Post on 05-Jan-2016
description
International Journal of Pharmaceutics 323 (2006) 5263
In situ coatingAn approach for parrinlqvSE-1580,May
006
Abstract
In this pa les inthe particle/p onalmethod used formand either o ) andpolymer (Po oscopconcentratio ted fras measured by the pendant drop method. Further, particle properties such as: size, dissolution time, powder flowability, and apparent particledensity, as measured by gas pycnometry, were affected by the type and concentration of the polymer. In addition, the particle surface morphologycould possibly be correlated to the surface elasticity of the droplet surface during drying. Moreover, an extensive investigation (Fourier transforminfrared spectroscopy, circular dichroism and size exclusion chromatography) of the structural effects of protein encapsulated in a polymeric coatingsuggested th 2006 Else
Keywords: Po
1. Introdu
Particleality or fordients, is iis an indusparticles (96MO
, Swe, WPEOonce
0.01uffe
oreCheEu
in ist. Loglassoluwitz
pray
roxufferhematic illustration of the formulation concept in situ coating.
mpetition during spray-drying implies adsorption ofive components to the air/liquid interface of drying
contrast to steady-state adsorption the time-scalee in spray-drying is restricted. The average life-plet surfaces, in a laboratory dryer was estimatedms (Elversson and Millqvist-Fureby, 2005b). Con-ransport and attachment to the interface is importantpetitiveness during spray-drying. At the surface, a
polymer that can rapidly rearrange and expose non-ns towards the air phase would be expected to be
studies establish the fact that the composition ofsurface is preserved during spray-drying (Faldt andl, 1994; Millqvist-Fureby et al., 1999; Landstrom etdler et al., 2000; Elversson and Millqvist-Fureby,
r example, in mixtures of BSA or sodium caseinate, protein is accumulated at the air/water interfaceying and thus appears on the powder surface (Faldtstahl, 1994). In contrast, neither of the components
tially accumulated in a mixture of glycine and lac-us, the surface composition of the powder reflects
sition of the spray solution (Faldt and Bergenstahl,hence possible to utilize surface competition dur-
drying in order to create desired powder proper-ttability, dissolution (Elofsson and Millqvist-Fureby,urface morphology (Elversson and Millqvist-Fureby,
tend to denature when exposed to an air/water inter-the protein is frequently unfolded. Considering the
ce area in a spray, the potential for surface-inducedn is substantial (Mumenthaler et al., 1994; Maa et
illqvist-Fureby et al., 1999), in particular at a low. Possibly, addition of a polymeric coating to proteins prepared by spray-drying can enhance protein sta-
eventing/reducing proteinsurface interactions. This
than stMillqvsitu cosimult
Conbilizatmethofied reformatconcer
homogimizedparticlformatcoating
2. Ma
2.1. M
Solmer w
tion VLouis,BuchswaukePPO27final cand 0,phate bMillip(Fluka(Merckorosce
Co., SAll
washwil, Swater.
2.2. S
Hydcold bserved for protein formulations with addition of low-eight surfactants such as polyoxy ethylene sorbitansorbate 20 (Maa et al., 1998) and Polysorbate 80ler et al., 1994; Broadhead et al., 1994; Millqvist-l., 1999), or sodium dodecyl sulphate (Adler et al.,ided the coating material, added to the protein for-
s the most efficiently adsorbing component it willtective layer at the droplet surface minimizing theic interaction between protein and air/liquid inter-
drying (Fig. 1). It has previously been observed
and once d24 h beforethe polymespray-drye(Elverssonand the ou5 ml/min, aflow was 0drying airat room temureby, 2005a). This formulation concept we call ing since both coating and particle formation occurusly, during spray-drying.ently, in situ coating is a formulation concept for sta-f protein formulations during spray-drying but thisalso be used for pure coating reasons, e.g. modi-
or oxidative protection. Since spray-drying enablesand coating of individual particles in a single step,n adhesion or wetting capability of coatings, coatingty and agglomeration of primary particles are min-e objective of this study was to influence specificperties and to reduce or prevent surface-induced con-l changes of protein during spray-drying by in situ
ls & methods
ials
s with different concentrations of non-ionic poly-prepared from stock solution of BSA (Cohn frac-% purity, Mw 67 kDa, Sigma Chemical Co., St.), d(+)-trehalose dihydrate (Fluka Chemie GmbH,itzerland) and HPMC (Mw 10 kDa, Aldrich, Mil-I) or Poloxamer 188 (Mw 76809510 Da, PEO80-80, Uniqema, Gouda, The Netherlands) to obtain antration of 0.5% (w/w) BSA, 9.5% (w/w) trehalose, 0.1 and 1% (w/w) polymer. A 10 mM sodium phos-r (pH 7.0) prepared from ultra-purified water (MilliQ,Systems, 18.2 MS resistivity), HNa2PO412H2Omie GmbH, Buchs, Switzerland) and NaH2PO4H2Orolab, Stockholm, Sweden) was used as solvent. Flu-othiocyanate (FITC) labeled BSA (Sigma Chemicaluis, MO) was used for CLSM imaging.sware was cleaned thoroughly by a surfactant-freetion (Deconex 20NS, Borer Chemie AG, Zuch-erland) and rinsed thoroughly with hot and cold
-drying
ypropyl methylcellulose (HPMC) was dispersed in(10 mM phosphate, pH 7.0) under vigorous stirringissolved the stock solution was kept for more thanuse to allow complete hydration and swelling of
r. Particles were prepared in a co-current lab-scaler (construction of the Institute for Surface Chemistry)et al., 2003). The inlet air temperature was 180 C
tlet temperature was kept at 70 C. Liquid feed wastomization airflow was 28 l/min and the drying air-.8 m3/min. The particles were separated from the
by a cyclone. Samples were stored in a desiccatorsperature and
54 J. Elversson, A. Millqvist-Fureby / International Journal of Pharmaceutics 323 (2006) 5263
Table 1Temperature cycle for DSC analysis of spray-dried powders
Segment Rate (C/min) Residence time (min)1 10 23456789
The nitrogen
2.3. Dynam
Surfacemer solutio(First TenVA). In thissize and shblunt-end mSyringes, Hwas calibra(72.3 mN/mtions werepoint was cdrop.
2.4. Scann
The parined with sCompany,ation voltaSEM-stub(SCD 050,of 640 A th
2.5. Confo
Coatedlaser-scannwith a 63was used tosion filter.the laser refirmed thatreflected li
2.6. Electr
The elewas assess
Kratos AnaAl K X-r
andprox
00 Aithisur
ativeientset alccor
sson
as p
apped wmerilledeforand
as 34
isso
n efpart
er (1otat
ific 1s detCycle Temperature cycle ( C)Heating 25140Cooling 14090Isothermal 90Heating 90160Isothermal 160Cooling 160140Heating 140240Isothermal 240Cooling 24025
flow was 40 ml/min.
ic surface tensiometry (pendant drop)
tensions at the air/water interface of protein and poly-ns were measured by the pendant drop technique
Angstroms AccuSoft, Version 1.961B, Portsmouth,technique, the surface tension is calculated from the
ape of a droplet hanging from a tip of a syringe with aetal or Teflon-coated needle (Hamilton MicrolitreTMamilton Bonaduz AG, Switzerland). The tensiometerted with 95% ethanol (23.1 mN/m) and MilliQ-water), at room temperature (23 C). The sample solu-monitored for approximately 14 s. The first dataollected approximately 0.07 s after formation of the
ing electron microscopy (SEM)
ticle shape and the surface morphology were exam-canning electron microscopy (SEM) (XL30TMP, FeiHillsboro, ON) in high vacuum mode. The acceler-ge was typically 25 kV. Powder was sprinkled on acovered by adhesive carbon tape and sputter coatedBalzers Union AG, Balzers, Lichtenstein) with Auickness (180 s).
cal laser-scanning microscopy (CLSM)
and uncoated particles were imaged in a confocaling microscope (Zeiss LSM 510 Meta, Germany)/1.3 oil objective. The 488 nm line of an Ar laser
cibles,was apthan 1spots w
Thethe relingred(Faldtlated a(Elver
2.7. G
TheanalyzMicrocell, fitimes bpurgesrate w
2.8. D
In acoatedof watously rScienttion, aimage FITC-labeled BSA, using a LP 505 nm emis-The pinhole was set to 0.7m. A higher effect ofsulted in photo bleaching of the particles, which con-the detected signal was from fluorescence and not
ght.
on spectroscopy for chemical analysis (ESCA)
mental surface composition of spray-dried particlesed by ESCA (AXIS HS photoelectron spectrometer,lytical, UK). The instrument used a monochromaticay light source. Powder was filled into DSC cru-
2.9. Differ
The thepowders w(DSC) (8225 mg of sampans. The ptransition tdeterminedbetween saindium (mp20 1
10 1
20 10
150
placed under vacuum overnight. The analysis areaimately 1 mm2 and the depth of analysis was less. Analysis was performed in triplicates at differentn a total area of 20 mm2.face composition of the powder was estimated fromamounts of carbon, oxygen and nitrogen in the pure(BSA and excipients) and in the spray-dried samples
., 1993). The percentage surface coverage was calcu-ding to a matrix model described in detail elsewhereand Millqvist-Fureby, 2005a).
ycnometry
arent particle density of spray-dried powders wasith nitrogen gas pycnometry (AccuPyc 1330,
tics, USA), with a 1-cm3 sample cell. The sampleto 2/3 with powder, was purged with nitrogen tene performing ten analysis runs. The pressure duringruns were 134.4 kPa (19.5 psig) and the equilibriumPa/min (0.005 psig/min).
lution properties
fort to illustrate the dissolution behavior of in situicles 50 mg of spray-dried sample was added to 1 ml8 C) in a 1.5 ml vial. The sealed vials were continu-ed on a Heidolph Duomax 1030 rocking table (Rose030, Edmonton, Canada) and the time for dissolu-ermined by visual inspection, was recorded.ential scanning calorimetry (DSC)
rmal properties of raw materials and spray-driedere examined with differential scanning calorimetrye, STARe System, Mettler Toledo, USA). Typicallyple was carefully weighted into 40l pinholed Al
owders were scanned twice around the expected glasso eliminate the effect of enthalpic relaxation. The Tgin the second heating scan was used for comparison
mples (Table 1). The instrument was calibrated using156.7 C, Hmelt = 28 J/g).
J. Elversson, A. Millqvist-Fureby / International Journal of Pharmaceutics 323 (2006) 5263 55
2.10. Fourier transform infrared spectrometry (FTIR)
Fourierspectrometdetector inpowder watein:KBr raat a load owas recons
BSA and apwas 4 cm
The abstion of abs(Dong et alfrom the prderivative sfunction (ODerivative(Kaleida Gamide I reparisons of
2.11. Circu
Far UVlarimeter (Jand 275 nmobtain a prto the concples were dof 0.12 mgspectrum w
2.12. Gel
Gel filtrumn with AUppsala, Swfiltered sama flow ratephosphate buble aggregafter appropeak perceinsoluble aUV absorbPerkin-Elm
3. Results
3.1. Dynam
Hydroxactive agen2003; Arboair/water in
Dynamic surface tension of BSA/trehalose solution after addition ofng amounts (indicated by darker spots) of: (a) 0.01%, 0.1% and 1%f HPMC, and (b) 0.01%, 0.1%, 1% (w/w) and 3.6% (w/w) of Poloxamer.mer added (squares). No BSA added (unfilled circles).
imately 44 mN/m (Machiste and Buckton, 1996). How-e time to reach equilibrium can be long, 25 h, dependingpolymer bulk concentration and polydispersity (Avranasou, 2003). Dynamic surface tension recorded by the pen-rop method demonstrated that complete adsorption oflymer to the air/water interface appeared already after547 mN/m) even at the lowest concentrations tested herea). Arboleya and Wilde (2005) obtained similar resultsring the surface tension after 1020 min, in solution of0.75% (w/w) of HPMC. However, for coating purposes,trations less than 0.001% (w/w) appeared insufficienthe surface tension remains unchanged for times as longin (induction phase) (Avranas and Iliou, 2003). Our mea-nts suggested that approximately 0.1% (w/w) (1% (w/w)weight) of HPMC was needed for efficient coating of
ince then the initial surface tension (
56 J. Elversson, A. Millqvist-Fureby / International Journal of Pharmaceutics 323 (2006) 5263
alone was considerably lower (>15 mN/m) than that of BSA andthe dynamic surface tension in the mixture of BSA and HPMCwas similar to that of HPMC alone (Fig. 2a). However, at aconcentration of 0.01% (w/w) of HPMC the initial surface ten-sion of the HPMCBSA mixture was only 5 mN/m lower thanthat of BSA alone and it was unclear whether this would besufficient for coating of BSA (Fig. 2a). Further, a polymer con-centration higher than 1% (w/w) does not necessarily lead toa more efficient coating. An unusual behavior with increasingdynamic surface tension with increasing concentration (15%,w/w) of HPMC was first reported by Machiste and Buckton(1996), which attracted attention to the influence of solution vis-cosity on the polymer diffusion rate. Later Arboleya and Wilde(2005) found a similar behavior even at concentrations as lowas 0.02 wt.%, possibly connected to the polydispersity or degree
Fig. 3. SurfaPercentage coof the polyme01% (w/w) d
of substitution. However, this was not confirmed in our experi-ments.
3.2. Dynamic surface tension of poloxamer
The equilibrium surface tension of poloxamer is compa-rable to that of HPMC (40 and 44 mN/m, respectively)(Alexandridis et al., 1994; Machiste and Buckton, 1996). How-ever, the dynamics of the polymers in relation to surface tensionappeared substantially different, at the times studied here. WhileHPMC gradually lowered the surface tension of the fresh sur-face, the initial adsorption of the poloxamer appeared muchfaster, observed as an instant reduction of the surface tension(Fig. 2b). The surface tension of HPMC then converged to near-equilibrium values while the poloxamer solutions appeared com-paratively slow in reaching the equilibrium surface tension. This
Surfage co
olyme/w) dce composition of in situ coated particles estimated by ESCA.verage by BSA (), trehalose (), and HPMC () as a functionr concentration: (a) 09% (w/w) dry weight of polymer, and (b)ry weight of polymer.
Fig. 4.Percentaof the p01% (wce composition of in situ coated particles estimated by ESCA.verage by BSA (), trehalose (), and poloxamer () as a functionr concentration: (a) 09% (w/w) dry weight of polymer, and (b)ry weight of polymer.
J. Elversson, A. Millqvist-Fureby / International Journal of Pharmaceutics 323 (2006) 5263 57
Table 2Atomic concentration determined by ESCA for spray-dried particles coated with HPMC and Poloxamer, respectively
Sample Atomic concentration (%)C (1s) O (1s) N (1s)
BSA 67.1 0.3 17.3 0.3 15.7 0.6Trehalose, spray-dried 57.1 0.0 42.9 0.0 HPMC 64.5 0.3 35.5 0.3 Poloxamer 188 71.7 0.4 28.3 0.4 BSA/trehalos, spray-dried (5:95) 61.6 02 28.5 0.3 10.0 0.2BSA/trehalos/HPMC, spray-dried (9.1% polymer) 63.6 0.1 36.5 0.1 0.00BSA/trehalos/poloxamer, spray-dried (9.1% polymer) 67.0 0.5 33.0 0.5 0.00Mean value S.D. (n = 3).
is in accordance with literature (Blomqvist et al., 2005), and thetime to reach equilibrium surface tension can be extremely long(days), due to polydispersity and slow exchange rates betweenthe adsorbed layer and the bulk. All formulations containingboth BSA and poloxamer followed the dynamics of the poly-mer which was expected from the reported solution diffusioncoefficients of Poloxamer 188 (9.2 109 m2/s) (Munoz et al.,2000b) and BSA (6.7 1011 m2/s) (Shen and Probstein, 1977).
3.3. Chemical surface composition of in situ coatedparticles
The atomic surface composition of pure materials andselected samples are shown in Table 2, and these data were used
to calculate the surface coverage of the different materials onsample powders as presented in Figs. 3 and 4. The percentagesurface coverage of polymer and protein correlated well withthe findings of dynamic adsorption as measured by the pendantdrop technique. In particles spray-dried from solutions of 1%(w/w) of dry weight of polymer no protein or only low levels(1.7% surface coverage) of protein were detected on the surfaceof the particles coated by HPMC and poloxamer, respectively(Figs. 3 and 4). At concentrations lower than 1% (w/w) of dryweight increasing levels of BSA were detected at the powder sur-face, and at the lowest amount of coating polymer used (0.1% insolids), it was not possible to calculate the surface coverage ofthe coating polymer. The reason for this is presumably that theC/O ratios in trehalose and polymer are not sufficiently differ-
Fig. 5. CLSMto the intensitcross-section illustrating the distribution of FITC-BSA in a particle before (a), and afy profile showed underneath.ter (b) in situ coating. The distribution of FITCBSA is correlated
58 J. Elversson, A. Millqvist-Fureby / International Journal of Pharmaceutics 323 (2006) 5263
ent to a give a low detection limit of either of these compoundswhen using the patch model. However, the surface coverage ofBSA could be estimated with good accuracy since N is exclu-sive to BSA. The decreasing level of BSA at the surface at eventhe lowest coating polymer concentration, and the appearanceof the corresponding placebo particles, suggest that the coat-ing polymer indeed was present at the powder surface, albeitat a low level. Uncoated particles displayed a mixed surfaceof BSA and trehalose, containing approximately 57% of BSA(Figs. 3 and 4). This level is lower compared to what has beenobserved for BSA spray-dried from other carbohydrate solutions(Faldt and Bergenstahl, 1994; Landstrom et al., 2000). Increas-ing the polymer content above 1% (w/w) resulted in higher levelsof polymer at the surface. However, at a polymer concentrationof 9.1% (w/w) of dry weight the surface coverage of polymer was30% higher with HPMC compared to poloxamer (Figs. 3 and 4).Even at a concentration as high as 26% (w/w) of dry weightpoloxamer the surface level was comparable to that of HPMCat 9.1% (w/w) of dry weight (data not shown). This can be pre-sumably explained by the poloxamer forming a thinner film thanHPMC, thus the ESCA signal is a combination of a (reasonably)complete surface film of poloxamer and the underlying mate-rial containing carbohydrate as well as protein. Interestingly, nosignal is detected from the protein. This might indicate a stericeffect from the adsorbed poloxamer layer, with the PEO-tails
pointing towards the solution at higher concentrations (Munozet al., 2000a; Blomqvist et al., 2005) Thereby, globular BSAbut not trehalose might be excluded from the (sub)surface layer.The CLSM pictures in Fig. 5 illustrate suppression of BSA fromthe particle surface by addition of HPMC. The concentrationof BSA was much higher near the particle surface in uncoatedparticles whereas the coated particles have a particularly highconcentration of BSA towards the centre of the particle.
3.4. Surface composition and correlations to particleproperties
As part of this investigation we wanted to investigate whetherspecific properties of the particles were affected at the polymerconcentrations appropriate for in situ coating. Indeed, variationin the surface composition induced changes in a number ofparticle properties such as the particle surface morphology,particle shape, particle size, dissolution time and powderflowability.
As expected, spray-dried particles containing proteins dis-played a characteristic raisin-like morphology (corrugated par-ticles), due to the adsorption of protein at the air/liquid interfaceof the droplets in the spray (Fig. 6a). During drying, proteinsform a visco-elastic adsorbed layer covering the surface of thedroplets. Addition of a low-molecular weight surfactant, such as
Fig. 6. Micro PMCweight.graphic pictures of BSA-containing in situ coated particles: (a) uncoated, (b) H , and (c) Poloxamer. The polymer content was 1% (w/w) of dry
J. Elversson, A. Millqvist-Fureby / International Journal of Pharmaceutics 323 (2006) 5263 59
Table 3Literature data on surface rheology of polymer and protein films
Polymer Surface concentration(mg/m2)
ElasticmodulusG(mN/m)
Dilatational modulusE (mN/m)
Reference
HPMC 120 Arboleya and Wilde (2005)
BSA1.37 59 Benjamins and Lucassen-Reynders (1998)1.95 69
60 Pereira et al. (2003)
PVA2.2 11 Benjamins and Lucassen-Reynders (1998)3.1 11
Poloxamer P85 0.1 2 Blomqvist et al. (2005)
polysorbateface; it alsosmooth sphIn terms ofsurface elasadsorbed lapolymers, sand the chato smooth,(Elverssonformer anddroplet as pa similar wsibly thereof the dropcles. To vein literaturlected andtheir flexibthe elasticWilde, 200and poloxaand Lucass2005), respof approxim1998; PerePVA. In adbates are vewhich mayafter additi2000). Conmers are likthe surfacehalose) wa
on owrinBSA
th anm c
es (ncoaties (Feolo
loxamd likat tsizeaffetennd Med aere
larlyand
gher), whsurfring
dieslymeay-dly m
polong tandpar
ionsion
Table 4Apparent part
Polymer (% o00.119
26
Mean value , not only expels protein molecules from the inter-produces a less cohesive film, which may result inerical particles (Maa et al., 1998; Adler et al., 2000).surface rheology, this means a film with negligibleticity due to the high mobility of the surfactant in theyer (Arboleya and Wilde, 2005). Linear film forminguch as polyvinyl alcohol (PVA), have a similar effectnge in particle surface morphology, from corrugatedis correlated to the PVA content of the particle surfaceand Millqvist-Fureby, 2005a). HPMC is a strong filmappears to form a similar type of film at the sprayroteins, since the particle morphology is affected inay as for protein containing particles (Fig. 6b). Pos-is a correlation between the visco-elastic propertieslet surface and the morphology of spray-dried parti-rify this hypothesis, data on surface viscosity founde for HPMC, BSA, PVA and poloxamer were col-it was possible to rank these polymers according toility and surface rheology (Table 3). For example,modulus of HPMC was 130 mN/m (Arboleya and5) whereas the dilatational modulus (E) of both PVAmer was comparatively lower, 11 mN/m (Benjaminsen-Reynders, 1998) and 2 mN/m (Blomqvist et al.,ectively. Consequently, with a dilatational modulusately 60 mN/m (Benjamins and Lucassen-Reynders,
ira et al., 2003) this places BSA between HPMC anddition, low-molecular surfactants, such as polysor-ry mobile and hence, without surface visco-elasticity,explain the frequent observations of smooth spheres
on of, e.g. polysorbates (Maa et al., 1998; Adler et al.,sequently, it might be possible to predict which poly-ely to change the surface morphology. For example,of particles without any BSA or HPMC (pure tre-
s smooth and particles were spherical (not shown).
Additihighlyulus ofbut wiogy froparticlin situparticlface rhthe poface anmobile
Thecan besurface(Kim auncoatticles wparticuweightand hishowninitialsize du
Stuthe pothe sprsiderabto theassessiextentcoatedreductcentraticle density by gas pycnometry of in situ coated particles
f dry weight) BSA/trehalose/HPMC (g/cm3) Trehalose/1.51 1.541.50 1.531.46 1.471.23 1.24
S.D.
60 J. Elversson, A. Millqvist-Fureby / International Journal of Pharmaceutics 323 (2006) 5263
Fig. 7. Dissoland the bulk c
spray-driedwith an incsity decreasThe increa(Elversson
A possimodified relow compaincrease intration spalated withUncoatedthe lowestdepended oin the bulkate dissoludesired dis
Interestipoloxamerdrying andtions wereshape andwater sorp(Kibbe, 20was not poand close iparticles in
erated into small aggregates, which may in part account for theimproved flow properties.
Inclusion of polymeric material can influence the re-crystallization behavior of, e.g. spray-dried lactose (Stubberudand Forbes, 1998; Corrigan et al., 2002; Berggren andAlderborn, 2004). However, it was unclear whether very lowconcentrations of polymer as in in situ coating would affect thesolid-state properties of the particles. The spray-dried particlesconsisted mainly of trehalose. Consequently, it was the transi-tions of trehalose that appeared from DSC analysis. Addition 5%of BSA induced a moderate (3 C) increase of the glass transi-tion temperature of the dried powder from 118.5 to 121.5 C(Table 5). However, addition of HPMC, which has a Tg around
, did not, at the studied concentrations, affect the Tg of thelationed oan ber t
merg trathe axam
ray-dst, adamo
each
tructes
inenati
as obromlix c, 19ble acom
etecPM
Table 5DSC results o
Polymer (% o
00.119
26
HPMC: Tg,ution time as a function of the level of HPMC at the particle surfaceoncentration of HPMC.
aqueous two-phase systems (ATPS). In formulationsreasing content of PVA the apparent particle den-ed linearly (Elversson and Millqvist-Fureby, 2005a).
sing surface load of PVA was confirmed by ESCAand Millqvist-Fureby, 2005a).ble application for in situ coated particles is forlease purposes. Although the amount of polymer isred to conventional coating solutions, a three-foldthe dissolution time was observed in the concen-
n investigated (Fig. 7). The dissolution time corre-the surface load of HPMC, as determined by ESCA.particles dissolved similar to particles coated withconcentration of HPMC. Hence, the dissolution timen the surface coverage rather than the concentration. Consequently, choosing a polymer with appropri-tion properties in situ coating can be used to obtain asolution profile of the powder.ngly, better flow properties of particles coated withcompared to HPMC were noticed during spray-powder handling. Possibly, interparticulate interac-affected by either surface composition or particle
morphology (Hickey et al., 1994). For example, the
175 Cformuobservamer c
i.e. lowpoloxameltinently,or poloing spcontraformstent of
3.5. Sparticl
Bovhad aBSA,plied fa -he(Petersof soluple. Inwere dwith Htion isotherm of poloxamer is below that of HPMC00), resulting in a less sticky powder. However, itssible to confirm these results by flowability tests
nspections of SEM images implied that the primarysamples with a high poloxamer content were agglom-
tion of natiFTIR was c(Table 6).FTIR but bthat a nativ
n thermal transitions in spray-dried powders coated with HPMC and poloxamer, resp
f dry weight) BSA/trehalose/HPMC Trehalose/HPMCTg (C) Tg (C)121.6 0.1 118.5 0.5121.1 0.3 122.1121.4 0.5 122.2 0.1121.6 0.2 123.0 1.5
170180 C (Kibbe, 2000); Poloxamer: Tm, 55 C, Tm, 115J/g, Tc, 30 C. Mean vas that remained amorphous. Neither was any effectn Tg from addition of poloxamer. The Tg of polox-e assumed to be similar to that of PEG (Kibbe, 2000),han all other constituents, and particles coated withshowed the presence of crystalline polymer, with ansition at approximately 5053 C (Table 5). Appar-ddition of an amphiphilic polymer, such as HPMCer result in phase separation close to the surface dur-rying, due to the surface activity of the polymer. Indition of, e.g. dextran, which has no surface activity,
rphous particles with a Tg dependant of the mass con-excipient (Elversson and Millqvist-Fureby, 2005a).
ural integrity of protein of in situ coatedFTIR, CD and gel ltration
serum albumin (BSA) in all in situ coated particlesve-like structure comparable to that of rehydratedserved from CD, FTIR and gel filtration. BSA sup-the commercial source and dissolved in buffer, hadontent of approximately 55%, which was expected95). Further, the content of 89% monomer and 11%ggregates were values typical of a commercial sam-parison, only slightly lower levels of -helix (47%)ted by FTIR in the liquid-state, for samples coatedC (Table 6). In the dried state, the high preserva-ve structure in in situ coated samples as displayed byonfirmed by oforandrade levels of soluble aggregatesPoloxamer-coated samples were not analyzed withoth CD (Fig. 8) and gel filtration (Fig. 9) confirmede structure was most likely. The structural integrity
ectively
BSA/trehalose/poloxamer
Tg (C) Tm (C) Hm (J/g)121.6 0.1 121.6 120.5 0 50.5 0.1 0.2 0.1121.4 51.1 0 4.1 0.1121.4 0.2 52.6 0.1 24.8 0.5
lue S.D. (n = 3).
J. Elversson, A. Millqvist-Fureby / International Journal of Pharmaceutics 323 (2006) 5263 61
Table 6Structural integrity of protein in in situ coated particles as analyzed with CD, FTIR and gel filtration
Excipients Polymer (% of dry weight) Protein:trehalose mss a b c
BSA, native
Trehalose, dextran 1:3.8Trehalose 0 1:19
In situ HPMC, trehalose0.1 1:191 1:199 1:19
In situ poloxamer, trehalose
0.1 1:191 1:199 1:1926 1:19
na, not analyzed; N, native BSA; -helix content (%); Aggr, soluble aggregates (insoluble); Frag,
of BSA in uncoated particles was higher than anticipated, andcomparable to that of BSA in coated particles. BSA spray-driedwith trehalose alone had a slightly lower -helix content (43%),and approximately 12% of soluble aggregates, compared to 11%in native or in situ coated samples (Table 6). In comparison, both-helix and34%, respetrehalose athe stabilizeffect, sincder only apexpected atprotein conin a higheras concentever, due toneeded forsolutions w
Fig. 8. CD spline), uncoateof dry weightmonomer content was substantially lower (33% andctively) as BSA was spray-dried in a formulation withnd dextran in ratio 1:5 (Table 6 and Fig. 9). Possibly,ing effect of trehalose dominated over the coatinge at a protein concentration of 5% (w/w) in the pow-proximately 4% of the total content of protein can bethe particle surface (Landstrom et al., 1999). A lowercentration in the spray solution would have resultedadsorbed fraction, and possibly a significant as well
ration dependent effect of the in situ coating. How-the limitations regarding the protein concentration
liquid FTIR, the protein concentration of the sprayas set to 5 mg/ml.ectra of in situ coated particles after rehydration. Native BSA (boldd (), 0.1% (w/w) ( ), 1% (w/w) ( ), 26% (w/w) (- - -)poloxamer.
Fig. 9. C
4. Conclu
Dynamisured by thdiction of tsurface. Duspray-dryinbecome mand hence,shell formathe correlatvery satisfyage of BSAof dry weigratio CD FTIR SEC
N Aggr Frag
N 55 89 11
33 34 2 (13) 51N 43 88 12
N 49 89 11 N 45 89 11 N 47 89 11
na na 86 14 N na 89 11 N na na na naN na na na na
fragments (%).ontent of monomeric and aggregated BSA by gel filtration.
ding remarks
c surface tension at the air/liquid interface, as mea-e pendant drop technique was very useful in pre-
he chemical composition of the spray-dried particlee to the very short lifetime of droplet surfaces duringg adsorption kinetics of surface-active components
ore important than the equilibrium surface tensionmolecules adsorbed at the surface in the moment oftion will remain there during drying. Consequently,ion between dynamic surface tension and ESCA wasing (Figs. 24). Efficient coating, with full cover-appeared at a polymer concentration 1% (w/w)
ht and the repression of BSA from the surface was
62 J. Elversson, A. Millqvist-Fureby / International Journal of Pharmaceutics 323 (2006) 5263
illustrated by CLSM of FITC-labeled BSA (Fig. 5). Further,changes in the surface morphology of particles (Fig. 6) mightbe correlatpolymers.several partime to dispycnometrcentrationpowder (Tameability oHPMC coamight indiAlthough imational stnot excludelow-dosageprotein is e
Acknowled
Financiagratefully aAndres Venology, Rouse of the A(AstraZenelarimeter,Chemistry,Institute ofmicrograph
Reference
Adler, M., Unof bovine
Alexandridis,Surface-ablock-poly
Arboleya, J.-methylcel485491.
Avranas, A., Iand the ansalts, as s102109.
Bayer, C., Kachemical vbed. Chem
Benjamins, J.proteins aProteins a
Berggren, J.,poly(vinyEur. J. Ph
Blomqvist, BPEOPPOof spread
Broadhead, Jprocess agalactosid
Caruso, F., 20
Columbano, A., Buckton, G., Wikely, P., 2003. Characterisation of surfacemodified salbutamol sulphate-alkylpolyglycoside microparticles preparedby spray-drying. Int. J. Pharm. 253, 6170.
n, D.Otionsical., H
r from330, U.,ueoun, J.,ulation, J.,
nge.n, J.,
particng spr., Beeinla, Bergders ad Stru
.F., Kindu
A.J., Cng the., Ichiate decoatin, PerrpatenA., 20maceY., Mom, Keen b
d Hydom, Ktitativ. B 12F., Ngitive re, E.Ohylce, K., 1Yorkt-Fur
sins243t-Furphase
rf. Scithalerng proe-typM.G.rs of sum prM.G.rs of skinetL.G.Clogy235.J., 1
licatioed to differences in surface visco-elasticity betweenIn addition, the in situ coating induced a change inticles properties of pharmaceutical relevance, such assolution and flowability. Moreover, studies with gasy, showed a correlation between the polymer con-and the apparent particle density of the spray-driedble 5), presumably due to changes in the gas per-f the surface. The effect was more pronounced forted particles than for poloxamer-coated particles andcate an effect of the thickness of the polymer film.t was not possible to detect any improved confor-ability of protein from in situ coating, the results dothat additional stabilization would be achieved in aformulation where the fraction of surface adsorbed
xpected to be higher (Table 6).
gements
l support from AstraZeneca R&D, Lund, Sweden iscknowledged. The authors also thank Assoc. Prof.
ide and Ulrika Eriksson (Deptartment of Biotech-yal Institute of Technology, Stockholm, Sweden) for KTA gel filtration equipment, Dr. Stefan Ulvenlundca R&D, Lund, Sweden) for use of the CD spectropo-
and Andreas Sonesson (YKI, Institute for SurfaceStockholm, Sweden/Deptartment of Physics, RoyalTechnology, Stockholm, Sweden) for the CLSM
s.
s
ger, M., Lee, G., 2000. Surface composition of spray-dried particlesserum albumin/trehalose/surfactant. Pharm. Res. 17, 863870.
P., Athanassiou, V., Fukuda, S., Hatton, T.A., 1994.ctivity of poly(ethylene oxide)-block-poly(propylene oxide)-(ethylene oxide) copolymers. Langmuir 10, 26042612.
C., Wilde, P.J., 2005. Competitive adsorption of proteins withlulose and hydroxypropyl methylcellulose. Food Colloids 19,
liou, P., 2003. Interaction between hydroxypropylmethylcelluloseionic surfactants hexane-, octane-, and decasulfonic acid sodiumtudied by dynamic surface tension. J. Colloid Interf. Sci. 258,
rches, M., Matthews, A., von Rohr, P.R., 1998. Plasma enhancedapor deposition on powders in a low temperature plasma fluidized. Eng. Technol. 21, 427430.
, Lucassen-Reynders, E.H., 1998. Surface dilatational rheology ofdsorbed at air/water and oil/water interfaces. In: Miller, R. (Ed.),t Liquid Interfaces, vol. 7. Elsevier, Amsterdam, pp. 284341.
Alderborn, G., 2004. Long-term stabilisation potential oflpyrrolidone) for amorphous lactose in spray-dried composites.arm. Sci. 21, 209215..R., Warnheim, T., Claesson, P.M., 2005. Surface rheology ofPEO triblock copolymers at the air/water interface: comparison
and adsorbed layers. Langmuir 21, 63736384.., Rouan, S.K.E., Hau, I., Rhodes, C.T., 1994. The effect ofnd formulation variables on the properties of spray-dried beta-ase. J. Pharm. Pharmacol. 46, 458467.01. Nanoengineering of particle surfaces. Adv. Mater. 13, 1122.
Corrigasoluchem
Dong, Awate3303
Elofssonin aq
Elverssoform
Elverssodryi2060
Elverssoandduri
Faldt, Pprot
Faldt, P.powFoo
Gibbs, Bfood
Hickey,enci
Jono, Kticulbed
Jung, J.and
Kibbe,Phar
Kim, K.Landstr
betwFoo
LandstrquanSurf
Maa, Y.sens
Machistlmet
MastersNew
Millqvistryp188,
Millqvistwo-Inte
Mumendryitissu
Munoz,layelibri
Munoz,layetion
Pereira,rheo2349
Peters, TApp., Healy, A.M., Corrigan, O.I., 2002. The effect of spray-dryingof polyethylene glycol (PEG) and lactose/PEG on their physico-properties. Int. J. Pharm. 235, 193205.uang, P., Caugney, W.S., 1990. Protein secondary structure in
second-derivative amide I infrared spectra. Biochemistry 29,8.Millqvist-Fureby, A., 2000. Drying of probiotic micro-organismss two-phase systems. Minerva Biotechnol. 12, 279286.Millqvist-Fureby, A., 2005a. Aqueous two-phase systems as an concept for spray-dried protein. Int. J. Pharm. 294, 7387.Millqvist-Fureby, A., 2005b. Particle size and density in spray-ffects of carbohydrate properties. J. Pharm. Sci. 94, 2049
Millqvist-Fureby, A., Alderborn, G., Elofsson, U., 2003. Dropletle size relationship and shell thickness of inhalable lactose particlesay-drying. J. Pharm. Sci. 92, 900910.rgenstahl, B., 1994. The surface composition of spray-driedctose powders. Colloid Surf. A 90, 183190.enstahl, B., Carlsson, G., 1993. The surface coverage of fat on foodnalyzed by ESCA (electron spectroscopy for chemical analysis).ct. 12, 225234.ermasha, S., Alli, I., Mulligan, C.N., 1999. Encapsulation in the
stry: a review. Int. J. Food Sci. Nutr. 50, 213224.oncessio, N.M., Van Oort, M.M., Platz, R.M., 1994. Factors influ-dispersion of dry powders as aerosols. Pharm. Technol., 5864.
kawa, H., Miyamoto, M., Fukomori, Y., 2000. A review of par-sign for pharmaceutical powders and their production by spoutedg. Powder Technol. 113, 269277.
ut, M., 2001. Particle design using supercritical fluids: literaturet survey. J. Supercrit. Fluids 20, 179219.00. Handbook of Pharmaceutical Excipients, 3rd ed. Americanutical Association and Pharmaceutical Press, New York.arshall, W.R., 1971. Am. Inst. Chem. Eng. J. 17, 5757.., Alsins, J., Bergenstahl, B., 2000. Competitive protein adsorptionovine serum albumin and -lactoglobulin during spray-drying.rocolloids 14, 7582..B.B., Alsins, J., Almgren, B., 1999. A fluorescence method fore measurements of specific protein at powder surfaces. Colloid, 429440.uyen, P.A.T., Hsu, S.W., 1998. Spray-drying of air/liquid interfaceecombinant human growth hormone. J. Pharm. Sci. 87, 152159.., Buckton, G., 1996. Dynamic surface tension of hydroxypropy-
llulose film-coating solutions. Int. J. Pharm. 145, 197201.991. Spray-drying Handbook. Longman Scientific & Technical,, pp. 193274.eby, A., Malmsten, M., Bergenstahl, B., 1999. Spray-drying ofurface characterisation and activity preservation. Int. J. Pharm.253.eby, A., Malmsten, M., Bergenstahl, B., 2000. An aqueous polymersystem as carrier in spray-drying of biological material. J. Colloid
. 225, 5461., M., Hsu, C.C., Pearlman, R., 1994. Feasibility study on spray-tein pharmaceuticals: recombinant human growth hormone and
e plaminogen activator. Pharm. Res. 11, 1220., Monroy, F., Ortega, F., Rubio, R.G., Langevin, D., 2000a. Mono-ymmetric triblock copolymers at the air/water interface. 1. Equi-operties. Langmuir 16, 10831093., Monroy, F., Ortega, F., Rubio, R.G., Langevin, D., 2000b. Mono-ymmetric triblock copolymers at the air/water interface. 2. Adsorp-ics. Langmuir 16, 10941101.
., Theodoly, O., Blanch, H.W., Radke, C.J., 2003. Dilatationalof BSA conformers at the air/water interface. Langmuir 19,6.995. All about albumin: biochemistry. In: Genetics and Medicalns. Academic Press.
J. Elversson, A. Millqvist-Fureby / International Journal of Pharmaceutics 323 (2006) 5263 63
Pfeffer, R., Dave, R.N., Wei, D., Ramlakhan, M., 2001. Synthesis of engineeredparticulates with tailored properties using dry particle coating. Powder Tech-nol., 117.
Shen, J.J.S., Probstein, R.F., 1977. Prediction of limiting flux in laminar ultra-filtration of macromolecular solutions. Ind. Eng. Chem. Fund. 16, 459465.
Stubberud, L., Forbes, R.T., 1998. The use of gravimetry for the study of theeffect of additives on the moisture-induced recrystallisation of amorphouslactose. Int. J. Pharm. 163, 145156.
Wan, L.S.C., Heng, P.W.S., Chia, C.G.H., 1991. Preparation of coated particlesusing a spray-drying process with an aqueous system. Int. J. Pharm. 77,183191.
Watano, S., Nakamura, H., Hamada, K., Wakamatsu, Y., Tanabe, Y., Dave, R.N.,Pfeffer, R., 2004. Fine particle coating by a novel rotating fluidized bedcoater. Powder Technol. 141, 172176.
In situ coating-An approach for particle modification and encapsulation of proteins during spray-dryingIntroductionMaterials & methodsMaterialsSpray-dryingDynamic surface tensiometry (pendant drop)Scanning electron microscopy (SEM)Confocal laser-scanning microscopy (CLSM)Electron spectroscopy for chemical analysis (ESCA)Gas pycnometryDissolution propertiesDifferential scanning calorimetry (DSC)Fourier transform infrared spectrometry (FTIR)Circular dichroism (CD)Gel filtration
Results and discussionDynamic surface tension of HPMCDynamic surface tension of poloxamerChemical surface composition of in situ coated particlesSurface composition and correlations to particle propertiesStructural integrity of protein of in situ coated particles-FTIR, CD and gel filtration
Concluding remarksAcknowledgementsReferences