pharmaceutical nanotechnology preparation of coated ... · controlled release vaccine delivery...

International Journal of Pharmaceutics 299 (2005) 155–166

Pharmaceutical Nanotechnology

Preparation of coated nanoparticles for a new mucosalvaccine delivery system

Olga Borgesa,b,∗, Gerrit Borcharda,1, J. Coos Verhoefa,Adriano de Sousab, Hans E. Jungingera

a Leiden/Amsterdam Center for Drug Research, Division of Pharmaceutical Technology,P.O. Box 9502, 2300 RA Leiden, The Netherlands

b Center for Pharmaceutical Studies, Laboratory of Pharmaceutical Technology, Faculty of Pharmacy,University of Coimbra, Rua do Norte, Coimbra 3000-295, Portugal

Received 2 December 2004; received in revised form 13 April 2005; accepted 23 April 2005

Abstract

It has been found that the adsorption of antigens onto chitosan particles is an easy and unique mild loading process suitable tobe used with vaccines. In order to increase the stability of this particles and to prevent an immediate desorption in gastrointestinalfluids, a coating process with sodium alginate was developed. One of the challenges of this developing process was to keep theparticles in the nanosized range in order to be taken up by M-cells of the Peyer’s patches. The observed inversion of the particles’zeta potential values after coating suggested the presence of an alginate coating layer. These results were confirmed by FTIR

e particlesintestinal

articles.

and DSC techniques. Additionally, in vitro release studies showed that the presence of the alginate layer around thwas able to prevent a burst release of loaded ovalbumin and to improve the stability of the nanoparticles in simulatedfluid at 37◦C. The optimisation of the coating process resulted in 35% (w/w) for the loading capacity of the coated pSEM investigations confirmed a suitable size of the coated nanoparticles for the uptake by M-cells.© 2005 Elsevier B.V. All rights reserved.

Keywords:Chitosan; Sodium alginate; Ovalbumin adsorption; Coated nanoparticles; Mucosal vaccination

1. Introduction

ingo itsTheacesainstuce

∗ Corresponding author. Tel.: +351 239859927;fax: +351 239827126.

E-mail address:[email protected] (O. Borges).1 Present address: Enzon Pharmaceuticals, 20 Kingsbridge Road,

Piscataway, NJ 08854, USA.

In recent years, mucosal vaccination is beconsidered as a subject of great interest due tadvantages above the i.m. or s.c. application.presence of specific antibodies in mucosal surfhas long been recognized as the first barrier agpathogens entrance. The most effective way to ind

0378-5173/$ – see front matter © 2005 Elsevier B.V. All rights reserved.doi:10.1016/j.ijpharm.2005.04.037

156 O. Borges et al. / International Journal of Pharmaceutics 299 (2005) 155–166

mucosal immunity (i.e., secretory IgA) is to administera vaccine directly to the mucosal surface. Additionally,the existence of a common mucosal immune systemallows successful targeting of vaccines to inductivecompartments within mucosa-associated lymphoidtissues, inducing local humoral responses in lymphoidtissues at distant mucosal loci (Alpar et al., 1998). Bothintranasal and oral routes have been used in severalstudies to achieve this goal. Particularly, the oraladministration permits targeting of a suitable vaccine-loaded delivery system to the ports of entry (so-calledM-cells) of the largest inductive lymphoid tissue inthe body, the intestine. The oral route is well acceptedand easily allows the vaccination of large populations.However, the acidic environment of the stomach andthe presence of enzymes make the oral delivery of vac-cines a challenge where is difficult to achieve high andreproducible effects. In order to solve these difficulties,a considerable number of polymeric microparticulatesystems are under investigation to deliver vaccinesto the intestine while protecting them from adverseconditions that could affect their bioactivity (Singh andO’Hagan, 1998). Another important aspect is that thesedelivery systems could act as imunostimulants or adju-vants, increasing the immunogenicity of poor immuneresponse antigens (Jabbal-Gill et al., 1999; Singh andO’Hagan, 1999).

Nevertheless, from a pharmaceutical perspec-tive, it became evident that further advances inthe formulation of delivery platforms needs to bei thea takeo neo thes oftt gut(p tioni ssfulc

asalv ourgL al.,2 by am Byt like

elevated temperatures, high shear rates or the pres-ence of organic solvents were avoided. This method hasalso been described by other groups that reported goodadsorption capacities for different substances (Mi et al.,1999; Hejazi and Amiji, 2002). In the case that the chi-tosan particles are not very porous, the antigen will bepreferentially adsorbed to the particle surface. This cancause stability problems because processes like desorp-tion or the attack of the antigens by enzymes or acidicsubstances from the body fluids may occur. Theseobstacles may be overcome by coating those particleswith an acid resistant polymer, like sodium alginate.

The two chosen polymers chitosan and sodiumalginate, for this novel delivery platform are naturallyoccurring polysaccharides. They are polyelectrolytepolymers of opposite charges, biocompatible andbiodegradable, and show a good safety profile.Furthermore they have been used as pharmaceuticalexcipients. Chitosan is the deacetylated form of chitincomprising copolymers of glucosamine andN-acetylglucosamine linked by�-(1-4) linkages. The primaryamino groups lead to special properties that make chi-tosan very interesting for pharmaceutical applications.Sodium alginate is also a hydrophilic polymer andcomprisesd-mannuronic (M) andl-guluronic acid(G) residues joined linearly by 1,4-glycosidic linkages(Johnson et al., 1997). The wide pharmaceutical appli-cability of alginates is, to a large extent, associated withtheir gel-forming capacity. Di- or polyvalent cations(calcium being the most widely studied example) cani nica al.,1 ringn lezF ;F 99;K al.,2bp ana veralp theya uner l andi ke

alsob ome

ntroduced in order to increase both the stability ofntigens in the gastro-intestinal tract and the upf antigen-containing particles by the M-cells. Of the parameters that should be addressed isize of the particles. It is well known that the sizehe particles should be below 10�m in order to beaken up by M-cells of the Peyer’s patches in theEldridge et al., 1991; Jani et al., 1992). Moreover, thereservation of antigen stability during encapsula

s also essential for the development of a succeontrolled release vaccine delivery platform.

Chitosan microparticles as an oral and intranaccine delivery system were already used inroup showing promising capabilities (Van derubben et al., 2001a,b, 2003; Bivas-Benita et003). In these studies, the vaccine was loadedild and simple but effective adsorption method.

his method, deleterious preparation conditions,

nduce the gelation by cross-linking of the gulurocid units (Rajaonarivony et al., 1993; Johnson et997). Sodium alginate has been used for prepaanoparticles (Rajaonarivony et al., 1993; Gonzaerreiro et al., 2002), microspheres (Wu et al., 1997undueanu et al., 1999; Takka and Acarturk, 19ulkarni et al., 2001; Chan et al., 2002; Coppi et002), microcapsules (Esquisabel et al., 2000) andeads (Kulkarni et al., 2001), for oral delivery. Inarticularly, the use of alginate microparticles asntigen delivery system has been described in seublications and there are some indications thatre able to induce a mucosal and systemic immesponse in a variety of animal species by both orantranasal administration (Cho et al., 1998; Bowersoct al., 1999; Rebelatto et al., 2001).

Over the last years, sodium alginate haseen used as a coating material for cells with s

O. Borges et al. / International Journal of Pharmaceutics 299 (2005) 155–166 157

advantages. It seems that the coating acts as a barrierto microbial contamination, and thus improvedsurvival prospects of the coated cells (Kampf et al.,2000). In another study, the coating is performed toprotect donor mammalian cells against antibodiesand cytotoxic cells of the host immune system,allowing the transplantation of cells in the absence ofimmunosuppression (de Vos et al., 2002).

This manuscript describes the development of atrue nanocoating procedure, whereas other publica-tions describing the entrapment of cells, liposomes(Machluf et al., 2000) or microspheres (Ramdas etal., 1999; Hejazi and Amiji, 2002) in an alginate gelmatrix. This is, as far as we know, the first time thatthe construction of a nanosized alginate-coated chi-tosan delivery system is described with the particularlyaspect that the antigen is adsorbed to the chitosan par-ticles surface.

2. Materials and methods

2.1. Materials

Chitosan was purchased from Primex BioChemicalsAS (ChitoClearTM, Avaldsnes, Norway). According tothe provider’s specifications, the degree of deacety-lation is 95% (titration method) and the viscosity is8 cP (1% solutions in 1% acetic acid). Low viscositysodium alginate was kindly donated by ISP Technolo-g n( omS ersr ionsw

2

cip-i sly( eda 2%(f ddi-t to2 adea m)a -W

model (output control “1”); Sonics & Materials,Inc., Danbury, USA). Sonication was maintained foradditional 15 min and the agitation for 60 min at roomtemperature (RT). The suspension was centrifuged for30 min at 3500 rpm (2800×g) and the supernatantwas discarded. The particles were re-suspended twicein Millipore water, centrifuged again for 30 min andthe supernatants were discarded. The particles werefrozen in liquid nitrogen and freeze-dried overnightusing a Christ freeze-dryer (Osterode am Harz, Ger-many). The dry powder was kept frozen until furtheruse.

2.3. Loading of the particles with ovalbumin

The first step of the loading procedure was thesuspension of the freeze-dried particles in a phosphatebuffer (pH 7.4) placed in an ultrasound bath for 30 minin order to disaggregate the particles. The loadingwas done by incubating a solution of ovalbuminwith chitosan particles under mild agitation at roomtemperature. The various concentrations used arepresented inTable 1.

The loading efficacy and the loading capacityof the uncoated particles were calculated by anindirect way, quantifying the protein that remained insolution. After incubation, an aliquot of the particlesuspension was centrifuged at 14,000 rpm for 30 minand the protein in supernatant was quantified byBCA-protein assay (PIERCE, Rockford, USA) using am del5 ancer ragea teina adedn ndi-t toc urvep mins

inge nge

ies Inc. (MANUCOL LB®, Surrey, UK). OvalbumiOVA; grade V; minimum 98%) was purchased frigma Chemicals (St. Louis, USA) and all the oth

eagents used were of analytical grade. All solutere prepared in Millipore water.

.2. Preparation of chitosan particles

Chitosan particles were prepared by the pretation/coacervation method described previouBerthold et al., 1996). Shortly, chitosan was dissolvt a concentration of 0.25% (w/v) in a solution withv/v) of acetic acid and 1% (w/v) of Tween® 80. Theormation of the particles was achieved after the aion of 3.5 ml of sodium sulfate solution (10%, w/v)00 ml of the chitosan solution. The addition was mt a rate of 1 ml/min under mild agitation (<50 rpnd continuous sonication (vibracell sonicator; 600

icroplate reader with a 590 nm filter (Bio-Rad mo50, Veenendaal, The Netherlands). The absorbeading value was corrected subtracting the avebsorbance reading obtained in the BCA-prossay from that one of the supernatants of unloanoparticles prepared exactly in the same co

ions. The corrected OD value was then usedalculate the concentration using the standard crepared at same time from individual ovoalbutandards.

The drug loading capacity (LC) and loadfficacy (LE) were calculated from the followiquations:

LC (%, w/w) = (total amount of ovalbumin− non-bound ovalbumin)/weight of the particles× 100LE (%) = (total amount of ovalbumin− non-boundovalbumin)/total amount of ovalbumin× 100


Tabl

e1

Com

posi

tion

ofth

edi

ffere

ntsy

stem

sus

edfo

rth

ede

velo

pmen

toft

heal

gina

teco

atin

gm

etho

dof

the

chito

san

nano

part

icle

s

AB

CD

E

Load

ing

1%pa

rtic

les;

0.5%

oval

bum

in(P

B;p

H7.

4)0.

5%pa

rtic

les;

0.25

%ov

albu

min

(PB

;pH

7.4)

0.37

5%pa

rtic

les;

0.19

%ov

albu

min

(PB

;pH

7.4)

0.4%

part

icle

s;0.

25%

oval

bum

in(P

B;p

H7.

4)0.

5%pa

rtic

les;

0.25

%ov

albu

min

(PB

;pH

7.4)

Coa

ting

No

0.12

5%pa

rtic

les;

0.5%

algi

nate

(wat

er)

0.18

8%pa

rtic

les;

0.5%

algi

nate

(pho

spha

tebu

ffer,

pH7.

4)0.

2%pa

rtic

les;

0.5%

algi

nate

(pho

spha

tebu

ffer,

pH7.

4)0.

25%

part

icle

s;1%

algi

nate

(pho

spha

tebu

ffer,

pH7.

4) 2.4. Coating of the nanoparticles with alginate

Various amounts of the ovalbumin-loaded particlesuspension were added under agitation to varioussolutions of sodium alginate (Table 1). The suspensionof the particles was maintained under agitation witha magnetic stirrer for 20 min at RT. The suspensionwas then centrifuged for 10 min at 1600 rpm and thesupernatant was discarded. To chemically cross-linkthe alginate at the particle’s surface, the particles werere-suspended in 0.524 mM CaCl2 solution and keptunder agitation for another 10 min. For the characteri-zation of the nanoparticles (Section2.5) the optimisedformulation batch was used as given inTable 1(system D).

2.4.1. Evaluation the desorption during thecoating procedure

During the incubation of the particles with sodiumalginate, aliquots of the particles suspension were col-lected, centrifuged at 14,000 rpm for 30 min and theprotein in the supernatant was assayed with a BCA-protein assay as described in Section2.3.

Statistical methods used in this section includedescriptive statistics (arithmetic mean and standarddeviation) and Student’st-test.

2.5. Characterization of the nanoparticles

2.5.1. Morphologypar-

t opy( acedo h atr nesso thing ater( M-6 ope(

2ated

b izer3 m,T ereb clesi ments

The morphology and surface appearance of theicles were examined by scanning electron microscSEM). One drop of the particles suspension was pln a gold-coated plate and maintained at least 12oom temperature in a desiccator for complete dryf the sample. The dry samples were coated with aold layer using Emitech K650X large sample coEmitech, Kent, UK) and observed with a Jeol JS700F field emission scanning electron microscJEOL BV, Schiphol-Rijk, The Netherlands).

.5.2. Size and zeta potential measurementsThe particle size and zeta potential were evalu

y a dynamic light scattering technique with a Zetas000HSA (Malvern Instruments, Bergen op Zoohe Netherlands). Zeta potential determinations wased on electrophoretic mobility of the nanoparti

n diluted aqueous suspensions. These measure


were performed at least in triplicate with independentparticle batches.

2.5.3. FT/IR studiesThe coated particles were washed with Millipore

water, centrifuged and the sediment was freeze-driedovernight (Labconco, Kansas City, USA). The coatedand uncoated particles were kept in desiccator atroom temperature until analysis. The IR spectra of thesamples were recorded using a Fourier-transformedinfrared spectrophotometer instrument FT/IR–420Jasco (Jasco Inc., Tokyo, Japan) with attenuated totalreflection (ATR).

2.5.4. Differential scanning calorimetry (DSC)DSC scans were recorded using a differential scan-

ning calorimeter (DSC-50, Shimadzu Co., Kyoto,Japan). Two to four milligrams of the dry particleswere accurately weighed into aluminium pans with-out seals and heated from 25 to 350◦C at a heating rateof 10◦C/min under a nitrogen flow of 20 ml/min.

2.6. In vitro release studies

The ovalbumin release from the coated and uncoatedparticles was performed in simulated intestinal fluid(SIF) as described in USP XXIV. The nanoparticlessuspensions were added (1:4) to individual tubescontaining the release medium previously equilibratedat 37◦C and placed in a shaker bath adjusted to5 ome ingfim edb theo teina lankn samec lanksf d inS

n-t andu of0 ra-s atha teinw sions

of unloaded particles were analysed under the sameconditions and were used as a blank for the correctionof the OD value of the samples analysed with BCA-protein assay as described in Section2.3. Additionally,the unbound ovalbumin in the particle’s suspension wasalso determined in order to calculate the amount of theovalbumin encapsulated in the beginning of the assay.

All experiments were performed at least in triplicate.

2.7. SDS-polyacrylamide gel electrophoretic(PAGE) analysis of released ovalbumin

Samples from the loading and coating particle sus-pensions were centrifuged at 14,000 rpm, the super-natants were collected and the concentration wasadjusted in order to have the same theoretical concen-tration (i.e., assumption of 0% of loading efficiency)of the protein. Alginate-coated chitosan nanoparticlesloaded with ovalbumin were diluted (1:5) in bufferphosphate pH 7.5 at 37◦C and incubated overnight at50 rpm. An aliquot was collected and centrifuged. Theovoalbumin samples were then suspended in the load-ing buffer and heated for 5 min at 100◦C just beforethe run.

The SDS-PAGE was performed with gels composedof 12% acrylamide, cast and run in Tris–glycine buffer.Gels were stained with 0.1% Coomassie brilliant bluein 10% of acetic acid in a solution of methanol:water(1:1).

3

3

inst sanp andfi iono withs thors( ne ionb , them thep thep vedb tein

0 rpm. At appropriate time intervals, samples frach tube were filtered with a low protein-bindlter (MILLEX ®GV—0.22�m; durapore PVDFembrane; Millipore, Malsheim, France) followy centrifugation for 20 min at 14,000 rpm andvalbumin in supernatant assayed with a BCA-prossay. Simultaneously, coated and uncoated banoparticles suspensions were submitted to theonditions and the filtered samples were used as bor correction of the BCA-protein assay as describeection2.3.For the determination of the total protein co

ent, 0.5 ml of the suspensions of the coatedncoated particles were incubated with 1.5 ml.085N hydrochloric acid solution (pH 1.2) in an ultound bath for 30 min, followed by 3 h in a water bt 37◦C. The samples were filtered and the proas assayed by the BCA-protein assay. Suspen

. Results and discussion

.1. Preparation of the vaccine delivery system

The preparation of the delivery system contahree main steps, the manufacturing of the chitoarticles, their ovalbumin loading by adsorptionnally the coating with sodium alginate. The formatf the chitosan particles by a precipitation processodium sulphate has been described by several auBerthold et al., 1996; Roy et al., 1999; Van der Lubbet al., 2001b). They all describe an existing correlatetween the necessary amount of sulphate ionsolecular weight and deacetylation degree ofolymer. The loading of the cationic particles withrotein is a very mild process, which can be achiey suspending the particles in a solution of the pro


Fig. 1. Loading capacity of the particles during the different stages of the coated particles preparation. Results for the different systems (A–E).

in an appropriate buffer. The adsorption of the proteinonto chitosan particles is mainly caused by ionic inter-action of the chitosan amino groups with the carboxylgroups of protein substrate in a buffer phosphate solu-tion of pH 7.4 with high buffer capacity. According toprevious work done in our laboratory (Van der Lubbenet al., 2001b), the loading capacity of ovalbumin ata chitosan particles concentration of 1% (w/v) is notsubstantially influenced in the range of 0.5–2% (w/v)ovalbumin. Therefore, an ovalbumin concentration of0.5% (w/v) was used as loading solution resulting ina sufficient high loading capacity. For instance, for thesystem A, as shown inFigs. 1 and 2, the loading capac-ity and the loading efficacy were about 40 and 80%,respectively. It was also found from the measurements

of the unbound protein at different time points duringthe loading process that the adsorption equilibrium israpidly reached. This means that the protein is boundto the surface of the particles and that the adsorptionprocess of the ovalbumin to the particles surfaceoccurs immediately after the addition of the proteinsolution to the particle suspension. Better loadingcapacity results (p< 0.00001) of the uncoated particleswere found when the ratio particles/ovalbumin in theloading suspension was decreased from 2 (Table 1systems A, B and E) to 1.6 (Table 1system D) whilethe loading efficacy (uncoated particles) was notdifferent (p= 0.15) between the systems.

The third step of the preparation was the coatingof the chitosan nanoparticles with alginate solution.

F es of t nt system(

ig. 2. Loading efficacy of the particles during the different stagA–E).

he alginate-coated particles preparation. Results for the differes


To our knowledge this coating process has not beendescribed in literature earlier. Consequently, the firstpart of this work is focused on the selection of theappropriate conditions and experimental methodologyfor the coating process with sodium alginate. For thatpurpose several systems with different ratios sodiumalginate/chitosan particles have been investigated withthe main objective to obtain a stable particle suspen-sion. In the majority of the first trials an immediateflocculation was observed, particularly in systems withhigher particles concentration. The formation of theseagglomerates was easily observed because of the for-mation of a precipitate and a clear supernatant. Systemswith a ratio alginate/particles > 2 have shown to be sta-ble. Thus, in the following optimization steps of thecoating methodology always ratios higher than 2 wereused.

Another important parameter studied was thedesorption of ovalbumin from the particles during thecoating process. The addition of the sodium alginatesolution to the suspension of the loaded particlesresulted in a new adsorption equilibrium characterizedby the different concentration of protein and by thepresence of alginate polymer that can compete withthe interaction of the charges at the particle’s surface.A significant decrease of the loading capacity ofthe coated particles was observed in all the systems(p< 0.05 for systems B–E). Furthermore, we haveobserved that modifications in the pH of the coatingmedium can also modify the adsorption equilibrium oft ro-c s tob chi-t d6 3.4a rvedm nots as itw weenl ess.T tednw dw isonw ,t thee cacy(

3.2. Characterization of the nanoparticles

3.2.1. Morphology, size and zeta potentialmeasurements

The precipitation of chitosan with sodium sulphateusing ultrasounds for homogenisation led to the for-mation of particles in the nanoscale size (Table 2).One of the major currently described drawbacks of thistechnique is the high polydispersity of the obtainednanoparticles (Tang et al., 2003). We have observednanoparticles with sizes ranging between 100 and1000 nm. The mean hydrodynamic diameter of theobtained particles after the precipitation process was684 nm and the size increased after the freeze-dryingprocess (Table 2). This is a direct consequence of parti-cle aggregation during the drying process. To overcomeparticle aggregation the use of trehalose as a lyopro-tectant was tried at concentrations of 3.3, 5 and 7%(w/v). A complete redispersibility of the freeze-driednanoparticles in all trehalose concentrations and themaintenance of the particle size could be observed bylight microscopy and dynamic light scattering tech-nique. However, we optimised the duration of thefreeze-drying process in order to avoid the use of cryo-protectants as they would interfere with the coatingprocess.

SEM observations of the coated particles (Fig. 3b)indicated that the size range of the particles remainedunchanged. This result indicates the feasibility ofcoating of ovalbumin-loaded chitosan nanoparticlesws arti-c isa is ac con-c

tlyr en-e , thec rti-c temw clesi tepso f thed nt ofe ed apt d in

he protein. In our first experiments, the coating pess was carried out at pH 5.5. This pH value seeme the most favourable for the interaction between

osan and alginate as the pKa of the chitosan is aroun.5 and the pKa of the sodium alginate is betweennd 4.4. However, at this pH value, we have obseore than 60% of ovalbumin desorption (data

hown). For that reason a pH 7.4 was adoptedas observed to have the better compromised bet

oading capacity and efficiency of the coating proche highest loading capacity for the alginate-coaanoparticles was achieved with system D (p= 0.086hen compared with system B andp= 0.032 compareith system E). On the other hand, the comparith system B (p= 0.001) and E (p= 0.004) showed

hat this option (system D) was achieved onxpenses of a slight decrease of the loading effiFig. 2).

ith a thin layer of alginate. SEM images (Fig. 3a)howed some small particles (<100 nm) and a ple’s agglomerate. However, it is not clear if thisconsequence of the freeze-dry process or if it

onsequence of the sample preparation (particle’sentration of the sample) for the SEM.

The stability of many colloidal systems is direcelated to the magnitude of their zeta potential. In gral, if the value of the particle zeta potential is largeolloidal system will be stable. Conversely, if the pale zeta potential is relatively small, the colloid sysill agglomerate. The surface charge of the parti

s of substantial importance in all the production sf these coated particles because the efficiency oifferent steps is directly related to the establishmelectrostatic interactions. Chitosan particles showositive value of about 37 mV (Table 3), which explains

he stability of the particle suspensions in water an


Table 2Size of the chitosan nanoparticles in different stages of the preparation process

Cumulant Z average (nm), mean± S.D. Polydispersity, mean± S.D.

After the precipitation process 643.2± 171.7 (16 batches) 0.379± 0.168 (16 batches)After lyophilization process 955.6± 161.0 (3 batches) 0.387± 0.129 (3 batches)

several buffer systems. After adsorption of ovalbumin,the particles are still positively charged with valueshigher than 30 mV. During the coating procedure aninversion of the surface charge of the particles to nega-tive values was observed. This zeta potential inversionexplains the difficulties that were found in preventingthe formation of particles agglomerates during the coat-ing process. After complete adsorption of the alginateto the particles, a surface charge of the particles of about−35 mV was found (Table 3). This zeta potential inver-sion is a strong indication of the presence of an alginatecoating on the surface of the particles.

F s. (A)A tingp

Table 3Zeta (ζ) potential of unloaded, ovalbumin-loaded chitosan nanopar-ticles and alginate-coated chitosan nanoparticles

Zeta potential (mV), meanaverage± S.D.

Empty chitosannanoparticles

+37.0± 3.6 (seven batches)

Ovalbumin-loadedchitosan particles

+41.3± 6.4 (five batches)

Alginate-coatedchitosan particles

−34.9± 8.3 (five batches)

3.2.2. Differential scanning calorimetryAs shown inFig. 4, the DSC scans of the chitosan

polymer exhibited an endothermic peak at about 66◦Cthat has been attributed to the evaporation of absorbedwater. The exothermic baseline deviation beginningaround 250◦C indicates the onset of chitosan degra-dation (Khalid et al., 2002). The analysis of the DSCcurves for chitosan particles showed two additionalendothermic peaks at about 237 and 275◦C. The peakat 237◦C is probably related to the breakdown of weakunspecific electrostatic interactions. The second peakis probably related to the cleavage of the electrostaticinteractions between the polymer and the sulphate ions.

Fig. 4. Differential scanning calorimetry curves of chitosan,unloaded chitosan particles and sodium sulphate.

ig. 3. Scanning electron micrographs of chitosan nanoparticlefter freeze-dry and ressuspending in water. (B) After the coarocedure with sodium alginate.


Fig. 5. Differential scanning calorimetry curves of sodium alginate,uncoated chitosan particles and alginate-coated particles. The DSCinvestigation was conducted using the unloaded particles.

After the coating process with sodium alginate, theparticles exhibited a completely different behaviouras shown inFig. 5, which means that an alterationoccurred in the composition of the particles. In the DSCscans, the two endothermic peaks disappeared and noendothermic peak was found in the temperature rangestudied. In fact, a gradual appearance of an exothermicbehaviour was detected starting around 200◦C thatcoincides with the exothermic behaviour of the sodiumalginate as referred to in several publications (e.g.,

Gonzalez-Rodriguez et al., 2002) as the decompositionof the polymer. These observations strongly supportthe presence of alginate molecules linked to thesurface of the chitosan particles: when the temperatureof the sample reaches values around 200◦C, the twophenomena, the exothermic contribution from thealginate and the endothermic contribution from thechitosan particles determine the shape of the curve.

3.2.3. FTIR characterizationThe FTIR spectra of chitosan and the chitosan par-

ticles, sodium alginate, and the coated particles areshown inFigs. 6 and 7, respectively.

In the chitosan spectra the strong and broad peaksin the 3400–3200 cm−1 ranges correspond to combinedpeaks of O H stretching and intermolecular hydrogenbonding. The NH stretching from primary amines areoverlapped in the same region.

The C O stretching (amide) peak near 1633 cm−1

and N H bending (amide and amine) peak near1542 cm−1 was observed as well. The intense peak at1414 cm−1 belongs to the NH stretching of the amideand ether bonds give the peaks in the fingerprint regionof the spectra, where the symmetric stretch of CO Cis found around wave numbers of 1065–1027 cm−1. Inthe chitosan particles the peak of 1558 cm−1 is shiftedto 1535 cm−1 and the relative intensity of this peak is

tosan a
Fig. 6. FTIR spectra of chi nd unloaded chitosan particles.


Fig. 7. FTIR spectra of sodium alginate, uncoated chitosan particles and alginate-coated chitosan particles. The FTIR investigation was conductedusing the unloaded chitosan particles.

reduced. In addition the peak related with CN stretch(1414 cm−1) has almost disappeared in the chitosanparticles. Similar observations are reported for chi-tosan particles prepared with tripolyphosphate (Xu andDu, 2003). These observations are in agreement withthe fact that the sulphate ions interact with the pri-mary amino groups of the chitosan, resulting in theformation of cross-linked chitosan particles (reticula-tion process).

Sodium alginate as a carboxylate salt showed astrong asymmetric stretch at 1605 cm−1. The frequencyof carbonyl absorption is lowered compared to thevalue found for the parent carboxylic acid due to aresonance phenomenon. The carboxyl and carboxy-late groups are present at wave number of about1000–1400 cm−1. From the FTIR analysis spectra ofthe coated particles we were able to distinguish thepresence of these three peaks that are different fromthose of the chitosan uncoated particles. As conclusionof these findings, the results clearly show the existenceof alginate coating layer around the chitosan particles.

3.3. Release studies

In a preliminary study, we observed that the tem-perature is an important determinant for the integrityof the chitosan particles placed in different pH buffers.The particle integrity was monitored by turbiditytransmission measurements at 500 nm. In fact at roomt sionso d in

simulated intestinal fluid (SIF). In contrast, when thechitosan particles were added to the same solutions(SGF and SIF) at 37◦C, a loss of integrity wasobserved translated with high values of transmissionmeasurements. Thus, the complete release of theovalbumin from the chitosan particles in SIF at 37◦C(Fig. 8) is directly related to the loss of integrity ofthe particles and less with a simple desorption phe-nomenon. In contrast, the coating of the particles withalginate increased the stability of the particles (lowervalues of transmission measurements) in SIF at 37◦Cresulting in a slower release rate of ovalbumin. After7 h more than 60% of ovalbumin was still found in thealginate-coated particles. These release studies alsosuggest that the formation of an alginate layer increases

Fig. 8. Release behaviour of ovalbumin from alginate-coated andu ◦ -s

emperature we were able to obtain stable suspenf the particles in simulated gastric fluid (SGF) an

ncoated chitosan particles in SIF at 37C. Individual points repreent the mean averages from three assays.


Fig. 9. Electrophoretic analysis on SDS-12% PAGE with Coomassiebrilliant blue staining. Shown are solutions of ovalbumin beforeconjugation to the nanoparticles (lane 1), ovoalbumin from the super-natant remaining after the conjugation (lane 2), ovoalbumin from thesupernatant remaining after the coating with alginate (lane 3), ovoal-bumin released from alginate-coated chitosan nanoparticles (lanes4–6).

the stability of the chitosan particles and reduces therelease of the adsorbed protein from these particles.

3.4. SDS-PAGE

The main task of this work was the developmentof an appropriate method for the encapsulation ofantigens to an optimised delivery platform. With thismethod the protein antigen was never exposed topotentially harsh conditions, such as the contact withorganic solvents or mechanical agitation or sonication.The SDS-PAGE (Fig. 9) of the released ovalbuminfrom the particles showed identical bands for theentrapped (lanes 4–6) and native ovalbumin (lane 1)and there were no additional bands to indicate thepresence of molecular weight aggregates or fragmentsgreater or less than 45 K (the molecular weight ofovalbumin). Hence, the data suggest that the structuralintegrity of ovalbumin was not significantly affectedby the entrapment procedure.

4. Conclusion

This work describes a new nanosized mucosal vac-cine delivery system, consisting of chitosan particlesthat are prepared by cross-linking with sodium sul-phate. The chitosan nanoparticles are loaded under

very mild conditions with a model antigen (ovalbu-min), which was negatively charged in the used buffersystems. The validity of an easy and economic loadingprocess was shown. In order to increase the stabil-ity of the loaded chitosan particles at physiologicaltemperature in SGF and SIF a coating process withsodium alginate was developed. This coating processwas optimized in such ways that only small amountsof ovalbumin were desorbed during the coating processand the antigen released from the coated nanoparticleswas strongly reduced in comparison to the uncoatedchitosan nanoparticles. Hence a nanosized deliveryplatform is described with improved features of anti-gen stability in simulated gastrointestinal fluids. In vivostudies are under way to show the efficacy of these sys-tems for mucosal vaccination.

Acknowledgements

This work was supported by a grant from thePortuguese Ministry of Science and Technology(SFRH/BD/5327/2001). We would like to thank Dr. H.K. Koerten (Faculty of Medicine, Leiden University)for his support with SEM, and to Dr. Alcino Leitaoand Dr. Joao Canotilho (Faculty of Pharmacy, Coim-bra University) for the discussions of FTIR spectra andDSC curves, respectively.

ISP Alginates (UK).

R

A asalro-

B char-pred-tory

B Deon ofchi-ry in

B ar-Oralinate

C icro-. Int.

eferences

lpar, H.O., Eyle, J.E., Williamson, E.D., 1998. Oral and nimmunization with microencapulated clinically relevant pteins. STP Pharm. Sci. 8, 31–39.

erthold, A., Cremer, K., Kreuter, J., 1996. Preparation andacterization of chitosan microspheres as drug carrier fornisolone sodium phosphate as model for anti-inflammadrugs. J. Control Release 39, 17–25.

ivas-Benita, M., Laloup, M., Versteyhe, S., Dewit, J.,Braekeleer, J., Jongert, E., Borchard, G., 2003. GeneratiToxoplasma gondii GRA1 protein and DNA vaccine loadedtosan particles: preparation, characterization, and preliminavivo studies. Int. J. Pharm. 266, 17–27.

owersock, T.L., HogenEsch, H., Suckow, M., Guimond, P., Mtin, S., Borie, D., Torregrosa, S., Park, H., Park, K., 1999.vaccination of animals with antigens encapsulated in algmicrospheres. Vaccine 17, 1804–1811.

han, L., Lee, H., Heng, P., 2002. Production of alginate mspheres by internal gelation using an emulsification methodJ. Pharm. 242, 259–262.


Cho, N.H., Seong, S.Y., Chun, K.H., Kim, Y.H., Kwon, I.C., Ahn,B.Y., Jeong, S.Y., 1998. Novel mucosal immunization withpolysaccharide-protein conjugates entrapped in alginate micro-spheres. J. Control Release 53, 215–224.

Coppi, G., Iannuccelli, V., Leo, E., Bernabei, M.T., Cameroni, R.,2002. Protein immobilization in crosslinked alginate micropar-ticles. J. Microencapsul. 19, 37–44.

de Vos, P., Hoogmoed, C.G., Busscher, H.J., 2002. Chemistry andbiocompatibility of alginate-PLL capsules for immunoprotectionof mammalian cells. J. Biomed. Mater. Res. 60, 252–259.

Eldridge, J.H., Stass, J.K., Meulbroek, J.A., McGhee, J.H., Tice, T.R.,Gilley, R.M., 1991. Biodegradable microspheres as a vaccinedelivery system. Mol. Immunol. 28, 287–294.

Esquisabel, A., Hernandez, R.M., Igartua, M., Gascon, A.R., Calvo,B., Pedraz, J.L., 2000. Effect of lecithins on BCG-alginate-PLLmicrocapsule particle size and stability upon storage. J. Microen-capsul. 17, 363–372.

Fundueanu, G., Nastruzzi, C., Carpov, A., Desbrieres, J., Rinaudo,M., 1999. Physico-chemical characterization of Ca-alginatemicroparticles produced with different methods. Biomaterials 20,1427–1435.

Gonzalez Ferreiro, M., Tillman, L., Hardee, G., Bodmeier, R., 2002.Characterization of alginate/poly-l-lysine particles as antisenseoligonucleotide carriers. Int. J. Pharm. 239, 47–59.

Gonzalez-Rodriguez, M.L., Holgado, M.A., Sanchez-Lafuente, C.,Rabasco, A.M., Fini, A., 2002. Alginate/chitosan particulate sys-tems for sodium diclofenac release. Int. J. Pharm. 232, 225–234.

Hejazi, R., Amiji, M., 2002. Stomach-specific anti-H. pylori therapy.I. Preparation and characterization of tetracyline-loaded chitosanmicrospheres. Int. J. Pharm. 235, 87–94.

Jabbal-Gill, I., Lin, W., Jenkins, P., Watts, P., Jimenez, M., Illum, L.,Davis, S.S., Wood, J.M., Major, D., Minor, P.D., Li, X., Lavelle,E.C., Coombes, A.G., 1999. Potential of polymeric lamellar sub-strate particles (PLSP) as adjuvants for vaccines. Vaccine 18,238–250.

J ande rate. 86,

J n ofs. J.

K of

K ze,rmalur. J.

K ki,iumBio-

Machluf, M., Apte, R.N., Regev, O., Cohen, S., 2000. Enhancingthe immunogenicity of liposomal hepatitis B surface antigen(HBsAg) by controlling its delivery from polymeric micro-spheres. J. Pharm. Sci. 89, 1550–1557.

Mi, F.L., Shyu, S.S., Chen, C.T., Schoung, J.Y., 1999. Porous chitosanmicrosphere for controlling the antigen release of Newcastle dis-ease vaccine: preparation of antigen-adsorbed microsphere andin vitro release. Biomaterials 20, 1603–1612.

Rajaonarivony, M., Vauthier, C., Couarraze, G., Puisieux, F., Cou-vreur, P., 1993. Development of a new drug carrier made fromalginate. J. Pharm. Sci. 82, 912–917.

Ramdas, M., Dileep, K.J., Anitha, Y., 1999. Alginate encapsultedbioadhesive chitosan microspheres for intestinal drug delivery. J.Biomat. Appl. 13, 290–296.

Rebelatto, M.C., Guimond, P., Bowersock, T.L., HogenEsch, H.,2001. Induction of systemic and mucosal immune response incattle by intranasal administration of pig serum albumin in algi-nate microparticles. Vet. Immunol. Immunopathol. 83, 93–105.

Roy, K., Mao, H.Q., Huang, S.K., Leong, K.W., 1999. Oral genedelivery with chitosan–DNA nanoparticles generates immuno-logic protection in a murine model of peanut allergy. Nat. Med.5, 387–391.

Singh, M., O’Hagan, D., 1998. The preparation and characterizationof polymeric antigen delivery systems for oral administration.Adv. Drug. Deliv. Rev. 34, 285–304.

Singh, M., O’Hagan, D., 1999. Advances in vaccine adjuvants. Nat.Biotechnol. 17, 1075–1081.

Takka, S., Acarturk, F., 1999. Calcium alginate microparticles fororal administration. I. Effect of sodium alginate type on drugrelease and drug entrapment efficiency. J. Microencapsul. 16,275–290.

Tang, E.S., Huang, M., Lim, L.Y., 2003. Ultrasonication of chitosanand chitosan nanoparticles. Int. J. Pharm. 265, 103–114.

Van der Lubben, I.M., Kersten, G., Fretz, M.M., Beuvery, C., CoosVerhoef, J., Junginger, H.E., 2003. Chitosan microparticles for

cacy

V .C.,iclesfocalDrug

V , G.,cina-take4.

W of theplant.

X pro-arm.

ani, P.U., McCarthy, D.E., Florence, A.T., 1992. Nanospheremicrosphere uptake via Peyer’s patches: observation of thof uptake in the rat after a single oral dose. Int. J. Pharm239–246.

ohnson, F.A., Craig, D.Q., Mercer, A.D., 1997. Characterizatiothe block structure and molecular weight of sodium alginatePharm. Pharmacol. 49, 639–643.

ampf, N., Zohar, C., Nussinovitch, A., 2000. Alginate coatingXenopus laevis embryos. Biotechnol. Prog. 16, 497–505.

halid, M.N., Agnely, F., Yagoubi, N., Grossiord, J.L., CouarraG., 2002. Water state characterization, swelling behavior, theand mechanical properties of chitosan based networks. EPharm. Sci. 15, 425–432.

ulkarni, A.R., Soppimath, K.S., Aminabhavi, T.M., RudzinsW.E., 2001. In-vitro release kinetics of cefadroxil-loaded sodalginate interpenetrating network beads. Eur. J. Pharm.pharm. 51, 127–133.

mucosal vaccination against diphtheria: oral and nasal effistudies in mice. Vaccine 21, 1400–1408.

an der Lubben, I.M., Konings, F.A., Borchard, G., Verhoef, JJunginger, H.E., 2001a. In vivo uptake of chitosan micropartby murine Peyer’s patches: visualization studies using conlaser scanning microscopy and immunohistochemistry. J.Target. 9, 39–47.

an der Lubben, I.M., Verhoef, J.C., van Aelst, A.C., BorchardJunginger, H.E., 2001b. Chitosan microparticles for oral vaction: preparation, characterization and preliminary in vivo upstudies in murine Peyer’s patches. Biomaterials 22, 687–69

u, J.X., Tai, J., Cheung, S.C., Tze, W.J., 1997. Assessmentprotective effect of uncoated alginate microspheres. TransProc. 29, 2146–2147.

u, Y., Du, Y., 2003. Effect of molecular structure of chitosan ontein delivery properties of chitosan nanoparticles. Int. J. Ph250, 215–226.

pharmaceutical nanotechnology preparation of coated ... · controlled release vaccine delivery...

Documents