Poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide)/poly(ϵ-caprolactone) (PCL) amphiphilic block copolymeric nanospheres: II. Thermo-responsive drug release behaviors

Journal of Controlled Release 65 (2000) 345–358www.elsevier.com/ locate / jconrel

Poly(ethylene oxide)-poly(propylene oxide)-poly(ethyleneoxide) /poly(e-caprolactone) (PCL) amphiphilic block

copolymeric nanospheresII. Thermo-responsive drug release behaviors

*So Yeon Kim, Jung Chul Ha, Young Moo LeeDepartment of Industrial Chemistry, College of Engineering, Hanyang University, Seoul 133-791, South Korea

Received 19 May 1999; accepted 22 September 1999

Abstract

Amphiphilic block copolymers composed of relatively hydrophilic PEO-PPO-PEO block copolymer (Pluronic) andpoly(´-caprolactone) with hydrophobic character were synthesized by ring-opening polymerization of ´-caprolactone in thepresence of PEO-PPO-PEO block copolymer using stannous octoate as a catalyst. Pluronic /PCL block copolymericnanospheres with core-shell structure were prepared by dialysis method. They showed the average diameter of 116–196 nmdepending on the type of copolymer. All the nanosphere samples exhibited a narrow size distribution. The critical micelleconcentrations of Pluronic /PCL amphiphilic block copolymers determined by fluorescence spectroscopy were lower than

1that of the common low molecular weight surfactant. Their core-shell structure was confirmed by H NMR spectroscopy.Pluronic /PCL block copolymeric nanospheres exhibited the reversible change of size depending on the temperature. Releasebehaviors of indomethacin from Pluronic /PCL block copolymeric nanospheres also showed temperature dependence and asustained release pattern. In addition, cytotoxicity test using an MTT assay method revealed that these indomethacin-loadedPluronic /PCL nanospheres could remarkably reduce the cell damage compared with the unloaded free indomethacin. 2000 Elsevier Science B.V. All rights reserved.

Keywords: Pluronic; Poly(´-caprolactone); Block copolymer; Nanospheres; Drug delivery system

1. Introduction cations. As the meaning of ‘drug delivery’ expandsto the targeting drug at a specific body site or

Over the past several decades much interest has releasing drugs when needed, as well as controllingfocused on designing new drug dosage forms in the release rate of drug, stimuli-sensitive drug deliv-order to enhance the effectiveness of existing medi- ery has been required depending on changes in

physiological signals in the body [1–5].Stimuli-sensitive systems using polymers can*Corresponding author. Tel.: 182-2-2291-9683; fax: 182-2-

change their volume and shape reversibly according2291-5982.E-mail address: [email protected] (Y.M. Lee) to various external physicochemical factors [1–5].

0168-3659/00/$ – see front matter 2000 Elsevier Science B.V. All rights reserved.PI I : S0168-3659( 99 )00207-2

346 S.Y. Kim et al. / Journal of Controlled Release 65 (2000) 345 –358

Namely, chemical signals, such as pH, metabolites, stimuli-responsive drug delivery systems [1–8].and ionic factors, will alter the molecular interactions However, only a few studies on stimuli-responsivebetween polymer chains or between polymer chain nanoparticulate systems have been performed.and solutes present in a system. Physical stimuli, We have already reported on amphiphilic blocksuch as temperature or electrical potential, may copolymeric nanospheres based on methoxy poly-provide various energy sources for molecular mo- (ethylene glycol) and polyester, such as poly(´-cap-tions and altering molecular interactions. These rolactone) [18,19], poly(D,L-lactide) [20] and poly-interactions will change the properties of polymer (glycolide) [21]. These block copolymeric nanos-materials, such as swelling, solubility, configuration pheres were less than 200 nm, had a relatively highor conformational change, redox states, and crys- loading content of a drug, and had the desired releasetalline /amorphous transition [3–5]. Such a system behaviors in vitro and higher cell viability comparedwill find potential applications to many clinical with unloaded free drug. More recently, we reportedsituations that require drug delivery at specific times, on the synthesis and characterization of amphiphilicin a pulsed manner, or depending on metabolite block copolymers composed of Pluronic series, ABAconcentration. triblock copolymers of poly(ethylene oxide)-poly-

Body temperature often deviates from normal (propylene oxide)-poly(ethylene oxide) (PEO-PPO-temperature, kept constant near 378C, by the physio- PEO), as a hydrophilic segment and poly(´-caprolac-logical presence of pathogens or pyrogens. This tone) (PCL) as a hydrophobic block copolymer [22].temperature change may be a useful stimulus that In addition, we prepared nanospheres using thesecan modulate the delivery of therapeutic drugs for synthesized copolymers by a dialysis process in adiseases with accompanying fever. In addition, exter- selective solvent and examined their properties as anally controlled temperature can also be used to drug carrier [22].modulate the release of drug. Therefore, temperature- The objective of the present study was to preparesensitive drug delivery systems have been extensive- temperature-responsive nanoparticles and to investi-ly studied [6–8]. Katono et al. investigated thermal gate their drug release behaviors. We introducedcollapse from IPN, which is composed of poly- indomethacin as a model drug into the Pluronic /PCL(acrylamide-co-butylmethacrylate) and poly(acrylic block copolymeric nanospheres and determined theiracid), and obtained on–off release profiles as a drug loading efficiency. Moreover, we performed thefunction of temperature [6]. Poly(N-iso- structural analysis of Pluronic /PCL block copoly-propylacrylamide) (PNIPAAm) hydrogels demonstra- meric nanosphere. The effect of temperature on theirted negative temperature sensitivity with lower criti- size and in vitro release behaviors of drug fromcal solution temperature (LCST) in an aqueous nanospheres depending on temperature are reportedsolution, which showed a decrease of swelling with here.increasing temperature above 30–328C [6–8]. LCSTresults from the influence of temperature on poly-mer–polymer interactions, such as hydrogen bonding 2. Experimentaland hydrophilic /hydrophobic interaction. By utiliz-ing PNIPAAm hydrogels, a temperature-controlled 2.1. Materialson–off drug delivery system for indomethacin wasachieved [8]. Pluronic series (poly(ethylene oxide)-poly-

In addition, one possible means of reaching the (propylene oxide)-poly(ethylene oxide) triblock co-above outlined goal may be a delivery via particulate polymers, Pluronic F-127, F-87, F-68, F-88) weredrug delivery systems. Particulate systems lend kindly supplied from BASF Corporation and usedthemselves to parenteral administration, and may be without further treatment. ´-Caprolactone monomeruseful as sustained-release injections or for the was purchased from Tokyo Kasei Organic Chemi-delivery of a drug to a specific organ or target site cals. Stannous octoate [Sn(OOCC H ) ] was ob-7 15 2

[9–17]. Until now, there have been a number of tained from Sigma (St. Louis, MO, USA). In-studies on polymeric materials of hydrogel type as domethacin (IMC) (1-(4-chlorobenzoyl)-5-methoxy-

S.Y. Kim et al. / Journal of Controlled Release 65 (2000) 345 –358 347

1H-indole-3-acetic acid) was obtained from Sigma. were collected by filtration and washed several timesNormal human fibroblast cells (KCLB No. 21947, with diethyl ether, and dried in a vacuum oven atCCD-98sk) were supplied from Korean Cell Line 408C for 3 days.Bank. Dulbecco’s modified eagle medium (DMEM),phosphate-buffered saline (PBS) and fetal bovine 2.3. Preparation of Pluronic /PCL nanospheresserum (FBS) were purchased from Life Technology and drug-loaded Pluronic /PCL nanospheres(Gibco, USA). Water was first treated with reverseosmosis system (Sambo Glove, Korea) and further Pluronic /PCL block copolymeric nanospheres anddeionized with a Milli-Q Plus System (Waters, hydrophobic drug-loaded nanospheres were preparedMillipore, USA). All other chemicals used were by a dialysis method [15,18–22]. Indomethacinreagent grade and used as purchased without further (IMC) which had hydrophobic properties was usedpurification. as a model drug. One hundred mg of Pluronic /PCL

block copolymer were dissolved in 10 ml organic2.2. Synthesis of Pluronic /PCL block copolymer solvent, dimethylformamide (DMF). In preparing

drug-loaded nanospheres, 100 mg of Pluronic /PCLAmphiphilic block copolymers containing block copolymer were dissolved in 10 ml DMF,

Pluronic and poly(´-caprolactone) were synthesized followed by the addition of IMC (100 mg) andwith various feed conditions of two components. As stirred at room temperature. To form amphiphilicdescribed in our previous study [22], copolymeriza- block copolymeric nanospheres and remove organiction was accomplished by ring opening polymeri- solvent, the solution was dialyzed for 24 h against 3 lzation of ´-caprolactone monomer in the presence of of water at 208C using cellulose membrane bagPluronic using stannous octoate as a catalyst. The (molecular weight cut-off, 12 000–14 000; size, 21 /Pluronics are PEO-PPO-PEO triblock copolymers in 35, Sigma). The water was exchanged at intervals ofwhich the hydrophobe consists of a propylene oxide 3–6 h. After dialysis, the micellar solution remainingblock, and hydrophile is made of ethylene oxide in the dialysis bag was collected and sonicated usingblocks. Branson 2210 sonicator (Brandon Ultrasonics, USA).

In the present study we used four kinds of Then, the solution was centrifuged (Jouan BP403,Pluronics: two samples (F127 and F87) consisted of France) at 1000 rpm for 3 min to eliminate aggre-approximately 70 wt% PEO moiety, and had total gated particles and unloaded IMC. The supernatants,molecular weights of 12 600 and 7700 (by BASF), micellar solutions, were frozen and lyophilized by arespectively. Eighty wt% of the other two Pluronics freeze-dryer system (Labconco, USA) for 2 days to(F88 and F68) were composed of PEO block in their obtain dried nanosphere products. After centrifuga-ends and their molecular weights were 11 400 and tion, nanospheres of more than 80% of the initial8400, respectively. The weighed amounts of PEO- feed for all Pluronic /PCL sample formulation couldPPO-PEO triblock copolymer and ´-caprolactone be obtained.monomer were added in a round-bottomed 100-mlflask connected with a vacuum joint, and the mixture 2.4. Characterization of Pluronic /PCL copolymericwas pre-heated in order to make well-mixed melting nanospheresphase. After mixing, the flask was cooled. Anappropriate amount of catalyst was added in the The molecular weight distribution of Pluronic /flask. The flask was degassed by connecting vacuum PCL block copolymers was characterized by thepumps and then sealed off and placed in a vacuum elution time relative to polystyrene monodisperseoven at 1408C. After 12 h, the reaction product was standards from the gel permeation chromatographycooled at an ambient temperature and dissolved in apparatus (GPC, Waters Model 510 HPLC pump,dichloromethane. The solution was extracted in an Milford, USA) with a Millennium software program.excessive amount of methanol to remove any un- Three StyragelE tetrahydrofuran (THF) columnsreacted PEO-PPO-PEO block copolymers, ´-cap- (each 30 cm37.8 mm I.D. HR-0.5, HR-4, HR-5, allrolactone monomers and a catalyst. The precipitates Waters) and a Waters R410 differential refractomet-


ric detector were used. The mobile phase was THF general formula for the photoelectron count-timewith a flow-rate of 1 ml /min. The mobile injection correlation function in the DLS measurement isvolume was usually 100 ml of stock solutions (0.1– expressed by the following equation.0.5%, w/v). The calibration curve was prepared

(1)before measurements by using five standard poly- g (t) 5E G(G ) exp(2Gt) dG (1)3styrenes with molecular weight of 1.285310 ,

3 4 4 42.965310 , 1.135310 , 2.855310 and 6.505310 , (1)where g (t) is the normalized first-order correlationrespectively (Shonex standard SM-105, Showa

function, t is the delay time, and G is the averageDenko, Japan).

characteristic line width and G(G ) is a distributionIn addition, the composition and number-average

function of G. The autocorrelation functions weremolecular weight (M ) of each copolymer in CDCln 3 analyzed using cumulants method in which1solution was determined by 500 MHz H NMR

1 (1) 2 2(Bruker AMX-500). Besides, H NMR measurement g (t) 5 exp[2Gt 1 (m /2)t 2 (m /3!)t 1 ? ? ? ]2 2was also carried out to investigate the core–shell

(2)structure of Pluronic /PCL block copolymeric nanos-pheres. Tetramethylsilane (TMS) was used as an yielding an average characteristic line width G, and a

2internal standard. variance (polydispersity index) m /G . The size2

The Pluronic /PCL nanospheres were observed distribution was also estimated from the correlationwith a field emission scanning electron microscope function profile using histogram analysis software. In(Jeol Model JSM-6340F). To prepare the sample for the histogram method, the normalized first-order

(1)FE-SEM measurement, the micellar solution of correlation function ( g (t)) is used to determine thenanospheres was placed onto a copper mount, and size distribution. G(G ) was determined using theevaporated slowly at 208C. Completely dried nanos- Marquart nonlinear least-squares routine and thenpheres were coated with gold. converted into the particle distribution.

Fluorescence spectroscopy measurements werecarried out to determine the critical micelle con- 2.6. Drug loading efficiency (DLE)centration (CMC) of Pluronic /PCL block copoly-meric nanospheres formed through solution behavior The amount of IMC loaded in the inner core ofof amphiphilic block copolymer in selective solvents. Pluronic /PCL block copolymeric nanospheres wasFluorescence excitation spectra of pyrene were mea- investigated using a UV-visible spectrophotometersured at various concentrations of Pluronic /PCL (Model UV-2101PC, Shimadzu, Japan). To evaluateblock copolymer using a spectrofluorophotometer the loading amount of drugs, freeze-dried nanos-(Model RF-5301PC, Shimadzu, Japan) [23,24]. pheres were disrupted by an addition of ethanol /THFEmission wavelength was 390 nm for excitation (1:1, v /v) cosolvent. Then, the amount of IMCspectra. Pyrene was chosen as a fluorescent probe introduced in nanospheres was determined bybecause of its photochemical properties suitable for measuring the UV absorbance at 319 nm. The DLEan effective probe [25]. was calculated by the weight ratio of IMC in

nanospheres to pre-weighed IMC-loaded nanos-2.5. Size and size distribution pheres using the following equation.

Amount of IMC in nanospheresAverage size and size distribution of Pluronic /]]]]]]]]]]DLE(%) 5 Amount of IMC-loaded nanospheresPCL block copolymeric nanospheres were measured

with a dynamic light scattering (DLS) spectrometer IMC]]]]]3 100 5 3 100 (3)(ELS-800, Photal, Otsuka Electronics, Japan) IMC 1 polymer

equipped with a He–Ne laser beam at a wavelengthof 633 nm at 208C. The intensity of the scattered 2.7. Effect of temperature on the size of Pluronic /light was detected at 908 to the incident beam. PCL nanospheresAnalyzing the DLS data using the cumulant method,we obtained the information about particle size. The In order to investigate the temperature sensitivity


of Pluronic /PCL block copolymeric nanospheres, maintained in Dulbecco’s modified Eagle mediumtheir size and size distribution were measured with (DMEM) supplemented with 10% fetal bovinechanging the temperature of nanosphere solutions. serum (FBS) and L-glutamine under 5% CO atmos-2

They were heated from 20 to 708C at a speed of phere at 378C. The cytotoxicity test was carried out58C/h. The size and size distribution of nanospheres while the cells were in a phase of exponential growthwere measured in every 108C by dynamic light with a doubling time of 48 h, and starting cell

3scattering measurement. After heating up to 708C, density in a 96-well microplate was 5310 /well.the solutions were then cooled to 108C at the same The viability was chosen as cytotoxicity parameterspeed while measuring their size and size distribution and determined using the 3-(4,5-dimethylthiazol-2-in every 208C increment. yl)-2,5-diphenyl tetrazolium bromide (MTT) assay

In addition, the response of nanospheres upon [26]. The fibroblast cells were incubated in DMEMrepetitive heating and cooling processes was ex- added with 10% FBS containing different concen-amined. Pluronic /PCL nanosphere solution was trations of IMC-loaded Pluronic /PCL nanospheres orheated from 25 to 408C, and the solution was cooled free IMC as a control, in a 96-well microplate for 72to 258C. That is, after altering the temperature at a h under 5% CO atmosphere at 378C. After the2

speed of 108C/h in range of 25–408C, we main- incubation period, MTT solution was added andtained that temperature for more than 30 min and further incubation for 4 h was performed. Thethen measured the size of Pluronic /PCL nanospheres Pluronic /PCL nanospheres–MTT mixture was re-at each temperature. These heating and cooling moved and isopropanol was added. After slowprocedures were repeated to monitor the reproduci- agitation for 5 min, the absorbance was determinedbility of their size and size distribution. at 540 nm using a microplate reader (EL310, Biotec).

The viability was expressed as a percentage com-2.8. In vitro drug release studies pared to a control untreated with IMC-loaded

Pluronic /PCL nanospheres or free drug by theThe release behaviors of IMC from Pluronic /PCL following equation.

block copolymeric nanospheres were investigated inViability (%) 5 (N /N ) 3 100 (4)t cvitro. An appropriate amount of IMC-loaded nanos-

phere was precisely weighed and suspended in 5 ml where N and N are the number of surviving cells int cof a phosphate-buffer solution (PBS, 0.1 M, pH 7.4). the treated group with IMC-loaded nanospheres orThe nanosphere solution was introduced into dialysis free IMC and in the untreated group, respectively.membrane bag and was placed in 250 ml of PBSrelease media with stirring. To investigate the tem-perature-responsive release profiles of drug from

3. Results and discussionnanospheres, the temperature of media was changedafter every 6 h from 15 to 458C until 24 h. After 24h, it was changed in every 24 h. At predetermined 3.1. Preparation of drug-loaded Pluronic /PCLtime intervals, 3-ml aliquots of the aqueous solution nanosphereswere withdrawn from the release media. After theconcentration of IMC released was monitored using We synthesized the Pluronic /PCL block copoly-UV spectrophotometric measurement, the solution mers by a ring-opening mechanism of ´-caprolactoneused as a sample was replaced with a fresh release unit in the presence of poly(ethylene oxide)-poly-media. (propylene oxide)-poly(ethylene oxide) triblock co-

polymer, Pluronic, using stannous octoate as a2.9. In vitro cytotoxicity test catalyst as described in our previous study [22]. Fig.

1 shows the mechanism for this coordination poly-The biological properties of Pluronic /PCL block merization and the chemical structure of the final

copolymeric nanospheres were determined by cyto- copolymer. The active hydrogen atom at one end oftoxicity test. Monolayer culture of normal human chains of Pluronic acted as an initiator and induced afibroblast cells (KCLB No. 21947, CCD-98sk) were selective acyl-oxygen cleavage of ´-caprolactone.


Fig. 2. Typical size distribution profile of Pluronic /PCL blockcopolymeric nanospheres (CF87-1) by dynamic light scattering:(a) unloaded nanospheres, (b) IMC-loaded nanospheres.

1sured by GPC and H NMR are listed in Table 1. Asalready reported in our previous study [22], weconfirmed the synthesis of Pluronic /PCL copolymer

1by structural analyses such as FT-IR, GPC, H NMR,wide-angle X-ray diffraction (WAXD), differentialscanning calorimetry (DSC) and thermogravimetricanalysis (TGA).

Drug-unloaded nanospheres and IMC-loadedFig. 1. Scheme of the synthesis of Pluronic /PCL block co- nanosphere with Pluronic /PCL block copolymerpolymer. synthesized were prepared by a dialysis method. The

size and size distribution of Pluronic /PCL blockcopolymeric nanospheres were determined by dy-

The type of Pluronic used, the feed ratio of namic light scattering (DLS). Fig. 2 shows thePluronic and ´-caprolactone monomer, and the com- typical size distribution of Pluronic /PCL blockpositions of Pluronic /PCL block copolymers mea- copolymeric nanospheres (CF87 sample) before and

Table 1Composition and molecular weight distribution of Pluronic /PCL block copolymers

Sample Feed weight composition Molecular weight Poly-dispersiya d(PCL/Pluronic /PCL) (M /M )b c d w nCalc. M Mn w

CF127 0.5:1:0.5 25 200 20 800 22 227 1.62CF87 0.5:1:0.5 15 400 12 786 14 004 1.44CF87-1 1.35:1:1.35 25 200 25 596 27 700 1.32CF88 0.5:1:0.5 22 800 17 844 19 412 1.46CF68 0.5:1:0.5 16 800 13 259 13 763 1.63CF68-1 0.86:1:0.86 22 800 28 808 24 438 1.32

b Calculated by feed compositions.c 1Determined by H NMR spectroscopy (CDCl ).3d Measured by GPC analysis.a Pluronic (PEO-PPO-PEO block copolymer): F-127, M 512 600, weight portion PEO block in total570%; F-87, M 57700, weightn n

portion PEO block in total570%; F-88, M 511 400, weight portion PEO block in total580%; F-68, M 58400, weight portion PEO blockn n

in total580%.


Table 2The size and drug loading efficiency of Pluronic /PCL block copolymeric nanospheres

b cSample Unloaded nanospheres Loaded nanospheres Drug loading efficiency (%)aSize (nm) S.D. (n55) Size (nm) S. D. (n55) Ave. S.D. (n53)

CF127 160.8 6.48 183.4 9.67 14.18 5.00CF87 132.6 2.70 174.4 11.28 17.15 1.47CF87-1 133.5 5.16 179.9 3.30 18.64 4.85CF88 140.4 11.67 152.8 6.16 16.27 2.95CF68 116.4 3.21 142.9 11.12 13.82 2.13CF68-1 192.0 9.17 195.0 6.80 14.41 3.04

a Determined by DLS measurement at 208C.b Feed weight ratio of polymer:indomethacin, 1.00:1.00.c DLE was calculated by Eq. (3).

after the loading of drug at 208C. Results are Fig. 3 shows that sub-micron sized particles of CF-summarized in Table 2. 87 sample had been prepared as evidenced by FE-

As shown in Fig. 2a, the average diameter of SEM image.drug-unloaded Pluronic /PCL nanospheres was less As can be seen in Fig. 2b and Table 2, the size ofthan 200 nm. Their size distribution is narrow nanospheres increases somewhat after IMC loadingindicating that all the block copolymer chains exist in the nanospheres. For example, the particle size ofas uniform nanospheres. Particle sizes of drug-un- IMC-loaded CF87-1 nanospheres (179.9 nm) isloaded Pluronic /PCL block copolymeric nanos- larger than that of unloaded CF87-1 nanospherespheres were in the range of 116–196 nm, and the (133.5 nm) obviously because of the loadingsize of nanospheres increased with the molecular amounts of drug. However, the size distribution wasweight of Pluronic /PCL block copolymers as de- not markedly influenced by the loading of a drug andscribed in Table 2. relatively similar to that before loading IMC.

Furthermore, Pluronic /PCL block copolymeric The amount of IMC incorporated in Pluronic /PCLnanospheres were observed with field emission scan- nanospheres was obtained from UV absorbance atning electron microscopy (FE-SEM) measurement. 319 nm. DLE calculated from Eq. (3) was summa-

Fig. 3. Field emission scanning electron microphotographs of Pluronic /PCL block copolymeric nanospheres (CF87): (a) 310 000; (b)350 000.


rized in Table 2. DLE values of Pluronic /PCL block as a hydrophobic group to Pluronic block as ananospheres were in the range of 13.8–18.6%. hydrophilic group is the same, it could be expected

In general, micellization of block copolymers because the PCL block length of CF127 (two PCLfollows the model of closed association, which block length6111 135; from the GPC results) isassumes a dynamic equilibrium between free co- absolutely greater than that of CF87 (two PCL blockpolymer chain and particles (micelles) with definite length67002, from the GPC results). From theseassociation degree (association number). Block co- results, it was concluded that the onset of micelliza-polymers dissolved in a selective solvent, thermo- tion was dominated mainly by the length of hydro-dynamically good solvent for one of the blocks and phobic block. These results are in a good agreementprecipitant for the other, form well-defined spherical with that of other studies on the thermodynamics ofassociates called micelles. The micelles consist of a micellization for block copolymers [23,24].

1relatively compact core formed by the least-soluble In addition, we carried out a H NMR analysis toblocks surrounded by a flexible and highly swollen estimate the structure of Pluronic /PCL block co-

1shell formed by the other blocks. polymeric nanospheres. High-resolution H NMRIn the present study, fluorescence spectroscopy spectroscopy has often been employed to analyze the

measurement was performed to investigate the for- microstructure of polymer chains in solution. Liu,mation of Pluronic /PCL block copolymeric nanos- Ando and others have investigated the interactionspheres. We measured the excitation spectra of pyrene between polymer chains and low-molecular weightat 390 nm of emission wavelength, and could materials in solution [27,28]. Nakamura et al. re-observe the shift of wavelength in the excitation ported the molecular motion of block copolymers in

1spectra with increasing concentration of block co- solution by high-resolution H NMR study [29].1polymer. In the plots, the intensity ratio of I /I Fig. 4 shows the H NMR spectra of IMC-loaded343 336

from pyrene excitation spectra against the log C of Pluronic /PCL block copolymeric nanospheres inPluronic /PCL block copolymers, CMC value was CDCl (a), IMC in CDCl (b) and IMC-loaded3 3

taken from the intersection of the tangent to the Pluronic /PCL block copolymeric nanospheres incurve at the inflection with the horizontal tangent D O (c). IMC-loaded Pluronic /PCL block copoly-2

through the points at low polymer concentrations. meric nanospheres were completely dissolved in24CMC values were calculated to be 1.630310 , CDCl solvent. Therefore, we could clearly observe3

25 266.289310 and 2.503310 M for CF87, CF127 the peaks assigned to methylene proton of poly(´-and CF87-1, respectively. It was observed that the caprolactone) units, peaks due to the PEO and PPOCMC value of CF87-1 was lower than that of CF87. blocks in Pluronic units, and peaks assigned to IMCSince the CF87-1 (feed weight composition, PCL/ in the spectrum (Fig. 4a) of IMC-loaded Pluronic /

1Pluronic /PCL51.35:1:1.35) has longer hydrophobic PCL nanosphere in CDCl . For the H NMR spec-3

chains than that of CF87 (feed weight composition, trum of IMC-loaded Pluronic /PCL block copoly-PCL/Pluronic /PCL50.5:1:0.5), the long PCL block meric nanospheres in D O (c), however, only chemi-2

of CF87-1 causes it to form micelle structures easier cal shifts due to PEO and PPO units in Pluronic werethan CF87. Therefore, CF87-1 block copolymer resolved. A sharp singlet at |3.63 ppm is due tocould be formed into micelles in much lower con- protons of CH CH units of PEO blocks. A doublet2 2

centration than CF87. Also, if we compare the at |1.13 ppm belongs to protons of CH groups in3

CF127 (M measured by GPC analysis522 227; PPO block. Broad peaks between 3 and 4 ppm comew

feed composition ratio, PCL/Pluronic /PCL5 from CH CH units of PPO block. From these results,2

0.5:1:0.5; as shown in Table 1) and CF87 (M it is presumed that the conformation of the am-w

measured by GPC analysis514 004; feed composi- phiphilic diblock copolymer in CDCl is of the3

tion ratio, PCL/Pluronic /PCL50.5:1:0.5), which mixed-microphase form in which both Pluronic andhave the same composition ratio of PCL to Pluronic poly(´-caprolactone) blocks are dissolved. In D O,2

within the copolymer and different molecular however, the amphiphilic block copolymer is in aweights, the CMC value of CF127 is lower than that separated-microphase form consisting of a solubleof CF87. Although the composition ratio of PCL Pluronic block part and insoluble poly(´-caprolac-


1Fig. 4. H NMR spectra of IMC-loaded Pluronic /PCL block copolymeric nanospheres (CF87) in CDCl (a), free IMC in CDCl (b),3 3

IMC-loaded Pluronic /PCL block copolymeric nanospheres (CF87) in D O (c).2

tone) block part. It was concluded that IMC-loaded hydrophilic water-soluble blocks and water-insolublePluronic /PCL block copolymeric nanospheres ex- block allows many variations of the total molecularisted as a core–shell-like structure composed of weight of the surfactants, as well as the ratio of thehydrophobic core of poly(´-caprolactone) and hydro- weight of the hydrophile to the hydrophobe [30–33].philic outer shell of Pluronic in water. At temperatures close to ambient, it is well known

that the highly concentrated polymer solutions ex-3.2. Effect of temperature on the size of Pluronic / hibit a dramatic change in viscosity, revealing aPCL nanospheres ‘thermoreversible gelation’. Particularly, aqueous

solutions of the polymer in concentrations aboveThe Pluronics are PEO-PPO-PEO triblock copoly- 20% (w/w) exhibit the phenomenon of reverse

mers in which the hydrophobe consists of a pro- thermal gelation, remaining as solutions at refriger-pylene oxide block and the hydrophile is made of ated temperatures and gelling upon warming totwo ethylene oxide blocks. This arrangement of the ambient levels [30–33].


A number of studies have been carried out in an tion behavior of surfactants in water has beenattempt to describe such aggregates of PEO-PPO- extensively investigated by a number of experimentalPEO block copolymers in water, and several mecha- techniques leading to a basic understanding ofnisms have been proposed as driving forces for this micellar size and shape effects. Especially, thethermal gelation. Rassing and Attwood related the structure changes of surfactant solutions as functionsgel transition to intrinsic changes in the micellar of temperature and concentration has been investi-properties [34]. Alexandridis et al. discussed the gated using the NMR self diffusion and protongelation in terms of entropic changes, involving relaxation studies. To examine the structure changelocally ordered water molecules close to the hydro- of Pluronic /PCL nanosphere depending on the tem-

1phobic units [30], whereas Mortensen and Pedersen perature, we performed the H NMR measurement atspeculated the possibility of an ordered three dimen- every 108C with varying the temperature from 20 tosional structured state or network [32]. 908C. NMR spectra for Pluronic /PCL nanospheres

To determine the thermal response behavior of the in D O at various temperature exhibited the constant2

present amphiphilic block copolymer, we studied the linewidths at half signal height for the methylenesize change of Pluronic /PCL block copolymeric signal of the PEO block and proton signals due tonanospheres as a function of temperature. Fig. 5 CH groups and CH CH units in PPO chain without3 2

shows the size of Pluronic /PCL block copolymeric broadening of peaks indicating the growth of aggre-nanospheres depending on the temperature. As gates. Also, these results were supported by theshown in Fig. 5, their size markedly decreased with studies about the thermodynamics and structure ofincreasing the temperature regardless of their drug micelle association in similar copolymer systemsloading. [36–38]. They reported that block copolymers under-

The size of micelles composed of PEO-PPO-PEO go closed association in dilute solutions to formblock copolymer generally increases with increasing micelles and that the association number of micellestemperature [32,35], the extent of the increase vary- is independent of temperature different from com-ing between different copolymers; indeed for some mon surfactant molecules.systems, the hydrodynamic size of the micelle re- Therefore, in the present Pluronic /PCL blockmains constant while the aggregation number in- copolymeric nanospheres, which are composed ofcreases with temperature probably as a result of relatively hydrophilic center of PEO-PPO-PEO blockdehydration of the PEO head groups. The aggrega- and two hydrophobic end blocks of poly(´-caprolac-

tone), it seems likely that the interaction betweenPPO block and poly(´-caprolactone) block shouldhave influenced the reverse tendency of size versustemperature in comparison with the Pluronic micelleitself. Namely, the solubility and hydrophilicity ofPPO block markedly decreases with increasing tem-perature, resulting in an increase of interactionbetween the PPO block in the outer shell structureand poly(´-caprolactone) in the inner micellar core.Owing to the increase of inter-molecular and intra-molecular interactions between PPO and PCL as afunction of temperature, the shrinkage and the thecompact structure by the strong chain–chain aggre-gation could be induced. As a result, it could beconsidered that the size of Pluronic /PCL blockcopolymeric nanospheres decreased with increasingtemperature. Further study on the exact mechanismFig. 5. Effect of temperature on the size of Pluronic /PCL blockon the temperature dependency of this Pluronic /PCLcopolymeric nanospheres (CF87-1): (a) IMC-loaded nanospheres,

(b) unloaded nanospheres. nanosphere is under way.


Temperature-sensitive drug delivery systems,which exhibited a slow release of drug at elevatedtemperature as opposed to ambient temperaturerange, have been studied. Okano and co-workershave reported on the thermo-sensitive release sys-tems with such temperature-dependent release ten-dency using interpenetrating polymer network com-posed of poly(N-isopropylacrylamide) [7,8].

From the distribution profiles of nanospheres ateach temperature, we can see that the ordinary sizedistribution is maintained. It indicates that all theblock copolymer chains exist as uniform nanospheresand the structures of nanospheres are not destroyedduring the change of environmental temperature.

Fig. 7. Release behaviors of indomethacin from Pluronic /PCLFig. 6 shows the change in the size of Pluronic /copolymeric nanospheres between 15 and 458C.PCL block copolymeric nanospheres (loaded CF68-

1) according to the repetitive thermal cycles. Duringthe change of temperature, nanospheres showed the significant sustained release characteristics of lessreversible change in size, which were similar to the than 30% for about 100 h.results in Fig. 5. Moreover, in spite of the repetition Fig. 7 describes a release profile of indomethacinof thermal processes, their particle size distribution (IMC) from Pluronic /PCL block copolymeric nanos-did not change much. pheres in vitro. It plots the relative release per-

centage of indomethacin based on the loading3.3. Release behaviors of drug depending on amount with altering temperature between 15 andtemperature 458C as a function of time. According to this figure,

the release amount of IMC from Pluronic /PCLIn vitro release experiment of IMC from Pluronic / nanosphere decreased with increasing temperature.

PCL nanospheres was performed. While the un- Generally, the drug release pattern depends onloaded free indomethacin showed a rapid release of many factors, including particle size, crystallinity,more than 97% within 36 h (not shown here), the surface character, molecular weight, polymer com-IMC loaded in the inner core of nanospheres showed position, swelling ratio, degradation rate, drug bind-

ing affinity and the rate of hydration of the polymericmaterials, etc. [39].

In the release behaviors of this Pluronic /PCLblock copolymeric nanosphere system, we can con-sider that the binding affinity of drug and polymer,and the degradation of polymer. However, the degra-dation of the present polymer might not largelyaffect the release behavior of a drug because poly(´-caprolactone) degrades quite slowly in this system[2]. It is concluded that the effect of binding affinityof IMC in nanospheres plays an important role invitro release profiles.

Because nanospheres show temperature sensitivi-ty, its particle size decreased at high temperatureresulting in a difficulty in drug diffusion throughnanospheres. In addition, the solubility and hydro-Fig. 6. Reversible temperature-responsive size change of

Pluronic /PCL copolymeric nanospheres (loaded CF68-1). philicity of PPO block remarkably decreased as the


temperature increased, resulting in an increase ofboth the interaction of PPO block composing outershell structure and poly(´-caprolactone) in the innermicellar core and the binding affinity between in-domethacin with moderate hydrophobic characterand polymer chain. Therefore, the decrease of therelease of IMC from nanosphere at high temperaturecould be induced. On the other hand, at lowertemperature, the reverse trend was observed and thereleasing amount of IMC was pretty much. As canbe seen in Fig. 7, this thermal responsive release ofIMC from Pluronic /PCL block copolymeric nanos-pheres exhibited a reversible behavior by repetitivetemperature change.

3.4. Cytotoxicity of Pluronic /PCL nanospheres inFig. 8. Cytotoxicity of IMC-loaded Pluronic /PCL block copoly-

vitro meric nanospheres after 3-day incubations on normal humanfibroblast cell cultures.Viability (%)5(N /N )3100, where N andt c t

N are the number of surviving cells in the treated group withThe biocompatibility of Pluronic /PCL block co- c

IMC-loaded nanospheres or free IMC and in the untreated group,polymeric nanospheres was evaluated in vitro byrespectively.cytotoxicity test using normal human fibroblast cells.

It compared with the cytotoxicity induced by freedrug which was not loaded into the polymeric composed of hydrophilic PEG block could reducenanospheres. The amount of surviving cells after the interaction between nanospheres themselves andincubation was estimated by MTT assay. The viabili- between nanospheres and cells by forming the stealthty was expressed in percent compared to a control surface. Second, the sustained release characteristicnot treated with IMC-loaded MePEG/glycolide of IMC from Pluronic /PCL block copolymericnanospheres or free drug. After 3-day incubations in nanospheres could help to maintain the viability ofnormal human cell cultures, viability of sample was cells.assayed. The viability of CF87 sample with 17.15%DLE and unloaded free IMC were exhibited in Fig.8. The free IMC exhibited higher cytotoxic activity 4. Conclusionsthan Pluronic /PCL block copolymeric nanospherescontaining IMC at the same concentration of IMC. Amphiphilic block copolymers based on PEO-This tendency was enhanced with the concentration PPO-PEO block copolymer (Pluronic) and poly(´-of IMC. Particularly, IMC-loaded nanospheres caprolactone) were synthesized by ring-opening(CF87 sample) showed high viability of about 80% polymerization of ´-caprolactone in the presence ofor above, whereas the free IMC exhibited the PEO-PPO-PEO block copolymer having hydroxylviability of less than 50% at the concentration of groups at two ends of chains using stannous octoate0.064 mg/ml. Its viability gradually decreased in as a catalyst. IMC-loaded and unloaded nanospheresproportion to the concentration. using Pluronic /PCL block copolymer with different

An improvement of viability by introducing drug composition were prepared by dialysis process ininto polymeric nanospheres could be explained with deionized water. Drug-unloaded Pluronic /PCL blockthe interaction of IMC with cells and the release copolymeric nanospheres showed an average diam-characteristic of IMC from polymeric nanospheres. eter of 116–196 nm, depending on the type ofFirst, by loading IMC into the core of polymeric copolymer. Their size increased with the molecularnanospheres, direct interactions between IMC and weight of block copolymers. All the nanospherecells could be reduced. Furthermore, the outer shell samples exhibited a narrow size distribution. After


IMC loading in the nanospheres, their size increases SYK and JCH are grateful to the Graduate School ofsomewhat. However, the size distribution was not Advanced Materials and Chemical Engineering atmarkedly influenced by the loading of drug and was Hanyang University for a fellowship.relatively similar to that before loading IMC. DLEvalues of Pluronic /PCL nanospheres were in therange of 13.8–18.6%. References

The critical micelle concentrations of Pluronic /PCL amphiphilic block copolymers determined by [1] A.G. Mikos, R.M. Murphy, H. Bernstein, N.A. Peppas,

Biomaterials for drug and cell delivery, in: Materials Re-fluorescence spectroscopy were lower than that of1 search Society Symposium Proceedings, Vol. 331, Materialscommon low molecular weight surfactant. From H

Research Society, USA, 1994.NMR analysis, it was confirmed that IMC-loaded [2] M. Chasin, R. Langer, Drug and the pharmaceutical sciences,Pluronic /PCL block copolymeric nanospheres ex- in: Biodegradable Polymers as Drug Delivery System, Vol. 3,isted as a core–shell-like structure composed of a Marcel Dekker, New York, 1990, pp. 71–120.

[3] R.L. Juliano, Targeted drug delivery, in: Handbook ofhydrophobic core of poly(´-caprolactone) and aExperimental Pharmacology, Vol. 100, Springer-Verlag, Ber-hydrophilic outer shell of Pluronic in water.lin, 1991.

Pluronic /PCL block copolymeric nanospheres ex- [4] H.E. Junginger, in: Drug Targeting and Delivery; Concept inhibited a change of size depending on the tempera- Dosage Form Design, Ellis Horwood, Chichester, UK, 1992.ture. As temperature increased, their size gradually [5] K. Park, Controlled drug delivery: challenges and strategies,

in: ACS Professional Reference Book, ACS, Washington,decreased. This could be induced by the strongDC, 1997, pp. 49–67, Ch. 4.chain–chain aggregation due to the increase of inter-

[6] H. Katono, A. Maruyama, K. Sanui, N. Ogata, T. Okano, Y.molecular and intra-molecular interaction due to Sakurai, Thermo-responsive swelling and drug releasechange of the solubility and hydrophilicity of the switching of interpenetrating polymer networks composed ofPPO block. Moreover, this change of size exhibited a poly(acrylamide-co-butylmethacrylate) and poly(acrylic

acid), J. Control. Release 16 (1991) 215–228.reversible tendency according to the repetitive ther-[7] T. Okano, Y.H. Bae, H. Jacobs, S.W. Kim, Thermally on–offmal cycles. Release behaviors of indomethacin from

switching polymers for drug permeation and release, J.Pluronic /PCL block copolymeric nanospheres also Control Release 11 (1990) 255–265.showed temperature dependence and sustained re- [8] K. Mukae, Y.H. Bae, T. Okano, S.W. Kim, A Thermo-lease pattern. The amount of IMC released from sensitive hydrogel: poly(ethylene oxide-dimethylsiloxane-

ethyloxide) /poly(N-isopropyl acrylamide) interpenetratingPluronic /PCL nanospheres decreased with increasingpolymer networks II. On–off regulation of solute releasetemperature. Particle size decreased at higher tem-from thermo-sensitive hydrogel, Polymer J. 2 (1990) 250–

peratures, resulting in a difficulty in drug diffusion 265.through nanospheres. In addition, the solubility and [9] L. Brannon-Peppas, Recent advances on the use of bio-hydrophilicity of PPO block significantly decreased degradable microparticles and nanoparticles in controlled

drug delivery, Int. J. Pharm. 116 (1995) 1–9.as the temperature increased. Also, in vitro cytotox-[10] J. Kreuter, Nanoparticle-based drug delivery systems, J.icity test using an MTT assay method showed that

Control. Release 16 (1991) 169–176.these IMC-loaded Pluronic /PCL nanospheres could [11] G. Cavallaro, M. Fresta, G. Giammona, G. Puglish, A.reduce cell damage compared with unloaded free Villari, Entrapment of b-lactams antibiotics in poly-IMC. Therefore, we could expect that these Pluronic / ethylcyanoacrylate nanoparticles: Studies on the possible in

vivo application of this colloidal delivery system, Int. J.PCL amphiphilic block copolymeric nanospheresPharm. 111 (1994) 31–41.would exhibit temperature-responsive behavior and

[12] T. Niwa, H. Takeuchi, T. Hino, M. Nahara, Y. Kawashima,could be useful as a drug carrier for targeted drug Biodegradable submicron carriers for peptide drugs: Prepara-delivery. tion of DL-lactide.glycolide copolymer (PLGA) nanospheres

with nafarelin acetate by a novel emulsion-phase separationmethod in an oil system, Int. J. Pharm. 121 (1995) 1123–1130.Acknowledgements

[13] M. Fresta, G. Cavallaro, G. Giammona, E. Wehrli, G. Puglisi,Preparation and characterization of polyethyl-2-

This research was supported by Korea Science and cyanoacrylate nanocapsules containing antiepileptic drugs,Engineering Foundation Grant [96-0300-12-01-3. Biomaterials 17 (1996) 751–758.


[14] A. Memoli, L. Palermiti, V. Travagli, F. Alhaique, Egg and [27] K. Liu, W. Burlant, High-resolution nuclear magnetic reso-soya phospholipids-sonication and dialysis: a study on nance of crosslinked polymers: Effects of crosslinked densityliposome characterization, Int. J. Pharm. 117 (1995) 159– and solvent interaction, J. Polym. Sci. Part A-1 5 (1967)163. 1407–1413.

[15] G. Kwon, S. Suwa, M. Yokoyama, T. Okano, Y. Sakurai, K. [28] I. Ando, A. Nishioka, Magnetic anisotropy effect on theKataoka, Enhanced tumor accumulation and prolonged circu- chemical shift in linear hadrocarbons, Macromol. Chem. 152lation times of micelle-forming poly(ethylene oxide-aspar- (1972) 7–14.tate) block copolymer-adriamycin conjugates, J. Control. [29] K. Nakamura, R. Endo, M. Takeda, Study of molecularRelease 29 (1994) 17–23. motion of block copolymers in solution by high-resolution

[16] C. Verdun, P. Couvreur, H. Vrancks, V. Lenaerts, M. Roland, proton magnetic resonance, J. Polym. Sci.: Polym. Phys. Ed.Development of a nanoparticle controlled release formulation 15 (1977) 2095–2101.for human use, J. Control. Release 4 (1986) 205–210. [30] P. Alexandridis, J. Holzwarth, T. Hatton, Micellization of

[17] K.C. Dowling, J.K. Thomas, A novel micellar synthesis and poly(ethylene oxide)-poly(propylene oxide)-poly(ethylenephotophysical characterization of water-soluble acrylamide- oxide) triblock copolymers in aqueous solutions: thermo-styrene block copolymers, Macromolecules 23 (1990) 1059– dynamics of copolymer association, Macromolecules 271064. (1994) 2414–2425.

[18] I.G. Shin, S.Y. Kim, Y.M. Lee, Methoxy poly(ethylene [31] M. Malmsten, B. Lindman, Self-assembly in aqueous blockglycol) /´-caprolactone amphiphilic block copolymeric mi- copolymer solutions, Macromolecules 25 (1992) 5440–5445.celle containing indomethacin: I. Preparation and characteri- [32] K. Mortensen, J.S. Pedersen, Structural study on the micellezation, J. Control. Release 51 (1998) 1–11. formation of poly(ethylene oxide)-poly(propylene oxide)-

[19] S.Y. Kim, I.G. Shin, Y.M. Lee, Methoxy poly(ethylene poly(ethylene oxide) triblock copolymer in aqueous solution,glycol) /´-caprolactone amphiphilic block copolymeric mi- Macromolecules 26 (1993) 805–812.celle containing indomethacin.: II. Micelle formation and [33] J. Gilbert, J. Richardson, M. Davies, K. Palin, J. Hadgraft,drug release behaviors, J. Control. Release 51 (1998) 13–22. The effect of solutes and polymers on the gelation properties

[20] S.Y. Kim, I.G. Shin, Y.M. Lee, Preparation and characteriza- of Pluronic F-127 solutions for controlled drug delivery, J.tion of biodegradable nanospheres composed of methoxy Control. Release 5 (1987) 113–118.poly(ethylene glycol) and D,L-lactide block copolymer as [34] J. Rassing, D. Attwood, Ultrasonic velocity and light-scatter-novel drug carriers, J. Control. Release 56 (1998) 197–208. ing studies on the polyoxyethylene-polyoxypropylene co-

[21] S.Y. Kim, I.G. Shin, Y.M. Lee, Amphiphilic diblock copolymer Pluronic F-127 in aqueous solution, Int. J. Pharm.polymeric nanospheres composed of methoxy poly(ethylene 13 (1983) 47–55.glycol) and glycolide: properties, cytotoxicity and drug [35] B. Lindman, A. Carlsson, G. Karlstrom, M. Malmsten,release behaviour, Biomaterials 20 (1999) 1033–1042. Nonionic polymers and surfactants-some anomalies in tem-

[22] J.C. Ha, S.Y. Kim, Y.M. Lee, Poly(ethylene oxide)-poly- perature dependence and in interactions with ionic surfac-(propylene oxide)-poly(ethylene oxide) (Pluronic) /poly(´- tants, Adv. Colloid Interfaces Sci. 32 (1990) 183–203.caprolactone) (PCL) amphiphilic block copolymeric nanos- [36] J.R. Quintana, M. Villacampa, I.A. Katime, Micellization ofpheres: I. Preparation and characterization, J. Control. Re- a polystyrene-b-poly(ethylene /propylene) block copolymerlease (in press). in n-dodecane /1,4-dioxane mixture. 1. Thermodynamics of

[23] I. Astafieva, X.F. Zhong, A. Eisenberg, Critical micellization micellization, Macromolecules 26 (1993) 601–605.phenomena in block polyelectrolyte solutions, Macromole- [37] C. Price, A.L. Hudd, R.B. Stubbersfield, B. Wright, A studycules 26 (1993) 7339–7352. of micelle formation by a poly(ethylene /propylene) block

[24] Z. Gao, A. Eisenberg, A model of micellization for block copolymer in a base lubricating oil, Polymer Commun. 21copolymers in solutions, Macromolecules 26 (1993) 7353– (1980) 9–10.7360. [38] J.R. Quintana, R.A. Salazar, I. Katime, Micelle formation

[25] K. Kalyanasundaram, J.K. Thomas, Environmental effects on and polyisobutylene solubilization by polystyrene-block-vibronic band intensities in pyrene monomer fluorescence poly(ethylene-co-butylene)-block-polystyrene block copoly-and their application in studies of micellar systems, J. Am. mers, Macromol. Chem. Phys. 196 (1995) 1625–1634.Chem. Sci. 99 (1977) 2039–2044. [39] B.D. Ratner, A.S. Hoffman, F.J. Schoen, J.E. Lemons, in:

[26] P. Cerrai, G. Guerra, L. Lelli, M. Tricoli, R. Guerra, M. Biomaterials Science, Academic Press, New York, 1996, pp.Cascone, P. Giusti, Poly(ester-ether-ester) block copolymers 347–356.as biomaterials, J. Mater. Sci. Mater. Med. 5 (1994) 33–39.

Documents

Poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide)/poly(ϵ-caprolactone) (PCL) amphiphilic block copolymeric nanospheres: II. Thermo-responsive drug release behaviors