Micellar formulations for drug delivery based on mixtures of hydrophobic and hydrophilic Pluronic® block copolymers

Journal of Controlled Release 94 (2004) 411 – 422 www.elsevier.com/locate/jconrel

Micellar formulations for drug delivery based on mixtures of hydrophobic and hydrophilic PluronicR block copolymers Kyung T. Oh, Tatiana K. Bronich, Alexander V. Kabanov * Department of Pharmaceutical Science, College of Pharmacy, 986025 University of Nebraska Medical Center, Omaha, NE 68198-6025, USA Received 18 August 2003; accepted 25 October 2003

Abstract Micelles formed by PluronicR block copolymers (PBC) have been studied in multiple applications as drug delivery systems. Hydrophobic PBC form lamellar aggregates with a higher solubilization capacity than spherical micelles formed by hydrophilic PBC. However, they also have a larger size and low stability. To overcome these limitations, binary mixtures from hydrophobic PBC (L121, L101, L81, and L61) and hydrophilic PBC (F127, P105, F87, P85, and F68) were prepared. In most cases, PBC mixtures were not stable, revealing formation of large aggregates and phase separation within 1 – 2 day(s). However, stable aqueous dispersions of the particles were obtained upon (1) sonication of the PBC mixtures for 1 or 2 min or (2) heating at 70 jC for 30 min. Among all combinations, L121/F127 mixtures (1:1% weight ratio) formed stable dispersions with a small particle size. The solubilizing capacity of this system was examined using a model water-insoluble dye, Sudan (III). Mixed L121/F127 aggregates exhibited approximately 10-fold higher solubilization capacity compared to that of F127 micelles. In conclusion, stable aqueous dispersions of nanoscale size were prepared from mixtures of hydrophobic and hydrophilic PBC by using the external input of energy. The prepared mixed aggregates can efficiently incorporate hydrophobic compounds. D 2003 Elsevier B.V. All rights reserved. Keywords: Block copolymer; Drug delivery; PluronicR; Poloxamer; Solubilization

1. Introduction Micelles formed from amphiphilic block copolymers have recently attracted significant attention in diverse fields of medicine and biology. In particular, polymeric micelles have been developed in pharmaceutics as drug and gene delivery systems [1 –3], as well as in diagnostic imaging techniques as carriers * Corresponding author. Tel.: +1-402-559-9915; fax: +1-402559-9543. E-mail addresses: [email protected] (T.K. Bronich), [email protected] (A.V. Kabanov). 0168-3659/$ - see front matter D 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.jconrel.2003.10.018

for various contrasting agents [4]. One of the examples, PluronicR block copolymer (PBC), consists of ethylene oxide (EO) and propylene oxide (PO) blocks that are arranged in a basic EOx – POy –EOx structure (often abbreviated as PEO – PPO – PEO). A prominent feature of PBC, specifically related to drug delivery applications, is the ability to self-assemble in aqueous solutions into multimolecular aggregates having spherical, rod-like or lamellar morphologies. The size and morphology of the PBC aggregates strongly depend on the block copolymer composition, specifically, the lengths of EO and PO units as well as the block copolymer concentration, and environmental

412

K.T. Oh et al. / Journal of Controlled Release 94 (2004) 411–422

parameters such as temperature and the quality of the solvent [2,5,6]. The hydrophobic core of such aggregates serves as a microenvironment for the incorporation of lipophilic compounds, while the hydrophilic corona maintains the dispersion stability of the PBC aggregates. Therefore, the noncovalent incorporation of drugs into the hydrophobic PO core of the PBC micelles results in increased solubility, increased metabolic stability and increased circulation time [3,5,7,8]. Although not systematically studied for drug delivery purposes, nonspherical polymeric micelles can provide potential benefits for formulation design. For instance, the lamellar aggregates formed by hydrophobic PBC are likely to exhibit a higher solubilization capacity than spherical micelles formed by hydrophilic PBC [6]. The obvious drawbacks of such systems are the formation of aggregates with a large size, which falls outside of the apparent preferred size range for drug delivery using nanoscale particles (10 – 200 nm) and lack of stability in aqueous dispersion leading to phase separation. Lamellar aggregates formed by PBC with long PO chains and short EO chains usually have larger sizes (ca. 1000 nm). Furthermore, even at low concentrations and at ambient temperatures these systems phase separate. Recently, Schillen et al. reported vesicles from one of the most hydrophobic PBC (L121) obtained by extruding the dispersion of the block copolymer through a 100 nm diameter membrane filter [9]. However, the PBC vesicles prepared in this way had low stability and eventually reverted to the phase separated state. One of the approaches to overcome these limitations is to combine a lamella-forming hydrophobic block copolymer with second hydrophilic block copolymers that can stabilize the mixed aggregates. The steric stabilization of a lipid dispersion by EO chains attached to the surface of lipid bilayer is very well-known [10]. Incorporation of hydrophilic PBC in lipid structures has also been shown to have a steric stabilizing effect. Indeed, previous works reported steric stabilization of lipid vesicles by incorporation of hydrophilic PBC F127 into the lipid membrane [11]. The incorporation of PBC with long EO chains prevents the stacking of lamellae, maintaining a dispersion of the lamellae and ultimately resulting in the formation of stable structures [11].

A mixture of hydrophobic PBC L61 and hydrophilic PBC F127 has already found practical application in pharmaceutics and is currently being evaluated in clinical trials as a doxorubicin delivery system [12]. However, in this mixture hydrophobic L61 was a minor component, which was solubilized in F127 micelles. In the current study, block copolymer mixtures composed of a hydrophobic lamella-forming PBC with very short EO chains and hydrophilic PBC with long EO chains, where hydrophobic PBC is a major or equal component of the mixture, are explored. These mixed PBC form small particles and display elevated stability in dispersion compared to the hydrophobic PBC alone. The factors governing the formation and colloidal stability of such systems and their capacity to solubilize a model hydrophobic compound (Sudan III) are evaluated and discussed.

2. Materials and methods 2.1. Materials PBC were commercially available from BASF Corporation (Parsipanny, NJ) and were used without additional purification. The molecular characteristics of the block copolymers used in this study are presented in Table 1. PBC are designated by their Table 1 Characteristics of PBC [13] Copolymer

MW a

Average number of PO units (NPO)b

Average number of EO units (NEO)b

HLBc

L61 F68 L81 P85 F87 L101 P105 L121 P123 F127

2000 8400 2750 4600 7700 3800 6500 4400 5750 12,600

31.03 28.97 42.67 39.66 39.83 58.97 56.03 68.28 69.40 65.17

4.55 152.73 6.25 52.27 122.50 8.64 73.86 10.00 39.20 200.45

3 29 2 16 24 1 15 1 8 22

a

The average molecular weights provided by the manufacturer. The average numbers of PO and EO units were calculated using the average molecular weights. c HLB values of the copolymers were determined by the manufacturer. b

K.T. Oh et al. / Journal of Controlled Release 94 (2004) 411–422

nomenclature indexes, e.g. ‘F127’ means ‘Pluronic F127.’ Sudan III (Matheson Coleman & BeU) was used as a hydrophobic model drug for solubility studies. All other chemicals used were of analytical grade. 2.2. Preparation of PBC mixtures The experiment design used for preparation of PBC mixtures is presented by the Scheme 1. Cold stock solutions (4 jC) of PBC in distilled water or phosphate buffer saline (PBS, pH 7.4) were mixed in the proportions indicated, and then left in the refrigerator overnight. The solutions were then transferred to room temperature for 12 h. Samples were divided in three parts (5 ml each). One part was used as a no energy input control. Another was sonicated for 1 or 2 min in polystyrene tubes (FALCONR, Becton Dickinson, NJ) using a probe sonicator (Sonicator XL, Misonix Farmingdale, NY) at 55 W. The third part was heated in a water bath at 70 jC for 30 min and then cooled at room temperature. The mixtures obtained were characterized as described below. 2.3. Physicochemical characterization 2.3.1. Turbidity measurements The turbidity experiments were carried out by measuring the transmittance of the mixtures using a Lambda 25 UV/Vis spectrophotometer (Perkin– Elmer

413

instrument Co.) at k = 520 nm. The data are reported as turbidity=(100 T)/100, where T is transmittance (%). 2.3.2. Size measurements The effective hydrodynamic diameter (Deff) of the particles was measured by photon correlation spectroscopy using a ‘‘Zeta-Plus’’ Zeta Potential Analyzer (Brookhaven Instrument Co.) equipped with the Multi Angle Sizing Option (BI-MAS). The sizing measurements were performed in a thermostatic cell at a scattering angle of 90j. Software provided by the manufacturer was used to calculate Deff values. The averaged Deff values were calculated from three measurements performed on each sample (n = 3). 2.4. Solubilization of water-insoluble dye Ten microliters of Sudan III stock solution (10 mg/ ml) in CHCl3 was added to empty vials and the solvent was evaporated. Two milliliter aqueous solutions of F127 and 1:1 wt. L121/F127 mixtures with total PBC concentrations of 0.02, 0.1, 0.2, 0.3 and 0.4 wt.% were individually prepared, sonicated for 2 min, added to the vials and then allowed to equilibrate in the shaker (100 rpm) at room temperature for 2 days. Absorption spectra of Sudan III in aqueous solutions at 25 jC were recorded on a Lambda 25 UV/Vis spectrophotometer. The data are reported as absorbance at 362 nm corresponding to the maximum in the Sudan III spectra.

Scheme 1. Experimental design illustration how the PBC mixtures were prepared and characterized.

414

K.T. Oh et al. / Journal of Controlled Release 94 (2004) 411–422

3. Results and discussions 3.1. Preparation of stable carrier from PBC mixture Binary mixtures were prepared from hydrophobic PBC (L121, L101, L81 and L61) and hydrophilic PBC (F127, P105, F87, P85 and F68) in PBS (pH 7.4). The study of mixture stability was carried out by measuring the size of the particles in the dispersions for several days. In most cases, the PBC mixtures were not stable, revealing formation of large aggregates (ca. 800 – 1000 nm) with a very wide particle size distribution and high turbidity. Ultimately, all samples phase separated within 1 – 2 day(s). However, stable aqueous dispersions of the particles were obtained in selected cases when the energy input (either sonication or heating) was applied during preparation (Table 2). For example, when PBC with long PO, such as L121 or L101,

was used as a hydrophobic component, addition of the hydrophilic PBC (P105 or F127) followed by energy input resulted in a decrease of the turbidity and formation of stable aqueous dispersions of mixed aggregates with rather small particle sizes (ca. 150– 200 nm). The processing parameters (i.e. sonication, temperature) did not significantly affect the polydispersity of the particles formed. The degree of polydispersity of the mixed aggregates varied in the range of 0.26 –0.3. Consequent measurements of these samples showed that the size of the particles remained practically unchanged for several days. Overall, the dispersion stability of the binary PBC systems was dependent on the structure of the PBC. In some cases, no precipitation in the solutions was observed for several days as indicated in Table 2. In other cases, for example, L121 and P85, although the size of particles significantly decreased after sonication, the resulting dispersions were not stable and precipitated

Table 2 The size and stability of the particles in PBC mixtures at various preparation conditions Components

L121 L121/F127 L121/P105 L121/P85 L101 L101/F127 L101/P105 L81 L81/F127 L81/F87 L61 L61/F127 L61/F68 a

% wt. (A/B)a

0.1 0.1/0.01 0.1/0.1 0.1/0.01 0.1/0.1 0.1/0.01 0.1/0.1 0.1 0.1/0.01 0.1/0.1 0.1/0.01 0.1/0.1 0.5 0.5/0.05 0.5/0.5 0.5/0.025 0.5/0.25 1 1.0/0.01 1.0/0.5 1.0/0.05 1.0/0.5

Size of aggregates upon preparations (nm)b

Effect of energy input

Control

Sonication

Heating

Decrease of sizec

1000 1193 1054 873 813 756 811 701 586 578 645 762 1151 416 767 999 964 1044 669 779 916 874

432 322 154 308 138 356 340 337 321 210 279 126 1038 321 812 753 781 976 659 795 611 592

1135 514 198 1018 157 676 244 979 584 213 440 171 801 399 723 1548 448 Precipitation Precipitation Precipitation Precipitation Precipitation

+ + ++ + ++ + ++ + + ++ + ++

Decrease of turbidity

Increase of stability

+

+ (7 days)

+

+ (7 days)

+

+ (5 days)

+

+ (4 days)

Hydrophobic PBC % (wt.) (A)/Hydrophilic PBC % (wt.) (B) (A/B: e.g. L121/F127). Size of the aggregates denote averaged effective hydrodynamic diameter calculated from three measurements performed on each sample (n = 3). c The first and second column indicates the effect of sonication and heating, respectively. b

K.T. Oh et al. / Journal of Controlled Release 94 (2004) 411–422

415

after 2 –3 days. Mixtures prepared with L61 or L81 displayed low stability and high turbidity under all conditions. In addition, the sonication time is an important factor affecting the size and stability of PBC mixtures. In most cases, an increase of sonication time resulted in a smaller size of the particles in the dispersions and an increased stability. However, L121/F127 mixtures sonicated for 3 min formed unstable dispersions, which precipitated after 2 days (data not shown). Sonication of L61 or L81 based mixtures for 2 min also resulted in liquid-phase separation. Based on above results, mixtures of L121 and F127 that were characterized as having both a small particle size and the highest degree of stability were selected for further study. 3.2. Mixture of L121 and F127 L121 and F127 copolymers have average compositions of EO5PO68EO5 and EO100PO65EO100, respectively. Both copolymers have a hydrophobic PPO block of practically the same length, while the length of the PEO blocks is drastically different. It is known that hydrophobic L121 does not form micelles. Even at low concentration and temperatures, unimers coexist with larger unimer aggregates (an L1 phase) [9]. At ambient temperature and low concentration (0.1 wt.%), L121 forms very turbid dispersions. The aggregates detected in such dispersion were approximately 1 Am in diameter (Fig. 1a). In contrast, F127 forms only small spherical aggregates in a wide range of concentrations. The mixtures of L121 and F127 were prepared using methods described above. The concentration of hydrophobic L121 was kept constant at 0.1 wt.%, while the concentration of hydrophilic F127 was varied from 0.01 to 0.1 wt.%. The L121/F127 mixtures prepared without energy input (control group) were very turbid in the entire range of compositions of the mixture. Conversely, in the mixtures prepared with energy input (sonication or temperature increase) the solution turbidity decreased as the concentration of hydrophilic F127 increased (Fig. 1a). The size of mixed PBC aggregates in the control group remained large (ca. 1000 nm), or even increased as the concentration of F127 was elevated. However, sonication or temperature elevation of these mixtures resulted in the formation of significantly smaller aggregates (Fig. 1b). Cryo-TEM images of

Fig. 1. (a) Turbidity of solution and (b) effective diameter of the particles formed in L121/F127 mixtures prepared (.) without any alteration, (n) with sonication for 1 min, (E) with sonication for 2 ) upon elevation of temperature. Deff values are within F min, and ( 20 nm in the cases when sonication and heating are applied.

L121/F127 (0.1%/0.1%) mixed aggregates prepared upon sonication revealed that these aggregates represent spherical particles (data not shown). The sizes of the particles calculated from the EM data were approximately 180 nm and are in a good agreement with those determined by dynamic light scattering. To characterize the stability of all the mixed samples prepared by the previous methods, the size measurements were repeated for several days for each sample. All solutions were stored at room tempera-

416

K.T. Oh et al. / Journal of Controlled Release 94 (2004) 411–422

ture. Once precipitation was detected visually, the sample was considered phase-separated. The shaking of those phase-separated PBC dispersions resulted in resuspension of the particles. However, the mixtures phase-separated again in few hours. Therefore, the first appearance of flakes was considered as the onset of precipitation. All mixtures prepared without energy input phase separated within 1 day. However, an increase in concentration of hydrophilic F127 and input of energy (sonication or temperature increase) resulted in formation of a stable dispersion (Fig. 2). Subsequent measurements of these samples revealed the particle size remained practically unchanged and no precipitation was observed for at least a week. It is important to note that these experiments were performed repeatedly over a period of a year using stock solutions prepared at various times. In all cases PBC mixed aggregates prepared under similar conditions (concentration of PBC, sonication time, annealing temperature) had practically the same sizes. For example, the variability in effective diameters of the particles in 0.1% L121/0.1% F127 mixture was of the order of 20 nm in the cases when sonication or heating was applied. Remarkably, the sonication of mixtures conducted at low temperature (ice bath) did not result in either a particle size decrease or stabilization of the dispersions. It is important to note that the results of the size measurement appeared to be dependent on the time of equilibration of PBC mixture at room temperature

before the input of energy. Specifically, incubation for at least 12 h was necessary in order to obtain reproducible data on the size of the particles formed. This is consistent with the prior observation by Schillen et al. [9], who noted that prolonged incubation of PBC prior to extrusion was essential for formation of PBC vesicles. 3.3. The effect of temperature on mixed PBC aggregates Temperature dependent hydration of PPO and PEO segments of the PBC is a key factor that determines the unique self-assembly behavior of PBC and rich phase diagrams of these copolymers in aqueous solutions. Therefore, the effect of the temperature on formation of the mixed PBC aggregates was also examined. Studies using L121 alone without adding F127 demonstrated that the size of the dispersion decreased as the temperature increased (Fig. 3). This was also accompanied by an increase in turbidity of the dispersion, so that at approximately 35 jC the solution became milky. This is consistent with the well-known low critical solution temperature (LCST) behavior of L121. Furthermore, the size of aggregates in L121/F127 dispersion prepared without sonication sharply decreased with a relatively small increase of temperature (from 23 to 27 jC) and then remained consistent up to 70 jC (Fig. 3). The size of the particles in the L121/F127 dispersion prepared using sonication was

Fig. 2. Stability of dispersion in the mixed L121/F127 solutions prepared using methods listed in Fig. 1.

K.T. Oh et al. / Journal of Controlled Release 94 (2004) 411–422

417

Fig. 3. Temperature dependence of the size of the particles formed in 0.1 wt.% L121 prepared without sonication (diagonal bar) and 0.1 wt.% L121/0.1 wt.% F127 mixtures: without sonication (filled bar) and with sonication at room temperature (opened bar) in PBS. Asterisk indicates the measurements in the dispersions that were first heated to 70 jC (30 min) and then cooled down to 23 jC.

already small at 23 jC and it did not change as the temperatures was elevated (Fig. 3, opened bars). It is noteworthy that above 25 jC, the error in size measurement for pure L121 increased significantly. That was not the case for the L121/F127 (0.1%/ 0.1%) mixture prepared with or without sonication step, which remained transparent in the entire range of temperature from 27 to 70 jC. Remarkably, in both cases, the sizes of the L121/F127 aggregates at elevated temperatures were practically the same, and did not change when the dispersions were cooled to ambient temperature. In contrast, the particle size in L121 dispersion increased after cooling. Overall, the energy input by a transient temperature increase can result in formation of stable binary dispersions of PBC. This procedure may be more useful in pharmaceutical formulation processes compared to sonication. 3.4. Possible explanation of the effect of energy input To explain the unusual behavior observed in the mixtures of hydrophobic and hydrophilic PBC, one should take into account the differences in the critical micelle temperature (CMT) of these polymers. Specifically, as schematically presented in Fig. 4, the following is proposed: (1) Hydrophobic PBC L121 at room temperature and the concentrations used (0.1%) is above its CMT

(2)

(3)

(4)

(5)

[9]. As a result it forms large aggregates, which were detected in the size measurement studies. Hydrophilic PBC F127 at room temperature and the concentration used (0.1% and less) is below its CMT [14]1. As a result it is not incorporated in the L127 aggregates and does not stabilize their dispersion. Once the temperature is increased, PBC form mixed aggregates. The resulting mixed aggregates have relatively small sizes due to steric stabilization effect of long PEO chains of F127 blended with L121. These aggregates may be micelles or vesicles (closed or ruptured). According to the literature data, the CMT of F127 at 0.1% is approximately 31 jC [14], which is generally consistent with the size decrease at approximately 30– 40 jC. Some (slow) interaction between L121 and F127 is possible even at room temperature, which can explain why incubation of the mixture at this temperature before sonication affects the sizes of the particles (as measured at high temperature).

1 CMT values for different PBC referenced here were determined using the pyrene solubilization technique [13,14] and surface tension measurements [15]. The values obtained using these techniques are in good agreement with each other as well as with CMT values determined using differential scanning calorimetry [16,17].

418

K.T. Oh et al. / Journal of Controlled Release 94 (2004) 411–422

Fig. 4. Schematic diagram of proposed mechanism for the formation of aggregates in the mixtures of hydrophilic and hydrophobic PBC. (a) At room temperature hydrophobic PBC is above its CMT forming large lamellar aggregates. Hydrophilic PBC is below CMT, it does not incorporate in the hydrophobic PBC aggregates. (b) Sonication or temperature increases results in dehydration of PO chains and incorporation of hydrophilic PBC into mixed aggregates, which are sterically stabilized by EO chains. (c) Once the sonication is stopped or temperature decreased the mixed aggregates become kinetically trapped and remain stable in dispersion for several days.

(6) After heating (or sonication) when the temperature is decreased (energy input stopped), the small mixed aggregates remain kinetically trapped. This explains why the small particle size is observed at room temperature for several days. The stabilization effect of long PEO chains of hydrophilic F127 blended with L121 in mixed PBC aggregates appears to be governed by kinetic factors. These aggregates are thermodynamically unstable and ultimately phase separate. These considerations have predictive ability and can, in particular, explain why mixtures containing either hydrophobic or hydrophilic PBC with shorter PPO (higher CMT) chains do not form small particles

under any condition. It is also predicted that hydrophilic PBC with a lower CMT can form stable dispersions even at lower temperature. To evaluate this prediction, mixtures of L121 with P123 (EO20PO69 EO20) were examined. As illustrated in Fig. 5, the results show that below 0.05% P123, all mixtures prepared without energy input produced aggregates of large size, which phase separated after 2 days. Conversely, above 0.05% P123, the dispersion was stable without precipitation for several weeks and the sizes of the aggregates were rather small. These results are consistent with the proposed mechanism. According to Alexandridis et al. [14], the CMT of P123 at 0.05% is 22.5 jC. Indeed, at 25 jC, i.e. above the CMT of 0.05% P123, the L121/P123 mixtures containing

K.T. Oh et al. / Journal of Controlled Release 94 (2004) 411–422

419

Fig. 5. Size of particles in L121/P123 mixtures prepared by mixing of PBC without energy input at room temperature. All samples were measured for several days. The shadow area corresponds to precipitation where no size measurement can be done.

0.05% (and more) P123 formed relatively small particles (250 nm). These mixtures remained stable without precipitation for about 4 weeks (not shown). Inputs of energy led to a further decrease of the particle size below 200 nm. Anomalous behavior associated with micellization process, especially in the context of CMT, has been observed in dilute solutions of various diblock and triblock copolymers including PBC [18 – 21]. For example, anomalous micellization of L64 in the unimer – micelle transition region was reported [19,20]. This effect was mainly attributed to composition heterogeneity of the PBC, particularly, the presence of more hydrophobic diblock copolymers. (The sizes of the aggregates detected were not affected by batch-to-batch variation.) However, it is unlikely that it could interfere with the effects reported in this study. Indeed, the anomalous micellization is always observed in the intermediate temperature regime between unimers and micelles, i.e. before the ‘‘real’’ CMT. It has been suggested that the large assembles formed in this region represent the phase separated droplets of minor insoluble components stabilized by the

adsorbed layer made up from the major component. As soon as the ‘‘real’’ CMT of a major component is reached, the insoluble components either solubilize into the core of the micelles or form mixed micelles with this major component (depending on their molecular characteristics). In our study, hydrophobic L121 at room temperature and the concentrations used (0.1%) was already above its CMT. Therefore, it is likely that hydrophobic minor contaminants such as diblock copolymers are already entrapped in the L121 aggregates and do not interfere with further micellization. Furthermore, anomalous micellization behavior has not been seen for hydrophilic PBC, F127 and F88 used in these studies [14,20]. Therefore, the effects of composition heterogeneity of copolymers on the behavior of the PBC mixtures under experimental conditions used (heating and sonication) are unlikely. 3.5. Solubilization of a hydrophobic dye The solubilization capacity of the mixed L121/ F127 aggregates was further evaluated using a model water-insoluble dye, Sudan III that has a maximum

420

K.T. Oh et al. / Journal of Controlled Release 94 (2004) 411–422

absorbance at 362 nm. The UV/Vis spectra of Sudan III in water and PBC solutions are shown in Fig. 6 as plot of UV absorbance (k = 362 nm), [A362 nm] versus total PBC concentration. As is seen in Fig. 6, mixed L121/F127 aggregates exhibit much greater solubilization capacity compared to F127 dispersions. The incorporation of insoluble dye into L121/F127 aggregates did not result in a change of the particle size and dispersions remained stable for at least a week (data not shown). It appeared that the lamella-forming L121 in the mixture provides for approximately 10-fold increase in the solubilization capacity compared to the spherical micelle forming F127 alone. Specifical-

ly, the calculated weight-to-weight ratio of Sudan III to copolymer in 0.1% L121/0.1% F127 mixed aggregates is approximately 105 mg/g (10.5%). Therefore, these mixtures of hydrophobic and hydrophilic PBC can be used to improve the solubilization of poorly soluble molecules.

4. Conclusions In previous studies it was reported that aggregates of hydrophobic lamella-forming PBC have higher solubilization capacity than aggregates of hydrophilic

Fig. 6. Solubilization of Sudan III in F127 micelles (o) and 1:1 (% wt.) L121/F127 mixtures (n) prepared with sonication for 2 min. (a) UV/Vis spectra of Sudan III containing in water (1), in 0.4% solution of F127 (2), and in mixture of 0.2% L121/0.2% F127 (3). (b) Amount of solubilized Sudan III depending on total concentration of PBC in dispersion.

K.T. Oh et al. / Journal of Controlled Release 94 (2004) 411–422

PBC that form spherical micelles. By properly selecting the mixtures of a hydrophilic and hydrophobic PBC, it has been demonstrated that relatively stable dispersions with relatively small particle sizes can be produced. These dispersions have a higher solubilization capacity with respect to a poorly soluble organic compound compared to a dispersion of hydrophilic PBC alone. A qualitative explanation of the dispersion behavior of PBC mixtures has been proposed, which facilitates selection of the mixture components displaying high colloidal stability, small particle size and good solubilization characteristics. These findings may be useful for pharmaceutical formulation development, particularly, for improvement of solubilization capacity and stability of polymer micelles. Furthermore, these mixtures can be useful for preparation of stable pharmaceutical formulation of hydrophobic PBC, which recently attracted significant attention as functional excipients in drug and gene delivery studies [5].

[6]

[7]

[8] [9]

[10]

[11]

[12]

Acknowledgements This work was in part supported by NSF DMR award (0071682) to A.V. Kabanov. K.T. Oh has been supported by a UNMC Fellowship. The authors would like to thank R. Nessler (University of Iowa, Central Microscopy Research Facility) for carrying out cryo-TEM experiments and Dr. V. Alakhov (Supratek Pharma Inc., Montreal, Canada) for valuable discussions.

References [1] G.S. Kwon, K. Kataoka, Block copolymer micelles as longcirculating drug vehicles, Adv. Drug Deliv. Rev. 16 (2 – 3) (1995) 295 – 309. [2] C. Allen, D. Maysinger, A. Eisenberg, Nano-engineering block copolymer aggregates for drug delivery, Coll. Surf. B: Biointerf. 16 (1999) 3 – 27. [3] K. Kataoka, A.V. Kabanov, Special Issue: Polymeric Micelles in Biology and Pharmaceutics, Colloids and Surfaces, B: Biointerfaces, Elservier, Amsterdam, Lausanne, New York, Oxford, Shannon, Tokyo, 1999. [4] V.P. Torchilin, Polymeric micelles in diagnostic imaging, Coll. Surf. B: Biointerf. 16 (1999) 305 – 319. [5] A.V. Kabanov, V.Y. Alakhov, Pluronic block copolymers in drug delivery: from micellar nanocontainers to biological re-

[13]

[14]

[15]

[16]

[17]

[18]

421

sponse modifiers, Crit. Rev. Ther. Drug Carrier Syst. 19 (1) (2002) 1 – 73. R. Nagarajan, Solubilization of hydrocarbons and resulting aggregate shape transitions in aqueous solutions of pluronic (PEO – PPO – PEO) block copolymers, Coll. Surf. B: Biointerf. 16 (1999) 55 – 72. V.Y. Alakhov, A.V. Kabanov, Block copolymeric biotansport carriers as versatile vehicles for drug delivery, Expert Opin. Invest. Drugs 7 (8) (1998) 1453 – 1473. G.S. Kwon, T. Okano, Soluble self-assembled block copolymers for drug delivery, Pharm. Res. 16 (5) (1999) 597 – 600. K. Schillen, K. Bryskhe, Y.S. Melnikova, Vesicles formed from a poly(ethylene oxide) – poly(propylene oxide) – poly (ethylene oxide) triblock copolymer in dilute aqueous solution, Macromolecules 32 (20) (1999) 6885 – 6888. I. Szleifer, O.V. Gerasimov, D.H. Thompson, Spontaneous liposome formation induced by grafted poly(ethylene oxide) layers: theoretical prediction and experimental verification, Proc. Natl. Acad. Sci. USA 95 (1998) 1032 – 1037. K. Kostarelos, T.F. Tadros, P.F. Luckman, Physical conjugation of (tri-)block copolymers to liposomes toward the construction of sterically stabilized vesicle systems, Langmuir 15 (2) (1999) 369 – 376. M. Ranson, D. Ferry, D. Kerr, J. Radford, D. Dichens, V. Alakhov, M. Brampton, J. Margison, Results of a cancer research campaign phase I dose escalation trial of SP1049C in patients with advanced cancer, The Fifth International Symposium on Polymer Therapeutics: From Laboratory to Clinical Practice, The Welsh School of Pharmacy, Cardiff University, Cardiff, UK, 2002, p. 15. E. Batrakova, S. Lee, S. Li, A. Venne, V. Alakhov, A. Kabanov, Fundamental relationships between the composition of pluronic block copolymers and their hypersensitization effect in MDR cancer cells, Pharm. Res. 16 (9) (1999) 1373 – 1379. P. Alexandridis, J.F. Holzwarth, T.A. Hatton, Micellization of poly(ethylene oxide) – poly(propylene oxide) – poly(ethylene oxide) triblock copolymers in aqueous solutions: thermodynamics of copolymer association, Macromolecules 27 (9) (1994) 2414 – 2425. A.V. Kabanov, I.R. Nazarova, I.V. Astafieva, E.V. Batrakova, V.Y. Alakhov, A.A. Yaroslavov, V.A. Kabanov, Micelle formation and solubilization of fluorescent probes in poly(oxyethylene-b – oxypropylene-b – oxyethylene) solutions, Macromolecules 28 (7) (1995) 2303 – 2314. P. Alexandridis, J.F. Holzwarth, Differential scanning calorimetry investigation of the effect of salts on aqueous solution properties of an amphiphilic block copolymer (poloxamer), Langmuir 13 (23) (1997) 6074 – 6082. J.K. Armstrong, J. Parsonage, B. Chowdhry, S. Leharne, J. Mitchell, A. Beezer, K. Loehner, P. Laggner, Scanning densitometric and calorimetric studies of poly(ethylene oxide)/poly (propylene oxide)/poly(ethylene oxide) triblock copolymers (poloxamers) in dilute aqueous solution, J. Phys. Chem. 97 (15) (1993) 3904 – 3909. I.W. Hamley, The Physics of Block Copolymers, Oxford University Press, Oxford, UK, 1998.

422

K.T. Oh et al. / Journal of Controlled Release 94 (2004) 411–422

[19] Z. Zhou, B. Chu, Anomalous association behavior of an ethylene oxide/propylene oxide ABA block copolymer in water, Macromolecules 20 (12) (1987) 3089 – 3091. [20] Z. Zhou, B. Chu, Anomalous micellization behavior and composition heterogeneity of a triblock ABA copolymer of (A)

ethylene oxide and (B) propylene oxide in aqueous solution, Macromolecules 21 (8) (1988) 2548 – 2554. [21] T.P. Lodge, J. Bang, K.J. Hanley, J. Krocak, S. Dahlquist, B. Sujan, J. Ott, Origins of anomalous micellization in diblock copolymer solutions, Langmuir 19 (6) (2003) 2103 – 2109.