Mixed-wall microcapsules made of cross-linked proteins and polysaccharides: preparation and properties.

J . MICROENCAPSULATION,

1991, VOL. 8, NO, 3, 335-347

Mixed-walled microcapsules made of cross-linked proteins and polysaccharides: preparation and properties M.-C. LEVY and M.-C. ANDRY

Journal of Microencapsulation Downloaded from informahealthcare.com by University of Newcastle on 01/06/15 For personal use only.

Laboratoire de Pharmacotechnie, URAjCNRS 492, FacultC de Pharmacie, Universiti: de Reims Champagne-Ardenne, 51, rue Cognacq- Jay, F51096 Reims Cedex, France

(Received 19 November 1990; accepted 12 December 1990) Microcapsules were prepared through an interfacial cross-linking process using terephthaloylchloride and applied to mixtures of a protein (human serum albumin or gelatin) and a polysaccharide. Their properties were compared with those of microcapsules prepared from the protein alone. Morphological characteristics of mixed-walled microcapsules were often modified, as seen by light and electron microscopy. Otherwise, they appeared to be more resistant to digestive media: they were gastroresistant, and their degradation time in pancreatin was prolonged upon raising the amount of polysaccharide. Moreover, the lysis time was shown to depend on the nature of the polysaccharide: microcapsules prepared from acidic polysaccharides at pH 9.8 were hydrolyzed faster. Lastly, the resistance increased upon decreasing the polymers/acylchloride ratio, or upon raising the reaction pH. Encapsulation assays were carried out with sodium salicylate, which was incorporated with a high efficiency. Mixed-walled microcapsules allowed a prolonged release of the tracer in vitro. As compared with protein microcapsules, the release profiles of batches prepared with hydroxyethylstarch exhibited only slight modifications of the initial part of the curve, while a significant burst effect was observed with carboxymethylcellulosecontaining microcapsules.

Introduction Over the past decade, considerable research efforts have been made to develop parenteral controlled-release systems (Langer 1989, Poznansky and Juliano 1984, Tice and Cowsar 1984). With this aim, various studies have been conducted dealing with microcapsules and microspheres prepared from proteins and polysaccharides (Davis et al. 1984, Gupta and Hung 1989, Maulding 1987). T h e difficulties of developing such particles lie in the numerous requirements that these particles must fulfil. One of the most important problems to be solved is obtaining controlled biodegradability (Sjoholm and Edman 1984): they must be destroyed in vivo but they have to remain stable in living tissues until complete release of the included drug is achieved. For several years we have been developing a microencapsulation process based on the interfacial cross-linking of biopolymers with acyldichlorides (Lkvy et al. 1982, Rambourg et al. 1982, Gukrin et al. 1983, Desoize et al. 1986, Livy and Gukrin 1987, Lkvy and Andry 1987). This technique has proved to be an efficient means of preparing stable and biodegradable microcapsules from various proteins such as serum albumin, haemoglobin, gelatin or milk proteins. T h e particles obtained using terephthaloylchloride were shown to be degraded in digestive media. Along with further studies in our laboratory, the process was applied to hydrosoluble polysaccharides such as dextran, soluble derivatives of cellulose and starch (Gourdier et al. 1983, Lkvy and Andry 1990a). T h e resulting microcapsules exhibited a complete 0265-2048491 $3.000 1991 Taylor & Francis Ltd.


336

M . - C . L t v y and M . - C . Andry

resistance to both gastric and intestinal media. These findings prompted us to associate the two kinds of biopolymers with the aim of developing particles with modulable biodegradability. Thus, microcapsules were prepared applying the interfacial cross-linking process to various mixtures of polysaccharides and proteins. Microcapsules with 'mixed walls' that exhibited interesting properties were then obtained: they were gastroresistant and enterosoluble. Moreover, under given reticulation conditions, the survival time in intestinal medium appeared to be related to the relative amount of the protein and the polysaccharide (Levy and Andry 1989, 1990a). In order to specify the properties of the mixed-walled microcapsules, two series of experiments were conducted using human serum albumin and gelatin, respectively. They were separately associated with various polysaccharides bearing acidic groups (alginate, gum arabic, carboxymethylcellulose) or not (dextran, hydroxyethylstarch). This paper reports on the effect of incorporating a polysaccharide into a cross-linked protein wall. Conclusions were drawn from comparative studies of mixed-walled versus pure protein microcapsules with regards to morphological characteristics, degradability and in vitro release profiles of encapsulated salicylate.

Materials and methods Materials The proteins used were human serum albumin (HSA) obtained from the Centre regional de Transfusion Sanguine (Reims, France), and type B 8O-Bloom gelatin (isoelectric point: 5 ) supplied by Mero-Rousselot-Satia (France). T h e polysaccharides used were gum arabic (Cooper, France), sodium alginate (Satialgine SLMK: Metro-Rousselot-Satia), carboxymethylcellulose (CMC) Blanose cellulose gum 12M8P, substitution degree: 1.2, Hercules, France). Dextran was obtained through lyophilization of RheomacrodexR saline solutions (Kabivitrum, France), and hydroxyethylstarch (HES) through lyophilization of PlasmasterilR (Fresenius, France). Terephthaloylchloride was purchased from Aldrich-Chimie (France). T h e surfactants were sorbitan trioleate and polysorbate (Seppic-Montanoir, France). Chloroform, cyclohexane and ethanol, analytical grade, were used without further purification (Prolabo, France). Several buffers were used for the preparation of the ~ buffer adjusted aqueous phase: phosphate buffers p H 7.2 and 8 and a 0 . 4 5 carbonate to p H 9.8 with HCl. Artificial digestive media were prepared according to U.S.P. XXI with pepsin from porcine stomach mucosa or pancreatin from porcine pancreas (Sigma). Preparation of microcapsules T h e interfacial cross-linking process was applied (Levy et al. 1982). Suitable amounts of protein and polysaccharide were dissolved in 3 ml of the selected buffer. Gelatin-containing solutions were prepared using a 30 min heating at 40°C, under magnetic stirring. T h e aqueous solution was emulsified in 15ml of an organic phase containing 5 per cent (v/v) sorbitan trioleate and consisting of chloroform : cyclohexane (1 :4, v/v) for HSA microcapsules, and pure cyclohexane for gelatin microcapsules. T h e interfacial cross-linking reaction was then started by the addition of 20 ml of a terephthaloylchloride solution in chloroform-cyclohexane (1 :4 v/v) and stirring was continued for 30 min. T h e reaction was ended by dilution with 30 ml of cyclohexane. T h e formed microcapsules were separated by centrifugation and washed several times, first with cyclohexane, then with a 5 per cent (v/v)

Mixed-walled microcapsules

337

polysorbate solution in 95 per cent ethanol, with 95 per cent ethanol (twice), and finally with water (twice). Variations were introduced in the p H of the aqueous phase, in the amounts of protein and polysaccharide, in the stirring speed, and in the terephthaloylchloride concentration. T h e reaction time was kept constant (30 min).


Morphological studies Microcapsules were examined by optical and scanning electron microscopy. T h e particle size range was assessed using conventional light microscopy. Enzymatic degradation Lyophilized microcapsules (50 mg) were rehydrated in a test tube with 2 ml of distilled water and then dispersed in 15 ml of an artificial digestive medium: pepsin p H 1.2 or pancreatin p H 7.5 and incubated at 37°C. T h e lysis time was evaluated by microscopic examination and was defined as the time for disappearance of all microcapsules. Encapsulation of salicylate and dissolution studies Control batches of microcapsules were prepared from 6 per cent (w/v) solutions of HSA in buffer p H 9.8 without the addition of polysaccharide and mixed-walled microcapsules from solutions containing 3 per cent HSA and 3 per cent H E S or CMC. Identical series were prepared from gelatin. T h e encapsulation process was conducted using a solution of 200mg salicylate in 6 m l of the buffered aqueous phase. This solution was emulsified in 30 ml of cyclohexane containing 2 per cent v/v sorbitan trioleate (5 min stirring at 2500 rpm). 40ml of a 2.5 per cent w/v solution of terephthaloylchloride were then added. After stirring for 30 min the reaction medium was diluted with 40 ml of cyclohexane. T h e sediment was washed four times with cyclohexane and further freed of the solvent under vacuum. Microcapsules were then congealed and lyophilized. An evaluation of drug loss was carried out for all batches. Preparation and washing media were evaporated and added with buffer p H 7.4. Salicylate was determined by means of a colorimetric method using reaction with Fe3+. For dissolution studies, a batch of lyophilized microcapsules was dispersed in 500ml of buffer pH7.4 at 37°C (Erweka dissolution apparatus, agitation: 40 rpm). Samples were withdrawn at intervals and filtered through a Millipore filter (0.22 pm) for salicylate determination. Each assay was triplicated.

Results Microcapsule feasibility was firstly investigated using equal concentrations of protein and polysaccharide in buffer p H 9.8. Complementary assays were performed for lower p H values and/or with various amounts of the two polymers. In the gelatin series of experiments, mixed-walled microcapsules were prepared from solutions containing 5 per cent (w/v) gelatin supplemented with 5 per cent polysaccharide (HES, gum arabic or alginate) in buffer p H 9.8, using a 5 per cent terephthaloylchloride solution (stirring speed: 2500 rpm). Control gelatin microcapsules were prepared from a 10 per cent gelatin solution under the same conditions. CMC-gelatin microcapsules were prepared using lower concentrations of the two polymers (3 per cent) and a higher stirring speed (3000 rpm), because of the viscosity of the solution. As a rule, stable microcapsules were obtained. They were easily separated and could be lyophilized, giving free-flowing powders. In all cases microscopic examination showed spherical particles with a distinct wall.

M.-C. L t v y and M.-C. Andry


338

Figure 1 . Optical photomicrographs (differential interference contrast): ( a ) control gelatin microcapsules (gelatin concentration: 10 per cent in buffer p H 9.8; terephthaloylchloride concentration: 5 per cent), ( b ) mixed-walled gum-gelatin microcapsules (gum arabic: 5 per cent; gelatin 5 per cent).

(4

(b)

Figure 2. Scanning electron micrographs: (a)control gelatin microcapsules, (b) mixedwalled gum-gelatin microcapsules, same magnification (manufacturing conditions: the same as in figure 1).



339

However, as compared with control gelatin microcapsules, mixed-walled microcapsules exhibited differences with regard to their morphology as shown by interference contrast microscopy and scanning electron microscopy. Whereas gelatin microcapsules showed a uniform aspect (figure 1 ( a ) ) and regular surfaces (figure 2 ( a ) ) , the wall of mixed-walled microcapsules often presented rough membranes. This was observed, for example, with gum-gelatin microcapsules. T h e granulous aspect, revealed by light microscopy (figure 1 (b)), was shown to correspond to irregularities of the surface (figure 2 (b)). CMC-gelatin microcapsules showed a similar aspect. As compared with gelatin microcapsules, the size range was unmodified for HES-gelatin and gum-gelatin microcapsules (10-3 5 pm), while the incorporation of CMC or alginate resulted in a significant increase in size (20-1 50 pm and 20-250 pm, respectively). Microcapsules were also obtained from equal amounts of gelatin and dextran or alginate at p H 8 o r 7.2. They were shown to be intact after lyophilization and to recover their spherical shape after rehydration. A similar series of assays was conducted with HSA used as 3 per cent (w/v) solutions in buffer p H 9.8 and supplemented with 3 per cent dextran, C M C or HES. The terephthaloylchloride concentration was 2.5 per cent (w/v) and the stirring speed was 2500rpm, except for the CMC-HSA batches which were prepared at 3000 rpm because of the viscosity of the aqueous phase. Control HSA microcapsules were prepared from a 6 per cent HSA solution. Stable mixed-walled microcapsules were obtained under these conditions. Like control microcapsules, they appeared as transparent spheres, as shown by microscopic examination. T h e size ranges were 5-25pm for HSA microcapsules, and 5-30pm, 10-25pm, and 30-120pm, for dextran-HSA, HES-HSA and CMC-HSA microcapsules, respectively. In another series of experiments conducted at p H 7.2, microcapsules were prepared from solutions of HSA and C M C using a 5 per cent (w/v) terephthaloylchloride solution. With 3 per cent of each polymer, as well as with 4 5 per cent HSA supplemented with 1.5 per cent CMC, satisfactory microcapsules were obtained which did not form aggregates and survived lyophilization. Figure 3 ( b ) presents a scanning electron micrograph of such mixed-walled microcapsules. They exhibited membranes with a rippled surface, thereby differing from smooth microcapsules obtained from pure HSA under the same conditions (figure 3 ( a ) ) . Concerning enzymatic degradation, the addition of a polysaccharide always resulted in an increased resistance of microcapsules. This phenomenon was observed with gelatin, as well as with HSA. Degradation in pepsin was investigated first. Control batches were prepared from 6 or 10 per cent solutions of HSA or gelatin in buffer p H 9.8, using 2.5 or 5 per cent terephthaloylchloride solutions. Under these conditions, microcapsules were shown to resist more than 24 h in gastric medium. No changes were observed upon replacing half of the protein amount by an equal amount of dextran or CMC. T h e reaction p H was then lowered to 8. Gelatin and HSA microcapsules were shown to be destroyed in pepsin (table 1). When half of the protein amount was replaced by an equivalent amount of polysaccharide, the microcapsules exhibited a resistance to pepsin which was complete with CMC-HSA microcapsules. With pancreatin, prolonged degradation times were also observed as compared with protein microcapsules. This study was mostly conducted using a reaction p H of 9.8, as gelatin and HSA microcapsules prepared at this p H value were shown to be degraded in pancreatin. Various batches were prepared at p H 9.8 from equal amounts of protein and polysaccharide. Results are shown in table 2. T h e influence

M . X . LCvy and M . - C . Andry


3 40

(4 (6) Figure 3 . Scanning electron micrographs: ( a ) control HSA microcapsules (HSA concentration: 6 per cent in buffer pH 7 2 ; terephthaloylchloride concentration: 5 per cent), ( b ) mixed-walled CMC-HSA microcapsules (CMC: 1.5 per cent; HSA 4 5 per cent), same magnification. Table 1 . Microcapsules prepared at pH 8; degradation in pepsin (terephthaloylchloride concentration: 5 per cent w/v). Concentrations wjv Protein

Polysaccharide

Gelatin 10 per cent 5 per cent

Gum arabic 5 per cent

HSA 6 per cent 3 per cent

CMC 3 per cent

0

0

Lysis time in pepsin lh 2h 1 h 30 min No lysis within 24 h

of the biopolymer(s) concentration was investigated first. T h e lysis time of control protein microcapsules was shown to depend on the protein concentration: the higher the protein concentration, the shorter the lysis time. Replacement of half of the protein amount by an equal amount of polysaccharide resulted in a significant increase of the lysis time. This effect was likewise all the more pronounced as the total concentration of the two polymers was lower. T h e influence of the nature of the polysaccharide, acidic or not, is also shown in table 2. In the two series of experiments conducted either with gelatin or with HSA, mixed-walled microcapsules prepared from dextran exhibited longer lysis times than batches prepared from CMC. Variations were then introduced concerning the relative amounts of protein and polysaccharide. T h e resistance to enzymatic digestion was shown to rise with the proportion of polysaccharide used. Table 3 illustrates the effect of increased amounts

341

Mixed-walled microcapsules Table

2. Microcapsule degradation in pancreatin: influence of biopolymer(s) concentration(s) and influence of the nature of the polysaccharide. (Reaction pH: 9.8; terephthaloylchloride concentration 2.5 per cent w/v.) Concentrations wjv


Protein

Table 3.

Polysaccharide

Gelatin 10 per 5 per 6 per 3 per 3 per

cent cent cent cent cent

HSA 20 per 10 per 10 per 5 per 6 per 3 per 3 per

cent cent cent cent cent cent cent

0 5 per cent dextran 0 3 per cent CMC 3 per cent dextran 0 10 per cent 0 5 per cent 0 3 per cent 3 per cent

HES dextran CMC dextran

Lysis time in pancreatin lh 5h 3h 5h 6h

30 min 2 h 30 min 5h 6 h 30 min 5 h 30 min 7h30rnin 9h

Microcapsule degradation in pancreatin: influence of the relative amounts of protein and polysaccharide." Concentrations w/v Protein

Polysaccharide

Gelatin 10 per cent 7.5 per cent 5 per cent 2.5 per cent

dextran 0 2.5 per cent 5 per cent 7.5 per cent

HSA 6 per cent 3 per cent 2.5 per cent

CMC 0 3 per cent 4.5 per cent

Lysis time in pancreatin lh 4h 5h 7h 1Ornin 20 min 60 rnin

'Manufacturing conditions: in the gelatin series, reaction pH: 9.8, terephthaloylchloride concentration: 2.5 per cent; in the HSA series, reaction pH: 7.2, terephthaloylchloride concentration: 5 per cent w/v. of dextran in mixed-walled microcapsules prepared at pH 9.8 from gelatin. T h e same observation was made with HSA a n d CMC cross-linked at p H 7.2. Finally, the survival time in pancreatin was shown to depend o n the polycondensation pH. As shown in table 4, mixed-walled microcapsules, prepared either from gelatin o r from H S A , exhibited longer lysis times in pancreatin upon raising the cross-linking p H . I n all cases, shorter degradation times were observed with batches prepared from the protein alone. T h e last part of t h e work was devoted to salicylate encapsulation and in vitro release studies. Salicylate-loaded microcapsules were all spherical in shape and well

342

M.-C. LCvy and M.-C. Andry

Table 4. Microcapsule degradation in pancreatin: influence of reaction pH. (Terephthaloylchloride concentration: 5 per cent w/v.) Components concentrations wjv


pH 7.2

PH 8 pH9.8

Gelatin 5 per cent

HSA 3 per cent

Gelatin 10 per cent

Dextran 5 per cent

HSA 6 per cent CMC 3 per cent

3 min 5-6 min lh

5-6 min 15 min 8h

+

+

10min 10-1 5 min

5h30min

20 min 45 min 9h

individualized, as seen by optical microscopy. T h e size range was 2-30 pm, except for CMC-HSA and CMC-gelatin microcapsules, which exhibited larger diameters, i.e., 5-60 pm and 5-1 50 pm, respectively. Mean weights of lyophilized batches were generally comprised between 1.1 and 1.22 g. However, lower yields were obtained from CMC-HSA (0.86 g) and CMC-gelatin (0.82 g) batches. T h e loss of salicylate in reaction and washing media was between 2 and 6 per cent for most batches, and it reached u p to 12 and 18 per cent for the CMC-HSA and CMC-gelatin batches, respectively. T h u s , the encapsulation percentages were comprised between 82 and 98 per cent which gave mean payloads of 16 to 20 per cent (calculated as the percentage of drug versus weight of dry microcapsules). I n most cases, salicylate release was shown to occur over a period of three days. However, it was achieved within only two days for CMC-gelatin, while it took four

SALICYLATE

so-.

RELEASED , %

-

A

C.M.C-GELATIN

-

GELATIN

A

A

30 -.

20-.

w

a

H.E.S - G E L A T I N


343

S A L I C Y L A T E RELEASED , %

3(1

.BUMIN 25


20

15

10

5

1

2

3

4

5

6

T I M E (hours)

Figure 5 . Release of sodium salicylate from cross-linked HSA microcapsules and mixedwalled microcapsules. days for HSA-HES batches. T h e release was not complete, an 18-25 per cent amount remaining encapsulated. Minimal retention (11 per cent) was observed with gelatin-CMC microcapsules. Figure 4 shows the release profiles obtained in the gelatin series. With the control gelatin microcapsules, a slow and regular release was observed after an initial one hour period of faster release. Only 10 per cent of the encapsulated drug could be found in the bath after 4 h. Mixed-walled microcapsules prepared from HES exhibited a slightly higher initial rate resulting in a release of 10 per cent of the drug within the first hour. Later on, the profile slope was comparable with that of the gelatin microcapsules. T h e most striking differences were noticed with the microcapsules prepared with CMC. They showed an initial ‘burst’ phase resulting in a 40 per cent release within the first 15 minutes. After the first two hours, the profile was unmodified. The release profiles obtained with HSA microcapsules are presented in figure 5. They show comparable results. Control HSA microcapsules exhibited a slow release rate with only 10 per cent of the encapsulated drug released after five hours. A similar profile, although slightly slower, was observed with mixed-walled HSA-HES microcapsules, while a marked increase of the release rate was observed during the first two hours for microcapsules prepared from CMC.

Discussion T h e above results show the particular properties of microcapsules prepared through interfacial cross-linking of proteins and polysaccharides. As compared with cross-linked protein microcapsules, they exhibited differences with regard to morphology and biodegradability which demonstrate the involvement of both


3 44

M.-C. Ldvy and M . - C . Andry

polymers in the walls. In the resulting ‘mixed-walled’ microcapsules, composite networks should be formed upon cross-linking of functional groups of the two polymers: hydroxy groups of the polysaccharide, involved in ester bonds, and acylable groups of the protein, mainly amino and hydroxy groups which form amide and ester linkages. As a matter of fact, we showed in a recent study concerning crosslinked HSA microcapsules that growing amounts of hydroxy groups were esterified upon raising the polycondensation p H (Livy et al. in press). For given reaction parameters, the membranes obtained from mixtures of a protein and a polysaccharide were then expected to involve variable amounts of amide and ester bonds, depending on the proportions of the two polymers in the solution. Actually, several parameters seem likely to interfere in the construction of the membrane. During the emulsification step, the formation of the interfacial film obviously depends on the behaviour of the two molecules at the oil/aqueous interface, i.e., on their respective adsorptive properties. In this respect, Stainsby (1986) pointed out that the composition of the interfacial film obtained from a mixture of two macromolecules would differ from the overall composition if the adsorptive properties of the components were different, as is the case with proteins and polysaccharides. Proteins are considered to be more surface active than polysaccharides, due to their greater hydrophobicities and better flexibilities. Mixtures of proteins and polysaccharides would then give rise to competitive adsorption phenomena at the interface, with an expected predominance of the protein. Otherwise, the possibility of interactions between the two polymers in the aqueous phase has to be considered. As a matter of fact, various proteins and polysaccharides have been shown to interact in aqueous solutions resulting in the formation of complexes, as reviewed for example by Ledward (1979) or Tolstoguzov (1 986). In some cases, rather than being considered separately, the behaviour of the two molecules at the interface might then be better described as the result of the formation of soluble complexes in the aqueous medium prior to the emulsification step and the subsequent adsorption of the complexes at the interface. In this respect, the nature of the polysaccharide, either acidic or neutral, the isoelectric point (i.e.p.) of the protein, and the reaction p H are all important factors that determine the interactions with the proteins. It may be pointed out that most of the work reported on the subject deals with acidic polysaccharides which are involved in electrostatic interactions with proteins. At p H values below the i.e.p. of the protein, insoluble electrostatic complexes can be formed, which are the basis of the well-known encapsulation procedures by complex coacervation. In the present study, the p H values of the aqueous phases ( 7 * 2 , 8or 9.8) were always above the i.e.p. of the proteins (i.e.p. HSA : 4.5; i.e.p. gelatin type B: 5 ) . Under these conditions it has been reported that acidic polysaccharides, other than sulphated compounds, cannot form complexes with globular proteins (Tolstoguzov 1986). There seemed then to be no chance of forming complexes from the negatively charged polysaccharides used in this study, as they all bore carboxylic groups. It might be different in the case of solutions containing neutral polysaccharides admixed with proteins. Actually, interactions between the two sorts of polymers have been described at pH values above the i.e.p. of the protein. For example, Ponder and Ponder (1959) reported on the formation of a complex between dextran and serum albumin at p H values between 9.6 and 6.6. It may be stressed that the reaction pH appears to be a determining factor in the construction of the wall. During the emulsification step it plays a role in the formation of the interfacial film by influencing the conformation of the protein, the



345

ionization state of the protein (and of the polysaccharide, if acidic), and possibly their interaction. It will further influence the polycondensation step, as alkalinization is known to favour acylation, and especially to increase esterification, as we have shown with HSA microcapsules upon raising the reaction pH. Concerning the polycondensation step, it seems reasonable to postulate mixed cross-linkages involving hydroxy groups of the polysaccharide on one side and available groups of the protein on the other side. This might account for the observed increased resistance of mixedwalled microcapsules to enzymatic degradation, as compared with microcapsules prepared from the protein alone. T h e establishment of mixed cross-linkages between the protein and the polysaccharide by terephthaloylchloride is assumed to protect the protein as a result of steric hindrance to protease attack. As a matter of fact, this mechanism has been proposed to account for the stabilization of various polypeptides against enzymatic degradation by covalent binding to polysaccharidic carriers (Lee 1986, Molteni 1982). In this context, intensification of covalent binding between the protein and the polysaccharide would result in an improved protective effect, as we observed with growing amounts of polysaccharide (table 3), or upon raising the reaction p H (table 4). Concerning this last point, control batches prepared from proteins also exhibited a progressive resistance to proteases upon raising the reaction p H as a result of a higher degree of cross-linking. They became resistant to pepsin, and then progressively resistant to pancreatin, as already observed (Lkvy and Andry, 1990 b). It may be emphasized that strong cross-linking conditions were used in this study, i.e., high terephthaloylchloride concentrations and a prolonged reaction time. They were responsible for the observed gastroresistance of control protein microcapsules prepared at pH 9.8. Otherwise, results presented in table 2 demonstrate the influence of the biopolymer(s) concentration(s) on the degradation properties of the resulting microcapsules for a given terephthaloylchloride concentration. Actually, rather than the acylchloride concentration, the polymer/ acylchloride ratio should be considered as one of the determining factors that influence the degree of cross-linking, as pointed out by Benita et al. (1984). A low ratio favours interfacial polycondensation, thereby reducing the degradation properties. This was observed with both protein microcapsules and mixed-walled microcapsules. Under all experimental conditions, mixed-walled microcapsules were found to be more resistant to enzymatic degradation than protein microcapsules. They then made it possible to obtain slowly degradable membranes under milder conditions, for example at lower p H values. Results presented in table 2 indicate that microcapsules prepared at p H 9.8 from a neutral polysaccharide were more resistant than batches obtained from an acidic polysaccharide. This might be accounted for by the ionization state of the negatively charged polysaccharide that would impair its adsorption at the interface and/or by easier acylation of primary alcohol functions in neutral polymers. It appears then that mixed-walled microcapsules make it possible to adjust biodegradability under different experimental conditions by combining variations of several parameters such as nature and amounts of protein and polysaccharide, polymers/acylchloride ratio, reaction pH. In addition, variations may be introduced in the reaction time, which had been kept constant in the study. Finally, the choice of the constituting polysaccharide was shown to influence the release kinetics of encapsulated salicylate. T h e release profiles presented in figures 4


346

M.-C. LCvy and M.-C. Andry

and 5 show that mixed-walled microcapsules allowed a prolonged release of salicylate. As compared with control batches prepared from the protein alone (gelatin or HSA), modifications were observed, mainly concerning the initial part of the curve. Only slight differences were noticed with microcapsules prepared from H E S whose initial rate was either increased (gelatin) or decreased (HSA). These variations may be related to differences in interactions between HES and the two proteins. T h e most striking differences were observed with microcapsules containing negatively charged CMC, which exhibited a significant burst effect. T h e hydrophilic properties of CMC may be involved here, as they could result in a faster rehydration of the lyophilized microcapsules, and subsequently in a faster dissolution of the encapsulated tracer. Differences in degree of cross-linking might also interfere, as already suggested. Lastly, an electrostatic repulse effect is assumed to account for this initial fast release of an anionic drug. I t would then be consistent with our other findings concerning the release profiles of salicylate from microcapsules prepared from three different gelatins with growing isoelectric points (Lkvy and Andry 1990b). I t has then been shown that the lower the isoelectric point, the higher the release rate. Furthermore, mixed-walled microcapsules prepared with B gelatin and alginate have been shown to retain a basic tracer (pilocarpine salt) more efficiently than capsules prepared from gelatin alone (Lkvy and Andry 1989). This possibility of modulating the release of charged drugs by using ionic groups of a polysaccharide appears to be a significant advantage. In addition, the incorporation of a polysaccharide in the wall might result in a decrease in immunogenicity, as the antigenic properties of various polypeptides and proteins have been reported to be decreased or abrogated by conjugating to polysaccharides, due to the masking of antigenic sites (Lee 1986, Molteni 1982). I n conclusion, mixed-walled microcapsules appear to be highly versatile carriers, which afford numerous possibilities of modulating the properties of the membrane and the release kinetics of encapsulated drugs.

References BENITA, S., FICKAT, R., BENOIT, J . P., BONNEMAIN, B., SAMAILLE, J. P., and MADOUI.E, P., 1984, Biodegradable cross-linked albumin microcapsules for embolization. Journal of Microencapsulation, 1, 317-327. DAVIS,S. S., IIUJM, L., McVie, J. G., and TOMLINSON, E., editors, 1984, Mzcrospheres and Drug Therapy. Pharmaceutical, Immunological and Medical Aspects (Amsterdam: Elsevier). DESOIZE, B., JARDILLIER, J. C., KANOUN, K., GUERIN,D., and LEVY,M.-C., 1986, In vitro cytotoxic activity of cross-linked protein microcapsules. Journal of Pharmacy and Pharmacology, 38, 8-1 3. GOURDIER, B., ANDRY,M.-C., and LEVY,M.-C., 1983, Microencapsulation VI: Microcapsules a paroi constitute de polyholosides rtticules. Proceedings of the 3rd Internatzonal Conference on Pharmaceutical Technology, A P G I , Paris, 3, 195-204. P., LEVY,M.-C., GAYOT,A., and TRAISNEL, M., 1983, MicroenGUERIN,D., RAMBOUHG, capsulation VIII: Microcapsules de charbon active i paroi de strum-albumine humaine rkticulte. Innovation et Technologie en Biologie et Midecine, 4, 24-32. C. T., 1989, Review. Albumin microspheres 11:applications in drug GUPTA, P. K., and HUNG, delivery. Journal of Microencapsulation, 6 , 463-472. LANCER,I>., 1989, Biomaterials in controlled drug delivery: new perspectives from biotechnological advances. Pharmaceutical Technology, 13, 18-30. LEDWARD, D. A., 1979, Protein-polysaccharide interactions. In Polysaccharides in Food, edited by J. M. V. Blanshard and J. R. Mitchell (London: Butterworth), pp. 205-217. LEE,H. L., 1986, Peptide and protein drug delivery: opportunities and challenges. Pharmacy International, 7 , 208-212.



347

LEVY,M.-C., and ANDRY, M.-C., 1987, Microencapsulation par rkticulation interfaciale de gklatine. Sciences Techniques et Pratiques Pharmaceutiques, 3, 644-65 1 . LEVY,M.-C., and ANDRY, M.-C., 1989, Microcapsules a proprietks modulables formkes de protkines et de polysaccharides corkticules. Thkrapie, 44, 365-369. LEVY,M.-C., and ANDRY,M.-C., 1990 a, Microcapsules prepared through interfacial crosslinking of starch derivatives. International Journal of Pharmaceutics, 62, 27-35. LEVY,M.-C., and ANDRY,M.-C., 1990 b, An evaluation of gelatin microcapsules prepared using an interfacial cross-linking process. Life Science Advances, 19, 21 9-227. LEVY,M.-C., and GUERIN, D., 1987, Microcapsules prkparhes par rkticulation interfaciale des protkines du lait. Pharmaceutica Acta Helvetiae, 8, 236-240. LEVY,M.-C., RAMBOURG, P., LEVY, J., and POTRON, G., 1982, Microencapsulation IV: Crosslinked haemoglobin microcapsules. Journal of Pharmaceutical Sciences, 71, 759-762. LEVY,M.-C., LEFEBVRE, S., RAHMOUNI, M., ANDRY,M.-C., and MANFAIT, M., in press, FT-IR spectroscopic studies of human serum albumin microcapsules prepared by interfacial cross-linking with terephthaloylchloride: influence of polycondensation pH on spectra and relation with microcapsule morphology and size. Journal of Pharmaceutical Sciences. MAULDING, H. V., 1987, Prolonged delivery of peptides by microcapsules. Journal of Controlled Release, 6 , 167-176. MOLTENI, L., 1982, Effects of the polysaccharidic carrier on the kinetic fate of drugs linked to dextran and inulin in macromolecular compounds. In Optimizationof Drug Delivery, A. Benzon Symposium 17, edited by H. Bundgaard, A. Bagger Hansen and H. Kofod (Copenhagen: Munksgaard), pp. 286-301. PONDER, E., and PONDER, R., 1959, The interaction of dextran with serumalbumin, Journal of General Physiology, 43, 753-758. POZNANSKY, M. J., and JULIANO, R. L., 1984, Biological approaches to the controlled delivery of drugs: a critical review. Pharmacological Reviews, 36, 277-336. RAMBOURG, P., LEVY,J., and LEVY,M.-C., 1982, Microencapsulation 111: Preparation of invertase microcapsules. Journal of Pharmaceutical Sciences, 71, 753-758. SJOHOLM, I., and EDMAN, P., 1984, The use of biocompatible microparticles as carriers of enzymes and drugs in vivo. In Microspheres and Drug Therapy, edited by S. S. Davis, L. Illum, J. G. McVie and E. Tomolinson (Amsterdam: Elsevier), pp. 245-262. STAINSBY, G., 1986, Foaming and emulsification. In Functional Properties of Food Macromolecules, edited by J. R. Mitchell and D. A . Ledward (London and New York: Elsevier), pp. 3 15-3 53. TICE, T. R., and COWSAR, D. R., 1984, Biodegradable controlled-release parenteral systems. Pharmaceutical Technology, 8, 26-36. TOLSTOGUZOV, V. B., 1986, Functional properties of protein-polysaccharide mixtures. In Functional Properties of Food Macromolecules, edited by J. R. Mitchell and D. A. Ledward (London and New York: Elsevier), pp. 3 8 5 4 1 5 .

The preparation and release properties of microcapsules [proceedings].

Protein microcapsules: preparation and applications.

Preparation and mechanical properties of photo-crosslinked poly(trimethylene carbonate) and nano-hydroxyapatite composites.

Preparation of microcapsules with the evaluation of physicochemical properties and molecular interaction.

Cellulose acetate phthalate microcapsules: method of preparation.

Preparation and characterization of crosslinked and polymerized hemoglobin solutions.

Preparation and in vitro, in vivo evaluation of clarithromycin microcapsules.

Preparation and in vitro evaluation of salbutamol sulphate microcapsules.

Preparation and characterization of vanillin-crosslinked chitosan therapeutic bioactive microcarriers.

Preparation and characteristics of gelatin sponges crosslinked by microbial transglutaminase.

Preparation of DNA-crosslinked polyacrylamide hydrogels.

Preparation and characterization of in-situ crosslinked pectin-gelatin hydrogels.

Preparation and characterizations of EGDE crosslinked chitosan electrospun membranes.

Preparation, properties, and plasma retention of human hemoglobin derivatives: comparison of uncrosslinked carboxymethylated hemoglobin with crosslinked tetrameric hemoglobin.

nanoparticles based on natural polysaccharides.

Preparation of thermally stable microcapsules with a chitosan-silica hybrid.

Synthetic biology towards the synthesis of custom-made polysaccharides.

Research of amoxicillin microcapsules preparation playing micro-jetting technology.

Preparation and characterisation of biodegradable pollen-chitosan microcapsules and its application in heavy metal removal.

Preparation and in vitro evaluation of slow release ketoprofen microcapsules formulated into tablets and capsules.

Preparation and physical evaluation of microcapsules of hydrophilic drug-cyclodextrin complexes.

Physiochemical and biological properties of phosphorylated polysaccharides from Dictyophora indusiata.

Preparation and characterization of starch crosslinked with sodium trimetaphosphate and hydrolyzed by enzymes.

Preparation and characterization of double crosslinked hydrogel films from carboxymethylchitosan and carboxymethylcellulose.