Oxidized Chondroitin Sulfate-Cross-Linked Gelatin Matrixes: A New

Biomacromolecules 2005, 6, 2040-2048

2040

Oxidized Chondroitin Sulfate-Cross-Linked Gelatin Matrixes: A New Class of Hydrogels S. Dawlee,† A. Sugandhi,† Biji Balakrishnan,† D. Labarre,‡ and A. Jayakrishnan*,† Polymer Chemistry Division, Biomedical Technology Wing, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Satelmond Palace, Poojapura, Trivandrum, Kerala 695 012, India, and Centre d'Etudes Pharmaceutiques, UMR CNRS 8612, Universite´ Paris-Sud, 5 rue J.B. Cle´ ment, 92296 Chatenay-Malabry Cedex, France Received January 8, 2005; Revised Manuscript Received March 5, 2005

A naturally occurring glycosaminoglycan such as chondroitin-6-sulfate was first converted in to its aldehyde derivative by periodate oxidation and used as a cross-linking agent for gelatin giving rise to a new class of hydrogels. Cross-linking was predominantly due to Schiff’s base formation between the -amino groups of lysine or hydroxylysine side groups of gelatin and the aldehyde groups in oxidized chondroitin sulfate. The hydrogels were prepared from chondroitin sulfate with different degrees of oxidation and gelatin. They were characterized for degree of cross-linking, cross-linking density, equilibrium swelling, water vapor transmission rate, internal structure, and blood-compatibility. Degree of cross-linking of the gels determined by trinitrobenzene sulfonic acid assay showed that, the higher the degree of oxidation of the polysaccharide, the higher the degree of cross-linking. Examination of the internal structure by scanning electron microscopy showed that the hydrogels were highly porous in nature with interconnecting pores ranging from 50 to 200 µm. Equilibrium swelling showed that the gels retained about 90% water and did not undergo dehydration rapidly. The hydrogels were nontoxic and blood-compatible. Since an important phase of early wound healing has been shown to involve secretion of glycosaminoglycans such as chondroitin sulfate by fibroblasts which form a hydrophilic matrix suitable for remodeling during healing, this new class of hydrogels prepared from chondroitin sulfate and gelatin without employing any extraneous cross-linking agents are expected to have potential as wound dressing materials. Introduction Although there has been considerable progress over the past 20 years in the technology for wound care management, there is still no ideal wound dressing material. Numerous natural and synthetic materials have been investigated and used as wound dressings.1 One of the earliest wound dressing based on the principles of wound healing was that of Yannas and Burke2 which consisted of a collagen-chondroitin sulfate (CS) sponge as the inner layer with an outer layer of silicone. CS was included in collagen to introduce a more porous structure in the dressing, to increase its elastic modulus as well as to impart resistance to degradation by collagenase. Similar dressings were developed later by a modification of the method of Yannas by Matsuda et al.3,4 They reported that the addition of CS reinforced the mechanical properties of collagen but decreased cell proliferation. CS is a glycosaminoglycan (GAG) in the extracellular matrix of all vertebrates and occurs both in the skeletal and soft connective tissue. It is a high viscosity mucopolysaccharide comprised of alternating units of β-1,4-linked glucuronic acid and β-1,3-N-acetyl galactosamine and is sulfated on either the 4 or 6 position of the galactosamine * To whom correspondence should be addressed. E-mail: dr_jkrishnan@ sify.com. Tel: +91-471-2340801. Fax: +91-471-2341814. † Sree Chitra Tirunal Institute for Medical Sciences and Technology. ‡ Universite ´ Paris-Sud.

residue. CS is the most abundant mucopolysaccharide in the body and is generally linked to a core protein thus producing a proteoglycan. Recent reports suggest that an important phase of early wound healing involves secretion of GAGs such as CS by fibroblasts which form a hydrophilic matrix suitable for remodeling during healing.5 In a recent study, it was shown that CS covalently cross-linked with collagen using a watersoluble carbodiimide or N-hydroxysuccinimide appeared to promote cellular in-growth and cartilage tissue formation in rabbits.6 Gelatin-CS gels cross-linked with similar crosslinking agents and impregnated with antibacterial proteins have been shown to reduce or prevent valve endocarditis when impregnated into the Dacron sewing ring of heart valve prostheses.7 CS has been reported to accelerate wound healing in sinonasal mucosa in a rabbit model.8 It has been suggested that CS acts as a surrogate extracellular matrix, serving as a repository for cytokines and growth factors produced by the regenerating mucosa. CS-coated poly(2-hydroxyethyl methacrylate) membranes were found to prevent adhesion in fullthickness tendon tears of rabbits.9 GAG hydrogels composed of poly(ethylene glycol) dialdehyde cross-linked with adipic dihydrazide derivatives of CS and hyaluronic acid have been evaluated as bio-interactive dressings for wound healing by Kirker et al.10 They found that wounds treated with such

10.1021/bm050013a CCC: $30.25 © 2005 American Chemical Society Published on Web 04/14/2005

Oxidized Chondroitin Gelatin Matrixes

films showed more dermal collagen regeneration and organization. It was shown that wounds exposed to hydrogel films containing CS were moist and hydrated, thus demonstrating that evaporative water loss and wound dehydration had been prevented. The acceleration of re-epithelialization with GAG indicated an enhancement of the overall healing process. CS is an endogenous substance present in different layers of the skin such as epidermis, dermis, and basement membrane. During the maturation of the wound, the concentrations of CS and dermatan sulfate tend to increase. These proteoglycans are reported to play an essential role at every stage of wound healing.11 Chitosan-CS and chitosan-hyaluronic acid polyelectrolyte complexes have also been evaluated as possible wound dressing materials.12 The chitosan-GAG complexes were found to inhibit cell proliferation in vitro which was attributed to the complete loss of charges borne by each polymer due to complexation. Wound healing was better due to chitosan alone as compared with the complexes in the short term, but the authors caution that long-term implantation studies are needed to reconfirm this observation. Choi et al.13 evaluated gelatin/hyaluronic acid, chitosan/hyaluronic acid, and gelatin/alginate gels prepared by cross-linking with the water-soluble carbodiimide N,N′-(3-dimethylaminopropyl)N′-ethyl carbodiimide (EDC) as wound dressings and found that the gelatin/hyaluronic acid gels showed highest healing performance, an indication of the positive role of GAG in the wound healing process. Gelatin has a long history of medical use as a plasma expander, as adhesive and absorbent pads, and as a wound dressing material.14,15 Dextran dialdehyde cross-linked gelatin hydrogels have been investigated as a bioactive wound dressing for delivery of growth factors.16,17 Therefore, hydrogel dressings containing CS would have many beneficial effects when used as a wound dressing material. In this work, we describe the preparation and characterization of cross-linked hydrogels from oxidized CS and gelatin without employing any extraneous cross-linking agents. CS was oxidized with sodium m-periodate to generate reactive aldehyde functions that would enter into Schiff’s reaction with amino groups present on gelatin to give a crosslinked gel. Although the exact role of GAG molecules in wound healing remains unresolved, we believe that such a composite matrix would have the haemostatic effect of gelatin and possibly the extracellular signaling and cell recognition properties of CS. Here, oxidized CS is made to react directly with gelatin to form the cross-linked gel, thus avoiding the use of toxic cross-linking agents such as carbodiimides, glutaraldehyde etc., thereby giving rise to potentially nontoxic gels which could find applications for wound management, drug delivery, tissue engineering, and other related applications. Materials and Methods Materials. Chondroitin-6-sulfate (CS) isolated from the cartilage of shark was procured from Fluka Chemie GmbH, Switzerland. Gelatin (type A, Bloom 300, molecular weight 100 000), 2,4,6-trinitrobezene sulfonic acid (TNBS), and

Biomacromolecules, Vol. 6, No. 4, 2005 2041

sodium m-periodate were obtained from Sigma Chemical Co, St. Louis, MO. Disodium hydrogen phosphate, monosodium hydrogen phosphate, sodium chloride, potassium iodide, and sodium thiosulfate were of analytical grade purchased from S. D. Fine Chemicals, Mumbai, India. Dialysis tubing of molecular weight (MW) cut off 3500 Da (SpectraPor) was from Spectrum Laboratories, CA. Cephalin reagent was from Diagnostica Stago, France. All other reagents were of analytical grade procured locally and water used was double distilled. Phosphate buffered saline (PBS, pH 7.4, 0.1 M) was prepared by dissolving 17.97 g of di-sodium hydrogen phosphate, 5.73 g of monosodium hydrogen phosphate and 9 g of sodium chloride in 1L distilled water. Methods. Periodate Oxidation of CS. Controlled periodate oxidation of CS leads to the opening of the vicinal hydroxyls of glucuronic acid ring resulting in the formation of two aldehyde functions per ring. Thus, 5 g of CS was dissolved in 100 mL of distilled water and reacted with stoichiometric amounts of NaIO4 at 20 °C in the dark with constant stirring for 6 h. The amount of periodate was varied to get different degrees of oxidation. The extent of oxidation was followed by determining the concentration of periodate left unconsumed by iodometry after 6 h.18 Briefly, a 5 mL aliquot of the reaction mixture was neutralized with 10 mL of sodium bicarbonate solution and iodine was liberated by the addition of 20% potassium iodide solution (2 mL). Liberated iodine was then titrated with standardized sodium thiosulfate solution using starch as the indicator. The oxidized CS was recovered from the reaction mixture by dialysis and lyophilization. Dialysis was done against distilled water (2 L) at room temperature for 2 days with several changes of water until it was periodate free. Removal of periodate was ascertained by the absence of turbidity of dialyzate with a solution of silver nitrate. The solution was then frozen and lyophilized to obtain a white cotton-like material. The final yield of the product depended on the extent of oxidation and ranged from 80% to 20% depending on the degree of oxidation (20% to 80%). Preparation of Chondroitin Dialdehyde-Gelatin Gel. Chondroitin sulfate dialdehyde (CSD) of different degrees of oxidation was made to react with gelatin to form Schiff’s linkage between free amino groups of the protein and the aldehyde groups of the polyaldehyde. A total of 1 mL of a 10% solution of CS of different degree of oxidation in PBS was added to 1 mL of a 20% aqueous solution of gelatin in glass vials of 10 mL capacity, and mixed using a Teflon magnetic stir bar. About 0.5 mL of this mixture was then added to different wells of a 4-well tissue culture plate (Nunc, Denmark). After 15 min at room temperature, the plates were kept in a refrigerator at 4 °C for 24 h to get gels of 15 mm diameter and 2 mm thickness. The surface of the gels was washed once with distilled water to remove crystallized buffer salts, frozen at -20 °C in a deep-freezer overnight and lyophilized to get dry gels which were stored in a desiccator at 4 °C in the refrigerator. Degree of Cross-Linking. The cross-linking degree of the hydrogels was determined by TNBS assay.19,20 About 5 mg of the lyophilized gels was weighed in to a test tube. Into this, 4 mL of distilled water was added, followed by 1 mL

2042

Biomacromolecules, Vol. 6, No. 4, 2005

Dawlee et al.

of 0.5% of TNBS solution, and 1 mL of 4% sodium bicarbonate (pH 8.5). It was then heated in a water bath maintained at 60 °C for 4 h. The unreacted gelatin in the hydrogel reacts with TNBS and forms a soluble complex. A total of 1 mL of this solution was further treated with 3 mL of 6 N HCl at 40 °C for 1.5 h, and its absorbance was determined at 334 nm after suitable dilution spectrophotometrically (Spectronic, Genesis 2, NY). A standard curve was plotted for non-cross-linked gelatin by treating various concentrations of gelatin by TNBS in a similar manner. Experiments were done in triplicate. The cross-linking degree was obtained from the differences between the absorbance values before and after cross-linking. It was determined by the following equation:

where χ1 is the interaction parameter, f is the cross-linking functionality, V1 is the molar volume of water (18.062 cm3/ mol), and υ2 is the volume fraction of polymer in the hydrogel when its reaches the equilibrium swelling state. The interaction parameter, χ1, was assumed to be 0.35 as it has been previously reported for similar interaction.20 It has been reported that the reactive functional groups present per 100 g of high quality gelatin are primarily hydroxyl, carboxyl, and amino at an amount of approximately 100, 75, and 50 meq of each of these groups, respectively.22 On this basis, the functionality of gelatin in terms of reactive amino groups was assumed to be 50. The molecular weight between crosslinks in the polymer M h c was determined by dividing the density of the polymer Fp by υe

degree of cross-linking (%) ) absorbance of cross-linked gel 1× 100 absorbance of non-cross-linked gel

M h c ) Fp/υe

{

}

Swelling Studies. The equilibrium weight swelling ratio (q) was experimentally determined using the following equation: q ) weight of swollen gel (Ws)/weight of dry gel (Wd) The percentage equilibrium water content or equilibrium hydration (H) of the hydrogel was calculated from the equation H (%) ) [1 - 1/q] × 100 Initially the sample gel was weighed in to a vial. To this was added 5 mL of PBS, and it was kept in water bath maintained at 37 °C. PBS was aspirated from the vials at different periods of time, and the gels were gently blotted prior to weighing. The experiment was repeated for a period of time until the weight varied less than 0.5%. The dry weight of the gel was determined by drying the gel in a vacuum oven at 50 °C till constant weight was obtained. Experiments were done in triplicate for each gel preparation. The degree of swelling of the gels is typically expressed as the equilibrium volume swelling ratio Q, defined as the volume of the swollen gel divided by the volume of the gel before swelling. The volume fraction of the polymer (υ2) in the hydrogel, which is a measure of interaction between polymer chains, is the reciprocal of Q. The volume of the hydrogel in the swollen state and in the dry state was obtained by determining their weights in air and in n-heptane, a nonsolvent for the polymer and calculated using the buoyancy principle with the following equation:21 V)

Ws - Wh Fh

where V is the volume of the polymer, Ws is the weight of the polymer in air, Wh is the weight of the polymer in n-heptane, and Fh is the density of n-heptane (0.684 g/cm3). Cross-linking density (υe, mol/cm3) of the hydrogels was subsequently calculated from the Flory- Rehner equation υe ) - [ln(1 - υ2) + υ2 + χ1υ22] [V1(υ21/3 - 2 υ2/f)]-1

Water Vapor Transmission Rate (WVTR). In most studies on wound coverings, water vapor transmission rate is defined as the steady flow of water vapor per unit area of surface in unit time at a specified humidity and temperature.23-25 The water vapor transmission rate is calculated from the weight loss of the reservoir that is filled with water and covered by the experimental membrane as stipulated by a modified ASTM standard method.25 A 10% solution of CDS in PBS was reacted with 20% solution of gelatin as before to obtain the membranes of 2 mm thickness and 15 mm diameter for water vapor transmission studies. The membrane was placed on the mouth of a cylindrical plastic vial of 1.4 cm diameter filled with 10 mL of distilled water, and the edges were sealed with Parafilm. The vials were kept in an incubator at 37 °C and 40% relative humidity. Evaporation of water through the test membrane was monitored by measurement of change in weight of the vial. A vial covered with Parafilm was used as the control. Two samples of each membrane were tested for a period of 7 h. WVTR was obtained by dividing the slope of water loss versus time curve by the transmission area. WVTR ) (Wi -Wt)/At where Wi is the initial weight of the assembly in grams, Wt is the weight of the assembly in grams after time t, A is the WVTR test area of the sample in m2, t was the time duration between Wt and Wi in h. EVaporation of Water from Gels. As a part of the characterization study, the rate of evaporation of water from hydrogels made from CS of different degrees of oxidation and gelatin was determined. For studying the evaporation rate, the gels were kept in an incubator maintained at 37 °C and 40% relative humidity and weighed at regular intervals. The percentage water loss was calculated by the following formula: water loss (%) )

(initial weight - final weight) × 100 initial weight

Scanning Electron Microscopy (SEM). SEM to examine the surface morphology and internal structure of the gels was done on lyophilized samples. For examining the internal structure, samples were cut with a razor blade; the cut surface



was coated with a thin layer of gold and examined in the microscope (Hitachi, model S-2400, Japan) in the usual way. Cytotoxicity EValuation. Cytotoxicity evaluation of CSDcross-linked gelatin gel was carried out by the direct contact assay with a monolayer of L929 mouse fibroblast cells according to ISO standards.26 Briefly L929 cells were subcultured from stock culture (National Centre for Cell Sciences, Pune, India) by trypsinization and seeded onto multi-well tissue culture plates (Nunc, Denmark). Cells were fed with Dulbecco’s minimum essential medium supplemented with bovine serum and incubated at 37 °C in 5% carbon dioxide atmosphere. When the cells attained a monolayer, the material was kept in contact with the cells in triplicate. After 24 h incubation at 37 ( 1 °C, the morphology of the cells was assessed using a phase contrast inverted microscope (Leica, WILD MPS32, Germany) in comparison with negative (high density polyethylene) and positive (copper wire) controls. Cellular responses were scored as 0, 1, 2 and 3 according to noncytotoxic, slightly cytotoxic, moderately cytotoxic, and severely cytotoxic. Blood-Compatibility EValuation. The haemolytic potential of the material is the measure of the extent of haemolysis that may be caused by the material when it comes in contact with blood. Haemolytic potential of hydrogels was determined according to O’Leary and Guess.27 Human blood (0.1 mL) anticoagulated with citrate was added to 7.5 mL of PBS containing gel (∼0.15 g) in different test tubes. A separate positive (100% haemolysis induced by replacing the PBS with 7.5 mL of 0.1% Na2CO3 solution) and a negative (0% haemolysis, PBS with no material added) control were also set up. Each set of experiments was done in duplicate. All the test tubes containing samples and the control were incubated for 1 h at 37 °C. After incubation, the tubes were centrifuged at 300 rpm for 5 min. The percentage haemolysis was calculated by measuring the optical density (OD) of the supernatant solution at 545 nm in a UV-vis spectrophotometer (Spectronic Genesys 2, NY) as per the following formula: haemolyis (%) ) OD of the test sample - OD of negative control 100 × OD of positive control The consumption of red blood cells (RBC) and platelets on contact with the hydrogels was analyzed using blood from human volunteers which was collected into sodium citrate as the anticoagulant in the ratio 9:1. The hydrogels were placed in wells of tissue-culture grade polystyrene Petri dishes and wetted using PBS. Each material was exposed to 1 mL of blood for 30 min under agitation at 75 ( 5 rpm using an environmental bath shaker (Labline Instruments Inc., Illinois, USA) thermostated at 37 ( 1 °C. Samples were withdrawn for analysis immediately after mixing and after 30 min. A well without any material was used as the reference. The consumption of RBC and platelets was analyzed using a hematology analyzer (Cobas Minos, Roche Diagnostics, France) calibrated using WHO-traceable standards. For evaluating the effect of material on blood coagulation, partial thromboplastin time (PTT) was determined on samples exposed to fresh citrated human blood immediately and after

Scheme 1. Periodate Oxidation of Chondroitin Sulfate

Table 1. Oxidation of CS with Sodium m-Periodate CS (g)

NaIO4 (g)

periodate equivalent (%)

oxidation (%)

yield (%)

5 5 5 5

2.070 1.543 1.038 0.513

79.98 59.61 40.08 19.75

79.3 ( 0.7 59.5 ( 0.4 39.6 ( 0.2 19.8 ( 0.4

19.99 ( 3.02 37.33 ( 2.51 68.60 ( 6.49 83.05 ( 3.61

30 min of exposure. Hydrogels were equilibrated for 1 h in PBS, exposed to blood under agitation at 100 rpm using the environmental bath shaker at 37 °C. After 1 and 30 min exposure, blood was centrifuged to obtain platelet poor plasma. Plasma was then mixed with Cephalin reagent and incubated for 3 min before adding CaCl2 solution to initiate clotting. Clotting time was noted using an automated coagulation analyzer (Diagnostica Stago, France). Bloodcompatibility evaluations were done according to ISO standards.28 Results and Discussion Periodate oxidation specifically cleaves the vicinal glycols in polysaccharides to form their dialdehyde derivatives. Each R-glycol group consumes one molecule of periodate, and under given conditions, the rate of the reaction is dependent principally on the stereochemistry of the R-glycol group. Periodate oxidation of CS cleaves only the vicinal hydroxyl groups on the C2 and C3 carbon atoms of β (1f4) linked glucuronic acid unit (Scheme 1). CS of different degrees of oxidation were prepared by employing different periodate equivalents to CS. The degree of oxidation and the yield of the product obtained are given in Table 1. As the amount of periodate is increased, the extent of oxidation is also increased, but the yield of the product obtained showed a systematic decrease with increase in the degree of oxidation. This points out to extensive degradation of the parent compound due to nonspecific oxidation cleaving the glycosidic bonds in the molecule. We employed dialysis membrane of MW cut off 3500, and therefore, it is evident high degrees of oxidation results in chain fragments having MW less than 3500 Da that seeps through the membrane during dialysis. Such nonspecific oxidation leading to cleavage of the backbone polymer has been reported in the case of periodate oxidation of alginates29 and possibly applicable to other polysaccharides similar in structure such as CS.

2044


Dawlee et al.

Table 2. Cross-linking Parameters of Hydrogels degree of oxidation of CS (%)

degree of cross-linking (%)

cross-linking density, υe, x 105 (mol/cm3)

molecular weight between cross-links, M h c (g/mol)

20 40 60

30.49 ( 1.00 37.65 ( 1.91 75.48 ( 2.18

4.30 ( 0.05 7.35 ( 0.03

6405.01 ( 75.01 5635.31 ( 24.61

Hydrogels were prepared by cross-linking gelatin with CSD. Cross-linking is predominantly due to Schiff’s base formation between the -amino groups of lysine or hydroxylysine side groups of gelatin and the aldehyde groups in CSD. Degree of cross-linking of the gels determined by measuring the amount of unreacted gelatin present by TNBS assay showed that, the higher the degree of oxidation, the higher the degree of cross-linking as the presence of a large number of aldehyde groups facilitates the gel formation (Table 2). Gelatin contains 50 meq of reactive amino groups per 100 g.22 Therefore, with equivalent volumes of 20% gelatin and 10% CSD, the amount of reactive aldehyde groups will be higher than the amount of reactive NH2 groups in gelatin in all cases except 20% CSD. However, the TNBS assay showed that only 75% cross-linking could be attained with 60% CSD. It has been reported that aldehyde groups of hexuronic acid forms hemiacetals with the hydroxyl groups of unoxidized residues that are situated in the remote portions in the same chain.29 Therefore, all of the available aldehyde functions will not be available for Schiff’s reaction and this could explain the incomplete reactivity of gelatin even in the presence of excess aldehyde groups theoretically present in the oxidized polysaccharide. SEM examination showed that the surface of gel prepared from 20% oxidized CS was very porous whereas gels prepared from 40 and 60% oxidized CS had more uniform and less porous appearance (Figure 1). It was seen that with 20% oxidized CS, the degree of cross-linking is less and therefore, less gelatin has been consumed in the cross-linking process. Washing the gel suface to remove crystallized buffer salts also removes uncross-lined gelatin on the surface resulting in a porous surface. With 40 and 60% oxidized CS, more gelatin is consumed in the cross-linking process resulting in a denser and less porous surface. Examination of the internal structure, however, showed that all of the hydrogels were highly porous in nature with interconnecting pores ranging from 50 to 200 µm (Figure 2). With an increase in the degree of oxidation of CS used for cross-linking gelatin, however, there was no discernible gross difference in the internal pore structure of the gels. The swelling properties of the hydrogels prepared from CSs of different degrees of oxidation and gelatin were examined. Gels prepared from CSs of different degrees of oxidation attained equilibrium swelling rapidly (Figure 3). Equilibrium fluid content was found to be between 88 and 92%. Statistical analysis (ANOVA) showed that there was no significant difference between the water uptake ability of gels prepared from 60% and 40% oxidized CS (p > 0.05). Although the degree of cross-linking is different for these two gels, the equilibrium swelling did not show any difference in the water uptake ability. However, water uptake

Figure 1. SEM of surface morphology of CSD-cross-linked gelatin hydrogels. 20% (a), 40% (b), and 60% oxidized CS (c).

ability of gels prepared from 20% oxidized CS was significantly different from other gels (p < 0.001). This lower water uptake ability is believed to be due to poor cross-linking and the resultant dissolution of the un-cross-linked gelatin in the medium since the aldehyde content in 20% oxidized CS itself is not theoretically sufficient to cross-link the entire gelatin unlike in the case of 40 and 60% oxidized CS. To examine this discrepancy further, the cross-linking density (υe) and the molecular weight between cross-links (M h c) of the gels were estimated. With gel prepared from 20% oxidized CS, reproducible results could not be obtained for both υe and M h c. This is attributed to the poor degree of cross-



Figure 3. Equilibrium swelling of CSD-cross-linked gelatin hydrogels. 20% (9), 40% ()), and 60% oxidized CS (1).

Figure 2. Internal structure CSD-cross-linked gelatin hydrogels. 20% (a), 40% (b), and 60% oxidized CS (c).

linking and the resulting dissolution of gelatin in the medium introducing significant errors in the swelling measurements. With 40 and 60% CSD cross-linked gels however, both υe and M h c were as expected, the 60% CSD-cross-linked gels showing a higher cross-linking density and lower molecular weight between cross-links as compared to the 40% CSDcross-linked gels (Table 2). Therefore, one would expect a higher degree of swelling for the 40% CSD-cross-linked gels as compared to the 60% CSD-cross-linked gels. However, this was not evident in swelling studies. Interstingly, a similar phenomenon was observed by Keuijpers et al. when gelatinCS gels were prepared by cross-linking with EDC.7 These authors found that a lower cross-linking density was not

associated with a higher degree of swelling for gelatin-CS gels for reasons which were not clear. In the present case, it is possible that, since these gels swelled to absorb large amount of water (∼90%), any small difference in swelling was not discernible due to the errors in measurement. Also, it has been shown that the gelation of gelatin aqueous solutions with oxidized dextran is governed by two strong interactions.17 One is associated with the chemical crosslinking of gelatin and the other is based on the ability of gelatin to form polymer network structures that are stabilized by physical cross-linking (gelatin-gelatin physical structuring). The lack of difference in swelling characteristics between these two gels could possibly be because such physical restructuring is similar for both gels and the change in the degree of cross-linking does not exert significant effect on swelling. Estimation of the cross-linking density of gelatin-CS gels prepared by EDC cross-linking by Kuijpers et al.7 showed that υe was of the order of 2.18 × 10-3 which is 2 orders of magnitude higher that the υe seen in the present investigation (Table 2). This was as expected since an activating agent such as EDC was employed for cross-linking gelatin-CS. The equilibrium water uptake of such gels was 70-75% as opposed to the nearly 90% value seen in the present case. The CSD-gelatin gels are therefore more loosely cross-linked, more porous, and more water-absorbing than those prepared by bifunctional cross-linking agents. One of the advantages of hydrogels as a wound dressing material is that it can provide a moist environment over the wound which is more conducive for healing. The moisture permeability of the hydrogel was determined by measuring the WVTR across the material. WVTR values were calculated by dividing the slopes of the initial linear portion of the water loss (g) versus time (h) curves by transmission area. The water loss versus time for gels prepared from 40% and 60% oxidized CS is plotted in Figure 4. Gels prepared from 20% oxidized CS were not subjected to this study as they were macroscopically very porous and poor in strength. There is linearity in the plot up to 4 h (R ) 0.94) and thereafter, the water loss tends to decrease. The WVTR of

2046


Dawlee et al.

Table 3. Platelet, RBC Counts, and PTT of Human Blood Immediately and after 30 min of Exposure to CS-Cross-Linked Gelatin Hydrogels RBC count (106/mm3)

platelet count (103/mm3)

PTT (s)

material

before

after

before

after

before

after

gelatin + 60% oxidized CS reference (tissue culture petri dish)

3.46 ( 0.2 3.34 ( 0.1

3.58 ( 0.5 3.40 ( 0.6

299 ( 6.1 280 ( 7.1

241 ( 5.5 239 ( 12.2

156 ( 1.6 123 ( 2.5

190 ( 2.1 139 ( 5.0

Figure 4. Water vapor transmission loss versus time for hydrogels prepared from 40% (b), and 60% oxidized CS (9).

hydrogels prepared from 40 and 60% oxidized CS were found to be 340.31 ( 28 and 339.97 ( 25 g/m2/h respectively which were not statistically different (p > 0.05). Lamke et al.30 reported the evaporative water loss for normal skin as 204 ( 12 g/m2/day and that for injured skin range from 279 ( 36 g/m2/day for a first degree burn to 5158 ( 202 g/m2/day for a granulating wound. Thus, the rate observed for the gels in the present study are higher than that of a granulating wound. However, similar values have been reported for commercial dressings such as Geliperm.23 The high WVTR observed may lead to the total dehydration of the wound surface. Gel dehydration may also result in the adhesion of the dressing to the wound surface. Previous studies have shown that the dehydration can be minimized by covering the hydrogels with a secondary covering.23 To examine the extent of water loss from the gels, we studied the rate of evaporation of water from the gels. The extent of evaporative loss of water with time from hydrogels examined at 37 °C and 40% relative humidity showed that the decrease in weight was linear with time up to 5 h and thereafter did not change significantly. The values are plotted in Figure 5. Evaporative loss was about 16% within 6 h. The rate of evaporation was calculated from the slope of the initial linear portion of the curve by linear regression (R ) 0.997). The rate of evaporative water loss was calculated to be 198.86 ( 4.5 and 190.88 ( 4.16 g/m2/h for gels prepared from 40 and 60% oxidized CS respectively which was also statistically not significantly different (p > 0.05). The fact that these gels were retaining 80% water even after 12 h at 37 °C is interesting in the sense that, when used as wound coverings, they will be able to provide a moist environment to the wound surface for prolonged periods. Even though the WVTR of these gels are high, since they retain the moisture within the matrix, by providing a secondary dressing on the surface of these gels, a well moist

Figure 5. Evaporative water loss from hydrogels prepared from 40% (b) and 60% oxidized CS (9).

Figure 6. Morphology of L929 fibroblasts cultured on the surface of hydrogels derived from 60% oxidized CS and gelatin showing absence of cytotoxicity (scored as 0).

dressing could possibly be maintained leading to better healing of the wounds. Preliminary cytotoxicity evaluation was done on gels cross-linked with 40% and 60% oxidized CS. The gels were found to be noncytotoxic in nature. Representative microphotograph of fibroblasts cells around hydrogel prepared from 60% oxidized CS (scored as zero) is shown in Figure 6. Although the aldehyde functions outnumber the amino groups in gelatin, no toxicity due to residual aldehyde was evident possibly due to the hemiacetal formation.29 Gelatin cross-linked with oxidized alginate also showed similar behavior when exposed to fibroblasts in vitro.31 In vitro screening of the hydrogels for blood-compatibility (Table 3) showed that there was no significant difference in count for RBC and platelets immediately and after 30 min of contact. Evaluations were confined to hydrogels obtained from gelatin and 60% oxidized CS. However, PTT values in the presence of these hydrogels before and after exposure


Oxidized Chondroitin Gelatin Matrixes Table 4. Haemolytic Potential of Hydrogels sample

haemolysis (%)

40% oxidized CS + gelatin 60% oxidized CS + gelatin

2.24 ( 1.81 1.70 ( 0.75

of blood samples indicated that the presence of these hydrogels prolongs the clotting time. Gelatin is known as a haemostatic agent14,32 and surfaces immobilized with GAGs such as CS are known to exert anti-thrombogenic activity.33 We reason that the anti-thrombogenic activity of CS is more influential than the haemostatic effect of gelatin in the composite matrix leading to prolongation of PTT. The haemolysis assay showed that the hydrogels were nonhaemolytic in nature. The haemolytic potential of the material is defined as the measure of the extent of haemolysis that may be caused by the material when it comes in contact with blood. Table 4 shows the percentage haemolysis of blood in contact with different samples at 37 °C for 60 min. All the samples were found to be nonhaemolytic, the extent of haemolysis being lower than the permissible level of 5%. Unlike chitosan or alginate, CS is an endogeneous substance in the skin and gelatin is a hydrolysis product of the connective tissue protein collagen. They are both biocompatible and completely biodegradable. Therefore, a composite matrix consisting of gelatin and CS would be wellsuited for applications such as wound dressing. Conclusions The polysaccharide CS was oxidized using periodate to generate aldehyde functions that wound enter into crosslinking with the amino groups of the protein, gelatin. This new class of hydrogels showed good equilibrium swelling properties, WVTR, porosity, and blood compatibility. Preliminary cytotoxicity screening demonstrated that they were not cytotoxic. The PTT assay showed that haemostatic effect of gelatin was suppressed by the anti-thrombogenic effect of CS leading to slightly prolonged blood clotting times in the presence of these hydrogels. Although CS-containing hydrogels with proteins such as gelatin cross-linked with extraneous cross-linking agents such as carbodiimides, glutaraldehyde etc., are known, the use of oxidized CS itself for protein cross-linking, we believe is novel and has the potential to eliminate the toxicity associated with many such agents. Even though the haemostatic activity of gelatin was not very evident in the present class of hydrogels, the extracellular signaling and cell recognition properties of CS coupled with the biodegradability of both CS and gelatin would possibly make them useful as materials for wound management, drug delivery and tissue engineering applications. Acknowledgment. Financial support from the IndoFrench Centre for Advanced Scientific Research, New Delhi (Project 2603-2) is gratefully acknowledged. References and Notes (1) Barnet, S. E.; Irving, S. J. Studies of wound healing and the effects of dressings. In High Performance Biomaterials, Szycher, M. Ed.; Technomic: Lancaster, 1991; pp 583-620.

(2) Yannas, I. V.; Burke, J. F. Design of an artificial skin. I. Basic design principles. J. Biomed. Mater. Res. 1980, 14, 65-81. (3) Suzuki, S.; Matsuda, K.; Isshiki, N.; Tamada, Y.; Ikada Y. Experimental study of newly developed bilayer artificial skin. Biomaterials 1990, 11, 356-60. (4) Matsuda, K.; Suzuki, S.; Isshiki, N.; Yoshioka, K.; Wada, R.; Okada, T.; Ikada, Y. Influence of glycosaminoglycans on the collagen component of a bilayer artificial skin. Biomaterials 1990, 11, 35155. (5) Kosir, M. A.; Quinn, C. C.; Wang, W.; Tromp, G. Matrix glycosaminoglycans in the growth phase of fibroblasts: more of the story in wound healing. J. Surg. Res. 2000, 92, 45-52. (6) Buma, P.; Pieper, J. S.; van Tienen, T.; van Susante, J. L.; van der Kraan, P. M.; Veerkamp, J. H.; van den Berg, W. B.; Veth, R. P.; van Kuppevelt, T. H. Cross-linked type I and type II collagenous matrixes for the repair of full-thickness articular cartilage defects-a study in rabbits. Biomaterials 2003, 24, 3255-63. (7) Kuijpers, A. J.; van Wachem, P. B.; van Luyn, M. J. A.; Brouwer, L. A.; Engbers, G. H. M.; Krijgsveld, J.; Zaat, S. A. J.; Dankert, J.; Feijen, J. In vitro and in vivo evaluation of gelatin-chondroitin sulphate hydrogels for controlled release of antibacterial proteins. Biomaterials 2000, 21, 1763-72. (8) Gilbert, M. E.; Kirker, K. R.; Gray, S. D.; Ward, P. D.; Szakacs, J. G.; Prestwich, G. D.; Orlandi, R. R.Chondroitin sulfate hydrogel and wound healing in rabbit maxillary sinus mucosa. Laryngoscope 2004, 114, 1406-9. (9) Gudemez E.; Eksioglu F.; Korkusuz, P.; Asan, E.; Gursel, I.; Hasirci, V.; Chondroitin sulfate-coated polyhydroxyethyl methacrylate membrane prevents adhesion in full-thickness tendon tears of rabbits. J. Hand. Surg. 2002, 27, 293-306. (10) Kirker, K. R.; Nielson, J. H.; Shelby, J.; Prestwich, G. D.; Glycosaminoglycan hydrogel films as biointeractive dressings for wound healing. Biomaterials 2002, 23, 3661-71. (11) Prathiba, V.; Gupta, P. D. Cutaneous wound healing: Significance of proteoglycans in scar formation. Curr. Sci. 2000, 78, 697701. (12) Denuziere, A.; Ferrier, D.; Damour O.; Domard, A. Chitosanchoindroitin sulphate and chitosan-hyaluronate polyelectrolyte complexes: biological properties. Biomaterials 1998, 19, 12751285. (13) Choi, Y. S.; Lee, S. B.; Hong, S. R.; Lee, Y. M.; Song, K. W.; Park, M. H. Studies on gelatin-based sponges. Part III. A comparative study of cross-linked gelatin/alginate, gelatin/hyaluronate and chitosan/ hyaluronate sponges and their application as a wound dressing in full thickness skin defect of rat. J. Mater. Sci., Mater. Med. 2001, 12, 67-73. (14) Kuijpers, A. J.; Engbers, G. H. M.; Krijgsveld, J.; Zaat, S. A.; Dankert, J,.; Feijen, J. Cross-linking and characterisation of gelatin matrixes for biomedical applications. J. Biomater. Sci. Polym. Ed. 2000, 11, 225-43. (15) Choi, Y. S.; Hong, S. R.; Lee, Y. M.; Song, K. W.; Park, M. H.; Nam, Y. S. Study of gelatin-containing artificial skin: I. Preparation and characteristics of a novel gelatin-alginate sponge. Biomaterials 1999, 20, 409-17. (16) Schacht, E.; Bogdanov, B.; Bulcke, A. V. D.; Rooze, N. D. Hydrogels prepared by cross-linking of gelatin with dextran dialdehyde. React. Funct. Polym. 1997, 33, 109-16. (17) Draye, J.-P.; Delaey, B.; Van de Voorde, A.; Van Den Bulke, A.; Bogdanov, B.; Schacht, E. In vitro release characteristics of bioactive molecules from dextran dialdehyde cross-linked gelatin hydrogel films. Biomaterials 1998, 19, 99-107. (18) Bassett, J., Denny, R. C., Jeffery, G. H.; Mendham, J. Vogel’s Text Book of QuantitatiVe Inorganic Analysis, 4th ed.; Longman: London, 1978; pp 370-375. (19) Bubnis, W. A.; Ofner, C. M. The determination of -amino groups in soluble and poorly soluble proteinaceous materials by a spectrophotometric method using trinitrobenzenesulfonic acid. Anal. Chem. 1992, 207, 129-133. (20) Lee, K. Y.; Bouhadir, K. H.; Mooney, D. J. Degradation behaviour of covalently cross-linked poly (aldehyde guluronate) hydrogels. Macromolecules 2000, 33, 97-101. (21) Berger, J.; Reist, M.; Mayer, J. M.; Felt, O.; Peppas, N. A.; Gurny, R. Structure and interactions in covalently and ionically cross-linked chitosan hydrogels for biomedical applications. Eur. J. Pharm. Biopharm. 2004, 57, 19-34. (22) Mark, H. F., Norman, G., Gaylord, N. G., Eds.; Encyclopaedia of Polymer Science and Technology; Interscience: New York, 1967; Vol. 7, pp 446-460.

2048


(23) Queen, D.; Gaylor, J. D. S.; Evans, J. H.; Courtney, J. M.; Reid, W. H. The preclinical evaluation of the water vapour transmission rate through burn wound dressings. Biomaterials 1987, 8, 367-71. (24) Wu, P.; Nelson, E. A.; Reid, W. H.; Ruckley, C. V.; Gaylor, J. D. S. Water vapour transmission rates in burns and chronic leg ulsers: influence of wound dressings and comparison with in vitro evaluation. Biomaterials 1996, 17, 1373-77. (25) Jonkman, M. F.; Molennar, I.; Nieuwenhuis, P.; Bruin, P.; Pennings, A. J. New method to assesss the water vapour permeance of wound coverings. Biomaterials 1988, 9, 263-67. (26) Biological evaluation of medical devices. Part 5. Tests for cytotoxicity: In vitro methods. ISO-10993-5; International Standards Organization: Geneva, Switzerland, 1992. (27) O’Leary, R. K.; Guess, W. L. Toxicological studies on certain medical grade plastics sterilized by ethylene oxide, J. Pharm. Sci. 1968, 57, 12-17. (28) Biological evaluation of medical devices-Part 4. Selection of tests for interactions with blood. ISO-10993-4, 2nd edition; International Standards Organization: Geneva, Switzerland, 2002.

Dawlee et al. (29) Painter, T.; Larsen B. Formation of hemiacetals between neighbouring hexuronic acid residues during the periodate oxidation of alginate. Acta Chem. Scand. 1970, 24, 813-33. (30) Lamke, L. O.; Nilsson, G. E.; Reithner, H. L. The evaporative water loss and the water permeability of grafts and artificial membranes used in the treatment of burns. Burns 1977, 3, 159-65. (31) B. Balakrishnan, A. Jayakrishnan, Self-cross-linking biopolymers as injectable, in situ forming biodegradable scaffolds. Biomaterials 2005, 26, 3941-3951. (32) Cenni, E.; Ciapetti, G.; Stea, S.; Corradini, A.; Carozzi, F. Biocompatibility and performance in vitro of a hemostatic gelatin sponge. J. Biomater. Sci. Polym. Ed. 2000, 11, 685-99. (33) Keuren, J. F. W.; Wielders, S. J. H.; Willems, G. M.; Morra, M.; Cahalan, L.; Cahalan, P.; Lindhout, T. Thrombogenicity of polysaccharide-coated surfaces. Biomaterials 2003, 24, 1917-24.

BM050013A

Oxidized Chondroitin Sulfate-Cross-Linked Gelatin Matrixes: A New

Recommend Documents