Full Paper
403
DOI: 10.1002/mabi.200600022
Summary: The preparation of an enzymatic controlled drug release system from blends of PVA/starch/aA, in the form of films, is described. It was shown that aA hydrolyses the starch within these films, resulting in a time-dependent change of the porosity in the matrix. Films were characterized by calorimetric analysis to study the interactions between the enzyme and the polymeric constituents at the molecular level. The presence of aA, in fact, influenced the PVA cristallization in the blends. Release tests and permeability experiments were carried out to evaluate the transport properties of the films. An increase in porosity and permeability was observed
by increasing aA content (16–28 wt.-%). Films loaded with theophylline and caffeine were also prepared to analyze drug release properties of the matrix. Drug release kinetics were coherent with the measured changes in porosity: at higher aA concentrations the amount of released drug increased under the influence of diffusion and erosion processes. The results obtained are promising for the realization of drug delivery devices for a rapid release or for the release of poorly soluble drugs which usually remain entrapped in the matrix.
SEM images of a PVA/starch/aA film before (A) and after (B) the erosion.
Enzymatic Erosion of Bioartificial Membranes to Control Drug Delivery Maria Laura Coluccio,1 Gianluca Ciardelli,*1,2 Franca Bertoni,1 Davide Silvestri,1 Caterina Cristallini,3 Paolo Giusti,1 Niccoletta Barbani1 1
Department of Chemical Engineering, Industrial Chemistry and Science of Materials, University of Pisa, Via Diotisalvi 2, 56126 Pisa, Italy E-mail:
[email protected] 2 Department of Mechanics, Politecnico in Turin 10129, Italy 3 Institute for Composite and Biomedical Materials (IMCB-CNR), Pisa 56126, Italy
Received: February 3, 2006; Revised: March 27, 2006; Accepted: March 29, 2006; DOI: 10.1002/mabi.200600022 Keywords: controlled release; enzymatic bioerosion; membranes transport properties
Introduction Recent development in therapeutic approaches based on drug delivery formulations require systems enabling a control of drug release dependent on time and on the site. In the first case the release must take place for a significant period (hours or days) and/or it must begin when the organism requires it. For a distribution control, the drug must be released at the target where the effect is required and its concentration must be within the window of
Macromol. Biosci. 2006, 6, 403–411
therapeutic levels. For example, drugs that are rapidly metabolized and eliminated from the body often require a continuous administration (as injections, etc.) or, better, a drug delivery system which assures an extended release duration; on the other hand drugs such as chemotherapeutics should be released to the precise site within the body.[1] Several approaches based on controlled release by polymer swelling, drug diffusion or polymer erosion have been proposed to achieve these goals.[2]
ß 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
404
M. L. Coluccio et al.
This study regards the drug release from a bioerodible matrix. A degradable controlled release system is based on a homogenous erosion of a polymer matrix through a physical or chemical mechanism (e.g., ionization, protonation, solubilization, hydrolytic or enzymatic degradation of the polymer). In the literature the erosion of biodegradable polysaccharide matrices has already been investigated; in particular starch and starch-based blends containing aamylase were prepared to obtain an accelerated drug release, useful, for example, for drugs with limited solubility or for drugs whose solubility can be influenced by the variation of gastro-intestinal pH. The introduction of matrices based on biological polymers is interesting since the degradation of natural products occurs naturally in the human body, producing no harmful metabolites; in fact a potential disadvantage of the biodegradable polymers is the eventual toxicity of the degradation products.[3–7] In this work, the preparation and characterization of bio-artificial polymeric materials, through blending synthetic and natural polymers, was studied in order to exploit the properties of a-amylase-controlled starch-based matrices and to combine the good mechanical properties of synthetic polymers with the biocompatibility of biopolymers and of their degradation products.[8–10] The idea is to obtain a material in which the porosity and, accordingly, the drug release are dependent from the amount of enzyme and its activity. In fact, the main release mechanism of a controlled drug delivery system is the drug diffusion through the matrix that is usually described by Fick’s law since, as the drug diffuses out, its concentration decreases with time. In the presence of a-amylase, both diffusion and erosion can operate simultaneously with a higher control of the release.[11] In most controlled drug release applications a zero-order kinetics is desirable and the appropriate combination of diffusion and erosion permits it. In particular the zero-order drug release is obtained with a surface-erosion controlled system, where the drug release rate is equal to the erosion rate of the delivery system, whilst the drug release rate from the delivery system undergoing bulk degradation is difficult to control because the release rate may change as the polymer degrades.[12] However, the bulk erosion of a matrix can be indicated, for example, to accelerate the drug release. This study refers to the use of starch as the biological component of bio-artificial materials with poly(vinyl alcohol) (PVA). The third component, a-amylase (aA), is an enzyme which is able to hydrolyze a-1–4 glucosidic bonds of starch. Starch is usually a combination of amylose, the nonbranched fraction, containing about 1 000 D-glucopyranosidic units, and amylopectin, the branched component. aA is an endoenzyme cleaving only the a-(1–4) D-glucopyranosidic linkages in starch. The degradation products of starch are mainly oligosaccarides, dextrin and maltose, which are harmless to the human body.[13–16] For this reason and also for its low cost, starch and cross-linked Macromol. Biosci. 2006, 6, 403–411
www.mbs-journal.de
starches are widely used for biomedical and environmentally friendly applications.[17,18] PVA is a hydrophilic synthetic polymer with reported biocompatibility and excellent mechanical properties. In particular, high molecular weight PVA is preferred because it does not release the monomer when it is incubated in an aqueous solution. This material eventually blended with its copolymers are used in several drug delivery applications, especially in the preparation of environment-sensitive hydrogels for ‘intelligent’ insulin release.[19–22] The enzymatic erosion of starch/PVA/aA films was investigated together with their drugs (theophylline and caffeine) release properties.
Experimental Part Materials PVA (Aldrich Chemie, Steinheim, Germany) (MW ¼ 85 000– 146 000, degree of hydrolysis: 99%), soluble starch (Carlo Erba, Italy), and aA from Bacillus subtilis (Sigma, St. Louis, MO, USA) (MW ¼ 96 000) were used as received. Film Preparation A 1% (w/v) starch solution was prepared at 80 8C in water, by stirring for 3–4 h and a 1% (w/v) PVA solution was prepared at 90 8C by stirring for 30 min. A starch/PVA blend was obtained mixing equal volumes of starch and PVA solutions. Then, starch/PVA/aA blends were prepared adding 20 mg and 40 mg of aA (16% or 28% in weight) into 10 ml of the starch/PVA solution. Blends with addition of theophylline (30 mg/10 ml) and caffeine (30 mg/10 ml) were also prepared. Starch/PVA, starch/PVA/aA, starch/PVA/drug and starch/PVA/aA/drug films (where ‘‘drug’’ indicated theophylline or caffeine) were obtained by evaporative casting at 37 8C in a ventilated oven (Table 1). Crosslinking by Dehydro-Thermal Treatment (DHT) DHT of all samples was carried out through two dehydration steps and a final cross-linking step. First, the films were placed in a vacuum oven at 50 8C and at a pressure of 0.1 mmHg for 2 h. Secondly, the temperature was then increased to 90 8C, and the films were treated for 3 h. Crosslinking was achieved by raising the temperature to 115 8C and keeping the samples at this temperature for 16 h under static vacuum.[23] Measurement of Enzyme Activity The effect of the DHT on enzyme activity was studied in our previous work.[23] The Km and Vmax values were calculated Table 1.
Acronyms used for the prepared films.
Film
Acronym
starch/PVA starch/PVA þ 20 mg aA starch/PVA þ 40 mg aA
TQ E20 E40
ß 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
405
Enzymatic Erosion of Bioartificial Membranes to Control Drug Delivery
using the Lineweaver and Burk diagram. The activity of the enzyme, after incubation in a water solution, was found to be similar for the crosslinked films compared with the untreated films. Thermal Analysis by DSC The thermal behavior of the films was studied using a PerkinElmer DSC7 differential scanning calorimeter. The samples were sealed into aluminium pans and scans were carried out at 10 K min1, under nitrogen flux. Analysis of Film Stability by Release Tests Release tests were carried out to evaluate the spontaneous leakage of PVA and starch from the films in aqueous solution, due to the solubility of materials in water. Both the release tests of PVA and those of starch were repeated three times. Each film was immersed in deionized water at 37 8C and the incubation solution was changed at fixed times. The removed solution was analyzed with a Shimadzu UV-VIS 2100 Scanning Spectrophotometer as follows:
a) the method of Buganda and Rudin was used to test the PVA release. This method is based on the use of iodine which generates a green complex when binding to PVA in the presence of boric acid. The absorbance of this compound is measured at 690 nm with an UV-Vis spectrophotometer.[24] b) the quantity of starch released was measured with an indirect method. The starch in solution was hydrolyzed adding aA and incubating the solution at 37 8C for 1 h. Then, the produced maltose was measured with the dinitrosalicylic acid (DNS) method. This method provides a reaction between DNS with the hydrolysis products of starch (maltose and dextrines) forming aminonitrosalicylic acid. The concentration of maltose was calculated from the amount of aminonitrosalicylic acid measured by spectrophotometry at 546 nm.[25]
Permeability Measurements and Morphological Characterization The starch hydrolysis influences the transport properties and the morphological characteristics of the films. Therefore the permeability, in convective or diffusive regime, of the E20 and E40 samples, before and after incubation in a buffer solution for 8–12 h, were investigated. The hydraulic permeability was measured with a permeation cell as shown in Figure 1(a), in which the driving force of the process was a pressure gradient DP. The relationship between the water volumetric flux (Jv) and DP may be expressed as follows: Jv ¼ Lp DP where Lp ¼ hydraulic permeability, a quantitative information on the solvent convective transport ([Lp] ¼ m2s kg1) DP ¼ P2 P1 P2 ¼ pressure imposed by pump G1 P1 ¼ atmospheric pressure. Permeability measurements were also performed at 37 8C using aqueous solutions containing solutes of different molecular weight, including sodium chloride (NaCl) (MW: 58 Da), vitamin B12 (MW: 1 355 Da) and bovine serum albumin (MW: 69 000 Da). The experimental apparatus used for solute permeability measurements is described in detail elsewhere.[26] It essentially consists of a stirred diffusion cell (Figure 1(b)) with two chambers separated by the membrane to be tested.[27,28] The
Leakage of aA The amount of aA leaking from E20 and E40 blends was measured. Each film was immersed in distilled water in a test tube at 37 8C. At different times 1 ml of the reaction solution was removed and incubated at 37 8C with 1 ml of a 1% substrate solution (Zulkowsky starch). After 10 min, the amount of maltose in solution was measured with the DNS method to calculate the amount of substrate digested by the enzyme. From the enzymatic activity the amount of leaked enzyme was deduced. The experiment was carried out in triplicate. Enzymatic Erosion The enzymatic erosion occurred when starch, contained into the films, was hydrolyzed from the endogenous enzyme in a buffer solution (pH ¼ 6.9) at T 37 8C. Three films of the E40 and three of the E20 samples were immersed each in 40 ml of buffer solution. At different times, 1 ml of the solution was removed and the produced maltose was measured with the DNS method, as described in the previous section. Macromol. Biosci. 2006, 6, 403–411
www.mbs-journal.de
Figure 1. (a) Apparatus for hydraulic permeability; (b) Apparatus for diffusive permeability. ß 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
406
M. L. Coluccio et al.
solute concentration on the permeate side of the membrane gradually increased until it became constant (i.e. until steadystate conditions were obtained). When the steady state is reached, the flux of solute across the membrane, Js, can be written in terms of solute concentrations in the two chambers of the diffusion cell as follows: Js ¼
PA ðCd CrÞ; h
where P is the solute permeability, A is the membrane area, h is the swollen membrane thickness, Cd and Cr are the solute concentrations in donor and receptor chambers of the diffusion cell, respectively. The above equation has been used to calculate P, provided that the concentration gradient, the flux under steady-state conditions and the swollen membrane thickness are known. Under a simultaneous action of a concentration and a pressure gradient (mixed condition), for analogy with the previous equation, the flux may be written as: Js ¼
P A ðCd CrÞ; h
where P* expresses the solute permeability in the mixed conditions. The morphological characteristics of the E20 and E40 films were analyzed by the Scanning electron microscopy (SEM), carried out using a Jeol JSM 300 microscope. Films were fractured in air, sputtered with gold and then observed at 10 kV. Analysis of Model Drugs Release Released theophylline and caffeine in solution were directly measured with the spectrophotometer by reading the absorbance at 296 and at 298 nm, respectively.
Figure 2. Macromol. Biosci. 2006, 6, 403–411
Table 2. Melting point and enthalpy of fusion of TQ, E20 and E40 samples compared to the data of crosslinked PVA. Film
crosslinked PVA TQ E20 E40
Tm
DH
8C
J g1
222 219 196 203
77 60 48 48
Results Thermal Analysis A decrease of PVA melting temperature and a lowering of crystallinity compared with samples without aA was observed (Figure 2, Table 2). However, these results can not be directly ascribed to the presence of the enzyme. In a previous study[23] we demonstrated, by means of DSC and X-ray diffractometry, that the crystalline structure of PVA increased in aA/PVA blends with aA content. The decrease of Tm and melting enthalpy of PVA in E20 and E40 samples could be explained with the presence of degradation products, occurring during film preparation, such as maltose or malto-dextrines that may affect the PVA crystallinity. A decrease of PVA melting temperature, related to an increase of the aA content, and a lowering of crystallinity compared with the samples without aA was observed (Table 2).
DSC thermograms of Starch/PVA and Starch/PVA/enzyme film. www.mbs-journal.de
ß 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
407
Enzymatic Erosion of Bioartificial Membranes to Control Drug Delivery
Analysis of Film Stability – PVA and Starch Release The percentages of leaked PVA and starch into the incubation solution at 37 8C, normalized to the initial PVA and starch content of the film, are reported as a function of time (Figure 3a and b). The released PVA amount from all samples is very low (
