Altering Associations in Aqueous Solutions of a ... - ACS Publications

Dec 10, 2005 - Nouha Sari-Chmayssem , Samir Taha , Hiba Mawlawi , Jean-Paul Guégan ... Teresa Calejo , Kaizheng Zhu , Bo Nyström , Sverre Arne Sande...
0 downloads 0 Views 114KB Size
190

J. Phys. Chem. B 2006, 110, 190-195

Altering Associations in Aqueous Solutions of a Hydrophobically Modified Alginate in the Presence of β-Cyclodextrin Monomers Ce´ line Galant,† Anna-Lena Kjøniksen,‡ Giao T. M. Nguyen,‡ Kenneth D. Knudsen,† and Bo Nystro1 m†,*,‡ Department of Physics, Institute for Energy Technology, Post Office Box 40, N-2027 Kjeller, Norway, and Department of Chemistry, UniVersity of Oslo, Post Office Box 1033, Blindern, N-0315 Oslo, Norway ReceiVed: April 11, 2005; In Final Form: NoVember 7, 2005

The formation of associative networks in semidilute aqueous solutions of hydrophobically modified alginate (HM-alginate) is dependent on intermolecular hydrophobic interactions. Addition of β-cyclodextrin (β-CD) monomers to the system provides decoupling of these associations via inclusion complex formation with the polymer hydrophobic tails. This results in a dramatic decrease in the viscoelastic response of the system and a more extended local structure of the polymer chains, as shown by small-angle neutron scattering (SANS) measurements. The zero-shear viscosity decreases about an order of magnitude when the β-CD concentration is increased from 0 to 12 mm. The lifetime of the associative network decreases strongly with increasing levels of β-CD addition. These findings clearly demonstrate that the hydrophobic association effect is efficiently reduced as the amount of β-CD is increased. In the framework of drug delivery, this effect may be useful to improve the release of therapeutic molecules that can be entrapped in the polymer matrix.

Introduction Hydrophobically modified polysaccharides are frequently used as stabilizers, binders, and thickening or gelling agents in biomedical, biotechnological, and pharmaceutical applications. In aqueous solutions, they show specific properties mainly due to their ability to develop intra- and/or intermolecular hydrophobic associations. These associations lead to the formation of aggregated structures and/or three-dimensional networks (gels) that show great potential for the entrapment of therapeutic active molecules.1 Generally, in the semidilute concentration regime, the hydrophobically modified polysaccharides exhibit viscosity values several orders of magnitude higher than their precursors. The gel formation, where hydrophobic interactions act as reversible cross-linkers, results in a radical change in dynamical and mechanical behavior and can thus be monitored by rheological dynamic measurements.2 Despite the importance of strong hydrophobic associations in promoting viscosity enhancement in such systems, a reduction of these interactions can also be powerful in many cases. For instance, the hydrophobic associations obstruct the characterization of a single polymer chain (e.g., the determination of the polymer molecular weight and the radius of gyration).3 Furthermore, to gain insights into features such as the microstructure and associating abilities of the hydrophobically modified polymer, its properties are usually compared with those of the unmodified analogue. However, such an assessment can encounter problems because modified and unmodified polymers may differ by more than just the hydrophobic substitutions.4 Comparing modified polymers with both active and deactivated hydrophobic groups is a more favorable basis for understanding their behavior. In the context of pharmaceutical applications, a modulation of the intensity of the hydrophobic associations may † ‡

Department of Physics, Institute for Energy Technology. Department of Chemistry, University of Oslo.

improve the release of active substances, such as therapeutic drugs that can be entrapped in the polymer matrix. In the present study, we investigate a potential powerful method to altering the associations in solutions of a hydrophobically modified (HM-) alginate (modified by octyl C8 groups) by forming inclusion complexes between the hydrophobic tails of the polymer and β-cyclodextrin (β-CD) molecules. As a reference in the evaluation of the hydrophobic effect, corresponding measurements have also been carried out on the unmodified analogue. The linear polysaccharide alginate is an anionic copolymer composed of residues of β-D-mannuronic acid and R-L-guluronic acid. It is found in nature as a structural component in brown algae, mainly extracted from the seaweed Laminaria hyperborea in a large-scale industrial production. It is widely used in the food industry and in drug delivery systems in pharmacy. β-CDs are produced by the action of the enzyme cyclodextrin glycosyltransferase (CGTase) upon a starch hydrolysate, thus forming a cyclic structure comprised of seven R-D-glucose units. The outside of the ring-shaped molecule is hydrophilic and its inner cavity is hydrophobic, thus making it possible for the hydrophobic C8 side chains of the polymer to be included inside the cavity (having an inner diameter of 7 Å and a depth of 7.8 Å). Such behavior is supported by previous studies, which reveal that CDs have superior tendencies to interact with the hydrophobic segments of other hydrophobically modified polymers, including derivates of poly(ethylene glycol)s,5,6 dextran,7 (hydroxyethyl)cellulose,4 (hydroxypropyl)methylcellulose,8 alkalisoluble emulsion polymers,9 and ethoxylated urethanes.10 By encapsulating polymer hydrophobic tails inside their cavity, β-CDs are thus expected to decouple polymer-polymer hydrophobic interactions and thereby alter the association intensity in HM-alginate solutions. In this work, we focus on investigating the effect of β-CD addition to HM-alginate solutions, evaluating the reversibility of the hydrophobic polymer interactions and the change in the

10.1021/jp0518759 CCC: $33.50 © 2006 American Chemical Society Published on Web 12/10/2005

β-Cyclodextrin Monomers in Alginate Networks polymer conformation. To our knowledge, this is the first work using a hydrophobically modified version of the highly important alginate polymer and its unmodified analogue in combination with cyclodextrin. To characterize the structure and interactions in aqueous solutions of alginate or HM-alginate in the presence of β-CD, we make use of different experimental techniques such as turbidimetry, small-angle neutron scattering (SANS), and rheology. This provides us with important information concerning the changes of the association strength and local structural rearrangements taking place during complexation. Experimental Section Materials. The unmodified alginate analogue, designated LF 10/60 LS (912912), supplied from FMC Biopolymers, Drammen, Norway, was used in this study. According to the specifications from the manufacturer, this sample has a weightaverage molecular weight of 152 000 and the guluronic acid to mannuronic acid ratio is 0.75. 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC-HCl) is a water-soluble derivative of carbodiimide and it is an activation reaction agent (purchased from Fluka) for the coupling between carboxylic function and primary amine. Octylamine (obtained from Fluka) is utilized for the hydrophobic (C8) modification of the polymer chains. The cosolute β-cyclodextrin was supplied by Fluka. All chemicals were used without further purification. Synthesis and Solution Preparation. An aqueous solution of sodium alginate (30 g; 3.0 wt %) was adjusted to pH ≈ 3.4 by addition of 0.1 M HCl, and the polymer concentration was diluted to 2.0 wt %. To this solution was added a certain amount of EDC-HCl (0.106 mol) dissolved in 100 mL of water. The concentration of EDC-HCl ([EDC-HCl]/[alginate unit] ) 0.7) was calculated from the molar quantity of the sodium alginate unit and carboxylate activation portion. After 5 min of reaction, octylamine (0.212 mol) ([octylamine]/[alginate unit] ) 1.4) was added and the mixture was stirred for 24 h at ambient temperature. The HM-polymer was isolated by precipitation in acetone and the polymer was collected by filtration. The reaction yield was 42.1%. To remove low molecular weight impurities, the polymer was thoroughly dialyzed against water (7 days) and isolated by freeze-drying. The NMR analysis of the freezedried sample did not reveal the presence of unreacted species. Alginate was thus hydrophobically modified by use of the coupling agent EDC-HCl to form amide linkages between amine-containing molecules and the carboxylate moieties on the alginate polymer backbone. This type of reaction has been described in detail elsewhere.11-13 The chemical structure and purity of the HM-alginate was ascertained by 1H NMR (D2O, 85 °C) with a 500 MHz Bruker DRX 500 spectrometer (Bruker Biospin, Fa¨llanden, Switzerland): δ (ppm) ) 5.08 (C1H, G unit), 4.68 (C1H, M unit), 0.88 (CH2CH3, hydrophobic part), 1.301.34 [(CH2)6CH3, hydrophobic part]; other protons of the alginate unit and octyl chains overlapped and were unresolved. As a reference for 1H NMR at 0 ppm, 3-trimethylsilylpropionic acid-d4 sodium salt was used. The hydrophobic modification degree was determined from the peak ratios between the anomeric protons and methyl protons of the octyl chain. The degree of modification determined from NMR analysis was 29.5 mol %. From element analysis of nitrogen contents, the degree of hydrophobic modification was 31.7 mol %. This suggests that the degree of hydrophobic modification is approximately 31 mol %. All the solutions were prepared in heavy water (D2O) by weighing the components. All the measurements were conducted

J. Phys. Chem. B, Vol. 110, No. 1, 2006 191 in the semidilute range at a fixed polymer concentration of 2.0 wt % and at a temperature of 25 °C. All experiments were carried out on solutions with no added salt, and the pH of the solutions was 3.8. Turbidimetry. The transmittances of 2.0 wt % solutions of alginate and HM-alginate in the presence of various levels of β-CD were measured with a temperature-controlled Helios Gamma (Thermo Spectronic, Cambridge, U.K.) spectrophotometer at a wavelength of 500 nm. The apparatus is equipped with a temperature unit (Peltier plate) that gives an efficient temperature control over an extended time. The turbidities τ of the samples are determined from the relationship τ ) (-1/L) ln (It/I0), where L is the light path length of the cell (1 cm), It is the transmitted light intensity, and I0 is the incident light intensity. The results from the spectrophotometer will be presented in terms of turbidity. Several samples were centrifuged prior to measurement, to remove air bubbles that had been formed during the dissolution process. Small-Angle Neutron Scattering. The SANS measurements were performed on the SANS installation at the IFE reactor at Kjeller, Norway. The wavelength was set by the aid of a selector (Dornier), with a high full width at half-maximum (fwhm) for the transmitted beam (∆λ/λ ) 20%), and maximized flux on the sample. The neutron detector was a 128 × 128 pixel, 59 cm active diameter, 3He-filled RISØ-type detector, which is mounted on rails inside an evacuated detector chamber. The investigated scattering vector q-range was defined by the neutron wavelengths λ ) 10.2 and 5.1 Å and the sample-to-detector distances D ) 3.4 and 1.0 m, covering the experimental q-range 8 × 10-3 e q e 0.3 Å-1. The scattering vector q is given by q ) (4π/λ) sin (θ/2), where θ is the scattering angle. The samples were investigated in 2 mm Hellma quartz cells, with D2O as a solvent to minimize incoherent scattering and to maximize the scattering contrast between the HM-alginate (scattering length density FHM-alginate ) 1.46 × 1010 cm-2) and the solvent (FD2O ) 6.41 × 1010 cm-2). The scattering length density calculated for the β-CD monomers is quite close to that of HM-alginate (Fβ-CD ) 0.88 × 1010 cm-2). For this reason, the contrast matching method14 has not been used to observe the scattering from the individual components within the mixtures. The abovementioned scattering length densities were calculated on the basis of the chemical composition of a β-CD unit, [C6H10O5]7, and tabulated values for the scattering lengths of the individual constituents, C, H, and O (0.665 × 10-12, -0.374 × 10-12, and 0.581 × 10-12 cm, respectively). Finally this was combined with the molecular weight of β-CD, 1135 g/mol, to obtain the average scattering length over a specific volume. A similar procedure was followed for HM-alginate, with the chemical composition of this particular alginate, as well as the level of addition of C8 side chains, taken into account. Standard reductions of the scattering data, including transmission corrections, were conducted by incorporating data collected from empty cell, beam without cell, and blocked-beam background to obtain the scattered intensities in absolute scale (centimeters-1). Subtraction of the solvent scattering was then done to obtain the coherent macroscopic scattering cross section dΣ/dΩ (q) of the system. Rheology. The steady and dynamic rheological measurements of the solutions were carried out on a Paar-Physica MCR 300 rheometer with a cone-and-plate geometry, a cone angle of 1°, and a diameter of 75 mm. This rheometer operates effectively with this geometry also on dilute polymer solutions, and even the viscosity of water can easily be measured over an extended shear rate domain. The samples were introduced onto the plate,

192 J. Phys. Chem. B, Vol. 110, No. 1, 2006

Galant et al.

Figure 1. Illustration of the effect of β-CD addition on the turbidity in 2.0 wt % solutions of alginate and HM-alginate. The turbidity of β-CD in D2O is also displayed.

and to prevent evaporation of the solvent, the free surface of the sample was always covered with a thin layer of low-viscosity silicone oil (the viscoelastic response of the samples is not observed to be affected by this layer). The values of the strain amplitude were checked to ensure that all oscillatory shear measurements were performed within the linear viscoelastic regime, where the dynamic storage modulus (G′) and loss modulus (G′′) are independent of the strain amplitude. The oscillating sweep measurements were conducted in the approximate angular frequency (ω) domain 0.1-100 rad/s, and the shear viscosity experiments were conducted over an extended shear rate range 10-4-103 s-1. Results and Discussion Turbidimetry. Figure 1 shows the effect of β-CD addition on the turbidity (τ) for 2.0 wt % solutions of alginate and HMalginate and the turbidity of β-CD without polymer. In the absence of polymer, a very low and constant value of the turbidity is found up to 8 mm, whereas at higher β-CD concentrations a progressively more pronounced rise of τ occurs. This probably reflects the low water solubility of β-CD and the formation of crystals.15 The stability problem in water solution of β-CD is a nontrivial issue that has been the subject of several publications in recent years.16-18 This problem is mainly related to the poor water solubility of β-CD, which is likely caused by an intramolecular hydrogen-bonding network between OH2 and OH3 of adjacent glucose residues, resulting in low exposure of hydrogen-bonding hydroxyl groups to the aqueous environment. In the presence of alginate and HMalginate, higher values of τ are observed due to the contribution of large-scale structures from the polymers through association. The turbidity of the alginate/β-CD and HM-alginate/β-CD mixtures is nearly constant at low levels of β-CD, whereas there is an overall increasing tendency in the value of τ at higher β-CD concentrations, probably due to the approach of the saturation level of free β-CD, and as will be discussed below, the interaction between alginate and β-CD may also contribute to this behavior. In the case of HM-alginate, there is a small (10%) drop of τ between 4 and 8 mm. The drop of τ can probably be ascribed to a reduced intensity of the hydrophobic associations because of encapsulation of hydrophobic moieties. This deactivation of hydrophobic groups leads to a lesser inclination to create association complexes. At higher β-CD concentrations, the free β-CD and the β-CD/HM-alginate complexes give rise to augmented values of τ. The higher values

Figure 2. (a) SANS (every second point is shown) from HM-alginate (2.0 wt %)/β-CD/D2O mixtures with the indicated concentrations of β-CD. The solid lines represent fits to eq 1. (b) Influence of β-CD concentration on the persistence length L and the correlation length ξ, both determined from fits with the aid of eq 1.

of the turbidity for HM-alginate as compared to the corresponding values for alginate at high levels of β-CD addition are attributed to residual hydrophobic interactions in solutions of HM-alginate. SANS Results. Figure 2a illustrates SANS spectra from HMalginate (2.0 wt %)/β-CD mixtures in D2O. The low-q features ( G′ over the entire concentration regime, which indicates that the viscous response dominates. In the case of HM-alginate, the value of G′ decreases monotonically as the β-CD concentration increases, suggesting a weakening of the network associations. The viscous counterpart exhibits only a slight decrease initially. At concentrations of β-CD up to approximately 10 mm, G′ > G′′. At higher levels of β-CD addition, a crossover is detected and G′ becomes smaller than G′′. This feature signalizes a transition from a dominating elastic behavior to a prevailing viscous response at higher amounts of β-CD. In the inset plot of Figure 5, the time of intersection τ* (τ* ) 1/ω*) is plotted as a function of the β-CD concentration. The parameter τ* has been estimated from the frequency dependence of the dynamic moduli and the observation of the frequency of intersection ω* (G′ ) G′′). The quantity τ* is related to the longest relaxation time and gives information about the lifetime of the associative network. The monotonic decrease in τ* with increasing levels of β-CD addition is probably another manifestation of weakened hydrophobic associations. According to the transient network theory, the elastic modulus is directly proportional to the number of active junctions.26 In view of this, the decreases of G′ and τ* observed for increasing β-CD concentrations reflect a reduction of the number of active junctions between HM-alginate chains. These findings are consistent with a deactivation of the hydrophobic polymer groups via the formation of inclusion complexes between these groups and the β-CD cavities. Conclusions The use of β-cyclodextrin monomers provides an efficient way of weakening the hydrophobic associations in HM-alginate solutions. The inclusion inside the β-CD cavities of the normally associating polymer side groups reduces the steady shear viscosity and the dynamic moduli of the HM-alginate solutions by 1 order of magnitude. The general picture that emerges is that the truncated cone structures from β-CD encapsulate hydrophobic tails on the HM-alginate chains and thereby reduce

the hydrophobic associations. On a nanoscopic level, the deactivation of the hydrophobic polymer groups by the β-CDs is reflected in a local stretching of the HM-alginate chains. A schematic illustration of the main effect of β-CD addition on the hydrophobic moieties in HM-alginate solutions is depicted in Figure 6. This drawing illustrates the partial decoupling of the hydrophobic tails at this temperature. Temperature is a variable that can be utilized to tune the strength of the hydrophobic interactions in HM-alginate systems, and a study in this direction is in progress. In view of these findings, it is reasonable to expect that modulation of the β-CD concentration in HM-alginate solutions may be a way of controlling the release of molecules that can be entrapped in the polymer network. These specific properties of the HM-alginate/β-CD system, including its biocompatibility, make it a good candidate for biomedical applications, such as controlled drug delivery. Acknowledgment. B.N., G.T.M.N., and K.D.K. gratefully acknowledge support from the Norwegian Research Council trough a NANOMAT Project (158550/431). K.D K. and C.G. thank the Marie Curie Industry Host Project (Contract G5TRCT-2002-00089) for support. References and Notes (1) Duval-Terrie´, C.; Cosette, P.; Molle, G.; Muller, G.; De´, E. Protein Sci. 2003, 12, 681. (2) Charpentier-Valenza, D.; Merle, L.; Mocanu, G.; Picton, L.; Muller, G. Carbohydr. Polym. 2005, 60, 87. (3) Islam, M. F.; Jenkins, R. D.; Bassett, D. R.; Lau, W.; Ou-Yang, H. D. Macromolecules 2000, 33, 2480. (4) Karlson, L.; Thuresson, K.; Lindman, B. Carbohydr. Polym. 2002, 50, 219. (5) Zhang, H.; Hogen-Esch, T. E.; Boschet, F.; Margaillan, A. Langmuir 1998, 14, 4972. (6) Karlson, L.; Thuresson, K.; Lindman, B. Langmuir 2002, 18, 9028. (7) Amiel, C.; David, C.; Renard, E.; Se´bille, B. Polym. Prepr. 1999, 40, 207. (8) Hu, X.; Zheng, P. J.; Zhao, X. Y.; Li, L.; Tam, K. C.; Gan, L. H. Polymer 2004, 45, 6219. (9) Abdala, A. A.; Tonelli, A. E.; Khan, S. A. Macromolecules 2003, 36, 7833. (10) Liao, D.; Dai, S.; Tam, K. C. Polymer 2004, 45, 8339. (11) Nakajima, N.; Ikada, Y. Bioconjugate Chem. 1995, 6, 123. (12) Hermanson, G. T. In Bioconjugate Chemistry; Academic Press: San Diego, CA, 1996; p 169. (13) Park, S. N.; Park, J.-C.; Kim, H. O.; Song, M. J.; Suh, H. Biomaterials 2002, 23, 1205. (14) Cotton, J.-P. J. Phys. IV France 1999, 9, 21. (15) Szejtli, J. Pure Appl. Chem. 2004, 76, 1825. (16) Coleman, A. W.; Nicolis, I.; Keller, N.; Dalbiez, J. P. J. Inclusion Phenom. Macrocyclic Chem. 1992, 13, 139. (17) Gaitano, G. G.; Brown, W.; Tardajos, G. J. Phys. Chem. B 1997, 101, 710. (18) Gonzales Gaitano, G.; Rodriguez, P.; Isasi, J. R.; Fuentes, M.; Tadayos, G.; Sanchez, M. J. Inclusion Phenom. Macrocyclic Chem. 2002, 44, 101.

β-Cyclodextrin Monomers in Alginate Networks (19) Cotton, J.-P. Introduction to scattering experiments, in Neutron, X-ray and light scattering; Lindner, P., Zemb, T., Eds.; North-Holland Delta Series: Amsterdam, 1991; p 3. (20) Horkay, F.; Hecht, A.-M.; Grillo, I.; Basser, P. J.; Geissler, E. J. Chem. Phys. 2002, 117, 9103. (21) Kjøniksen, A.-L.; Hiorth, M.; Roots, J.; Nystro¨m, B. J. Phys. Chem. B 2003, 4, 337. (22) Tho, I.; Kjøniksen, A.-L.; Nystro¨m, B.; Roots, J. Biomacromolecules 2003, 4, 1623. (23) Kjøniksen, A.-L.; Hiorth, M.; Nystro¨m, B. Eur. Polym. J. 2005, 41, 761.

J. Phys. Chem. B, Vol. 110, No. 1, 2006 195 (24) (a) Groot, R. D.; Agterof, W. G. M. J. Chem. Phys. 1994, 100, 1649 and 1657. (b) Groot, R. D.; Agterof, W. G. M. Macromolecules 1995, 28, 6284. (25) Leibler, L.; Rubinstein, M.; Colby, R. H. Macromolecules 1991, 24, 4701. (26) Tanaka, F.; Edwards, S. F. Macromolecules 1992, 25, 1516. (27) Semenov, A. N.; Joanny, J.-F.; Khokhlov, A. R. Macromolecules 1995, 28, 1066. (28) Rubinstein, M.; Dobrynin, A. V. Trends Polym. Sci. 1997, 5, 181. (29) Rubinstein, M.; Dobrynin, A. V. Curr. Opin. Colloid Interface Sci. 1999, 4, 83.