Ultrathin Antibiotic Walled Microcapsules - Biomacromolecules (ACS

Biomacromolecules 2005, 6, 229-234

229

Ultrathin Antibiotic Walled Microcapsules Ajay J. Khopade,*,†,‡ N. Arulsudar,‡ Surekha A. Khopade,‡ and J. Hartmann† Max Planck Institute of Colloids and Interfaces, Am Mu¨hlenberg 1, D-14476 Golm, Germany, and Sun Pharma Advanced Research Centre, Tandalja, Baroda-390020 Gujarat, India Received August 3, 2004; Revised Manuscript Received September 29, 2004

Ultrathin microcapsules comprised of anionic polyelectrolytes (PE) and a polycationic aminoglycoside (AmG) antibiotic drug were prepared by depositing PE/AmG multilayers on zinc oxide (ZnO) colloid particles using the layer-by-layer self-assembly technique and subsequently dissolving the ZnO templated cores. The polyelectrolytes, dextran sulfate sodium (DxS) and poly(styrenesulfonate) (PSS), were selected owing to their different backbone structure. An aminoglycoside, tobramycin sulfate (TbS), was used for studying DxS/TbS or PSS/TbS multilayer films. The multilayer growth on ZnO cores was characterized by alternating zeta potential values that were different for the DxS/TbS and PSS/TbS multilayers due to the PE chemistry and its interaction with Zn2+ ions. Transmission and scanning electron microscopy provide evidence of PE/TbS multilayer coating on ZnO core particles. The slow acid-decomposition of the ZnO cores using weak organic acids and the presence of sufficient quantity of Zn2+ in the dispersion were required to produce antibiotic multilayer capsules. There was no difference in the morphological characteristics of the two types of capsules; although, the yield for [PSS/TbS]5 capsules was significantly higher than for [DxS/TbS]5 capsules which was related to the physicochemical properties of DxS/TbS/Zn 2+ and PSS/TbS/Zn2+ complexes forming the capsule wall. The TbS quantity in the multilayer films was determined using a quartz crystal microbalance and high performance liquid chromatography techniques which showed less TbS loading in both, capsules and multilayers on planar gold substrate, than the theoretical DxS:TbS or PSS:TbS stoichiometric ratio. The decomposition of the [PE/TbS]6 multilayers was fastest in physiological buffer followed by mannitol and water. The decomposition rate of the [PSS/TbS]6 multilayers was slower than [DxS/TbS]6 monolayers. The incomplete decomposition of DxS/TbS under saline conditions suggests the major role of hydrogen bonding for stability of DxS/TbS multilayers. A combination of hydrogen bonding and hydrophobic interaction between phenyl rings in PSS was responsible for PSS/TbS multilayer stability. In vivo studies in rabbits highlight the safety and sustained drug delivery potential of the PE/AmG microcapsules. The antibiotic walled ultrathin capsules presented here are suitable for sustained ophthalmic antibiotic delivery. Introduction The extension of the layer-by-layer (LbL) technique1 to sacrificial spherical templates produces novel ultrathin capsules after core removal that can be utilized for a number of applications.2 Ultrathin capsules are prepared via the LbL self-assembly method on spherical templates which may be comprised of permutation and combination of oppositely charged polyelectrolytes (PE),2 dendrimers,3 bio-polyelectrolytes (polynucleotides, proteins and polysaccharides),4 polyvalent small molecular weight organic compounds,5 inorganic ions,6 metal coordinate complex,7 and inorganic nanoparticles.8 Capsules with different pairs as building blocks can be designed to have significantly different characteristics, functions and applications.2-8 An attractive application of capsular colloids is drug delivery, because of their distinctive physicochemical properties, such as, ultrathin capsular membrane (i.e., nanometer thick film),9 membrane elasticity,10 stimuli responsive membrane permeability,11 * To whom correspondence should be addressed. Fax: +91 265 2354897. E-mail: [email protected]. † Max Planck Institute of Colloids and Interfaces. ‡ Sun Pharma Advanced Research Centre.

encapsulation/entrapment of a range of chemical compounds, solvents and biological macromolecules in the hollow core or membrane12-14 and their subsequent slow release,15 and the possibility of surface modification and/or functionalization.16 This application demands that both core and coating material be biocompatible. Biocompatibility of the core is an important issue because of the presence of the traces of the core decomposition products in the multilayer capsules.17 The most commonly reported decomposable particle supports for ultrathin capsule preparation are nonbiocompatible, decomposable melamine formaldehyde and polystyrene that require a dilute acid (HCl) or organic solvent (tetrahydrofuran), respectively, for their removal.2-10 To a lesser extent, inorganic particles such as gold and silica nanoparticles,18 metal carbonates19 microparticles (MnCO3, CdCO3 and CaCO3), polyester (polylactide and lactide-co-glycolide),20 and biological cells21 are also used as core materials for ultrathin capsule preparation. Some of these cores involve extremely cautious coredissolution processes (removal of silica with HF),18 use of organic solvents20 (acetone, N-methyl pyrrolidone mixture for polyester cores), or oxidizing agents21 with limited bio-acceptability. Another major constraint in the practical utility of ultrathin capsules is the cationic

10.1021/bm049554a CCC: $30.25 © 2005 American Chemical Society Published on Web 12/07/2004

230

Biomacromolecules, Vol. 6, No. 1, 2005

membrane building material since most of them are objectionably toxic especially from the routes other than oral, such as ophthalmic, pulmonary, nasal, and injectable routes. Recently, our focus is to study ultrathin capsules that are practically relevant for drug delivery; hence, we carefully select purely pharmaceutically acceptable, biocompatible materials, especially the positively charged membrane building material. For example, our recent report on core-assisted microcapsules uses an acceptable, biocompatible divalent Ca2+ as membrane building cationic component generated from the CaPO4 core.22 In the current work, we examine the ultrathin capsule formation using one of the several aminoglycoside antibiotics that are polycationic in nature and hence are suitable as the membrane components for the preparation of multilayer capsules.2-8 These antibiotic capsules could themselves serve as delivery systems where there is a need to control infections, for example, external ocular infections caused by S. epidermidis, S. aureus, and P. aeruginosa. Current treatment of such infections is primarily based on the topical administration of aminoglycoside anti-infective agents such as tobramycin sulfate (TbS), which is also a drug of choice for the penicillin, gentamycin, and kanamycin resistant bacterial strains.23 TbS eye-drops are available in the market as 0.3%w/v aqueous solution (Tobrex, Alkon). The successful ophthalmic delivery of antibiotic capsules would require additional considerations such as the prolonged retention/delivery of capsules at the site of action (cul-dusac), sustained drug release at hours scale, nonirritability to the corneal membrane, and the stability of capsules in the medium. The present study is aimed at fulfilling these objectives. The antibiotic capsules are prepared using biocompatible zinc oxide particles as core and dextran sulfate (DxS) and poly(styrenesulfonate) (PSS) as a capsule shell forming material in combination with TbS. The Zn2+ ions produced from core decomposition are already approved for ophthalmic and pulmonary use. They enhance retention of amino glycoside antibiotics in the biological membranes.24 Polymeric sulfonates form an important component of donoreye (cornea) preserving solutions25 and ophthalmic formulations (Betaxon, Alcon Laboratories). This paper reports on fully biocompatible, pharmaceutically acceptable, and application defined ultrathin capsules. Experimental Section Materials. Dextran sulfate sodium (DxS), average Mw ∼ 100 kD, poly(ethylene)amine (Mw ∼ 25 kD) and poly(styrene sulfonate) sodium (PSS, Mw ∼ 70 kD) was purchased from Sigma-Aldrich Chemical Co. and was used without further purification. TbS was a kind gift from Sun Pharma Advanced Research Centre, Baroda, India. All experiments were carried out at room temperature (25 °C). Zinc acetate and diethylene glycol were purchased from SD fine chemicals, Pune, India. Parenteral grade mannitol was from Akzo Nobel, Italy. Ultrapure water (Millipore) with a specific resistance of more than 18 MΩ cm was used. All of the other reagents used for synthesis or HPLC were of analytical reagent grade. Preparation and Coating of ZnO Core Particles. Zinc oxide cores were synthesized in-house by a reported method,26

Khopade et al.

i.e., by hydrolyzing zinc acetate in diethyleneglycol at 160 °C without any particle size control step. The nanometer range particles present in the supernatant after mild centrifugation were removed. Ethylene glycol was washed off with water before the coating steps. The bare, neutral to weakly charged ZnO particles (0.2 mL of ∼5-7 wt % dispersion) were incubated with 1 mL of polyelectrolyte (PE) solution (1 mg mL-1, DxS or PSS) for 10 min, followed by three centrifugation (160g for 5 min)/wash cycles, and finally dispersed in 0.5 mL of water. PE-coated ZnO particles were then incubated with 1 mL of a TbS solution (1 mg mL-1) for another 10 min and three centrifugation/wash cycles performed (as above). The PE and TbS adsorption steps were repeated to build multilayers on the ZnO particles. Ultrathin capsules were finally obtained by dissolving the ZnO core by the addition of 1 mL of 0.01 M acetic acid solution for 10 min, followed by centrifugation (1000g for 10 min) and washing three times with water. Microscopy. Transmission electron microscopy (TEM) measurements were performed with a Zeiss EM912 Omega microscope operated at 120 kV (equipped with an EDX attachment). Scanning electron microscopy (SEM) was done at a low accelerating voltage of 3 kV on a Zeiss Gemini 1550, a high-resolution instrument. Microelectrophoresis. The electrophoretic mobility of the coated particles was measured to follow DxS/TbS multilayer growth on ZnO particles by using a Malvern Zetasizer 4. The mobilities were converted to the electrophoretic potential using the Smoluchowski relation ζ ) uη/, where u and ζ are the mobility and zeta potential of the particles and, η and  are the viscosity and permittivity of the solution, respectively. All measurements were performed in airequilibrated pure water without added electrolyte. Quartz Crystal Microgravimetry. A QCM setup using a frequency counter (HP 53131A) and a home-built oscillator for 9 Hz quartz crystal microbalance electrodes was used for measurements. The amount of material deposited on a QCM electrode is directly proportional to the observed QCM frequency shift, and can be calculated using the Sauerbrey equation,27a as described in previous work.27b In our systems, a frequency decrease of 5 Hz corresponds to a mass uptake of 4.3 ng (electrode area ) 0.16 cm2). High Performance Liquid Chromatography. High performance liquid chromatography (Waters, USA) was performed on a Hypersil BDS C-18, 250 × 4.6 mm, 5 µm column, using a mixture of 1.0 g of Tris in 420 mL of water, 10 mL of 1.0 N H2SO4 and 570 mL acetonitrile (total volume 1000 mL). The injection volume was 20 µL and the flow rate of the mobile phase was 1.5 mL/min having a total run time of 40 min. The eluted drug was detected at 365 nm using a UV detector. The working reference standard was 20 µg/mL of TbS solution in 1:1:3 Tris in 90% aqueous DMSO, alcoholic dinitroflurobenzene solution and acetonitrile mixture, which was prepared by heating to 60 °C for 50 min and cooling to room temperature before injection. The analytical method has excellent specificity, sensitivity and reproducibility. Animal Studies. Three rabbits were used at a time for each study. For the antibiotic clearance study, a single 50

Ultrathin Antibiotic Walled Microcapsules

Biomacromolecules, Vol. 6, No. 1, 2005 231

Figure 1. Chemical structures of (a) DxS, (b) PSS, and (c) TbS.

Figure 2. ζ-potential versus layer number for the LbL self-assembly of PE (DxS, dots and PSS, solid triangles) and TbS on ZnO particles. The odd layer numbers correspond to PE deposition and the even layer numbers to TbS adsorption. Layer number equal to zero corresponds to the bare ZnO particles.

µL dose of the DxS/TbS capsular suspension (equivalent to 0.3% w/v TbS) was instilled in the right eye cul-du-sac using a micropipet (inside the center of the lower cul-de-sac). Care was taken not to touch the corneal surface to avoid irritation. An aqueous suspension of 0.3% w/v TbS marketed eye drops was instilled as a standard in the right eye of another rabbit. The contra lateral eye served as a control for all rabbits and received STF (pH 7.4). After a definite period of time, eyelids were closed together and 20 µL of tear sample between the eyelids was collected with a calibrated capillary. This collected tear was diluted to 100 µL and injected into the HPLC to determine the drug concentration. The eye was examined after 1 h instillation for any manifestation of irritation caused by formulation under investigation such as swelling, redness and lacrimation. Results and Discussion Formation of DxS/TbS Microcapsules. In the current study, ZnO core particles are used for the first time for LbL assembly of poly-ionic materials. Hence, electrophoresis experiments were conducted to follow multilayer growth. The ζ-potential values for the PE and TbS adsorption steps are shown in Figure 2. The ζ-potential of bare ZnO particles was + 3.5 ( 1 mV. The ζ-potential of ca. -15 ( 2 mV was obtained when the first PE (DxS or PSS) layer was

Figure 3. (a) Transmission electron micrograph of ZnO core particles. Nanocrystallites of 20-50 nm size are visible on the surface of the particles shown in the inset. (b) Scanning electron micrograph of [DxS/TbS]5@ZnO. The [PSS/TbS]5@ZnO particles bear same morphological characteristics.

formed on ZnO core particles. Adsorption of the cationic TbS over the PE layers did not cause a reversal of sign of the ζ-potential. Alternating ζ-potential values are excellent indicator of multilayer film growth on particles, even though they alternated between higher and lower negative values for the PE and TbS adsorption steps, respectively.28 The innermost layer of DxS bears the ζ-potential of -13 ( 2 mV which is notably less than for DxS on the outermost layer (ca. -44 ( 3 mV) in [DxS/TbS]5 multilayers. This difference between ζ-potential of the innermost (-3 ( 1 mV) and outermost layer (-29 ( 3 mV) was lower for [PSS/TbS]5 multilayers. A probable explanation for this difference is that there is a considerable fraction of the PSS and DxS charges that are compensated by Zn2+ ions from the core in the film. It has been previously noted that the ζ-potential values vary due to the different nature of, the particle surface e.g. ion eluting inorganic particles or fixed surface charge density of polymeric particles onto which the layers were deposited and small molecular weight polyions as layer materials.5-6,22 The ion eluting inorganic cores and lower molecular weight coating materials usually lowers the ζ-potential values in water. On the whole, the alternating ζ-potentials observed with each coating step strongly suggest the sequential adsorption of PEs and TbS on ZnO core particles ([PE/TbS]n@ZnO). The uncoated particles used for the coating were spherical and 0.3-0.9 µm in size (Figure 3, parts a and b). The particles consisted of ∼20-50 nm nanocrystallites that are clearly visible on the surface of the particles in TEM (Figure

232

Biomacromolecules, Vol. 6, No. 1, 2005

Figure 4. Transmission electron micrograph of [DxS/TbS]5 multilayer capsules.

3a, inset) and SEM images (Figure 3b). Wide-angle X-ray diffraction measurements (not shown) also revealed hexagonal nanocrystallites of the same size without preferential growth direction. The [PE/TbS]5@ZnO particles showed a smoothened surface in the SEM, which indicates multilayer coating. Exposing [PE/TbS]5@ZnO particles to 0.01 M acetic acid resulted in dissolution of the core templates. The dissolved ZnO was expelled from the core via permeation of soluble Zn2+ and CH3COO- through the polyelectrolyte multilayers upon dissolution and washing.29 Evidence of capsule formation was directly obtained from TEM (Figure 4) indicating that PE/TbS complex was stable under acidic treatment condition. PSS and DxS are known to interact with Zn2+ so it was possible that the divalent zinc ions would displace some of the TbS molecules that bridges PSS or DxS chains.3b,22 Needlelike crystalline structures and fine ripples were observed on the capsule membrane surface under TEM, suggesting that ternary insoluble complexes of PE/TbS/Zn2+ formed the capsular membrane. The fine ripples could result from drying effects on the capsules resulted from the membrane formed on the irregular contours of the surface nanocrystallites (Figure 3a, inset) on ZnO template. The capsules were different compared to the conventional linear polyelectrolyte capsules or polyelectrolyte/dye capsules in that the majority of the capsules possess ripples rather than large folds on drying. At higher magnifications the membrane surface look crumbled. Energy-dispersive X-ray (EDX) analysis also revealed the presence of Zn in the capsules (data not shown). Although the two types of capsules (PSS/ TbS and DxS/TbS) were almost similar in morphological characteristics (Figure 4), the yield was considerably different. The DxS/TbS capsules showed much more destruction under TEM depending on the coredissolution conditions. Mild core dissolution conditions such as incubation of [PE/TbS]n@ZnO with weak organic acids (0.01 M acetic or citric acid) to dissolve the core improved the yield. It was absolutely necessary to have at least 15 mM Zn2+ ions in the dispersion medium. As Zn2+ ions were washed off from the dispersion medium (using dialysis) containing suspended capsules, the ripples disappeared and the capsular membrane collapsed into DxS/TbS complex, which looked like a circular patch or typical vesicles under the TEM. Although the PSS/TbS capsules were relatively stable, vesicular and droplets like structures were also observed under Zn2+

Khopade et al.

Figure 5. (a) Loading of TbS (% w/w) in PE/TbS multilayer film determined by QCM (dots) and HPLC (open circles) methods. (b) Less amount of TbS loading than the theoretical stoichiometric ratio for DxS/TbS and PSS/TbS is expressed as percentage. Triangles, DxS/TbS; squares, PSS/TbS; solid legends, QCM method; open legends, HPLC method.

removal conditions. To understand this difference, we studied the comparative DxS/TbS and PSS/TbS complex properties. The two complexes phase separate in water as high-density liquid and viscous liquid droplets that converts to a sticky, plastic, and semisolid/solid complex, respectively, on addition of Zn2+ ions. The two complexes air-dry as a glassy material with different properties described elsewhere.30 Analogously, variations in the yield of the capsules were obtained after decomposition of [PE/TbS]n@ZnO cores. Thus, it can be concluded that the yield of the capsules was a result of the different physicochemical properties of the DxS/TbS, PSS/ TbS, DxS/TbS/Zn2+ and PSS/TbS/Zn2+ complexes. TbS Loading. Loading of TbS in DxS/TbS and PSS/TbS multilayers from aqueous solutions was studied by QCM. The results are shown in Figure 5. A priming layer of a linear poly(ethylene)amine (Mw ∼ 25 000) was deposited to aid the adsorption of the first PE layer. A large shift, -δF ) 126 ( 40 Hz, corresponding to adsorption of the DxS or PSS, followed with a smaller -δF ) 89 ( 15 Hz for deposition of TbS. The results are expressed as the mean of three QCM measurements each for the two PEs adsorption steps, since the differences were not too large. The frequency change data correspond to 108 ( 34 ng of PE and 76 ( 13 ng of TbS deposition per cycle (PE/TbS ratio, 1:0.7), which is 30% and 37% lower than the theoretical stoichiometric ratio of 1:2.5 and 1:3.6 for PSS/TbS and DxS/TbS, respectively.31 This explains the negative ζ-potential values obtained even for the particles coated with a positively charged TbS outer layer. The TbS loading was calculated as 41 ( 2.5%w/w for three-step deposition. The TbS loading in the multilayers was found to increase up to 52 ( 2.8%w/w with increasing number of deposition cycles suggesting simultaneous desorption and adsorption of the two multilayer building species (Figure 5).32 Loading was also determined by HPLC because the QCM does not take into account the adsorption/desorption phenomenon that occurs during the multilayer buildup. The multilayer adsorbed QCMs (3 Nos. from the triplicate study) were pooled in a 1 cm quartz cuvette containing 0.5 mL of mobile phase and sonicated for 30 min to extract drug from the multilayers into the

Ultrathin Antibiotic Walled Microcapsules

Figure 6. Decomposition behavior of (a) [DxS/TbS]6 film and (b) [PSS/TbS]6 film in Millipore water (dots), isotonic mannitol (squares) and physiological buffer (triangles) solution. The two films were prepared under the same conditions. The percentage film remaining was measured as % mass loss from a film of the sample.

mobile phase (see experimental for mobile phase composition). The loading values determined by HPLC were higher (52 ( 2.5% to 63 ( 3.5%) than that obtained from QCM measurements (Figure 5) confirming desorption of PEs during TbS adsorption steps. The polyelectrolyte desorption could be reduced by an intermittent treatment of multilayers with ZnCl2.3b The drug loading was 48 ( 4.2%w/v in the capsule samples. Values from the two multilayers (planar and capsular) are fairly in mutual agreement considering the replacement of TbS by Zn2+ ions during core-removal from the capsules. Decomposition of [PE/TbS]n Films. The capsules were proposed for an ophthalmic use therefore, the [PE/TbS] multilayer stability with respect to isotonic physiological buffer and isotonic sugar solution was examined. Although multilayers of strong polyelectrolytes with appreciable molecular weights (> 10,000) are typically stable with respect to decomposition when immersed in salt solutions, high ionic strength solutions are shown to cause the removal of dye molecules from multilayers composed of polyelectrolyte and small, charged multivalent dyes assembled on planar supports.5-6,32 The weight loss of [DxS/TbS]6 multilayer film on gold substrate dipped in Millipore water, physiological buffer, isotonic mannitol solution was measured using a QCM. As shown in Figure 6, the [DxS/TbS]6 multilayers were stable in water or mannitol solution but were found to readily decompose when exposed to physiological buffer at pH 7.4. About 40% of the film remained on the gold substrate suggesting the role of hydrogen bonding in addition to the electrostatics in the formation of [DxS/TbS]6 multilayers. The fastest decomposition in buffer indicates that inspite of hydrogen bonding, electrostatic remains the major force for driving LbL assembly of DxS/TbS films. In the case of [PSS/ TbS]6 films, the rate of decomposition was slower. More than 80% of the film remained on the substrate during the study period. This shows in addition to hydrogen bonding, increased hydrophobic interactions between phenyl groups in the presence of salt (organizational effects30) also play a major role in [PSS/TbS]6 film stability. After similar treatments, i.e., when the dispersion containing the antibiotic

Biomacromolecules, Vol. 6, No. 1, 2005 233

Figure 7. Concentration profile of TbS in lachrymal fluid over time from TbS ultrathin capsules (dots) and commercially available eye drops Tobrex, 0.3% w/v TbS solution (solid squares). Approximately 150 µg (50 µL) of TbS was instilled in the eyes of rabbits. Inset shows the area under the curve of the drug concentration profile. The area denoted in the legend is calculated using a trapezoidal rule from time, t ) 0.5-8 h. The diagonally shaded portion is the difference in area and indicates an improvement in bioavailability from TbS ultrathin capsules compared to the TbS solution eye drops.

walled capsules was exposed to isotonic buffer solution or isotonic mannitol solution for the same periods and aliquots then dried on solid substrates for analysis, no intact capsules were detected by electron microscopy in the buffer treated dispersion. In addition, no pellet was obtained after centrifugation, pointing toward the decomposition of the capsules. Unlike conventional polyelectrolyte multilayers decomposition into individual components or complexes, the decomposition of PE/TbS capsules occur largely via the formation of liquid PE/TbS complexes indicated by the turbidity of the decomposition medium and observation of the aqueous biphasic liquid droplets under the optical or confocal laser scanning microscope. This is also in agreement with the TEM observations taken to optimize capsule decomposition. TbS Clearance and Corneal Irritation in Rabbits. The application of eye drops usually provokes tearing reflexes due to the burning, itching, and stinging sensation caused by the drops. This in turn causes an increased tear flow and consequently a dilution of the applied medication. With the subsequent blinking the formulation drains off leading to significant drug loss. Our study shows that the drug (TbS) concentration in lachrymal fluid from the ultrathin antibiotic capsule formulation was above the minimum inhibitory concentration for over 6 h compared to the commercial eye drop (Tobrex) formulation in which TbS was completely absent from the rabbit tears after 2 h. The TbS concentration profile in the lachrymal fluid over an 8 h period is shown in Figure 6. The area under the lachrymal TbS concentration versus time curve (AUC0.5-8 h) was 4 times higher for the antibiotic capsule formulation, which suggests improved bioavailability from these formulations compared to the commercially available TbS solution eye drops. Even though ultrathin capsular membranes offer limited drug loading and control on drug release, they can still be successfully utilized for effective drug delivery application through the judicious

234

Biomacromolecules, Vol. 6, No. 1, 2005

selection of the capsular components and route of drug administration. As a general response stated above, mild lacrimation was produced by all formulations for a transient period in the beginning including DxS/TbS capsules. However no sign of swelling and redness was observed in the eye of the rabbit during the entire period of study (8 h). This suggests that neither the ultrathin capsules nor the polymers cause irritation to the rabbit eyes. Though the ultrathin capsules are of sufficiently larger in size (0.3-0.9 µm), their ultrathin, flexible nature and decomposition behavior as described earlier fails to produce any irritation or adverse reaction to the eye. Conclusions We have shown that PE/TbS multilayer films can be coated on ZnO core particles and ultrathin antibiotic walled microcapsules can be obtained after acid-decomposition and removal of the ZnO core from PE/TbS multilayer-coated particles. The Zn2+ ions from ZnO core enhance the PE-PE binding interactions within the multilayers, forming insoluble PE/TbS/Zn2+ ternary complexes, and thus improving their stability. Depending on the number of layers and the type of polyelectrolyte, ultrathin capsules with different drug loading, yield and decomposition characteristics could be obtained. The DxS/TbS capsular films are nonirritating to the rabbit eyes and show sustained drug level in cul-du-sac of the rabbit eye demonstrating its utility for sustained release application. Acknowledgment. Rona Pitschke and Dayang Wang are thanked for assistance with TEM measurements. H. Mo¨hwald is thanked for supporting the work within the MPI of Colloids and Interfaces. A.J.K. acknowledges the financial assistance by Max Planck Research Foundation during the research stay at MPI of Colloids and Interfaces. Note Added after ASAP Publication. This manuscript was initially published ASAP on 12/7/2004. Changes were made to the third paragraph in the Experimental Section and the third paragraph in the Results and Discussion Section. The correct version was posted on 12/14/2004. References and Notes (1) (a) Decher, G.; Hong, J. D. Makromol. Chem. 1991, 46, 321. (b) For a review, see: Decher, G. Science 1997, 277, 1232. (c) Decher, G.; Eckle, M.; Schmitt, J.; Struth, B. Curr. Opin. Colloid Interface Sci. 1998, 3, 32. (2) Donath, E.; Sukhorukov, G. B.; Caruso, F.; Davis, S. A.; Mo¨hwald, H. Angew. Chem., Int. Ed. Engl. 1998, 37, 2201. (3) (a) Khopade, A. J.; Caruso, F. Nano Lett. 2002, 2, 415. (b) Khopade, A. J.; Caruso, F. Biomacromolecules 2002, 3, 1154. (4) (a) Schu¨ler, C.; Caruso, F. Biomacromolecules 2001, 2, 921. (b) Caruso, F.; Trau, D.; Mo¨hwald, H.; Renneberg, R. Langmuir 2000, 16, 1485. (c) Lvov, Y.; Caruso, F. Anal. Chem. 2001, 73, 4212. (d) Tiourina, O. P.; Sukhorukov, G. B. Int. J. Pharm. 2002, 242, 155. (e) Balabushevich, N. G.; Tiourina, O. P.; Volodkin, D. V.; Larionova, N. I.; Sukhorukov, G. B. Biomacromolecules 2003, 4, 1191. (5) Dai, Z. F.; Voigt, A.; Leporatti, S.; Donath, E.; Dahne, L.; Mo¨hwald, H. AdV. Mater. 2001, 13, 1339. (6) (a) Radtchenko, I. L.; Sukhorukov, G. B.; Leporatti, S.; Khomutov, G. B.; Donath, E.; Mo¨hwald, H. J. Colloid Interface Sci. 2000, 230, 272.

Khopade et al. (7) Caruso, F.; Schu¨ler, M.; Kurth, D. G. Chem. Mater. 1999, 11, 3309. (8) (a) Mamedov, A. A.; Kotov, N. A. Langmuir 2000, 16, 5530. (b) Voigt, A.; Buske, N.; Sukhorukov, G. B.; Antipov, A. A.; Leporatti, S.; Lichtenfeld, H.; Baumler, H.; Donath, E.; Mo¨hwald, H. J. Magn. Magn. Mater. 2001, 225, 59. (b) Antipov, A. A.; Sukhorukov, G. B.; Fedutik, Y. A.; Hartmann, J.; Giersig, M.; Mo¨hwald, H. Langmuir 2002, 18, 6687. (c) Shchukin DG, Sukhorukov GB, Mo¨hwald H. Angew. Chem., Int. Ed. 2003, 42, 4472. (9) (a) Leporatti, S.; Voigt, A.; Mitlohner, R.; Sukhorukov, G.; Donath, E.; Mo¨hwald, H. Langmuir 2000, 16, 4059. (b) Gao, C. Y.; Leporatti, S.; Moya, S.; Donath, E.; Mo¨hwald, H. Langmuir 2001, 17, 3491. (10) (a) Leporatti, S.; Gao, C.; Voigt, A.; Donath, E.; Mo¨hwald, H. Eur. Phys. J. E 2001, 5, 13. (b) Lulevich, V. V.; Andrienko, D.; Vinogradova, O. I. Condensed Matter 2003, 1-4. (c) Lvov, Y.; Antipov, A. A.; Mamedov, A.; Mo¨hwald, H.; Sukhorukov, G. B. Nano Lett. 2001, 1, 125. (11) (a) Sukhorukov, G. B.; Antipov, A. A.; Voigt, A.; Donath, E.; Mo¨hwald, H. Macromol. Rapid Commun. 2001, 22, 44. (b) Glinel, K.; Sukhorukov, G. B.; Mo¨hwald, H.; Khrenov, V.; Tauer, K. Macromol. Chem. Phys. 2003, 204, 1784. (12) Sukhorukov, G. B.; Dahne, L.; Hartmann, J.; Donath, E.; Mo¨hwald, H. AdV. Mater. 2000, 12, 112. (13) Moya, S.; Sukhorukov, G. B.; Auch, M.; Donath, E.; Mo¨hwald, H. J. Colloid Interface Sci. 1999, 216, 297. (14) Ibarz, G.; Dahne, L.; Donath, E.; Mo¨hwald, H. AdV. Mater. 2001, 13, 1324. (15) (a) Antipov, A. A.; Sukhorukov, G. B.; Leporatti, S.; Radtchenko, I. L.; Donath, E.; Mo¨hwald, H. Colloids Surf. A 2002, 198-200, 535. (b) Antipov, A. A.; Sukhorukov, G. B.; Donath, E.; Mo¨hwald, H. J. Phy. Chem. B 2001, 105, 2281. (16) Khopade, A. J.; Caruso, F. Langmuir 2003, 19, 6219. (17) Gao, C. Y.; Moya, S.; Lichtenfeld, H.; Casoli, A.; Fiedler, H.; Donath, E.; Mo¨hwald, H. Macromol. Mater. Eng. 2001, 286, 355. (18) (a) Gittins, D. I.; Caruso, F. AdV. Mater. 2000, 12, 1947. (b) PastorizaSantos, I.; Scho¨ler B.; Caruso, F. AdV. Funct. Mater. 2001, 11, 122. (19) (a) Antipov, A. A.; Shchukin, D.; Fedutik, Y.; Petrov, A. I.; Sukhorukov, G. B.; Mo¨hwald, H. Colloids Surf. A 2003, 224, 175. (b) Gaponik, N.; Radtchenko, I. L.; Gerstenberger, M. R.; Fedutik, Y. A.; Sukhorukov, G. B.; Rogach, A. L. Nano Lett. 2003, 3, 369. (c) Silvano, D.; Krol, S.; Diaspro, A.; Cavalleri, O.; Gliozzi, A. Microscopy Res. Technol. 2002, 59, 536. (20) Shenoy, D. B.; Antipov, A. A.; Sukhorukov, G. B.; Mo¨hwald, H. Biomacromolecules 2003, 4, 265. (21) (a) Ai, H.; Fang, M.; Jones, S. A.; Lvov, Y. M. Biomacromolecules 2002, 3, 560. (b) Donath, E.; Moya, S.; Neu, B.; Sukhorukov, G. B.; Georgieva, R.; Voigt, A.; Baumler, H.; Kiesewetter, H.; Mo¨hwald, H. Chem. Eur. J. 2002, 8, 5481. (22) Khopade, A. J.; Caruso, F. Chem. Mater. 2004, 16, 2107. (23) Chambers H. F. Goodman & Gilman’s The Pharmacological Basis of Therapeutics; Hardman, J. G., Limbird, L. E., Gilman, A. G., Eds.; McGraw Hill: New York, 2001. (24) (a) Sweren R. J. J. Am. Acad. Dermatol. 1991, 24, 664. (b) Schachner, L.; Pestana, A.; Kittles, C. J. Am. Acad. Dermatol. 1990, 22, 489. (25) (a) Chen, C. H.; Chen, V. N.; Chen, S. C.; Cintron, C.; Covington, H. I.; Gregory, J. D. Curr. Eye Res. 1988, 7, 303. (b) Moreau, J. M.; Green, L. C.; Engel, L. S.; Hill, J. M.; O’Callaghan, R. J. Curr. Eye Res. 1998, 17, 808. (26) Seelig, E. W.; Yamilov, A.; Cao, H.; Chang, R. P. H. Mater. Chem. Phys. 2002, 9172, 1. (27) (a) Sauerbrey, G. Z. Phys. 1959, 155, 206. (b) Caruso, F.; Niikura, K.; Furlong, D. N.; Okahata, Y. Langmuir 1997, 13, 3422. (28) (a) Okubo, T.; Suda, M. J. Colloid Interface Sci. 1999, 213, 565. (b) Okubo, T. Suda, M. Colloid Polym. Sci. 1999, 277, 813. (c) Netz, R. R.; Joanny, J. F. Macromolecules 1999, 32, 9013. (29) Petrov, A. I. Gavryushkin, A. V. Sukhorukov, G. B. J. Phys. Chem. B 2003, 107, 868. (30) Khopade, A. J.; Arulsudar, N.; Khopade, S. A.; Mo¨hwald, H. 2004 (Unpublished results, manuscript under preparation.) (31) Theoretical calculation: Mw TbS ) 467.52, Mw DxS ) 100 000, thus the maximum possible theoretical TbS/DxS (w/w) ratio ) 3.5:1, assuming 1:1 TbS:sulfonate binding and 2.3 sulfonate per dextran monomer. TbS/PSS (w/w) ratio ) 2.5:1. (32) Tedeschi, C.; Caruso, F.; Mo¨hwald, H.; Kirstein, S. J. Am. Chem. Soc. 2000, 122, 5841.

BM049554A