116 (2006) 159–169www.elsevier.com/locate/jconrel

Journal of Controlled Release

Layer-by-layer coating of degradable microgels for pulsed drug delivery

B.G. De Geest a, C. Déjugnat b,c, E. Verhoeven a, G.B. Sukhorukov c,d, A.M. Jonas e, J. Plain e,f,J. Demeester a, S.C. De Smedt a,⁎

a Department of Pharmaceutics, Faculty of Pharmaceutical Sciences, Ghent University, Harelbekestraat 72, 9000 Ghent, Belgiumb Institut de Chimie Séparative de Marcoule (ICSM) CEA/CNRS 2926, 30207 Bagnols-sur-Cèze Cedex, France

c Max Planck Institute of Colloids and Interfaces, Am Muehlenberg 1, D-14476 Potsdam, Germanyd IRC/Materials, Queen Mary University of London, Mile End Road, E1 4NS, London, UK

e Unité de Physique et de Chimie des hauts Polymères, Université catholique de Louvain, Croix du Sud, 1-B-1348 Louvain-la-Neuve, Belgiumf Laboratoire de Nanotechnologie et d'Instrumentation Optique, Université de Technologie de Troyes 12, rue Marie Curie, 10010 Troyes cedex, France

Received 29 March 2006; accepted 8 June 2006Available online 21 June 2006

Abstract

Recently, we reported on “self-rupturing” microcapsules which consist of a biodegradable dextran-based microgel surrounded by apolyelectrolyte membrane. Degradation of the microgel increases the swelling pressure in the microcapsules which, when sufficiently high,ruptures the surrounding polyelectrolyte membrane. The membrane surrounding the microgels is deposited using the layer-by-layer (LbL)technique, which is based on the alternate adsorption of oppositely charged polyelectrolytes onto a charged substrate. In this paper, we characterizewith confocal microscopy, electrophoretic mobility, scanning electron microscopy and atomic force microscopy in detail the deposition and theproperties of the LbL coatings on the dextran microgels. We show that by fine-tuning the properties of both the microgel core and the LbLmembrane the swelling pressure which is evoked by the degradation of the microgel is indeed able to rupture the surrounding LbL membrane.Further, we show that the application of an LbL coating on the surface of the microgels dramatically lowers the burst release from themicrocapsules and results in massive release at the time the microcapsules rupture.© 2006 Elsevier B.V. All rights reserved.

Keywords: Pulsed release; Layer-by-layer; Capsules; Microgels; Polyelectrolytes

1. Introduction

Due to the ever increasing amount of biopharmaceuticals,there is a growing need for advanced drug delivery systems [1].Instead of sustained drug release, pulsed drug delivery could beattractive to deliver certain therapeutics [2]. With the termpulsed drug release, we mean that after administration initiallyno drug release occurs. Only after a well defined period thetherapeutics are released. Pulsed release could be advantageousfor, e.g., drugs that develop biological tolerance when they areconstantly present at their target site or for drugs that requiredosing at night [3–6]. Also, microparticles that could suddenlyrelease antigens at well defined times after injection could showpotential for ‘single-shot vaccination’ [7]. For this purpose, the

⁎ Corresponding author. Tel.: +32 9 264 80 76; fax: +32 9 264 81 89.E-mail address: stefaan.desmedt@ugent.be (S.C. De Smedt).

0168-3659/$ - see front matter © 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.jconrel.2006.06.016

injectable device should consist of different types of micro-particles, each type of particle suddenly releasing its encapsu-lated antigen at a well defined time after injection. Deviceswhich show pulsed release upon applying an external triggersuch as pH [8,9], electric field [10,11], IR-light [12–14], etc.have been described. Also implants for pulsed drug delivery[15,16] have been reported; however, injectable microparticlescould offer substantial advantages towards patient complianceas they are less invasive.

Recently, our group introduced a new type of microparticleswhich we termed “self-rupturing microcapsules” [17]. Thesemicrocapsules are able to rupture without the use of any externaltrigger. This is advantageous compared to devices which needsan external trigger to induce drug release. Our ultimate goal is todevelop a formulation consisting of different populations ofself-rupturing microcapsules, each population exploding at adifferent time after injection. This would offer the possibility to

160 B.G. De Geest et al. / Journal of Controlled Release 116 (2006) 159–169

generate multiple drug pulses after a single injection. The self-rupturing microcapsules consist of a biodegradable microgelcore surrounded by a suitable membrane (Fig. 1). It was shownthat upon degradation of the microgels the swelling pressure ofthe degrading microgels was able to suddenly rupture themembrane [18]. In this system, the membrane surrounding themicrogels plays a major role as it has to be (i) permeable towater, (ii) impermeable to the degradation products of themicrogels (upon degradation the gels turn into a polymer so-lution) and (iii) rupture when the swelling pressure reaches acritical value.

We applied a membrane at the surface of the microgels by“layer-by-layer electrostatic self-assembly (LbL-ESA)” which isa very promising technology for the coating of microparticles[19–21]. First developed by Decher et al. on planar substrates[22,23], this technique has been extended to the coating ofcolloidal particles by Sukhorukov et al. [24]. Briefly explained, asillustrated in Fig. 1, this approach is based on the alternate ad-sorption of oppositely charged polyelectrolytes (PE's) on acharged surface, driven by the electrostatic interaction at each stepof adsorption. It has been shown that a wide variety of colloidalparticles such as organic latex particles [24,25], inorganic parti-cles [26,27], dye and drug crystals [28,29], protein aggregates[30,31] and biological cells [32] can be coated by polyelectrolytemultilayer deposition. Also LbL coating of respectively alginateand poly(N-isopropylacrylamide) (pNIPAAm) microgels hasbeen reported [33]. A major advantage of the LbL technique isthe possibility to tune the layer thickness on the nanometer scaleand thus control themechanical properties and the permeability ofthe polyelectrolyte shell.

As the properties of the LbL membrane of the self-rupturingmicrocapsules has a major impact on the release features of thisnew type of microparticles, this paper especially aimed to

Fig. 1. Schematic representation of a microgel coated by the LbL technique. Before dthree-dimensional network by chemical cross-links (•). (I), (II) and (III) show tpolyelectrolytes. The microgels described in this paper degrade by hydrolysis of thepolymer (dextran) chains arise. (IV) At the end of the degradation process, the core ocore exceeds a critical value, the membrane suddenly ruptures (V).

characterize in detail the LbL films deposited on the surface ofthe microgels. We also aimed to show that choosing an optimalLbL coating significantly reduces burst release and makesrelease to occur at the time the microcapsules rupture.

2. Materials and methods

2.1. Materials

Dextran (Mw 19 kDa), fluorescein isothiocyanate-dextrans(FITC-dextran, Mw respectively 20 and 150 kDa), tetramethylrhodamine B isothiocyanate-dextrans (TRITC-dextran, Mw158 kDa), N,N,N′,N′-tetramethylenediamine (TEMED),methacrylic acid (MAA) and dimethyl aminoethyl methacrylate(DMAEMA), rhodamine B isothiocyanate (RBITC), sodiumpoly(styrene sulfonate) (PSS, 70 kDa) and poly(allylaminehydrochloride) (PAH, 70 kDa) were purchased from Sigma-Aldrich-Fluka. Potassium peroxodisulphate (KPS) and poly(ethyleneglycol) (PEG, 20 kDa) were purchased from Merck.RBITC labelled PAH was synthesized as reported in literature[34]. The buffers at pH 7 were 0.1 M phosphate buffers and thebuffers at pH 9 were 0.1 M carbonate buffers.

2.2. Synthesis of dex-HEMA

Dextran-hydroxyethyl methacrylate (dex-HEMA) was pre-pared and characterized according to a method described else-where [35]. Dextran with a number average molecular weight of19 kDa was used. The degree of substitution (DS, the number ofHEMA groups per 100 glucopyranose residues of dextran) wasdetermined by proton nuclear resonance spectroscopy (1H-NMR)in D2O with a Gemini 300 spectrometer (Varian) [36]. The DS ofthe dex-HEMA used in this study was 2.5.

egradation, the polymer chains (dextran chains in this study) are connected into ahe coating of the microgels by sequential adsorption of oppositely chargedcross-links. As degradation proceeds, the cross-link density decreases and freef the particles consists of a dextran solution. When the swelling pressure of the

161B.G. De Geest et al. / Journal of Controlled Release 116 (2006) 159–169

Preparation of dex-HEMA microgels. Dex-HEMA micro-gels, with an initial water content of 70% (w/w), were preparedaccording to Franssen and Hennink [37]. In detail, 71 mg dex-HEMA was dissolved in 1.577 ml water and subsequentlyemulsified by vortexing with a 3.35 ml 24% (v/v) aqueous PEGsolution. Radical polymerisation of the aqueous dex-HEMAemulsion droplets was initiated by adding TEMED (100 μl pHneutralized with 4 N HCl) and KPS (9 mg). The reaction wascarried out at room temperature for 1 h. Afterwards, the obtainedmicrogels were washed three times with pure water to removePEG, KPS and TEMED. Finally, the microgels were suspendedin 5 ml pure water and stored at −20 °C. The size distributionwas measured by laser diffraction (Mavern Mastersizer). Toprepare negatively and positively charged dex-HEMA micro-gels [38], respectively methacrylic acid (MAA, 25 μl) ordimethyl aminoethyl methacrylate (DMAEMA, 35 μl) wasadded to the dex-HEMA/PEG mixture just before vortexing it.In this paper, “dex-HEMA-MAA microgels” and “dex-HEMA-DMAEMA microgels” refer to negatively charged and posi-tively charged microgels, respectively. FITC/TRITC-dextranswere incorporated in the microgels by adding 100 μl of a FITC/TRITC-dextran solution (50 mg/ml) to the dex-HEMA/PEGmixture just before the vortexing step.

2.3. LbL coating of the dex-HEMA microgels

Dex-HEMA microgels were coated by the consecutiveadsorption of oppositely charged polyelectrolytes using thecentrifugation technique [24]. The microgels (500 μl from theoriginal suspension) were dispersed in 1 ml of polyelectrolytesolution (2 mg/ml in 0.5 M NaCl). The polyelectrolytes wereallowed to adsorb for 15 min, under continuous shaking. Thedispersion was then centrifuged at a speed of 300×g for 3 min.Subsequently, the supernatant was removed and the microgelswere redispersed in Milli-Q water to remove the non-adsorbedpolyelectrolytes. This washing was repeated twice before thesecond polyelectrolyte solution was added. The process wasrepeated until the desired LbL coating was reached.

2.4. Release experiments

TRITC-dextran containing dex-HEMA-DMAEMA micro-gels were prepared and LbL coated as described in the para-graphs above. However, the scale on which the LbL coating wasperformed was 10 ml instead of 1 ml. After preparation, theuncoated and (PSS/PAH)3 coated dex-HEMA-DMAEMAmicrogels were centrifuged, the supernatant was removed andthe microgels/microcapsules were redispersed in centrifugationtubes containing 50 ml 0.1 M carbonate buffer (pH 9). Thecentrifugation tubes were thermostatisized at 37±0.5 °C and themicrogels/microcapsules were kept in suspension by mechan-ical agitation. 1 ml samples were withdrawn at 2.5 h timeintervals (by a VenKel Industries 800 automatic samplingstation) and filtered to remove the microcapsules. The fluo-rescence intensity of the samples was measured with a WallacVictor 2 (Perkin Elmer) plate reader. The measured fluorescencevalues were normalized against the fluorescence values mea-

sured at the end of the release experiments. It was verified thatthe measured fluorescence values belonged to the range inwhich a linear relation exists between the concentration of theTRITC-dextran solutions and their fluorescence.

2.5. Confocal laser scanning microscopy

Confocal micrographs of the LbL coatedmicrogels were takenwith a MRC1024 Bio-Rad confocal laser scanning microscope(CLSM) equipped with a krypton-argon laser. An inverted micro-scope (Eclipse TE300D, Nikon) was used which was equippedwith a water immersion objective lens (Plan Apo 60X, NA 1.2,collar rim correction, Nikon).

2.6. Electrophoretic mobility

The electrophoretic mobility of the LbL coated microgels wasmeasured using aMalvern Zetasizer 2000 (Malvern Instruments).The ζ-potential was calculated from the electrophoretic mobility(μ) using the Smoluchowski relation: ζ=μη/ε where η and ε arethe viscosity and permittivity of the solvent, respectively. Sincethe largest microgels did not allow ζ-potential measurements(they sedimented too quickly in the cuvette of the instrument andmay disturb the measurement), wemeasured the ζ-potential of thesmallest microgels (approximately 1 μm). Therefore, the dex-HEMA microgel dispersions were centrifuged (1 min at lowspeed (100 g)) and the ζ-potential measurements were done thedex-HEMA microgels which remained in the supernatant.Therefore, we took 50 μl sample of the supernatant and dilutedit with 2 ml water.

2.7. Scanning electron microscopy (SEM)

Scanning electron microscopy (SEM) measurements on(LbL coated) microgels were carried out using a Zeiss DSM 40instrument operating at an accelerating voltage of 3 kV.

2.8. Atomic force microscopy (AFM)

Experiments were performed on air-dried (LbL coated)microgels deposited onto microscope glass slides. Images wereobtained in tapping mode under ambient conditions with anAutoProbe CP system (Park Scientific Instruments) using a100 μm scanner. Si3N4 cantilevers (spring constant about0.1 Nm−1) with integrated pyramidal tips were used. The AFM-tip was positioned on top of rather large microgels (diameter of15 μm) in order to avoid the effect of the curvature on theimage.

3. Results and discussion

3.1. Preparation of neutral and charged dex-HEMA microgels

To synthesize biodegradable dextran hydrogels [39], meth-acrylate moieties were linked to the dextran backbone byhydrolysable carbonate esters (Fig. 3A). Radical polymerizationof the methacrylate moieties cross-links the dextran chains. The

Fig. 2. Size distribution of the dex-HEMA microgels as measured by laserdiffraction.

162 B.G. De Geest et al. / Journal of Controlled Release 116 (2006) 159–169

fabrication of dex-HEMA microgels using the water-in-wateremulsion technique is based on the immiscibility of the PEGand dex-HEMA solutions. Microgels with an average diameterof 7 μm were obtained (Fig. 2 gives the size distribution of theobtained microgels. It could be expected that charged dex-HEMA microgels would be more suitable for LbL coating thanneutral ones. Van Tomme et al. recently reported that chargeddex-HEMA microgels can be prepared by copolymerization ofdex-HEMAwith MAA (pKa 4.5) (Fig. 3B) or DMAEMA (pKa

8.4) (Fig. 3C) [38]. The incorporation of these charged groupsinto the microgels was verified by measuring the ζ-potential ofthe dex-HEMA microgels (Fig. 4A). Indeed, a negative ζ-po-tential (−30 mV) was measured in case MAAwas used, while apositive ζ-potential (+28 mV) was observed in case DMAEMAwas used, indicating the successful charge loading of the dex-HEMA microgels.

Fig. 3. Chemical structure of the monomer in dex-HEMA (A), which isglucopyranose substituted with HEMA, MAA (B) and DMAEMA (C).

We observed that the charge of the microgels stronglyinfluences their degradation rate. While the positively chargeddex-HEMA-DMAEMA (70%, DS 2.5) microgels were com-pletely degraded within 5 days (at 37 °C and pH 7.4), it took upto 30 days to completely degrade the negatively charged dex-HEMA-MAA (70%, DS 2.5) microgels. It has been reportedthat the degradation rate of alkaline ester hydrolysis can beinfluenced by neighbouring ternary amine groups [40,41]. Afternucleophilic attack of the hydroxyl ion on the ester group, theintermediate is stabilized by resonance stabilisation whichpromotes the alkaline hydrolysis of the ester group thus in-creasing the degradation rate of the dex-HEMA-DMAEMAmicrogels. On the other hand, in case of dex-HEMA-MAAmicrogels, the presence of the carboxyl group of the MAAwilldestabilize the intermediate and thus defavourize the alkalinehydrolysis of the ester group, resulting in a slower degradationof the dex-HEMA-MAA microgels, which was indeedobserved.

3.2. LbL coating of the dex-HEMA microgels

To perform the LbL coating of the microgels we used poly(allylamine hydrochloride) (PAH) as polycation and sodiumpoly(styrene sulfonate) (PSS) as polyanion. This polyelectrolytepair is well studied for the coating of both flat as well as

Fig. 4. Evolution of the ζ-potential, as a function of layer number, during theLbL coating of (A) positively (dex-HEMA-DMAEMA) and negatively (dex-HEMA-MAA) charged microgels and (B) neutral dex-HEMAmicrogels (n=5).

Fig. 5. Confocal images and fluorescence profiles along the yellow lines indicated on the confocal images of (A) dex-HEMA-DMAEMAmicrogels coated with (PSS/PAH)3, (B) dex-HEMA-MAA microgels coated with (PAH/PSS)3 and (C) dex-HEMA microgels coated with (PSS/PAH)3. The PAH was fluorescently labelled withRITC. The scale bar represents 10 μm.

Fig. 6. SEM images of (A) uncoated and (B) (PSS/PAH)3 coated dex-HEMA-DMAEMA microgels.

163B.G. De Geest et al. / Journal of Controlled Release 116 (2006) 159–169

colloidal templates [28,42–44]. Initially, we deposited threePSS/PAH bilayers onto the microgels. Fig. 4A–B show theresults of ζ-potential measurements on uncoated and LbL coat-ed dex-HEMA microgels. Before coating, the ζ-potential ofneutral, dex-HEMA-MAA and dex-HEMA-DMAEMA micro-gels were respectively 0 mV, −30 mV and 28 mV. The ζ-po-tential values proved the successful incorporation of respectivelyMAA and DMAEMA. Fig. 4A and B clearly show that thecharge of the microgels changes upon submerging them in PSSand PAH solutions, indicating that multilayer build-up takesplace. The ζ-potential profile observed upon exposure of themicrogels to PSS/PAH solutions agrees well with literature dataon PSS/PAH coating of other types of particles [24].

Fig. 4B shows that LbL coating of the neutral dex-HEMAmicrogels is also possible. The obtained ζ-potential profile ishowever quite irregular. For charged microgels (Fig. 4A), elec-trostatic interactions between the microgels and the polyelec-trolytes are the main driving force for polyelectrolyte adsorption.However, for neutral microgels other interactions such ashydrophobic interactions [45–47] and physical entanglements[48,49] between the polyelectrolytes and the microgels mostlikely play a role. These interactions seem to be strong enoughbecause otherwise the adsorbed layers would be removed uponadsorption of the next polyelectrolyte layer.

3.3. Microscopy on LbL coated dex-HEMA microgels

Fig. 5 shows confocal microscopy images of (A) dex-HEMA-DMAEMA microgels coated with (PSS/PAH)3, dex-HEMA-MAA microgels coated with (PAH/PSS)3 and dex-HEMAmicrogels coated with (PSS/PAH)3. Rhodamine labelled PAH(PAH-RITC), which is positively charged, was used to visualizethe LbL coating. In case of dex-HEMA-DMAEMA microgels

(Fig. 5A) a distinct ring PAH-RITC is observed while in case ofdex-HEMA-MAA microgels (Fig. 5B) both a ring as well asinwards diffusion of PAH-RITC is observed. In case of neutraldex-HEMA microgels (Fig. 5C), a homogeneous filling withPAH-RITC is observed. These observations indicate that duringthe LbL coating of the chargedmicrogels only the polyelectrolytes

164 B.G. De Geest et al. / Journal of Controlled Release 116 (2006) 159–169

oppositely charged to the microgels will diffuse inside themicrogels, forming a complex with the MAA, respectivelyDMAEMA groups, while the polyelectrolytes with equal chargeas the microgels will only adsorb on the microgel surface duringthe multilayer build-up.

To get a rough estimation on the thickness of the LbLcoating, scanning electron microscopy (SEM) and atomic forcemicroscopy (AFM) were performed on both the coated anduncoated microgels. The SEM images in Fig. 6 reveal that thesurface of the uncoated particles is rather smooth compared tothe coated ones, which show a more granular structure. Espe-cially when PAH was the outermost polyelectrolyte layer aremarkable “brain-like” structure appeared. A similar morphol-ogy was observed using AFM by McAloney et al. [50,51] whenstudying the deposition of PSS/PAH multilayers on planarsubstrates at high salt concentrations. The difference betweenuncoated and PSS/PAH coated dex-HEMA-DMAEMA micro-gels was confirmed by AFM measurements. Uncoated micro-gels (Fig. 7A) show irregularities ranging from 2 to 5 nm, whilePSS/PAH coated microgels (Fig. 7B) show irregularitiesranging from 40 to 100 nm. The irregularities in the morphologyof the LbL coating are highly likely due to the differentialdrying between the gel core and the LbL coating. From these

Fig. 7. AFM images of uncoated (left) en (PSS/PAH)3 coated dex-HEMA-DMAEMAthe surface. (A2) and (B2) show the roughness of the surface along the line marked

data, we estimated the thickness of the LBL membrane to beseveral tens of nanometers.

3.4. Permeability of the LbL coating

As outlined in Fig. 1, the rupturing of the microcapsules istriggered by the swelling pressure of the degradingmicrogels. Toexert a sufficient pressure, it is important that the degradationproducts of the microgels, i.e. 19 kDa dextran chains, do notdiffuse trough the coating during the degradation of themicrogels. In a previous paper, we reported that the permeabilityof the (PSS/PAH)3 membrane seems to be pH-dependent as self-rupturing capsules could only be obtained upon incubation of themicrocapsules at pH 9 (thus suggesting that at pH 9 the coating isimpermeable for 19 kDa dextran chains), while upon incubationat pH 7 the microcapsules' membrane remained intact (thussuggesting that at pH 7 the coating is permeable for 19 kDadextran chains) [17]. To further investigate the permeability ofthe (PSS/PAH)3 coating, we used 20 kDa FITC-dextrans as itsmolecular weight corresponds well to the molecular weight ofthe dextran chains the microgels are composed of.

The (PSS/PAH)3 coated dex-HEMA-DMAEMA microgelswere incubated in a 1 mg/ml FITC-dextran solution. The

microgels (right) recorded in tapping mode. (A1) and (B1) show the topology ofon (A1) and (B1). (A3) and (B3) are the 3D images of the surface.

Fig. 8. Permeability of the LbL coating surrounding the dex-HEMA-DMAEMAmicrogels to 20 kDa FITC-dextrans. The LbL coated microgels are immersedinto a buffered solution containing 1 mg/ml FITC-dextrans and are visualized byconfocal microscopy.

165B.G. De Geest et al. / Journal of Controlled Release 116 (2006) 159–169

experiment was performed at pH 7 and pH 9. From the confocalimages in Fig. 8, it is clear that at pH 7 the capsules' wall ispermeable to the 20 kDa FITC-dextrans while at pH 9 it isimpermeable. When the number of polyelectrolyte bilayers isincreased from 3 to 6, we observed that also at pH 7 the coating

Fig. 9. Confocal images after degradation of the microgel core at pH 7 of (A) emptFITC-dextran (green colour) loaded dex-HEMA-DMAEMA microgels coated withscale bar represents 10 μm. (C) and (D) show the SEM images corresponding to, re

becomes impermeable. As a control, when non-coated dex-HEMA-DMAEMAmicrogels were incubated in a FITC-dextransolution they appeared to be permeable to the FITC-dextrans(data not shown). As PAH is a weak polyelectrolyte (pKa 8.5[52]), it is not surprising that the properties of PSS/PAHpolyelectrolyte multilayers are pH-dependent. The pH-depen-dent stability [53] and permeability [42] of PSS/PAH-basedpolyelectrolyte capsules has indeed been shown. Antipov et al.reported that PSS/PAH-based capsules become impermeable tohigh molecular weight species upon increasing the pH from 7 toabove 8 [42]. Apparently, an analogue phenomenon is observedin our case.

3.5. Behaviour of the microcapsules during degradation of themicrogel core

To evaluate the effect of the degradation of the microgel coreon the integrity of the polyelectrolyte membrane, the micro-capsules were incubated at 37 °C and pH 7 (for 10 days) and atpH 9 (overnight). It was observed with confocal microscopythat upon degradation of the microgel core at pH 9 only rem-nants of broken (PSS/PAH)3 microcapsules could be detected,while the (PSS/PAH)6 microcapsules appeared to maintain theirintegrity upon degradation of the microgel core at pH 9.Apparently, the increase in number of bilayers from 3 to 6 notonly decreases the permeability but also increases the mech-anical strength of the microcapsules, preventing them fromrupturing upon degradation of the microgels core. The (PSS/PAH)3 and (PSS/PAH)6 microcapsules which were degraded at

y dex-HEMA-DMAEMA microgels coated with (PSS/PAH)3 and (B) 150 kDa(PSS/PAH)3. The PAH was fluorescently labelled with RITC (red colour). Thespectively, (A) and (B).

Fig. 10. (A1–A5) CLSM images of (PSS/PAH)3 coated dex-HEMA-DMAEMA microgels during the degradation at pH 11. The dex-HEMA-DMAEMA microgelswere loaded with 150 kDa FITC-dextran (green colour) and the PAHwas labelled with RITC (red colour). The scale bar represents 10 μm. (B) Increase in microcapsulediameter during the degradation of the microgel core (n=10).

Fig. 11. Cummulative release curves of 158 kDa TRITC-dextrans from uncoateddex-HEMA-DMAEMA (open symbols) and (PSS/PAH)3 coated dex-HEMA-DMAEMA microgel (closed symbols). The data points are interconnected witha cubic B-spline. The experiments were run in duplicate.

166 B.G. De Geest et al. / Journal of Controlled Release 116 (2006) 159–169

pH 7 also maintained their integrity and none of them rupturedupon degradation of the microgel core. The higher permeabilityof the LbL membrane at pH 7 most likely causes the outwardsdiffusion of the degradation products of the microgel core whichprevent the build-up of a swelling pressure, thus preventing therupturing of the microcapsules. Fig. 8A–B show confocalmicroscopy images of (A) hollow and (B) FITC-dextran (i.e.150 kDa FITC-dextrans were used as model drug) filled (PSS/PAH)3 microcapsules obtained after the degradation of themicrogel core at pH 7. Note that in case of FITC-dextran filledmicrocapsules (Fig. 8B) the microgels were loaded with FITC-dextrans during their synthesis and thus before the LbL coating.The fact that the FITC-dextrans remain inside the capsulesproves that no rupturing of the LbL coating occurred during thedegradation, as this would lead to the release of the FITC-dextrans as shown in our previous work [17]. Fig. 9C and D areSEM images corresponding to the microcapsules shown inrespectively Fig. 9A and B. The collapsed state of the micro-capsules clearly proves that indeed the microgel core isdegraded. In case FITC-dextrans were encapsulated, the micro-capsules showed a more granular structure due to the pre-cipitation of the FITC-dextrans upon drying of the sample. Asimilar morphology of filled capsules after drying has beenobserved by Sukhorukov et al. [54].

To visualize the rupturing of the (PSS/PAH)3 coated dex-HEMA-DMAEMA microgels, it was investigated up to whichpH the (PSS/PAH)3 coating was stable. Dejugnat andSukhorukov have reported on the pH responsive properties of

hollow PSS/PAH-based capsules and from there results it can beconcluded that the PSS/PAH coating remains intact until a pH of11 whereas at a pH above 12 the multilayer membrane becomesirreversibly destroyed [53]. Therefore, the PSS/PAH)3 coateddex-HEMA-DMAEMA microgels were incubated in a solutionbuffered at pH 11 and the degradation process was monitored byconfocal microscopy. Fig. 10A1–A5 shows snapshots of the

167B.G. De Geest et al. / Journal of Controlled Release 116 (2006) 159–169

microcapsules during the degradation. Initially, they start toswell and at a certain moment they rupture leading to the releaseof the encapsulated material, which was 150 kDa FITC-dextran.Fig. 10B shows the evolution of the microcapsules' diameterduring the degradation of the microgel core. As can be ob-served, the microcapsules gradually swell until an almost two-fold increase in diameter followed by the rupturing of themembrane.

3.6. Release from uncoated and (PSS/PAH)3 coated dex-HEMA-DMAEMA microgels

In a next step, the effect of the (PSS/PAH)3 coating on therelease of 158 kDa TRITC-dextrans from dex-HEMA-DMAEMA microgels, degrading at pH 9, was investigated.We expected to observe a pulsed release for the followingreasons. First, at pH 9, the (PSS/PAH)3 coating surroundingdex-HEMA-DMAEMA microgels should not be permeable to158 kDa TRITC-dextrans as it is even impermeable to 20 kDadextrans (Fig. 8). In other words, the 158 kDa TRITC-dextransshould not leak through the membrane during degradation of themicrogels. Second, at pH 9, the (PSS/PAH)3 microcapsules areself-rupturing, i.e. the membrane breaks due to the increase inswelling pressure of the microgels.

Fig. 11 shows that the release of 158 kDa TRITC-dextransfrom uncoated and (PSS/PAH)3 coated dex-HEMA-DMAEMAmicrogels differs significantly . Uncoated microgels show asubstantial burst release followed by a continuous release untilthe microgels are completely degraded after 20 h. A burst releasedoes not occur in case of (PSS/PAH)3 coated dex-HEMA-DMAEMA microgels, the 158 kDa TRITC-dextran releaseremains even low during the first 10 h of degradation. After thisinitial phase, the majority of the encapsulated 158 kDa TRITC-dextran molecules are released in a couple of hours allowing toconclude that (PSS/PAH)3 coating of the dex-HEMA-DMAEMA microgels causes the release of the 158 kDaTRITC-dextrans to be much more pulsatile.

One could, however, wonder why the observed release pulseis less steep than one would expect when all the microcapsuleswould simultaneously rupture. This may be partially explainedby Laplace's law:

p ¼ 2gr

With p the pressure, γ the membrane tension and r the radius.Microcapsules will rupture when the swelling pressure of themicrogels evokes a tension in the membrane which exceeds thetensile strength of the membrane. Laplace's law states that, byincreasing the microcapsules' radius, the pressure required toevoke the same membrane tension, decreases. This implies thatlarger microcapsules require less swelling pressure to ruptureand will thus rupture earlier than smaller ones. As the microgelsreported in this paper are polydisperse in size, one could indeedexpect that not all the microcapsules will self-rupture at thesame time but the time of rupturing will show a distribution.Recently, we reported on the fabrication of highly monodispersedex-HEMA microgels using a microfluidic emulsification

device [55]. Future work will focus on the use of thesemonodisperse microgels to produce a highly uniform popula-tion of microcapsules. Another aspect which should beaddressed in future work is how to modify the properties ofthe LbL coating surrounding the microgels in order to renderthis coating sufficiently impermeable under physiologicalconditions without increasing the mechanical strength of thiscoating to a point where it can no longer be ruptured by theswelling pressure of the degrading microgel.

4. Conclusions

Dextran-based microgels, with different surface charges, wereused as template for the LbL assembly of the polyelectrolytes PSSand PAH. ζ-potential measurements and CLSM proved thatpolyelectrolytes can be sequentially adsorbed onto the surface ofneutral, positively as well as negatively charged dextran micro-gels leading to microcapsules. It was observed that the positivelycharged dex-HEMA-DMAEMA microgels were the most pro-mising as template for LbL assembly. The permeability of PSS/PAH-based LbL coatings surrounding dex-HEMA-DMAEMAmicrogels was investigated. It was found that three bilayers ofPSS/PAH rendered the microcapsules impermeable to 20 kDaFITC-dextrans at pH 9 while they were still permeable at pH 7.Increasing the number of polyelectrolyte bilayers to 6 renderedthe microcapsules impermeable also at pH 7.

We showed that upon degradation at pH 9 the dextranmicrogels were able to rupture their surrounding (PSS/PAH)3coating, resulting in ‘self-rupturing microcapsules’, while whendegraded at pH 7 the (PSS/PAH)3 coating did not rupture,leading to hollow (PSS/PAH)3 capsules. This was explained bythe pH-dependent permeability of the (PSS/PAH)3 coating tothe degradation products of the microgels. When degraded atpH 9, the release of high molecular weight TRITC-dextran,encapsulated in (PSS/PAH)3 coated microcapsules, was signif-icantly more pulsatile compared to the TRITC-dextran releasefrom uncoated dextran microgels.

The concept presented in this paper may be promising to-wards biomedical applications, especially in the field of pulseddrug delivery, as the time of rupturing is determined by thedegradation rate of the microgels, which can be tailored fromdays to several weeks by varying the cross-link density of themicrogels [39]. Our further research will focus on LbL coating ofdegradable microgels making use of biocompatible polyelec-trolytes with the aim to obtain microcapsules which are underphysiological conditions impermeable to both the encapsulateddrugs as well as the degradation products of the dextran gels, twomajor requirements to obtain pulsed delivery from this type ofmicromaterials.

Acknowledgements

Ghent University is thanked for a scholarship BOF 01112403.Mies van Steenbergen and Sophie Van Tomme are gratefullythanked for helpful discussions. Bernard Nysten is thanked forgranting the access to the AFM instrument. Michelle Prevot isthanked for the SEMmeasurements. Christophe Déjugnat thanks

168 B.G. De Geest et al. / Journal of Controlled Release 116 (2006) 159–169

the European project “Nanocapsules” for funding. Pr. Dr. H.Möhwald is greatly thanked for reading the manuscript.

References

[1] R. Langer, D.A. Tirrell, Designing materials for biology and medicine,Nature 428 (6982) (2004) 487–492.

[2] B.G. Stubbe, S.C. De Smedt, J. Demeester, “Programmed polymericdevices” for pulsed drug delivery, Pharm. Res. 21 (10) (2004) 1732–1740.

[3] G.W. Creasy, M.E. Jaffe, Pulsatile delivery systems, Ann. N.Y. Acad. Sci.618 (1991) 548–557.

[4] P.E. Lazzerini, P.L. Capecchi, S. Bisogno, M. Galeazzi, R. Marcolongo, F.L.Pasini, Reduction in plasma homocysteine level in patients with rheumatoidarthritis given pulsed glucocorticoid treatment, Ann. Rheum. Dis. 62 (7)(2003) 694–695.

[5] B. Lemmer, Chronopharmacology—time, a key in drug-treatment, Ann.Biol. Clin. 52 (1) (1994) 1–7.

[6] T.F. Nielsen, P. Ravn, Y.Z. Bagger, L. Warming, C. Christiansen, Pulsedestrogen therapy in prevention of postmenopausal osteoporosis. A 2-yearrandomized, double blind, placebo-controlled study, Osteoporosis Int. 15(2) (2004) 168–174.

[7] J. Hanes, J.L. Cleland, R. Langer, New advances in microsphere-basedsingle-dose vaccines, Adv. Drug Deliv. Rev. 28 (1) (1997) 97–119.

[8] C. Dejugnat, F. Halozan, G.B. Sukhorukov, Defined picogram doseinclusion and release of macromolecules using polyelectrolyte micro-capsules, Macromol. Rapid Commun. 26 (2005) 961–967.

[9] D.M. Lynn, M.M. Amiji, R. Langer, pH-responsive polymer microspheres:rapid release of encapsulated material within the range of intracellular pH,Angew. Chem., Int. Ed. 40 (9) (2001) 1707–1710.

[10] P.F. Kiser, G. Wilson, D. Needham, A synthetic mimic of the secretorygranule for drug delivery, Nature 394 (6692) (1998) 459–462.

[11] P.F. Kiser, G. Wilson, D. Needham, Lipid-coated microgels for thetriggered release of doxorubicin, J. Control. Release 68 (1) (2000) 9–22.

[12] A.G. Skirtach, C. Dejugnat, D. Braun, et al., The role of metal nano-particles in remote release of encapsulated materials, Nano Lett. 5 (7)(2005) 1371–1377.

[13] B. Radt, T.A. Smith, F. Caruso, Optically addressable nanostructuredcapsules, Adv. Mater. 16 (23–24) (2004) 2184–2189.

[14] A.S. Angelatos, B. Radt, F. Caruso, Light-responsive polyelectrolyte/goldnanoparticle microcapsules, J. Phys. Chem., B 109 (7) (2005) 3071–3076.

[15] A.C.R. Grayson, I.S. Choi, B.M. Tyler, et al., Multi-pulse drug deliveryfrom a resorbable polymeric microchip device, Nat. Mater. 2 (11) (2003)767–772.

[16] J.T. Santini, M.J. Cima, R. Langer, A controlled-release microchip, Nature397 (6717) (1999) 335–338.

[17] B.G. De Geest, C. Dejugnat, G.B. Sukhorukov, K. Braeckmans, S.C. DeSmedt, J. Demeester, Self-rupturing microcapsules, Adv. Mater. 17 (19)(2005) 2357–2361.

[18] http://www.wiley-vch.de/contents/jc_2089/2005/ c1951_s.html.[19] A.M. X.J. Arys, A. Laschewsky, R. Legras, Supramolecular Polyelectro-

lyte Assemblies, Marcel Dekker, New York, 2000.[20] P. Bertrand, A. Jonas, A. Laschewsky, R. Legras, Ultrathin polymer coatings

by complexation of polyelectrolytes at interfaces: suitablematerials, structureand properties, Macromol. Rapid Commun. 21 (7) (2000) 319–348.

[21] G.Decher, J. Schlenoff,Multilayer Thin Films,Wiley-VCH,Weinheim, 2002.[22] G. Decher, Fuzzy nanoassemblies: toward layered polymeric multi-

composites, Science 277 (5330) (1997) 1232–1237.[23] G. Decher, J.D. Hong, Buildup of ultrathin multilayer films by a self-

assembly process: 1. Consecutive adsorption of anionic and cationicbipolar amphiphiles on charged surfaces, Makromol. Chem., Macromol.Symp. 46 (1991) 321–327.

[24] G.B. Sukhorukov, E. Donath, H. Lichtenfeld, et al., Layer-by-layer selfassembly of polyelectrolytes on colloidal particles, Colloid Surf., APhysicochem. Eng. Asp. 137 (1–3) (1998) 253–266.

[25] E. Donath, G.B. Sukhorukov, F. Caruso, S.A. Davis, H. Mohwald, Novelhollow polymer shells by colloid-templated assembly of polyelectrolytes,Angew. Chem. Int. Ed. 37 (16) (1998) 2202–2205.

[26] F. Caruso, R.A. Caruso, H. Mohwald, Nanoengineering of inorganic andhybrid hollow spheres by colloidal templating, Science 282 (5391) (1998)1111–1114.

[27] B.G. De Geest, R.E. Vandenbroucke, A.M. Guenther, et al., Adv. Mater. 18(2006) 1005–1009.

[28] A.A. Antipov, G.B. Sukhorukov, E. Donath, H. Mohwald, Sustainedrelease properties of polyelectrolyte multilayer capsules, J. Phys. Chem., B105 (12) (2001) 2281–2284.

[29] F. Caruso, W.J. Yang, D. Trau, R. Renneberg, Microencapsulation ofuncharged low molecular weight organic materials by polyelectrolytemultilayer self-assembly, Langmuir 16 (23) (2000) 8932–8936.

[30] N.G. Balabushevitch, G.B. Sukhorukov, N.A. Moroz, et al., Encapsulationof proteins by layer-by-layer adsorption of polyelectrolytes onto proteinaggregates: factors regulating the protein release, Biotechnol. Bioeng. 76(3) (2001) 207–213.

[31] F. Caruso, D. Trau, H. Mohwald, R. Renneberg, Enzyme encapsulation inlayer-by-layer engineered polymer multilayer capsules, Langmuir 16 (4)(2000) 1485–1488.

[32] B. Neu, A. Voigt, R. Mitlohner, et al., Biological cells as templates forhollow microcapsules, J. Microencapsul 18 (3) (2001) 385–395.

[33] N. Greinert, W. Richtering, Influence of polyelectrolyte multilayeradsorption on the temperature sensitivity of poly(N-isopropylacrylamide)(PNiPAM) microgels, Colloid Polym. Sci. 282 (10) (2004) 1146–1149.

[34] H. Jeffery, S.S. Davis, D.T. Ohagan, The preparation and characterizationof poly(lactide-co-glycolide) microparticles: 1. Oil-in-water emulsionsolvent evaporation, Int. J. Pharm. 77 (2–3) (1991) 169–175.

[35] W.N.E. vanDijkWolthuis, S.K.Y. Tsang, J.J. KettenesvandenBosch, W.E.Hennink, A new class of polymerizable dextrans with hydrolyzablegroups: hydroxyethyl methacrylated dextran with and without oligolactatespacer, Polymer 38 (25) (1997) 6235–6242.

[36] W.N.E. Vandijkwolthuis, O. Franssen, H. Talsma, M.J. Vansteenbergen,J.J.K. Vandenbosch, W.E. Hennink, Synthesis, characterization, andpolymerization of glycidyl methacrylate derivatized dextran, Macromol-ecules 28 (18) (1995) 6317–6322.

[37] O. Franssen, W.E. Hennink, A novel preparation method for polymeric micro-particles without the use of organic solvents, Int. J. Pharm. 168 (1) (1998) 1–7.

[38] S.R. Van Tomme, M.J. van Steenbergen, S.C. De Smedt, C.F. vanNostrum, W.E. Hennink, Self-gelling hydrogels based on oppositelycharged dextran microspheres, Biomaterials 26 (14) (2005) 2129–2135.

[39] W.N.E. vanDijkWolthuis, J.A.M. Hoogeboom, M.J. vanSteenbergen, S.K.Y.Tsang, W.E. Hennink, Degradation and release behavior of dextran-basedhydrogels, Macromolecules 30 (16) (1997) 4639–4645.

[40] P. van de Wetering, N.J. Zuidam, M.J. van Steenbergen, O. van derHouwen, W.J.M. Underberg, W.E. Hennink, A mechanistic study of thehydrolytic stability of poly(2-(dimethylamino)ethyl methacrylate), Macro-molecules 31 (23) (1998) 8063–8068.

[41] S.R. Van Tomme, C.F. Van Nostrum, S.C. De Smedt, W.E. Hennink,Degradation behavior of dextran hydrogels composed of positively andnegatively charged microspheres, Biomaterials 27 (22) (2006) 4141–4148.

[42] A.A. Antipov, G.B. Sukhorukov, S. Leporatti, I.L. Radtchenko, E. Donath,H. Mohwald, Polyelectrolyte multilayer capsule permeability control,Colloid Surf., A Physicochem. Eng. Asp. 198 (2002) 535–541.

[43] A.A. Antipov, G.B. Sukhorukov, H. Mohwald, Influence of the ionicstrength on the polyelectrolyte multilayers' permeability, Langmuir 19 (6)(2003) 2444–2448.

[44] C. Gao, E. Donath, S. Moya, V. Dudnik, H. Mohwald, Elasticity of hollowpolyelectrolyte capsules prepared by the layer-by-layer technique, Eur.Phys. J., E 5 (1) (2001) 21–27.

[45] M. Castelnovo, J.F. Joanny, Formation of polyelectrolyte multilayers,Langmuir 16 (19) (2000) 7524–7532.

[46] N.A. Kotov, Layer-by-layer self-assembly: the contribution of hydropho-bic interactions, Nanostruct. Mater. 12 (5–8) (1999) 789–796.

[47] T. Lojou, P. Bianco, Buildup of polyelectrolyte–protein multilayerassemblies on gold electrodes. Role of the hydrophobic effect, Langmuir20 (3) (2004) 748–755.

[48] T. Serizawa, H. Sakaguchi, M. Matsusaki, M. Akashi, Polyelectrolytemultilayers prepared on hydrogel surfaces, J. Polym. Sci. Part A, Polym.Chem. 43 (5) (2005) 1062–1067.

169B.G. De Geest et al. / Journal of Controlled Release 116 (2006) 159–169

[49] H. Sakaguchi, T. Serizawa, M. Akashi, Layer-by-layer assembly onhydrogel surfaces and control of human whole blood coagulation, Chem.Lett. 32 (2) (2003) 174–175.

[50] R.A. McAloney, V. Dudnik, M.C. Goh, Kinetics of salt-induced annealingof a polyelectrolyte multilayer film morphology, Langmuir 19 (9) (2003)3947–3952.

[51] R.A. McAloney, M. Sinyor, V. Dudnik, M.C. Goh, Atomic forcemicroscopy studies of salt effects on polyelectrolyte multilayer filmmorphology, Langmuir 17 (21) (2001) 6655–6663.

[52] K. Itano, J.Y. Choi, M.F. Rubner, Mechanism of the pH-induceddiscontinuous swelling/deswelling transitions of poly(allylamine hydro-

chloride)-containing polyelectrolyte multilayer films, Macromolecules 38(8) (2005) 3450–3460.

[53] C. Dejugnat, G.B. Sukhorukov, PH-responsive properties of hollowpolyelectrolyte microcapsules templated on various cores, Langmuir 20(17) (2004) 7265–7269.

[54] G. Sukhorukov, L. Dahne, J. Hartmann, E. Donath, H. Mohwald,Controlled precipitation of dyes into hollow polyelectrolyte capsulesbased on colloids and biocolloids, Adv. Mater. 12 (2) (2000) 112–115.

[55] B.G. De Geest, J.P. Urbanski, T. Thorsen, J. Demeester, S.C. De Smedt,Synthesis of monodisperse biodegradable microgels in microfluidicdevices, Langmuir 21 (2005) 10275–10279.