- Email: [email protected]
Low-degree oxidized scleroglucan and its hydrogel
International Journal of Biological Macromolecules 28 (2001) 351– 358 www.elsevier.com/locate/ijbiomac
Low-degree oxidized scleroglucan and its hydrogel Hitoshi Maeda a, Giuseppe Rambone b, Tommasina Coviello a,*, Yoshiaki Yuguchi a, Hiroshi Urakawa a, Franco Alhaique b, Kanji Kajiwara a b
a Faculty of Engineering and Design, Kyoto Institute of Technology, Kyoto, Sakyo-ku, Matsugasaki, 606 -8585 Japan Department of Chemistry and Technology of Biologically Acti6e Compounds, Faculty of Pharmacy, ‘La Sapienza’ Uni6ersity, Piazzale A. Moro 5, I-00185 Rome, Italy
Received 28 July 2000; received in revised form 23 January 2001; accepted 25 January 2001
Abstract A controlled oxidation of scleroglucan was performed with sodium periodate to prepare aldehyde derivatives (scleraldehyde) with a low degree of oxidation (10 and 20%), which were utilized for crosslinking reactions with hexamethylenediamine. The structural characterization of scleraldehydes and their corresponding hydrogels was attempted by small-angle X-ray scattering (SAXS). While scleraldehyde with a higher degree of oxidation ( ] 50%), according to an earlier research, was found to disentangle into single chains as the degree of oxidation increases; scleroglucan bearing a low percentage of aldehydic groups (up to 20%) retains mainly the conformation of the natural polysaccharide, thus the system can be represented as composed of triple helices with only minor disentanglements at the sites where the aldehyde groups are present. The hydrogel prepared from scleraldehyde with a low degree of oxidation is brittle and fragmented, in contrast to the elastic/homogeneous hydrogel earlier prepared from scleraldehyde with a high degree of oxidation. The hydrogel from scleraldehyde with a low degree of oxidation was found to possess a network structure that consisted mostly of the triple helices crosslinked in specific points where the triple helices are disentangled into single chains because of the presence of the aldehyde groups. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Scleroglucan; Hydrogel; SAXS
1. Introduction Scleroglucan, a neutral extracellular polysaccharide secreted by fungi of the genus sclerotium, has a primary structure which consists of a linear (1 3)-b-D-glucan to which single b-D-glucopyranosyl groups are linked (1 6) at every third unit of the main chain. It is known that scleroglucan assumes a triple-stranded helical conformation in aqueous solution and a single coiled disordered conformation in methylsulphoxide or at high pH values (NaOH\ 0.2 M) [1 – 4]. Polysaccharides containing this repeating unit and with a molecular weight higher than 104 are known to possess an anti-tumor activity against Sarcoma 180 [5]; furthermore, scleroglucan has been proposed for the formulation of sustained delivery dosage forms [6–9]. * Corresponding author. Tel.: +39-0-64-9913300; fax: + 39-0-649913133. E-mail address: [email protected] (T. Coviello).
Carboxylated and crosslinked derivatives of scleroglucan were also studied for the preparation of controlled release matrices capable to respond to different environmental conditions (e.g. pH or ionic strength) [10]. The b-D-glucosyl side chains can be mildly oxidized with sodium periodate [11,12] to yield the aldehyde derivative (scleraldehyde, Fig. 1), which is converted into hydrogel by a crosslinking reaction with diamine [13 –15]. While the complete oxidation disentangles the triple-stranded helical conformation, scleraldehyde with a lower degree of oxidation (up to 20%) was found to retain mainly the triple-stranded helices. In an earlier paper [16], hexamethylenediamine was used for the crosslinking reaction on scleraldehyde with a high degree of oxidation (40 and 100%) but not all diamines were actually involved in crosslinking; here the crosslinked domains were found to consist of parallel-aligned multiple single chains tied with the diamine. In the present study small-angle X-ray scattering (SAXS) technique was used to investigate the structural
0141-8130/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S0141-8130(01)00126-X
352
H. Maeda et al. / International Journal of Biological Macromolecules 28 (2001) 351–358
characteristics of scleraldehyde with a lower degree of oxidation, and of its gel obtained with the same amine as crosslinking agent. The preparation of scleraldehyde with a 10 and 20% degree of oxidation [11,12] and the subsequent crosslinking reaction are also described. Obtained information on the structure of the polymeric networks can be useful both for the interpretation of the behavior of the gels, when used as drug delivery systems, and for the optimization of controlled release dosage form formulations.
2. Experimental
2.1. Materials and methods Scleroglucan (Actigum CS11) was kindly provided by Mero-Rousselot-Satia (France). Hexamethylenediamine (HMD) was purchased from Aldrich (Germany). All other chemicals were of analytical grade and used as received. Scleroglucan was dissolved in water (0.5% w/v) and then sonicated at 20 MHz and 100 W for 2 h (Probe Ultrasonic Processor, Sonics and Materials VC300) to
Fig. 1. Oxidation reaction scheme of scleroglucan.
H. Maeda et al. / International Journal of Biological Macromolecules 28 (2001) 351–358
353
Fig. 2. SAXS profiles observed from scleroglucan, scleraldehyde 10%, scleraldehyde 20% and scleraldehyde 100% in aqueous solution. Ordinates are given in arbitrary units.
reduce the original molecular weight, that from a value of 1.4×106 became about 500000. The molecular weight of the sample was determined before and after sonication by means of viscosity data. Viscosity measurements were carried out at 25°C ( 90.01°C) using a Schott– Gerate equipment and an Ubbelhode capillary viscometer with internal diameter of 0.53 mm. Inserting the values of the extrapolated intrinsic viscosity in the Mark–Houwink –Sakurada equation valid for scleroglucan [17], the molecular weights of native and sonicated samples were obtained. The sonicated scleroglucan solution was filtered through a 0.8 mm pore membrane and the polysaccharide was precipitated with isopropyl alcohol. The obtained polymer was re-dissolved in distilled water and dialyzed exhaustively against distilled water. The solution was finally freeze-dried to yield purified scleroglucan. A weighed amount of purified scleroglucan was dissolved in water (0.5% w/v) and an appropriate amount of sodium periodate was added. The solution was magnetically stirred and kept in the dark for 24 h before the purification by exhaustive dialysis against distilled water
at 4°C. By means of this method scleroglucan was oxidized to scleraldehyde [11,12] of two different degrees of oxidation (i.e. 10 and 20%) by adjusting the stoichiometric ratio [periodate]/[scleroglucan]. The actual degree of oxidation was confirmed by potentiometric titration of a small amount of the carboxylated samples obtained by further oxidation of the aldehyde groups by sodium chlorite [18]. Those samples are referred to as scleraldehyde 10% and scleraldehyde 20%, respectively. The crosslinking reaction was performed directly on the solution containing scleraldehyde 10 and 20%, by addition, while stirring, of an excess of HMD at 40°C (Fig. 1). Unlike in the case of scleraldehyde 40 and 100% gels, no homogeneous phase was constituted, and the new gels were fragmented and floated in the solution. They were kept overnight at the same temperature and then dialyzed against distilled water, at 4°C, in order to remove the excess of HMD before the SAXS measurements. A very low percentage of diamine reacted only with one of its amine group, since a negligible amount of free NH2 groups was still present in the crosslinked polymer, as indicated by the colorimetric test with sodium nitroprusside.
354
H. Maeda et al. / International Journal of Biological Macromolecules 28 (2001) 351–358
2.2. Small-angle X-ray scattering The SAXS measurements were performed at BL-10C of the Photon Factory, Tsukuba, Japan. An incident X-ray was monochromatized to u= 0.149 nm and focused to the position of the detector with a bent focusing mirror. The scattered X-ray was detected by a one-dimensional position sensitive proportional counter (PSPC) positioned at the distance of about 1 m from the sample holder. A flat sample cell of 0.2 cm pathlength made of glass with 20 mm thick quartz windows was used for the solutions. Gels were cut in an appropriate size and sandwiched with 20 mm thick quartz plates inserted in an empty cell. The temperature of the cell was controlled by circulating water of a constant temperature through the cell holder. The solutions were directly injected into the cell, which was placed in the cell holder at least 10 min prior to the SAXS measure-
ments. The SAXS intensities were accumulated for a total of 600 s, in order to assure enough statistical accuracy without degrading the polysaccharide samples by X-ray irradiation. The scattering intensities were corrected with respect to the variation of the incident X-ray flux, by monitoring with an ion chamber installed in front of the cell holder. The X-ray absorption of the solution was compensated by measuring the incident and transmitted X-ray intensities. The excess scattering intensities were calculated by subtracting the scattering intensities of the solvent from those of sample solutions or gels.
3. Results and discussion Gels prepared from the scleraldehyde with a low degree of oxidation (up to 20%) showed a completely
Fig. 3. Molecular models for scleroglucan triple helix and single coil.
H. Maeda et al. / International Journal of Biological Macromolecules 28 (2001) 351–358
355
Fig. 4. Observed SAXS profiles (open symbols) and the scattering profiles calculated (solid lines) from the triple helix (Fig. 3) for scleroglucan, scleraldehyde 10, 20 and 100%. Ordinates are given in arbitrary units.
different appearance from those prepared from scleraldehyde with a high degree of oxidation (] 50%). The former gels are brittle and fragmented, whereas the latter ones are elastic and homogeneous. The SAXS was observed from the scleraldehyde samples with the two degrees of oxidation, i.e. 10 and 20%. Fig. 2 shows typical SAXS results obtained from the solutions of 10 and 20% scleraldehyde in the double logarithmic plots; for an appropriate comparison, also the data from scleroglucan and scleraldehyde 100% are shown. Here I(q) denotes the scattered intensity and q is the magnitude of the scattering vector given by (4y/u)sin(q/2) with q and u being the scattering angle and the wavelength of an incident beam, respectively. The conformation of scleraldehyde seems not to change when the degree of oxidation is low (samples 10 and 20%) as seen from the observed SAXS profiles. The observed scattering profiles were analyzed by assuming that most part of the scleraldehyde chain is composed of triple-stranded helix in terms of a modified broken rod model [19,20], which is expressed as
q 2I(q) :% q 2wi ML,i Ui (q)+ const. c i
(1)
Here the constant term takes into account the spatial correlation of the components [21], and accounts for the structural characteristics of the boundary of triplestranded helices to disentangled chains. ML,i is a linear molar mass of the component i. In the present case, i denotes either a triple-stranded helix component or a single chain component, and Ui (q) is a corresponding particle scattering factor calculated from the atomic coordinates of the respective molecular models (see Fig. 3) according to the Debye formula n
n−1
Ui (q)= % f 2j g 2j (q)+ 2 % j=1
n
%
j=1 k=j+1
sin(djkq) djkq (2)
fj fkgj (q)gk (q)
where fj and djk denote the atomic scattering factor of the atom j and the distance between the jth and kth atoms, respectively. The molecular model for a triple helix was adapted from the crystallographic data [22,23] and that for a single coil was generated by the Monte Carlo method according to the energy map of lami-
356
H. Maeda et al. / International Journal of Biological Macromolecules 28 (2001) 351–358
naribiose [24]. The form factor gj (q) for a single atom is assumed to be represented by the form factor of a rigid sphere having the radius equivalent to the van der Waals radius Rj of the jth atom as gj (q) =
3[sin(Rj q)− (Rj q)cos(Rj q)] (Rj q)3
(3)
Rj was set to 0.501 nm or 0.450 nm for a carbon or an oxygen atom, respectively, in the present calculation, which is three times larger than the respective van der Waals radius in order to take into account the effect of hydrated water along the polysaccharide chain. The molecular models of a triple-stranded helix and a single chain are composed of 80 and 100 residues, respectively. Scleraldehydes 10 and 20% seem to assume a triplestranded helical conformation in aqueous solution, as a broken rod model with a single triple-helix component fits well with the observed SAXS profiles. The fitting example will be seen from Fig. 4, which shows the observed SAXS profiles and the calculated scattering profiles for scleraldehyde 10 and 20% in aqueous solution in terms of the Kratky plots (q 2I(q) plotted against q); for an appropriate comparison, also the data obtained with scleroglucan and 100% scleraldehyde are shown. Scleraldehyde 100% consists of single coils, while about 50% of the chain was found to be involved in a single coil in the case of scleraldehyde 40% [16]. Thus one can assume that oxidation of side groups causes disentangling the triple-stranded helix as found for the carboxylated scleroglucan (sclerox) [19]. However, a degree of oxidation up to 20% will not cause remarkable conformational changes from a triple-
Fig. 6. Molecular model for the network of scleraldehyde hydrogel tied with hexamethylenediamine. Scleraldehyde 10% gel (top) and scleraldehyde 20% gel (bottom).
Fig. 5. SAXS profiles from the hydrogels of scleraldehyde 10, 20 and 100%. Ordinates are given in arbitrary units.
stranded helix to a single coil, and the observed SAXS profiles from scleraldehyde 10 and 20% are almost identical to the one of scleroglucan (Fig. 2). The aldehyde groups can react with amino groups, and a hydrogel can be formed by coupling scleraldehyde with diamine. Since the coupling reaction takes place in alkaline conditions (pH\ 12), all triplestranded helices are disentangled into single chains as it occurs for scleroglucan at these pH conditions. When the degree of oxidation is high (over 40%), scleraldehyde is prevented from reforming a triple-stranded helix after the reaction with HMD [16]. On the contrary, when the degree of oxidation is lower than 20%, scleraldehyde may resume a triple-stranded helix after the
H. Maeda et al. / International Journal of Biological Macromolecules 28 (2001) 351–358
reaction, during the dialysis process that decreases remarkably the environmental pH value. The SAXS profiles obtained with the hydrogels prepared from scleraldehyde 40 and 100% revealed a characteristic peak at q3 nm − 1, which was attributed to the fine structure in the crosslinked domain composed of scleraldehyde chains aligned in parallel [16]. Once HMD ties two scleraldehyde chains, further crosslinking will take place around that tie because more chance will be available for finding a counterpart. When the degree of oxidation is low, no aldehyde group will be available in the proximity of a crosslinked site. As seen from Fig. 5, no characteristic peak at q 3 nm − 1, due to scleraldehyde chains aligned in parallel, was observed from the hydrogel of scleraldehyde 10 and 20%. That is, the lower degree of HMD tie would not prevent scleraldehyde 10 and 20% gel from resuming a triple-helical conformation, although no triple helices are involved in the crosslinked domain of scleraldehyde 40 and 100% gels. Taking into account that scleraldehyde with a low
357
degree of oxidation resumes a triple-helical conformation in gel and that the crosslinking reaction takes place where aldehyde groups are available, the resulting network may look like the model proposed in Fig. 6. Here the model network is built by assuming a random orientation of triple helices and a crosslinking at a site where two aldehyde groups meet. A scleraldehyde triple helix disentangles into single chains at the position where an aldehyde group is introduced. Since the probability of two or more consequent residues being oxidized is extremely small for scleraldehyde with a degree of oxidation 20% or less, the network is thought to be constituted mostly of rigid triple helices. The disentangled part (composed of 6 or 12 residues) was introduced according to the degree of oxidation. Several network fragments were generated where the fragment was assumed to consist of four triple helices (in the case of scleraldehyde 10% gel) or eight triple helices (in the case of scleraldehyde 20% gel). The SAXS profiles were calculated according to Eq. (1) with the network fragments (represented in Fig. 6)
Fig. 7. The SAXS profiles observed from scleraldehyde 10% hydrogel, 20% hydrogel, 40% hydrogel and 100% hydrogel. The solid lines indicate the scattering profiles calculated according to the respective molecular models (Fig. 6) from Eq. (1). Ordinates are given in arbitrary units.
358
H. Maeda et al. / International Journal of Biological Macromolecules 28 (2001) 351–358
and single coils (see Fig. 3). Fig. 7 shows the comparison of the observed SAXS profiles with the profiles calculated by Eq. (1) for scleraldehyde 10% hydrogel, 20% hydrogel, 40% hydrogel and 100% hydrogel. The calculated profiles fit well with the observed SAXS profiles from scleraldehyde 10 and 20% gels, indicating that the model network composed of triple helices represents at least qualitatively a real network structure in those hydrogels. The use of Eq. (1) implies the random spatial correlation between the network fragments. It can then be assumed that those network fragments are linked in a random fashion to yield the calculated scattering profiles. In either case, the fraction of the single coil component was small and less than 9%.
4. Conclusions Scleraldehyde with a low degree of oxidation (10 and 20%), prepared by a controlled oxidation of scleroglucan, was found to retain essentially a triple-stranded helical conformation, while the triple-stranded chain disentangles in single chains with increasing the degree of oxidation (40 and 100%) [14]. The hydrogel prepared from scleraldehyde with a low degree of oxidation, according to SAXS profiles analysis, can be represented by a network composed of randomly oriented triple helices interlinked at the sites where the aldehyde groups are present. Hydrogels are often used for the formulation of controlled release formulations and it is well known that delivery profiles from polymeric networks can be remarkably affected by the actual structure of the polymeric system; thus the structural characterization of these scleroglucan derivatives appears to be very important as they can be proposed as new matrices for sustained release of bioactive materials and the possibility to shed some light upon their structure can lead to an improvement of the performances of new delivery systems.
Acknowledgements Y.Y. acknowledges a financial support of JSPS Research Fellowship for Young Scientists. K.K. is indebted to ‘La Sapienza’ University for a visiting
.
professorship, and JSPS/CNR for a personal grant. This work was carried out under the approval of the Photon Factory Advisory Committee (Proposal No. 94G-291) and was supported in part by MURST funds.
References [1] Norisuye T, Yanaki T, Fujita H. J Polym Sci Polym Phys Ed 1980;18:547. [2] Yanaki T, Kojima T, Norisuye T. Polym J 1981;13:1135. [3] Sato T, Norisuye T, Fujita H. Carbohydr Res 1981;95:195. [4] Sato T, Norisuye T, Fujita H. Macromolecules 1983;16:185. [5] Norisuye T. Makromol Chem Suppl 1985;14:105. [6] Touitou E, Alhaique F, Riccieri FM, Riccioni G, Santucci E. Drug Des Deliv 1989;5:141. [7] Alhaique F, Carafa M, Riccieri FM, Santucci E, Touitou E. Pharmazie 1993;48:432. [8] Rizk S, Duru C, Gaudy D, Jacob M, Ferrari F, Bertoni M, Caramella C. Int J Pharm 1994;112:125. [9] Crescenzi V, Dentini M. In: Ottenbrite RM, Huang SJ, Park K, editors. Hydrogels and Biodegradable Polymers for Bioapplications. ACS Symposium Series 627. Washington, DC: American Chemical Society, 1996:221 – 32. [10] Coviello T, Grassi M, Rambone G, Santucci E, Carafa M, Murtas E, Riccieri FM, Alhaique F. J Controlled Release 1999;60:367. [11] Rankin JC, Jeanes A. J Am Chem Soc 1954;76:4435. [12] Jeanes A, Wilham CA. J Am Chem Soc 1950;72:2655. [13] Alhaique F, Coviello T, Carafa M, Murtas E, Riccieri FM, Santucci E. Third Jerusalem Conference on Pharmaceutical Sciences and Clinical Pharmacology-JC, Jerusalem, 1996; p. 20. [14] Coviello T, Rambone G, Alhaique F, Carafa M, Murtas E, Riccieri FM, Santucci E. Polymer Networks 96 Doorn, The Netherlands,1996; 17. [15] Alhaique F, Coviello T, Carafa M, Murtas E, Rambone G, Riccieri FM, Santucci E, Paci M. Proc Int Symp Control Rel Bioact Mater 1997;24:1043. [16] Maeda H, Coviello T, Yuguchi Y, Urakawa H, Rambone G, Alhaique F, Kajiwara K. Polym Gels Networks 1998;6:355. [17] Yanaki T, Norisuye T, Fujita H. Macromolecules 1980;13:1466. [18] Hofreiter BT, Wolff IA, Mehltretter CL. J Am Chem Soc 1957;79:6454. [19] Coviello T, Maeda H, Yuguchi Y, Urakawa H, Kajiwara K, Dentini M, Crescenzi V. Macromolecules 1998;31:1602. [20] Guenet JM. Thermoreversible Gelation of Polymers and Biopolymers. London: Academic Press, 1992. [21] Luzzati V, Benoit H. J Appl Cryst 1961;14:297. [22] Bluhm TL, Deslandes Y, Marchessault RH, Pe´ rez S, Rinaudo M. Carbohyd Res 1982;100:117. [23] Chuah CT, Starko A, Deslandes Y, Marchessault RH. Macromolecules 1983;16:1375. [24] Kitamura S, Minami T, Nakamura Y, Isoda H, Takeo K, Kobayashi H, Mimura M, Urakawa H, Kajiwara K, Ohno S. THEOCHEM 1997;395-396:425.
Low-degree oxidized scleroglucan and its hydrogel Hitoshi Maeda a, Giuseppe Rambone b, Tommasina Coviello a,*, Yoshiaki Yuguchi a, Hiroshi Urakawa a, Franco Alhaique b, Kanji Kajiwara a b
a Faculty of Engineering and Design, Kyoto Institute of Technology, Kyoto, Sakyo-ku, Matsugasaki, 606 -8585 Japan Department of Chemistry and Technology of Biologically Acti6e Compounds, Faculty of Pharmacy, ‘La Sapienza’ Uni6ersity, Piazzale A. Moro 5, I-00185 Rome, Italy
Received 28 July 2000; received in revised form 23 January 2001; accepted 25 January 2001
Abstract A controlled oxidation of scleroglucan was performed with sodium periodate to prepare aldehyde derivatives (scleraldehyde) with a low degree of oxidation (10 and 20%), which were utilized for crosslinking reactions with hexamethylenediamine. The structural characterization of scleraldehydes and their corresponding hydrogels was attempted by small-angle X-ray scattering (SAXS). While scleraldehyde with a higher degree of oxidation ( ] 50%), according to an earlier research, was found to disentangle into single chains as the degree of oxidation increases; scleroglucan bearing a low percentage of aldehydic groups (up to 20%) retains mainly the conformation of the natural polysaccharide, thus the system can be represented as composed of triple helices with only minor disentanglements at the sites where the aldehyde groups are present. The hydrogel prepared from scleraldehyde with a low degree of oxidation is brittle and fragmented, in contrast to the elastic/homogeneous hydrogel earlier prepared from scleraldehyde with a high degree of oxidation. The hydrogel from scleraldehyde with a low degree of oxidation was found to possess a network structure that consisted mostly of the triple helices crosslinked in specific points where the triple helices are disentangled into single chains because of the presence of the aldehyde groups. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Scleroglucan; Hydrogel; SAXS
1. Introduction Scleroglucan, a neutral extracellular polysaccharide secreted by fungi of the genus sclerotium, has a primary structure which consists of a linear (1 3)-b-D-glucan to which single b-D-glucopyranosyl groups are linked (1 6) at every third unit of the main chain. It is known that scleroglucan assumes a triple-stranded helical conformation in aqueous solution and a single coiled disordered conformation in methylsulphoxide or at high pH values (NaOH\ 0.2 M) [1 – 4]. Polysaccharides containing this repeating unit and with a molecular weight higher than 104 are known to possess an anti-tumor activity against Sarcoma 180 [5]; furthermore, scleroglucan has been proposed for the formulation of sustained delivery dosage forms [6–9]. * Corresponding author. Tel.: +39-0-64-9913300; fax: + 39-0-649913133. E-mail address: [email protected] (T. Coviello).
Carboxylated and crosslinked derivatives of scleroglucan were also studied for the preparation of controlled release matrices capable to respond to different environmental conditions (e.g. pH or ionic strength) [10]. The b-D-glucosyl side chains can be mildly oxidized with sodium periodate [11,12] to yield the aldehyde derivative (scleraldehyde, Fig. 1), which is converted into hydrogel by a crosslinking reaction with diamine [13 –15]. While the complete oxidation disentangles the triple-stranded helical conformation, scleraldehyde with a lower degree of oxidation (up to 20%) was found to retain mainly the triple-stranded helices. In an earlier paper [16], hexamethylenediamine was used for the crosslinking reaction on scleraldehyde with a high degree of oxidation (40 and 100%) but not all diamines were actually involved in crosslinking; here the crosslinked domains were found to consist of parallel-aligned multiple single chains tied with the diamine. In the present study small-angle X-ray scattering (SAXS) technique was used to investigate the structural
0141-8130/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S0141-8130(01)00126-X
352
H. Maeda et al. / International Journal of Biological Macromolecules 28 (2001) 351–358
characteristics of scleraldehyde with a lower degree of oxidation, and of its gel obtained with the same amine as crosslinking agent. The preparation of scleraldehyde with a 10 and 20% degree of oxidation [11,12] and the subsequent crosslinking reaction are also described. Obtained information on the structure of the polymeric networks can be useful both for the interpretation of the behavior of the gels, when used as drug delivery systems, and for the optimization of controlled release dosage form formulations.
2. Experimental
2.1. Materials and methods Scleroglucan (Actigum CS11) was kindly provided by Mero-Rousselot-Satia (France). Hexamethylenediamine (HMD) was purchased from Aldrich (Germany). All other chemicals were of analytical grade and used as received. Scleroglucan was dissolved in water (0.5% w/v) and then sonicated at 20 MHz and 100 W for 2 h (Probe Ultrasonic Processor, Sonics and Materials VC300) to
Fig. 1. Oxidation reaction scheme of scleroglucan.
H. Maeda et al. / International Journal of Biological Macromolecules 28 (2001) 351–358
353
Fig. 2. SAXS profiles observed from scleroglucan, scleraldehyde 10%, scleraldehyde 20% and scleraldehyde 100% in aqueous solution. Ordinates are given in arbitrary units.
reduce the original molecular weight, that from a value of 1.4×106 became about 500000. The molecular weight of the sample was determined before and after sonication by means of viscosity data. Viscosity measurements were carried out at 25°C ( 90.01°C) using a Schott– Gerate equipment and an Ubbelhode capillary viscometer with internal diameter of 0.53 mm. Inserting the values of the extrapolated intrinsic viscosity in the Mark–Houwink –Sakurada equation valid for scleroglucan [17], the molecular weights of native and sonicated samples were obtained. The sonicated scleroglucan solution was filtered through a 0.8 mm pore membrane and the polysaccharide was precipitated with isopropyl alcohol. The obtained polymer was re-dissolved in distilled water and dialyzed exhaustively against distilled water. The solution was finally freeze-dried to yield purified scleroglucan. A weighed amount of purified scleroglucan was dissolved in water (0.5% w/v) and an appropriate amount of sodium periodate was added. The solution was magnetically stirred and kept in the dark for 24 h before the purification by exhaustive dialysis against distilled water
at 4°C. By means of this method scleroglucan was oxidized to scleraldehyde [11,12] of two different degrees of oxidation (i.e. 10 and 20%) by adjusting the stoichiometric ratio [periodate]/[scleroglucan]. The actual degree of oxidation was confirmed by potentiometric titration of a small amount of the carboxylated samples obtained by further oxidation of the aldehyde groups by sodium chlorite [18]. Those samples are referred to as scleraldehyde 10% and scleraldehyde 20%, respectively. The crosslinking reaction was performed directly on the solution containing scleraldehyde 10 and 20%, by addition, while stirring, of an excess of HMD at 40°C (Fig. 1). Unlike in the case of scleraldehyde 40 and 100% gels, no homogeneous phase was constituted, and the new gels were fragmented and floated in the solution. They were kept overnight at the same temperature and then dialyzed against distilled water, at 4°C, in order to remove the excess of HMD before the SAXS measurements. A very low percentage of diamine reacted only with one of its amine group, since a negligible amount of free NH2 groups was still present in the crosslinked polymer, as indicated by the colorimetric test with sodium nitroprusside.
354
H. Maeda et al. / International Journal of Biological Macromolecules 28 (2001) 351–358
2.2. Small-angle X-ray scattering The SAXS measurements were performed at BL-10C of the Photon Factory, Tsukuba, Japan. An incident X-ray was monochromatized to u= 0.149 nm and focused to the position of the detector with a bent focusing mirror. The scattered X-ray was detected by a one-dimensional position sensitive proportional counter (PSPC) positioned at the distance of about 1 m from the sample holder. A flat sample cell of 0.2 cm pathlength made of glass with 20 mm thick quartz windows was used for the solutions. Gels were cut in an appropriate size and sandwiched with 20 mm thick quartz plates inserted in an empty cell. The temperature of the cell was controlled by circulating water of a constant temperature through the cell holder. The solutions were directly injected into the cell, which was placed in the cell holder at least 10 min prior to the SAXS measure-
ments. The SAXS intensities were accumulated for a total of 600 s, in order to assure enough statistical accuracy without degrading the polysaccharide samples by X-ray irradiation. The scattering intensities were corrected with respect to the variation of the incident X-ray flux, by monitoring with an ion chamber installed in front of the cell holder. The X-ray absorption of the solution was compensated by measuring the incident and transmitted X-ray intensities. The excess scattering intensities were calculated by subtracting the scattering intensities of the solvent from those of sample solutions or gels.
3. Results and discussion Gels prepared from the scleraldehyde with a low degree of oxidation (up to 20%) showed a completely
Fig. 3. Molecular models for scleroglucan triple helix and single coil.
H. Maeda et al. / International Journal of Biological Macromolecules 28 (2001) 351–358
355
Fig. 4. Observed SAXS profiles (open symbols) and the scattering profiles calculated (solid lines) from the triple helix (Fig. 3) for scleroglucan, scleraldehyde 10, 20 and 100%. Ordinates are given in arbitrary units.
different appearance from those prepared from scleraldehyde with a high degree of oxidation (] 50%). The former gels are brittle and fragmented, whereas the latter ones are elastic and homogeneous. The SAXS was observed from the scleraldehyde samples with the two degrees of oxidation, i.e. 10 and 20%. Fig. 2 shows typical SAXS results obtained from the solutions of 10 and 20% scleraldehyde in the double logarithmic plots; for an appropriate comparison, also the data from scleroglucan and scleraldehyde 100% are shown. Here I(q) denotes the scattered intensity and q is the magnitude of the scattering vector given by (4y/u)sin(q/2) with q and u being the scattering angle and the wavelength of an incident beam, respectively. The conformation of scleraldehyde seems not to change when the degree of oxidation is low (samples 10 and 20%) as seen from the observed SAXS profiles. The observed scattering profiles were analyzed by assuming that most part of the scleraldehyde chain is composed of triple-stranded helix in terms of a modified broken rod model [19,20], which is expressed as
q 2I(q) :% q 2wi ML,i Ui (q)+ const. c i
(1)
Here the constant term takes into account the spatial correlation of the components [21], and accounts for the structural characteristics of the boundary of triplestranded helices to disentangled chains. ML,i is a linear molar mass of the component i. In the present case, i denotes either a triple-stranded helix component or a single chain component, and Ui (q) is a corresponding particle scattering factor calculated from the atomic coordinates of the respective molecular models (see Fig. 3) according to the Debye formula n
n−1
Ui (q)= % f 2j g 2j (q)+ 2 % j=1
n
%
j=1 k=j+1
sin(djkq) djkq (2)
fj fkgj (q)gk (q)
where fj and djk denote the atomic scattering factor of the atom j and the distance between the jth and kth atoms, respectively. The molecular model for a triple helix was adapted from the crystallographic data [22,23] and that for a single coil was generated by the Monte Carlo method according to the energy map of lami-
356
H. Maeda et al. / International Journal of Biological Macromolecules 28 (2001) 351–358
naribiose [24]. The form factor gj (q) for a single atom is assumed to be represented by the form factor of a rigid sphere having the radius equivalent to the van der Waals radius Rj of the jth atom as gj (q) =
3[sin(Rj q)− (Rj q)cos(Rj q)] (Rj q)3
(3)
Rj was set to 0.501 nm or 0.450 nm for a carbon or an oxygen atom, respectively, in the present calculation, which is three times larger than the respective van der Waals radius in order to take into account the effect of hydrated water along the polysaccharide chain. The molecular models of a triple-stranded helix and a single chain are composed of 80 and 100 residues, respectively. Scleraldehydes 10 and 20% seem to assume a triplestranded helical conformation in aqueous solution, as a broken rod model with a single triple-helix component fits well with the observed SAXS profiles. The fitting example will be seen from Fig. 4, which shows the observed SAXS profiles and the calculated scattering profiles for scleraldehyde 10 and 20% in aqueous solution in terms of the Kratky plots (q 2I(q) plotted against q); for an appropriate comparison, also the data obtained with scleroglucan and 100% scleraldehyde are shown. Scleraldehyde 100% consists of single coils, while about 50% of the chain was found to be involved in a single coil in the case of scleraldehyde 40% [16]. Thus one can assume that oxidation of side groups causes disentangling the triple-stranded helix as found for the carboxylated scleroglucan (sclerox) [19]. However, a degree of oxidation up to 20% will not cause remarkable conformational changes from a triple-
Fig. 6. Molecular model for the network of scleraldehyde hydrogel tied with hexamethylenediamine. Scleraldehyde 10% gel (top) and scleraldehyde 20% gel (bottom).
Fig. 5. SAXS profiles from the hydrogels of scleraldehyde 10, 20 and 100%. Ordinates are given in arbitrary units.
stranded helix to a single coil, and the observed SAXS profiles from scleraldehyde 10 and 20% are almost identical to the one of scleroglucan (Fig. 2). The aldehyde groups can react with amino groups, and a hydrogel can be formed by coupling scleraldehyde with diamine. Since the coupling reaction takes place in alkaline conditions (pH\ 12), all triplestranded helices are disentangled into single chains as it occurs for scleroglucan at these pH conditions. When the degree of oxidation is high (over 40%), scleraldehyde is prevented from reforming a triple-stranded helix after the reaction with HMD [16]. On the contrary, when the degree of oxidation is lower than 20%, scleraldehyde may resume a triple-stranded helix after the
H. Maeda et al. / International Journal of Biological Macromolecules 28 (2001) 351–358
reaction, during the dialysis process that decreases remarkably the environmental pH value. The SAXS profiles obtained with the hydrogels prepared from scleraldehyde 40 and 100% revealed a characteristic peak at q3 nm − 1, which was attributed to the fine structure in the crosslinked domain composed of scleraldehyde chains aligned in parallel [16]. Once HMD ties two scleraldehyde chains, further crosslinking will take place around that tie because more chance will be available for finding a counterpart. When the degree of oxidation is low, no aldehyde group will be available in the proximity of a crosslinked site. As seen from Fig. 5, no characteristic peak at q 3 nm − 1, due to scleraldehyde chains aligned in parallel, was observed from the hydrogel of scleraldehyde 10 and 20%. That is, the lower degree of HMD tie would not prevent scleraldehyde 10 and 20% gel from resuming a triple-helical conformation, although no triple helices are involved in the crosslinked domain of scleraldehyde 40 and 100% gels. Taking into account that scleraldehyde with a low
357
degree of oxidation resumes a triple-helical conformation in gel and that the crosslinking reaction takes place where aldehyde groups are available, the resulting network may look like the model proposed in Fig. 6. Here the model network is built by assuming a random orientation of triple helices and a crosslinking at a site where two aldehyde groups meet. A scleraldehyde triple helix disentangles into single chains at the position where an aldehyde group is introduced. Since the probability of two or more consequent residues being oxidized is extremely small for scleraldehyde with a degree of oxidation 20% or less, the network is thought to be constituted mostly of rigid triple helices. The disentangled part (composed of 6 or 12 residues) was introduced according to the degree of oxidation. Several network fragments were generated where the fragment was assumed to consist of four triple helices (in the case of scleraldehyde 10% gel) or eight triple helices (in the case of scleraldehyde 20% gel). The SAXS profiles were calculated according to Eq. (1) with the network fragments (represented in Fig. 6)
Fig. 7. The SAXS profiles observed from scleraldehyde 10% hydrogel, 20% hydrogel, 40% hydrogel and 100% hydrogel. The solid lines indicate the scattering profiles calculated according to the respective molecular models (Fig. 6) from Eq. (1). Ordinates are given in arbitrary units.
358
H. Maeda et al. / International Journal of Biological Macromolecules 28 (2001) 351–358
and single coils (see Fig. 3). Fig. 7 shows the comparison of the observed SAXS profiles with the profiles calculated by Eq. (1) for scleraldehyde 10% hydrogel, 20% hydrogel, 40% hydrogel and 100% hydrogel. The calculated profiles fit well with the observed SAXS profiles from scleraldehyde 10 and 20% gels, indicating that the model network composed of triple helices represents at least qualitatively a real network structure in those hydrogels. The use of Eq. (1) implies the random spatial correlation between the network fragments. It can then be assumed that those network fragments are linked in a random fashion to yield the calculated scattering profiles. In either case, the fraction of the single coil component was small and less than 9%.
4. Conclusions Scleraldehyde with a low degree of oxidation (10 and 20%), prepared by a controlled oxidation of scleroglucan, was found to retain essentially a triple-stranded helical conformation, while the triple-stranded chain disentangles in single chains with increasing the degree of oxidation (40 and 100%) [14]. The hydrogel prepared from scleraldehyde with a low degree of oxidation, according to SAXS profiles analysis, can be represented by a network composed of randomly oriented triple helices interlinked at the sites where the aldehyde groups are present. Hydrogels are often used for the formulation of controlled release formulations and it is well known that delivery profiles from polymeric networks can be remarkably affected by the actual structure of the polymeric system; thus the structural characterization of these scleroglucan derivatives appears to be very important as they can be proposed as new matrices for sustained release of bioactive materials and the possibility to shed some light upon their structure can lead to an improvement of the performances of new delivery systems.
Acknowledgements Y.Y. acknowledges a financial support of JSPS Research Fellowship for Young Scientists. K.K. is indebted to ‘La Sapienza’ University for a visiting
.
professorship, and JSPS/CNR for a personal grant. This work was carried out under the approval of the Photon Factory Advisory Committee (Proposal No. 94G-291) and was supported in part by MURST funds.
References [1] Norisuye T, Yanaki T, Fujita H. J Polym Sci Polym Phys Ed 1980;18:547. [2] Yanaki T, Kojima T, Norisuye T. Polym J 1981;13:1135. [3] Sato T, Norisuye T, Fujita H. Carbohydr Res 1981;95:195. [4] Sato T, Norisuye T, Fujita H. Macromolecules 1983;16:185. [5] Norisuye T. Makromol Chem Suppl 1985;14:105. [6] Touitou E, Alhaique F, Riccieri FM, Riccioni G, Santucci E. Drug Des Deliv 1989;5:141. [7] Alhaique F, Carafa M, Riccieri FM, Santucci E, Touitou E. Pharmazie 1993;48:432. [8] Rizk S, Duru C, Gaudy D, Jacob M, Ferrari F, Bertoni M, Caramella C. Int J Pharm 1994;112:125. [9] Crescenzi V, Dentini M. In: Ottenbrite RM, Huang SJ, Park K, editors. Hydrogels and Biodegradable Polymers for Bioapplications. ACS Symposium Series 627. Washington, DC: American Chemical Society, 1996:221 – 32. [10] Coviello T, Grassi M, Rambone G, Santucci E, Carafa M, Murtas E, Riccieri FM, Alhaique F. J Controlled Release 1999;60:367. [11] Rankin JC, Jeanes A. J Am Chem Soc 1954;76:4435. [12] Jeanes A, Wilham CA. J Am Chem Soc 1950;72:2655. [13] Alhaique F, Coviello T, Carafa M, Murtas E, Riccieri FM, Santucci E. Third Jerusalem Conference on Pharmaceutical Sciences and Clinical Pharmacology-JC, Jerusalem, 1996; p. 20. [14] Coviello T, Rambone G, Alhaique F, Carafa M, Murtas E, Riccieri FM, Santucci E. Polymer Networks 96 Doorn, The Netherlands,1996; 17. [15] Alhaique F, Coviello T, Carafa M, Murtas E, Rambone G, Riccieri FM, Santucci E, Paci M. Proc Int Symp Control Rel Bioact Mater 1997;24:1043. [16] Maeda H, Coviello T, Yuguchi Y, Urakawa H, Rambone G, Alhaique F, Kajiwara K. Polym Gels Networks 1998;6:355. [17] Yanaki T, Norisuye T, Fujita H. Macromolecules 1980;13:1466. [18] Hofreiter BT, Wolff IA, Mehltretter CL. J Am Chem Soc 1957;79:6454. [19] Coviello T, Maeda H, Yuguchi Y, Urakawa H, Kajiwara K, Dentini M, Crescenzi V. Macromolecules 1998;31:1602. [20] Guenet JM. Thermoreversible Gelation of Polymers and Biopolymers. London: Academic Press, 1992. [21] Luzzati V, Benoit H. J Appl Cryst 1961;14:297. [22] Bluhm TL, Deslandes Y, Marchessault RH, Pe´ rez S, Rinaudo M. Carbohyd Res 1982;100:117. [23] Chuah CT, Starko A, Deslandes Y, Marchessault RH. Macromolecules 1983;16:1375. [24] Kitamura S, Minami T, Nakamura Y, Isoda H, Takeo K, Kobayashi H, Mimura M, Urakawa H, Kajiwara K, Ohno S. THEOCHEM 1997;395-396:425.