Formulation and Drying of Alginate Beads for Controlled Release and

ARTICLE pubs.acs.org/Biomac

Formulation and Drying of Alginate Beads for Controlled Release and Stabilization of Invertase Patricio R. Santagapita, M. Florencia Mazzobre, and M. Pilar Buera* Industry Department and Organic Chemistry Department, Faculty of Exact and Natural Sciences, University of Buenos Aires (FCEyN-UBA) and National Council of Scientific and Technical Research (CONICET), Buenos Aires, Argentina

bS Supporting Information ABSTRACT: Several alternatives to the conventional alginate beads formulation were studied for encapsulation of invertase. Pectin was added to the alginate/enzyme solution while trehalose and β-cyclodextrin were added to the calcium gelation media. The effect of composition changes, freezing, drying methods (freeze, vacuum, or air drying), and thermal treatment were evaluated on invertase stability and its release kinetics from beads. The enzyme release mechanism from wet beads depended on pH. The addition of trehalose, pectin, and β-cyclodextrin modified the bead structure, leading in some cases to a release mechanism that included the relaxation of the polymer chains, besides Fickian diffusion. Enzyme release from vacuum-dried beads was much faster than from freeze-dried beads, probably due to their higher pore size. The inclusion of β-cyclodextrin and especially of pectin prevented enzyme activity losses during bead generation, and trehalose addition was fundamental for achieving adequate invertase protection during freezing, drying, and thermal treatment. Present results showed that several alternatives such as drying method, composition, as well as pH of the relese medium can be managed to control enzyme release.

’ INTRODUCTION Encapsulation is a process by which certain active substances (flavors, drugs, enzymes, vitamins, essential oils, etc.) are included in a matrix or wall system with the purpose to control their release and protect them from deteriorative processes like oxidation, evaporation, degradation, or thermal denaturation.1,2 In search of suitable matrices for biomolecule stabilization, delivery, and controlled release, ionically cross-linked hydrogels have been thoroughly investigated.1,3,4 Alginate is one of the most popular anionic polyelectrolytes used for bead preparation because it is nontoxic, biodegradable, and biocompatible.5 Consisting of mannuronic and guluronic acid blocks, alginate solution turns into a hydrogel matrix through the exchange of sodium ions from guluronic acid with divalent cations, such as calcium, barium, or strontium, and through the formation of electrostatic binding between guluronic residues.6 The immobilization procedure on alginate beads is not only inexpensive but also very easy to carry out and provides extremely mild conditions, so that the potential for industrial application is considerable.7 However, some disadvantages are often associated with this carrier, including high biomolecule leakage, low mechanical strength and large pore size.8 To optimize the encapsulation efficiency and controlled release of enzymes from the gel matrix, new formulations incorporating chitosan, polymers, or starch have been studied.4,9
11 Although several proteins have been encapsulated in alginate beads, the release mechanisms and protein stability, which are fundamental aspects of the potential applications, have not been analyzed simultaneously. r 2011 American Chemical Society

Pectin is a polygalacturonic polymer partially methoxylated and shares some characteristics with alginate: it is nontoxic, biodegradable, and biocompatible.12 It is also used as encapsulating agent because the low methoxyl pectins can form gels with divalent cations by ionotropic gelation.6 Another common approach for increasing protein stability is to dehydrate or freeze proteins in the presence of saccharides, trehalose being one of the most used excipients. The action of sugars can be ascribed to both kinetic and specific effects. At the kinetic level, they promote the formation of amorphous systems. At the specific-interaction level, sugars and particularly trehalose interact by hydrogen bonding with biological structures, stabilizing them during drying.13,14 Several enzymes have been stabilized through immobilization in glassy sugar matrices;13
15 however, they are not effective at controlling the release or the delivery of the encapsulated compound. Therefore, the incorporation of sugars to polyelectrolyte beads emerges as an interesting option. The use of cyclodextrins arose in the last years as an approach for increasing protein stability. β-Cyclodextrin (β-CD) is a cyclic nonreducing oligosaccharides composed of 7 glucopyranose units with a hydrophobic central cavity and a hydrophilic outer surface.16 CDs are capable to include a variety of hydrophobic guest compounds, such as aromatic amino acids located at the proteins surface.17 Thermal stability of enzymes could thus be improved by chemical conjugation of the enzyme with several Received: May 3, 2011 Revised: August 1, 2011 Published: August 02, 2011 3147

dx.doi.org/10.1021/bm2009075 | Biomacromolecules 2011, 12, 3147–3155

Biomacromolecules β-CD derivatives which induced multipoint supramolecular advantageous interactions at the surface of the enzyme.17 A synergistic effect was previously observed on invertase stability subjected to thermal and hydric stresses in dehydrated formulations containing trehalose and β-cyclodextrin.14 The function of CDs in the formulation of microparticles for bioactive compounds delivery involves: (i) its use as an additive to protect the encapsulated molecule,18
21 (ii) its incorporation as drugCD complexes for controlled release,22 and (iii) their chemical cross-linking for obtaining microparticles to decrease the release rate of microencapsulated compounds.22
24 Often, more than one ot these three function of the CDs are combined in the same microparticles.21,22 Invertase (β-fructofuranosidase) is a glycoenzyme of high technological impact in the production of invert sugar, widely used in the pharmaceutical and food industries, and also for the development of analytical devices.25 For these applications, a highly stable and reusable invertase form is desired, to reduce costs and increase the productivity of the overall process. The purpose of present work was to study the effects of different alginate beads formulations and drying methods on invertase encapsulation to provide new alternatives for enzyme stabilization and for managing enzyme release.

’ MATERIALS AND METHODS Materials. Sodium alginate (A) (from Laminaria hyperborean, molecular mass of 1.97  105 Da, mannuronate/guluronate (M/G) ratio = 0.6) was from BDH, Poole, United Kingdom. Pectin (P; low methoxyl amidated pectin, with esterification degree between 26 and 31%; amidation degree between 16 and 19%) was obtained from Cargill Inc., Minneapolis (MN), U.S.A. The enzyme invertase from Saccharomyces cerevisiae (β-fructofuranosidase, E.C. 3.2.1.26, 1840 U/mg, molecular mass 270 kDa) was from Fluka, Buchs, Switzerland. R,R-Trehalose dihydrate (T), was from Hayashibara Co, Ltd., Japan, and β-cyclodextrin (β-CD) from Amaizo, Hammond (IN), U.S.A. All other reagents were commercially available and used as received. Gel Beads Preparation. Beads were prepared by ionotropic gelation according to the drop method described elsewhere,5,9 with variations in both alginate and gelling calcium chloride solutions. Sodium alginate solution (1% w/v) was prepared in acetate buffer 50 mM pH 3.8 and invertase solution was added to a final concentration of 117 U/mL. Acetate buffer pH 3.8 was selected to guarantee an electrostatic interaction between the polymer and the enzyme. Because the isoelectric point of invertase is between 4.0 and 4.526 and the pKa values of alginate are 3.38 and 3.65,27 alginate would have a net negative charge at pH 3.8 and the enzyme a net positive charge. Four types of beads were prepared, with the following composition: alginate (A); alginatetrehalose (A-T); alginate-trehalose-CD (A-TCD); alginate-pectintrehalose (AP-T). A peristaltic pump was used to drop 10 mL of the alginate
enzyme mixture into 100 mL of the gelling solution. As gelling solutions, 2.5% w/v CaCl2 solution was prepared in acetate buffer 50 mM pH 3.8 and was used as it for the preparation of A beads, or was supplemented with 20% w/v trehalose for A-T beads. Also, 1.5% w/v βCD was added to CaCl2
trehalose solution for preparing A-TCD beads. For AP-T preparation, a 1% w/v alginate/pectin (3:1 mass fraction) solution containing the enzyme was dropped into the CaCl2
trehalose solution using the same procedure described previously. Pectin was negatively charged during AP-T beads generation since it isoelectric point is around 3.5.9 Beads generation was performed in a cold bath of water and ice in equilibrium (0 °C) under constant stirring. This temperature was chosen to avoid enzyme denaturation

ARTICLE

during beads generation. A needle size of 0.25 (diameter)  6 (length) mm (Novofine 31 G, Novo Nordisk A/S, Bagsvaerd, Denmark) was used, the distance of the needle above the solution was 6 cm, and the speed of the pump was fixed at 9.0 ( 0.1 rpm. After generation, the beads were hardened 15 min in the CaCl2 solution,1 washed 5 times with cold water, and stored in microcentrifuge tubes at 4 °C until their corresponding treatments. Preliminary studies showed that to achieve adequate enzyme protection during freezing or drying, trehalose and β-CD should be included in the gelling media, not in the alginate (or alginate
pectin) solution. This can be explained as a consequence of a dilution effect when the sugars were incorporated only in the initial alginate solution during bead generation, and they diffuse out of the beads to the CaCl2 solution. Loading Efficiency of the Enzyme in the Beads. The amount of enzyme loaded in the beads after generation was determined through the loading efficiency parameter (L.E.) defined in eq 1. L:E: ¼

L  100 L0

ð1Þ

where L is the amount of enzyme in the beads and L0 is the initial amount of enzyme in the alginate solution. The Bradford method28 was used to determine the amount of protein in each fraction. For L determination, 15 beads were dissolved in 0.25 mL of 10% w/v sodium citrate (NaCit) solution during 1 h in constant sink. Protein concentration was normalized for a volume factor and for bead. Physical Treatment of the Beads. Dehydration. Beads dehydration was performed by three different methods: vacuum drying (VD), air drying (AD), and freeze-drying (FD). VD was performed in an oven operating at a chamber pressure of 113 mbar during 4 h at 25 °C. AD was made at 40 ( 1 °C for 4 h in a forced air circulation oven. FD of beads (frozen 24 h at
18 °C and exposed to liquid nitrogen) was performed during 12 h in a Heto Holten A/S, cooling trap model CT 110 freeze-dryer (Heto Lab Equipment, Denmark) operating at a condenser plate temperature of
110 °C and a minimum chamber pressure of 4  10
4 mbar. The main drying was performed without shelf temperature control. Secondary drying was performed at 25 °C. After dehydration, the beads were maintained in vacuum desiccators until their corresponding treatments or property determinations. Freezing. Wet beads were placed in microcentrifuge tubes and frozen at
18 °C (conventional freezer) during 24 h. Thermal Treatment. Wet and dried beads were exposed to thermal treatment at 50 °C in a thermostatic bath for 24 and 96 h, respectively. Beads Characterization. Digital Image Analysis. The size and shape evaluations of the beads were carried out by analyzing the digital images captured by a digital camera (Canon PowerShot A70 3.2 Mpix, Canon Inc., Malaysia; with zoom fixed at 3.0) installed on a binocular microscope (magnification 7, Unitron MS, Unitron Inc., New York, U.S.A.). The pictures were analyzed with the software IMAGE J. Wet beads were dyed with 9% v/v methylene blue solution for 15 min to increase the contrast between the bead and background, which is critical for the analysis. Dried beads were measured as obtained. Area, perimeter, Feret’s diameter, and circularity were analyzed for at least 50 (wet systems) or 40 beads (dried systems) by applying the “analyze particle” command of the software. The Feret’s diameter corresponds to the longest distance between any two points along the bead boundary. Circularity is defined as a value between 0 and 1 indicating how closely the shape of the particle resembles a circle. The effect of treatment (wet or drying) and the effect of composition on Feret’s diameter and circularity were analyzed by 1 way ANOVA with Bonferroni post test using GraphPad Prism v 5. The Image J software was calibrated to transform the measured pixels in length units (mm) by taking pictures of a caliper section. 3148

dx.doi.org/10.1021/bm2009075 |Biomacromolecules 2011, 12, 3147–3155

Biomacromolecules Water Content and Water Activity. The total water content of the beads was determined gravimetrically by difference in weight before and after drying in vacuum oven for 48 h at 96 ( 2 °C. These drying conditions were selected in previous studies14 and they were adequate to determine water content in the studied systems with a confidence interval of 6% for 95% certainty. For wet beads, the water content is expressed in wet basis (amount of water related to the total amount of sample). Instead, for dried beads, the water content was expressed in dry basis (amount of water related to the dried matter). The effect of drying method and the effect of composition on water content were analyzed by t test using GraphPad Prism v 5. Water activity (aw) was determined by means of an Aqualab instrument (Decagon Devices, Inc., U.S.A.). A special sampler holder was used to reduce the number of beads to be placed, and the corresponding calibration curve was performed with salts of known aw.29 Scanning Electron Microscopy (FEG-SEM). The morphology of the beads dried through the different methods was observed by a scanning electron microscopy with field emission gun (FEG-SEM) model supra 40 (Carl Zeiss SMT Inc., U.S.A.) with InLes detector. This detector allowed observation of the bead structure without the need to apply a metal coating to the samples, allowing a pore size determination to be performed. Thermal Transitions by Differential Scanning Calorimetry. Glass transition temperature (Tg) and endothermal peaks were determined by differential scanning calorimetry (DSC) by means of a Mettler Toledo 822 DSC (Mettler Toledo AG, Switzerland) and STARe Thermal Analysis System version 3.1 software (Mettler Toledo AG). For each system the endothermal baseline shift represents the glass transition. The instrument was calibrated using standard compounds (indium and zinc) of defined melting point and heat of melting. All measurements were made in duplicate with 5
10 mg sample mass (25
30 dried beads), using hermetically sealed aluminum pans of 40 μL inner volume (Mettler), heated from
20 to 340 °C at 10 °C/min; an empty pan was used as a reference. The confidence intervals estimated for temperature values and for enthalpy values were 2 °C and 10 mJ, respectively. The effect of the drying method and the effect of composition on glass transition temperature (Tg), change in heat capacity at Tg, and enthalpy relaxation were analyzed by t test using GraphPad Prism v 5. Enzyme Release. Enzyme release was studied as a function of time in wet and dried beads for all analyzed compositions. Since the quantity of enzyme present in each bead was low (0.5
0.4 μg protein/bead) to quantify the protein content for any of the traditional methods, the enzyme activity determination was chosen as the most sensitive method to follow the release of the enzyme. Eight beads were placed in microcentrifuge tubes (in duplicate) with 120 μL of cold (4 °C) sodium acetate buffer 50 mM (pH 3.8, 4.5, or 5.5, as indicated in each case), and stirred. Then, after appropriate times, enzyme activity was determined in 100 μL of the release medium and inside the beads (see next section). The sum of both values at each time corresponded to the initial enzyme activity in the beads. Control systems were also prepared to confirm that there was no enzyme inactivation inside or outside the bead during the release experiment. The release curves were modeled by Peppas equation (see eq 3 in Supporting Information) by using GraphPad Prism v5 software, and the corresponding parameters were calculated. An F test was used to evaluate significant differences between the fitted release curves and between the calculated parameters (n and k) for different beads compositions and pH levels of the release medium. Invertase Activity. One enzymatic unit was defined as the amount of invertase needed to hydrolyze 1.0 μmol of sucrose per minute at pH 4.6 at 37 °C. After each treatment, invertase activity in wet or dehydrated beads and in the release media was determined spectrophotometrically by the dinitrosalicylic acid method.30 The enzyme activity of invertase was determined on four beads or on 100 μL of the release medium by

ARTICLE

Figure 1. Feret’s diameter (a) and circularity (b) of wet and dried alginate beads (A) in the presence of trehalose (A-T) and β-CD (A-TCD) or pectin (AP-T). Wet beads (striped bars); freeze-dried beads (FD, grey bars); vacuum-dried beads (VD, white bars); air-dried beads (AD, black bars). Standard deviation values are included. The number of replicates was 50 for wet beads and around 40 for dried beads. For a given treatment, significant differences due to beads composition are indicated with different letters (a
d; P < 0.05). For beads of same composition, significant differences due to the treatment are indicated with different numbers (1
3; P < 0.05). adding 400 μL of sucrose solution (200 mM in sodium acetate buffer 50 mM pH 4.6). After 10 min at 37 °C, the reaction was stopped by enzyme thermal denaturation (10 min at 100 °C) and the reducing sugars were determined spectrophotometrically at 546 nm using 3,5dinitrosalicylic acid method.30 For each system, the amount of hydrolyzed sucrose after a given treatment (S) was related to the amount of sucrose hydrolyzed before the treatment (S0), and the remaining activity (R.A.) was expressed as a percentage: R.A. = 100S/S0. Two samples were taken for each condition and duplicate measurements were performed. The confidence interval was 7%, calculated by measuring four samples of the same run. To evaluate the effect of the different additives on enzyme stability after beads generation, independently of the loading efficiency, an activity index was calculated as shown in eq 2: activity index ¼

act:=L:E: ðact:=L:E:ÞA

ð2Þ

where act./L.E. is the ratio between the enzyme activity and the loading efficiency of a given bead formulation after generation and (act./L.E.)A is the ratio between the enzyme activity and the loading efficiency of the alginate beads after generation. The effect of treatment (freezing, thermal treatment or drying methods) and the effect of composition on remaining invertase activity were analyzed by 1 way ANOVA with Bonferroni post test using GraphPad Prism v5. A t test was used to evaluate significant differences between the activity indexes for different beads compositions. 3149

dx.doi.org/10.1021/bm2009075 |Biomacromolecules 2011, 12, 3147–3155

Biomacromolecules

’ RESULTS AND DISCUSSION Beads Characterization. Alginate macrobeads of different formulations obtained by the drop method were dehydrated by freeze-drying (FD), vacuum drying at 25 °C (VD), or drying in a forced air circulation oven at 40 °C (AD). The size and morphological characterization of the beads are of interest because they could have impact on the stability of the encapsulated compounds and on textural properties of the beads.11,31 Size (determined as Feret’s diameter), circularity, area, and perimeter of the wet and dried beads were analyzed by optical microscopy and a digital image processing technique. Figure 1 shows the Feret’s diameter (Figure 1a) and circularity (Figure 1b) obtained for wet and dried alginate beads (A), generated in the presence of trehalose (A-T), and β-cyclodextrin (A-TCD) in the gelation media or pectin (AP-T) in the core gel. Changes in perimeter and area (data not shown) followed the same trend shown in Figure 1 for the Feret’s diameter. Wet beads had a diameter close to 1.5 mm, which was slightly modified by the presence of trehalose and β-CD and showed high circularity for all the studied compositions. A-TCD beads had a drop-like form (as observed by visual inspection), which was reflected in the higher diameter and lower circularity of these beads. Upon drying, beads had drastically reduced their size (Figure 1a, except FD A-TCD) and circularity (Figure 1b), and higher dispersion of the measured parameters was observed, as reflected in the higher standard deviation values. Freeze-drying was the method that less affected the size, independently of the composition. The size reduction was up to 40% for the freeze-dried beads and between 40 and 60% for vacuum- or air-dried ones in relation to wet beads. In the presence of pectin or β-CD, the diameter of the dried beads was significantly higher than that of A or A-T. For any of the drying methods, the addition of trehalose improved the circularity of the beads, further addition of β-CD or pectin seems to counteract the effect of the sugar. Freeze-drying was the method that most affected the circularity of the beads except in presence of pectin which allowed better maintaining this property regardless of the type of drying. No significant differences were observed in diameter or circularity of the beads dried either by VD or AD. Considering the high influence of water on stability when formulations containing active compounds are stored, the analysis of the water content and water activity becomes of fundamental importance to define the appropriate processing and storage conditions. Particularly, after drying additives like trehalose can crystallize as dihydrate at aw values between 0.43 and 0.90 (modifying the water content and biomolecule stability in the system).14,32 The aw values determined for wet beads were between 0.994 and 0.997 and were similar to previously reported data for alginate beads.1 Alginate wet and dried beads showed higher water content than the beads containing trehalose, β-CD or pectin. The aw values of dried beads were between 0.181 and 0.330; the lower values were obtained for the freeze-dried systems. A-TCD beads had the lower water content and aw values, in both wet and dried states. The maintenance of an amorphous matrix is generally desired in order to assess physicochemical stability of dehydrated systems.14,15 Therefore, glass transition temperature (Tg) was determined by DSC for the dried beads studied. Table S1 (see Supporting Information) shows the water content, Tg, Δcp values (difference in specific heat capacity, related to the baseline shift), and enthalpy relaxation values obtained for the different beads

ARTICLE

and drying methods. The Tg could only be determined for beads containing trehalose (Table S1), with values being similar to those previously reported for this sugar.13,32 Tg could not be observed in alginate beads without trehalose (A) in our experimental conditions. The Tgs of polymers are often more difficult to observe in comparison with sugars, especially in low moisture systems, because of their low Δcp and a wide temperature range of the glass transition.33 Tg values obtained for the combined systems indicated that beads were in the amorphous state at ambient temperature. The protective effect of trehalose on biomolecules is not only explained by the capacity of the sugar to form glassy structures, in which biomolecules are kinetically stabilized, but also by the specific hydrogen bonds established between the sugar and the biomolecule.13 As also shown in Table S1, both vacuum or air-dried beads presented higher Δcp and enthalpy relaxation associated to Tg, which indicates the formation of more stable glasses during the aging (20 days at temperature below Tg prior to DSC determination).34 It is important to note that beads dried by freeze-drying are subject to a rapid cooling under liquid nitrogen and are dried in a very high vacuum. These treatments under extreme conditions generate glasses with an excess of energy,35 which have lower (or do not have) enthalpy relaxations in the Tg zone. Vacuum or airdried beads are subjected to milder conditions than freeze-dried ones, and the corresponding glasses showed higher enthalpy relaxation (Table S1), as shown also by Surana et al.36 for trehalose amorphous systems. Furthermore, higher enthalpy relaxation values were obtained when β-CD or pectin were included in the formulation. The magnitude of enthalpy relaxation is related to the mobility of the molecules in the glassy state and could directly affect the physical stability of the glasses.37,38 It is interesting to note that trehalose systems presenting high enthalpy relaxations showed less tendency of the sugar to crystallize,34 which is known to negatively affect the stability of several enzymes and proteins.14,15 Enzyme Release and Transport Mechanism Modeling. Release from Wet Beads. The release of the enzyme is an important parameter for potential applications of encapsulated systems. It is well-known that diffusion of small molecules such as glucose or ethanol from alginate beads is not affected by the alginate network, while the diffusion of proteins from the gel is dependent on their molecular weight.39 Also, the charge of the protein, the concentration of alginate, and the ratio of mannuronic/guluronic acids influenced protein release.5 The electrostatic interactions between alginate (and pectin) and the enzyme could define the release mechanism, since the pH of the medium determines the global charge of the enzyme and of the polymers. The isoelectric point of invertase is between 4.0 and 4.526 and pKa values of alginate are 3.38 and 3.65 for mannuronate and guluronate residues,27 respectively. Beads were generated at pH 3.8 to ensure the maximum electrostatic interaction between alginate and the enzyme. Although the selected pH is very close to the pKa of guluronate residues of alginate, due to the high affinity of calcium ions for guluronate, a gel was adequately formed, and the enzyme was successfully immobilized, as confirmed by the high loading efficiency parameter (L.E.). The L.E. evaluates the invertase content inside the beads after generation by determining the ratio between the amount of protein present in the beads and the enzyme concentration in the alginate solution. The L.E. of the alginate beads (A) was 83 ( 1% and was comparable to reported L.E. values for encapsulated bovine serum albumin.40 Trehalose incorporation leads to a greater loading efficiency 3150

dx.doi.org/10.1021/bm2009075 |Biomacromolecules 2011, 12, 3147–3155

Biomacromolecules

ARTICLE

Figure 2. Enzyme release from wet beads exposed to acetate buffer at pH 3.8, 4.5, or 5.5 as a function of incubation time at 4 °C. The release is expressed as the ratio between the activity of the enzyme released at each time (At) with respect to the maximum invertase activity correspondent to all the invertase present in the beads (A∞). Lines show the fitting obtained using Peppas equation (eq 3, Supporting Information). Standard deviation values (n = 4) are included.

(L.E. was 92 ( 2% in A-T beads), while the incorporation of β-CD or pectin offset the positive effect of the sugar, being L.E. values 82 and 73 ( 1%, for A-TCD and AP-T beads, respectively. Pectin was also negatively charged during the AP-T beads generation, because its isoelectric point is at 3.5,9 similar to alginate. By changing the pH of the solution, different enzyme
alginate
pectin electrostatic interactions can be generated. Figure 2 shows the enzyme release from wet beads exposed to acetate buffer at pH 3.8, 4.5, or 5.5. At the selected pH levels, alginate and pectin were always negatively charged, while the global charge of the enzyme was different, being positive at pH 3.8, neutral at pH 4.5, or negative at pH 5.5. The enzyme release was evaluated as a function of time of storage by the relation At/A∞, where At is the activity of the invertase released at time t and A∞ is the maximum invertase activity correspondent to all the invertase present in the beads. The general behavior of enzyme release was similar to previously reported data for other types of compounds41
43 in which at short times (t < 4 h) a rapid enzyme release was observed, while at long times (t > 4
5 h) the rate of release decreased, and finally a plateau was reached (data not shown). At pH 3.8, the release of the enzyme was much slower than at the other studied pH levels. At this pH, invertase and the polymers alginate or pectin are oppositely charged, and therefore, electrostatic interactions between the enzyme and the biopolymers are favored. At pH 5.5, both enzyme and alginate/pectin were negatively charged, leading to repulsion forces between them, and higher release of the protein was observed. However, the highest release rate was observed at pH 4.5. Because pH could also influence the release mechanism,5 a mathematical modeling was applied to analyze changes in the release mechanism due to changes in the environmental conditions. Many models have been developed on the basis of the

Fick’s equation. Because the exact solution for the Fick equation is rather complex, in the past years some models that simplify the analysis were studied. In particular, Peppas equation relates the fractional release of a drug with time (eq 3, Supporting Information)1,44
48 and it is possible to adapt it to our material (see Supporting Information). The general expression of eq 3 was previously employed to study the release of several substances from dry and wet beads.1,11,44 For Fickian conditions, Ritger and Peppas47,48 simplified solutions to Fick’s second law for many geometries, different boundary conditions, and short times (which correspond to a fractional release of 0.6), allowing a simple calculation of the diffusion coefficient D, according to eq 4 (see Supporting Information). The parameters n, k, and D, obtained by applying eqs 3 and 4 to data displayed in Figure 2, are shown in Table S2 (see Supporting Information). The obtained R2 were higher than 0.96. The enzyme release from the A beads at pH 3.8 and 4.5 showed n values of 0.43, which implies that the main transport mechanism is Fickian and is limited by diffusion, so D could, thus, be calculated using eq 4. The D value was 10 times higher at pH 4.5 than at pH 3.8, according to the stronger electrostatic interactions promoted at pH 3.8. Then, the electrostatic interactions between alginate and enzyme could define the release mechanism. In systems at pH 5.5, the n values were between 0.55 and 0.75, which correspond to anomalous transport,49 in which both diffusion and polymer chains relaxation influence the enzyme release.50
52 Because n values are quite different between systems, it is unadvised to perform a direct comparison of k values. The presence of trehalose, β-CD, or pectin influenced the release mechanism by changing the n value, indicating that the polymer chains relaxation, besides a diffusion mechanism, 3151

dx.doi.org/10.1021/bm2009075 |Biomacromolecules 2011, 12, 3147–3155

Biomacromolecules influenced the release on wet beads.1,53 As shown in Figure 2b,c, enzyme release was very fast from systems containing β-CD, as will be discussed later.

Figure 3. Enzyme release from freeze- (FD) and vacuum-dried (VD) A-T beads (a) and from the FD beads with further pectin or β-CD addition (b) as a function of incubation time in acetate buffer 50 mM pH 4.5 at 4 °C. The release is expressed as the ratio between the activity of the enzyme release at each time (At) with respect to the maximum invertase activity correspondent to all the invertase present in the beads (A∞). Lines show the fitting obtained by using Peppas equation (eq 3, Supporting Information). Standard deviation values (n = 4) are included.

ARTICLE

The D values obtained for the alginate beads (A) were reasonable according to the size of the enzyme involved. Tanaka et al.43 reported values of D for release from wet alginate beads between 10
7 and 10
8 cm2/s for γ-globulin (154 kDa) and fibrinogen (341 kDa), respectively, which are enzymes of a similar molecular weight to the one used in our work (invertase, 250 kDa). Other authors reported D values between 10
6 and 10
7 cm2/s for different compounds (of minor molar mass) encapsulated in wet alginate beads.41
43 Release from Dried Beads. Figure 3 shows the release of the enzyme depending on the type of drying (Figure 3a) and on the composition of the freeze-dried systems (Figure 3b), as a function of incubation time in acetate buffer pH 4.5. Only beads containing trehalose were analyzed since drying of alginate beads without trehalose (A) resulted in a significant loss of the enzyme activity, as will be discussed further in the next section. Figure 3a shows that the release was much faster in the vacuum-dried A-T beads than in the freeze-dried ones. Figure 3b shows that the rate of invertase release in freeze-dried beads greatly increased when pectin and mainly β-cyclodextrin were included. The release kinetics shown in Figure 3 were fitted employing eq 3, and the obtained parameters are shown in Table S3 (see Supporting Information). The n values were