a high degree of conversion (i.e., in the presence of an excess of amine, sample 4). (ii) G residues in alternating sequences present reactivity higher than G in homopolymeric sequences; as a consequence, the introduction of galactose will first affect the former ones. In fact, only G residues neighboring an M residue bear a galactose moiety at low degree of derivatization, as in sample 2. With an increase in the amount of galactosylamine, the reactivity on G residues in homopolymeric sequences becomes relevant (samples 3). By performing the derivatization reaction on preformed calcium alginate beads (sample 5), the total derivatization is lowered, as a result of the protecting effect of calcium ions on the carboxylic groups, namely, those of the homopolymeric sequences. Surprisingly, a low extent of derivatization was obtained with G-enriched alginate (sample 10). This finding could be connected with the presence of extended homopolymeric G blocks (NG>1 ∼ 33), which impair the reactivity and accessibility of the G residues. b. Properties. b1. Bead-Forming Ability and Dimensions. Particular attention has been addressed to the ability of galactose-substituted alginates (2) and (3) to form calcium beads. It was found that both samples, when letting a 2% (w/v) aqueous solution drip into 50 mM calcium chloride, can form stable beads. The size of the beads was compared to that of calcium beads of unmodified alginate (1) under the same conditions. Interestingly, the presence of galactose, and the consequent reduction of the number of carboxylic groups, affects the size of the beads that can be obtained, as shown in Figure 4. The effect on the dimension of the hydrated beads formed in 50 mM calcium chloride is not very marked, with a relative diameter increase of 4% and 10% for samples 2 and 3, respectively. This suggests that the presence of galactose brings about but a small interference with the packing of the polymer chains in the gel-

Galactose-Substituted Alginate

Biomacromolecules, Vol. 4, No. 3, 2003 629

Figure 5. Swelling kinetics of dry calcium beads reported as diameter increase against time: O, sample 1; 4, sample 2; 0, sample 3. Results are presented as mean ( standard deviation. Lines are drawn to guide the eye and are obtained with an empirical double exponential growth fit.

Figure 6. Stability of calcium beads as function of the diameter increase every change of saline solution: O, sample 1; 4, sample 2; 0, sample 3. (Dimensions of the beads are controlled by an electrostatic bead generator.) Results are presented as mean ( standard deviation. Lines are drawn to guide the eye.

forming process. The presence of galactose was relatively more effective on the dimensions of the dry beads. In the case of sample 2 a decrease of 15% in the diameter is achieved, while with sample 3 the decrease reaches 30% with respect to dry calcium beads of the unmodified polymer 1. The observed behavior can be explained by taking into account: (i) the larger value of hydration of a uronate salt with respect to a neutral sugar; (ii) the larger number of free guluronate residues in unsubstituted sample 1 and the prevailing fraction of substituted G groups in the whole of substituted uronates in samples 2 and 3. Finally, side-chain substitution was already mentioned as an important triggerer of hydration, namely, an enhancer.13,14 In the dry state the effect of (i) is prevailing, barely stemming from the larger number of particles (counterions) and electric charges. This hydration is the innermost one (“crystallization” water). In the hydrated form, an osmotic swelling originating from the same elements dominating in the dry state is still present. However, important conformational effects from the galactose side chains becomes dominant, likely related to the interruption of long G-block sequences able to bind gelforming divalent counterions. The formation of the latter ones is well known to be accompanied by an extensive dehydration, much larger than that of sequences not producing gel junctions.32,33 As a whole, in the swollen beads the “conformationally disordering” effect of the galactose side chains prevails over that deriving from the loss of ionic charges and mobile counterions producing a net gain of hydration. b2. Swelling of the Beads. Swelling of calcium alginate beads of samples 1, 2, and 3, dried as described in the Experimental Section, was studied by suspending the beads in deionized water. The analysis by optical microscopy showed that both dry and reswollen gel beads retained their spherical form. The increase in diameter as a function of time is shown in Figure 5. The reswelling turned out to be fast, and the equilibrium dimension was reached after about 2 h (the diameter was further measured after 24 h, but no substantial difference was observed). The rate and the equilibrium value of the swelling were affected by the degree

of substitution. Thus, the higher the substitution, the higher the volume increase. In fact, samples 2 and 3 showed, at equilibrium, a relative diameter increase, with respect to the dry bead, of ∼1.6 and ∼2.3 times, respectively, compared to the 1.3-fold increase of sample 1. The volume of an ionic gel is governed mainly by a positive osmotic pressure (swelling) which is counterbalanced at equilibrium by a negative pressure due to the elasticity of the network.34 The elasticity of calcium alginate gels is mostly enthalpic rather than entropic;35,36 it largely depends on the number and the strength of the cross-links. As noted above, the presence of the galactose moiety linked via an amide bond to the polymer chain impairs the cooperative binding of calcium ions. The role of the cooperative calcium binding by specific sites on the polymer and its impairment by galactose substitution can be properly assessed by successive rinsing of the beads with 0.9% (w/v) aqueous NaCl (saline). Competition between monovalent and divalent counterions depends on the relative affinity for the polymer, the polyelectrolyte charge density, and the stoichiometric ratio of the two mobile ion species. Despite the well-known higher affinity for calcium ions, the latter ions can be displaced by an overwhelming amount of Na+ counterions. The lower the affinity for Ca2+, the lower the concentration of Na+ able to displace the divalent ions and the lower the number of saline changes necessary to produce a given value of the swollen bead. The displacement of calcium ions by sodium ions leads to a decrease of the number and length of the G junctions and to an increase of the osmotic swelling; hence a less stable gel will be formed with an increase of the dimensions of the gel bead. The dimensional variation is related to the relative affinity of the polyelectrolyte toward calcium ions. Figure 6 shows the difference in behavior for the beads of alginate samples 1, 2, and 3, respectively. The effect of the presence of galactose on the dimensions and the swelling of the beads must be interpreted as leading to polymers with lower affinity toward calcium. In fact it can be observed that in the case of sample 2, the diameter is increased ∼2 times after four saline changes in comparison

630

Biomacromolecules, Vol. 4, No. 3, 2003

Figure 7. Cytotoxicity of samples 1, 2, and 3 against PK15-cell line compared to a positive control (untreated cells) and negative control (human serum). Results are presented as mean ( standard deviation (n ) 6).

with beads of sample 1, in which the increase after four changes is just ∼1.3. An even lower affinity toward calcium is shown by sample 3, in which the increase in diameter is ∼2 times just after two saline changes. b3. Cytotoxicity. Since alginate is known to be biocompatible,37-41 it is important to determine the effect on a cell line of the chemically modified polymers. The presence of small amounts of chemical reagents, in fact, could lead to cytotoxicity that could hamper its use in the encapsulation of cells. Figure 7 shows the effect of solutions of samples 1, 2, and 3 against PK15 cells. It can be noticed that all samples had no cytotoxic effects at a concentration of 200 µg/mL.42 As a negative control, the cytotoxicity of human serum (100 µL) on PK15 cells was tested,43 evidencing no survival of cells after the treatment. This preliminary result shows the lack of cytotoxicity of the galactose-substituted alginate. Conclusions Alginate was coupled with galactosylamine obtaining a novel and new derivative. The modified polymer was studied by means of 1H NMR spectroscopy and the presence of galactose on M or G residues determined, making the present sample the first galactose-substituted alginate for cell encapsulation to be characterized at the molecular level.12 The calcium beads showed a reduced dimensional stability due to a lower affinity for calcium. The presence of the introduced sugar affected both dimensions and swelling, but no cytotoxic effects on PK15 cells were detected. By presenting galactose moieties as the ASGPR ligands, this polymer could find an application in the encapsulation of hepatocytes as a good bead-former to achieve an effective protection barrier. In a possible combination with polymer-clustered antigens,43 it can be proposed as a bioactiVe biomaterial in xenografts. Acknowledgment. The authors thank Ing. Wenche Strand for the expert assistance in running NMR spectra. This work was supported by a research grant of F.B.C. S.r.l. to the Department of Biochemistry, Biophysics and Macromolecular Chemistry of the University of Trieste. References and Notes (1) Painter, T. J. In The Polysaccharides; Aspinall, G. O., Ed.; Academic

Donati et al. Press: New York, 1983; Vol. 2, pp 195-285. (2) Gorin, P. A. J.; Spencer, J. F. T. Can. J. Chem. 1966, 44, 993. (3) Govan, J. R. W.; Fyfe, J. A. M.; Jarman, T. R. J. Gen. Microbiol. 1981, 125, 217. (4) Skja˚k-Bræk, G. Biochem. Soc. Trans. 1992, 20 (1), 27. (5) Draget, K. I.; Skja˚k-Bræk, G.; Smidsrød, O. Int. J. Biol. Macromol. 1997, 21, 47. (6) Strand, B. L.; Mørch, Y. A.; Skja˚k-Bræk, G. MinerVa Biotec. 2000, 12, 223. (7) Thu, B.; Bruheim, P.; Espevik, T.; Smidsrød, O.; Soon-Shiong, P.; Skja˚k-Bræk, G. Biomaterials 1996, 17, 1031. (8) Rowley, J. A.; Madlabayan, G.; Mooney, D. J. Biomaterials 1999, 20, 45. (9) Lopina, S. T.; Wu, G.; Merril, E. W.; Cima, L. G. Biomaterials 1996, 17, 559. (10) Matsuda, T.; Moghaddam, M. J. J. Polym. Sci., Polym. Chem. 1993, 31, 589. (11) Smidsrød, O.; Skja˚k-Bræk, G. TIBTECH 1990, 8, 71. (12) Yang, J.; Goto, M.; Ise, H. Cho, C.; Akaike, T. Biomaterials 2002, 23, 471. (13) Skja˚k-Bræk, G.; Zanetti, F.; Paoletti, S. Carbohydr. Res. 1989, 185, 131. (14) Skja˚k-Bræk, G.; Paoletti, S.; Gianferrara, T. Carbohydr. Res. 1989, 185, 119. (15) Strand, B. L.; Mørch, Y. A.; Syvertsen, K. R.; Espevik, T.; Skja˚kBræk, G. J. Biomed. Mater. Res., in press. (16) Kobayashi, K.; Kobayashi, A.; Akaike, T. Methods in Enzymology; Lee, Y. C., Lee, R. T., Eds.; Academic Press: San Diego, 1994; Vol. 247; p 409. (17) Dixit, V.; Arthur, M.; Gitnick, G. Biomater. Artif. Cells, Immobilization Biotechnol. 1993, 21 (2), 119. (18) Lim, F.; Moss, R. D. J. Pharm. Sci. 1981, 70 (4), 351. (19) Lee, Y. C.; Lee, R. T. Neoglycoconjugates: preparations and applications; Academic Press: San Diego, CA, 1994. (20) Kobayashi, A.; Akaike, T.; Kobayashi, K.; Sumitomo, H. Macromol. Chem. Rapid. Commun. 1986, 7, 645. (21) Skja˚k-Bræk, G.; Larsen, B.; Smidsrød, O. Int. J. Biol. Macromol. 1986, 8, 330. (22) Ertesva˚g, H.; Skja˚k-Bræk, G. Methods in Biotechnology-Carbohydrate Biotechnology Protocols; Walker and Bucke, Eds.; Humanoria Press: London, 1999; p 71. (23) Haug, A.; Larsen, B.; Smidsrød, O.; Painter, T. Acta Chem. Scand. 1969, 23, 2955. (24) Ramstad, M. U.; Ellingsen, T.; Høydal, H. K.; Valla, S.; Skja˚k-Bræk, G.; Levine, D. W. Enzyme Microb. Technol. 1999, 24, 636. (25) Campa, C.; Donati, I.; Vetere, A.; Gamini, A.; Paoletti, S. J. Carbohydr. Chem. 2001, 20 (3&4), 263. (26) Lubineau, A.; Auge´, J.; Droillat, B. Carbohydr. Res. 1995, 266, 211. (27) Grasdalen, H.; Larsen, B.; Smidsrød, O. Carbohydr. Res. 1979, 68, 23. (28) Klein, J.; Vorlop, K. D.; Kluge, H. M.; Washausen, P. DECHEMAMonog. 1979, 84, 274. (29) Strand, B. L.; Ga˚serød, O.; Kulseng, B.; Espevik, T.; Skja˚k-Bræk, G. Submitted for publication in J. Microencapsul. (30) Sun, J. H.; Wang, L. S. H.; Waring, M. A.; Wang, A. M.; Woodman, K. K.; Sheil, A. G. R. Artif. Organs 1997, 21 (5), 408. (31) Grasdalen, H. Carbohydr. Res. 1983, 118, 255. (32) Cesaro, A.; Delben, F.; Paoletti, S. J. Chem. Soc., Faraday Trans. 1 1988, 84 (8), 2573. (33) Paoletti, S.; Cesaro, A.; Ciana, A.; Delben, F.; Manzini, G.; Crescenzi, V. Solution properties of Polysaccharides; Brant, David A., Ed.; ACS Symposium Series 150; American Chemical Society: Washington, DC, 1981; p 379. (34) Tanaka, T. In Structure and Dynamics of Biopolymers; Nicolini, C., Ed.; NATO Advanced Study Institutes Series; Boston, 1987; p 237. (35) Andresen, I.-L.; Smidsrød, O. Carbohydr. Res. 1977, 58, 271. (36) Lapasin, R.; Paoletti, S.; Zanetti, F. Progress and Trends in Rheology II; supplement to Rheologica Acta; Giesekus, H., Hibberd, M. F., Mitschka, P., Riha, P., Sestak, J., Eds.; Steinkopff Verlag: Darmstadt, 1988; p 422. (37) De Vos, P.; De Haan, B. J.; Van Schilfgaarde, R. Biomaterials 1996, 18 (3), 273. (38) De Vos, P.; De Haan, B. J.; Wolters, G. H. J.; Strubbe, J. H.; Van Schilfgaarde, R. Diabetologia 1997, 40, 262. (39) Otterlei, M.; Østgaard, K.; Skja˚k-Bræk, G.; Smidsrød, O.; SoonShiong, P.; Espevik, T. J. Immunother. 1991, 10, 286. (40) Smith, N. A.; Goosen, M. F. A.; King, G. A.; Faulkner, P.; Daugulis, A. J. Biotechnol. Technol. 1989, 3, 61.

Galactose-Substituted Alginate (41) Kulseng, B.; Thu, B.; Espevik, T.; Skja˚k-Bræk, G. Cell Transplantation 1997, 6, 387. (42) Strand, B. L.; Ryan, L.; Veld, P. I.; Kulseng, B.; Rokstad, A. M.; Skja˚k-Bræk, G.; Espevik, T. Cell Transplantation 2001, 10, 263.

Biomacromolecules, Vol. 4, No. 3, 2003 631 (43) Vetere, A.; Donati, I.; Campa, C.; Semeraro, S.; Gamini, A.; Paoletti, S. Glycobiology 2002, 12 (4), 283.

BM020114Y