Biomacromolecules 2008, 9, 2283–2291
2283
Articles Morphological Control of Poly(ethylene glycol)-block-poly(ε-caprolactone) Copolymer Aggregates in Aqueous Solution Nichole Fairley,† Bryan Hoang,† and Christine Allen*,†,‡ Department of Pharmaceutical Sciences, Leslie Dan Faculty of Pharmacy, and Department of Chemistry, Faculty of Arts and Science, University of Toronto, 144 College Street, Toronto, Ontario, Canada M5S 3M2 Received December 13, 2007; Revised Manuscript Received July 5, 2008
In aqueous solution, it was found that the amphiphilic copolymer poly(ethylene glycol)-b-poly(caprolactone) (PEG5000-b-PCL4100) formed different morphologies, including long rod-like, short rod-like, or spherical aggregates, when the copolymer concentration was increased. Nearly identical morphologies were observed with the addition of increasing amounts of PEG2000-distearoylphosphoethanolamine (PEG2000-DSPE) to the copolymer. The morphologies of the aggregates in solution were confirmed by negative stain transmission electron microscopy (TEM) and cryogenic-TEM (cryo-TEM). The critical micelle concentrations of the PEG5000-b-PCL4100 copolymer, PEG2000-DSPE and a mixture of the two materials (PEG5000-b-PCL4100/PEG2000-DSPE) were evaluated to determine the thermodynamic stability of the aggregates. Differential scanning calorimetry was performed to gain insight into the degree of mixing of PEG5000-b-PCL4100 and PEG2000-DSPE. Overall, combining PEG5000-b-PCL4100 and PEG2000-DSPE produced a single population of mixed micelles with rod-like or spherical morphologies depending on the material composition and concentration.
Introduction Amphiphilic molecules, such as phospholipids and block copolymers, have been reported to self-assemble into a diverse array of morphologies in aqueous media. The evolution of distinct morphologies has been reported to depend on a wide range of factors, including the nature and concentration of the material, solvent conditions, salt concentrations, and the presence of small molecule solubilizates.1-5 Importantly, the morphology of these nanosized aggregates has been shown to influence their performance in applications such as drug delivery. Indeed the physical properties of micelles including size, size distribution, and morphology impact their stability, drug loading, and release properties as well as in vivo pharmacokinetics and biodistribution.6 The majority of the research on morphogenesis of amphiphilic block copolymer systems in aqueous media has been performed on nonbiocompatible and nonbiodegradable systems.7,8 For example, seminal research by Eisenberg’s group demonstrated the formation of a hierarchy of morphologies from poly(acrylic acid)-b-polysytrene (PAA-b-PS), poly(ethylene oxide)-b-polystyrene (PEO-b-PS), and poly(acrylic acid)-b-polybutadiene (PAA-b-PB) “crew-cut” copolymers, wherein the coronaforming block is significantly shorter than the core-forming block. Alterations in the copolymer composition, copolymer concentration, and solution conditions (ion content and pH) were shown to result in the controlled preparation of spheres, rods, vesicles, lamellae, large compound micelles, and tubules.2,4,9,10 Won et al. also investigated the influence of copolymer composition on the morphology of copolymer aggregates. Specifically, they showed that increasing the PEO content in PEO-b-PB copolymers resulted in a transition from bilayers to cylinders to spheres.8 * To whom correspondence should be addressed. Phone: (416) 946-8594. Fax: (416) 978-8511. E-mail:
[email protected]. † Department of Pharmaceutical Sciences. ‡ Department of Chemistry.
More recently, Geng and Discher reported the formation of worm-like micelles from the biocompatible copolymer poly(ethylene oxide)-b-poly(caprolactone) (PEO-b-PCL).11 They found that over time the giant worm-like micelles prepared from PEO2000-b-PCL2770 and PEO5000-b-PCL6500 copolymers spontaneously shorten to form spherical micelles. It was proposed that end-hydrolysis of the PCL block lead to an increase in the weight fraction of PEO resulting in the worm-to-sphere transition.11 Later, Cai et al. produced spherical and giant flexible worm micelles from the same PEO-b-PCL copolymers (i.e., PEO5000-b-PCL6500). Importantly, the worm-like micelles, termed filomicelles, were found to load twice as much of the hydrophobic drug paclitaxel as spherical micelles.12 Du et al. also demonstrated that PEOm-b-PCLn (m ) 44, 113; n ) 21-232) copolymers form micelles with a range of morphologies depending on the lengths of the copolymer blocks. Successive increases in the PCL block length were shown to result in a change in the morphology from spheres, to rod-like, to wormlike and finally to lamellar aggregates.13 In addition, Zupancich et al. demonstrated that spherical, cylindrical, and bilayered vesicle structures could be formed from poly(ethylene oxide)b-poly(γ-methyl-ε-caprolactone) in water.14 The capability to produce variations in morphology in a controlled manner is of great interest due to the potential advantages and applications of nonspherical copolymer aggregates in the field of drug delivery.15-17 As greater control over morphology is achieved, efforts can be directed toward elucidating the influence of morphology on biodistribution, transport, and targeting of nonspherical drug delivery vehicles. To this point, the influence of copolymer concentration, temperature, and pH on the morphology of biocompatible and biodegradable systems has only been explored to a limited extent.18,19 This study focused on establishing the relationship between copolymer concentration and morphology for a PEG5000b-PCL4100 copolymer system in aqueous solution. In addition, the effect of adding a PEG-phospholipid, specifically PEG2000DSPE, to the copolymer was investigated as a means to gain
10.1021/bm800572p CCC: $40.75 2008 American Chemical Society Published on Web 08/15/2008
2284
Biomacromolecules, Vol. 9, No. 9, 2008
further compositional control of morphology. Furthermore, the solution properties of the block copolymer PEG5000-b-PCL4100 and the PEG-lipid, PEG2000-DSPE, were investigated, as well as the physicochemical characteristics of the PEG5000-b-PCL4100/ PEG2000-DSPE mixed micelle system.
Experimental Section Materials. Methoxy poly(ethylene glycol) (MePEG, Mn ) 5000, Mw/Mn ) 1.06), toluene, were purchased from Sigma (Sigma-Aldrich, Oakville, ON, Canada). Tetrahydrofuran (THF) was purchased from Caledon Laboratories (Georgetown, ON, Canada). ε-Caprolactone (εCL, 99%) was purchased from Acros Organics (Fisher Scientific, Pittsburgh, PA). Toluene was dried under calcium hydride and distilled prior to use. All other chemicals were obtained from Sigma and used as received. 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-methoxy (poly(ethylene glycol)) (PEG2000-DSPE) was obtained from Genzyme Pharmaceuticals (Cambridge, MA). 1,6-Diphenyl-1,3,5hexatriene (DPH) was purchased from Fluka (Fluka-Sigma-Aldrich, Oakville, ON, Canada). Synthesis of PEG-b-PCL Copolymer. The copolymer was synthesized via metal-free cationic ring opening polymerization of ε-CL in the presence of MePEG with hydrogen chloride (HCl) in ether as the catalyst.20,21 A typical procedure was carried out as follows: 1 g of MePEG (0.2 mmol, Mn ) 5000, Mw/Mn ) 1.06) was introduced to a flame dried, round-bottom flask and dried twice by azeotropic distillation of toluene. A 10 mL volume of dried dichloromethane and 1 g of ε-CL (8.76 mmol, dried and distilled over calcium hydride) were subsequently added to the reaction vessel. A total of 0.6 mL (1 M, 0.6 mmol) of HCl in ether was added at 25 °C and maintained at this temperature overnight with continuous stirring. A 0.1 mL aliquot of triethylamine was added to terminate the reaction. Precipitated triethylamine-HCl salt was removed by filtration and the methoxy poly(ethylene glycol)-b-poly(caprolactone) (abbreviated as PEG-b-PCL) diblock copolymer was collected by precipitation in an ether/hexane mixture. Purification of the copolymer was achieved by the dissolution/ precipitation method with dichloromethane and ether/hexane (1:1, v/v), respectively, followed by filtration and vacuum drying. Characterization of PEG-b-PCL. 1H NMR spectra were obtained using a Varian Gemini 200 spectrometer (200 MHz for 1H) with deuterated chloroform (CDCl3) as the solvent and internal standard. Gel permeation chromatography (GPC) was performed on a Waters 590 liquid chromatography system equipped with three Waters Styragel HR 4E columns and a 410 differential refractometer at room temperature. Samples were measured at 30 °C with THF as solvent at a flow rate of 1.0 mL/min. The molecular weight was calibrated using polystyrene standards and data analysis was performed using Windowsbased Millenium 2.0 software package (Waters Inc., Milford, MA). Proton assignments determined from the 1H NMR spectrum were as follows: signals for the MePEG unit appeared at 3.36 ppm (3H, CH3-O-), 3.51 ppm (2H, CH3-O-CH2-CH2), 3.62 ppm (4H, -O-CH2-CH2-), 4.20 ppm (2H, O-CH2-CH2-O-CO); while signals for the PCL block appeared at 1.41 ppm (2H, CO-CH2-CH2-CH2-CH2-CH2-CO), 1.63 ppm (4H, CO-CH2-CH2-CH2-CH2-CH2-CO), 2.28 ppm (2H, CO-CH2CH2-CH2-CH2-CH2-CO), 3.85 ppm (2H, CH2-CH2-OH), 4.04 ppm (2H, CO-CH2-CH2-CH2-CH2-CH2-CO), respectively. Critical Micelle Concentration (CMC) Measurements. The CMC values for PEG5000-b-PCL4100, PEG2000-DSPE, and PEG5000-b-PCL4100/ PEG2000-DSPE were determined by an established fluorescence-based method and surface tension measurements.22,23 Aliquots of stock solutions of PEG5000-b-PCL4100, PEG2000-DSPE, or PEG5000-b-PCL4100/ PEG2000-DSPE in chloroform were added to glass vials, such that the concentrations ranged from 0.5 to 500, 0.05 to 875, and 0.05 to 875 mg/L, respectively, with at least 10 concentrations analyzed in each series. An aliquot of a stock solution of DPH in chloroform was then added to each vial such that the concentration of DPH was maintained at 1 mg/L in each solution. The solutions were stirred overnight and the solvent was evaporated under nitrogen. The dried vials were heated to 65 °C and 1 mL of double distilled water was added slowly to each vial. The solutions were equilibrated by stirring overnight at room temperature, followed by measurement of the fluorescence emission at 430 nm with excitation at 350 nm (Spectra GeminiXS dual-scanning microplate spectrofluorometer, Molecular Devices, Sunnyvale, CA). The CMC was defined as the point of intersection of the two lines in the plot of fluorescence versus concentration.
Fairley et al. For the surface tension measurements (Sigma700 Tensiometer, KSV Instruments, Monroe, CT), individual stock solutions of PEG5000-bPCL4100 and PEG2000-DSPE were prepared in THF. In short, aliquots of copolymer, PEG-lipid, or a combination of the two were added to glass vials. The solutions were vortexed and the solvent was evaporated under nitrogen. The dried films were heated to 65 °C and 20 mL of double distilled water was added to each vial. The solutions were left to equilibrate overnight at room temperature prior to using the Wilhelmy plate method for surface tension measurements.24 The CMC was defined as the point of intersection of the two lines in the plot of surface tension versus concentration. Preparation of Aqueous Micelle Solutions. The thin-film hydration method was used to prepare PEG5000-b-PCL4100, PEG2000-DSPE, and PEG5000-b-PCL4100/PEG2000-DSPE micelles.20 Individual stock solutions of PEG5000-b-PCL4100 and PEG2000-DSPE were prepared in THF. Specific volumes of the stock solutions were then added to vials to vary the mole ratio of block copolymer to PEG-lipid. Solutions containing only PEG-lipid or both copolymer and PEG-lipid (micelle solutions with equal volumes but varying mole ratios (95:5, 90:10, 80: 20, 70:30)) were mixed to achieve a final material concentration of 10 mg/mL. Solutions containing only block copolymer were mixed to achieve final material concentrations of 0.5, 2, 5, 10, 25, and 50 mg/ mL. All solutions were then stirred overnight, thoroughly dried under nitrogen, and left overnight under vacuum. Following solvent removal, the dry films obtained for the copolymer and copolymer/lipid mixtures were thicker in appearance than the film obtained for the PEG-lipid. Double distilled water at 65 °C (above the melting temperature of PCL) was added to rehydrate the dried films. The solutions were then vortexed for 5 min, left to stir for 5 days at room temperature, and sonicated for 45 min prior to characterization to prevent the aggregation of micelles particularly at higher copolymer concentrations. Transmission Electron Microscopy (TEM). Micelle morphology was evaluated by TEM with a Hitachi 7000 microscope operating at an acceleration voltage of 75 kV (Schaumburg, IL). The micelle solutions were diluted in double distilled water and negatively stained with a 2% aqueous solution of uranyl acetate. The negative stain surrounds the material with an electron dense layer generating reversecontrast, negative electron images.25 The sample solutions were deposited on copper grids that had been precoated with carbon and negatively-charged (Ted Pella Inc., Redding, CA). The copper grids were then blotted and left to stand to allow solvent evaporation. The average diameters of the copolymer aggregates in the TEM images were obtained using SigmaScan Pro software (Jandel Scientific). Cryogenic Transmission Electron Microscopy (cryo-TEM). Samples for cryo-TEM were prepared by applying 5 µL of negativelystained (2% v/v uranyl acetate) sample on a Quantifoil R2/2 (Quantifoil Micro Tools GmbH, Jena, Germany) holey carbon film supported by a copper grid. The thin layer of the micelle solution on the grid was immediately frozen by plunging into liquid ethane at -180 °C and then transferring to liquid nitrogen at approximately -185 °C. Images were obtained at a temperature of approximately -170 °C and with a 200 kV acceleration voltage, using a FEI Tecnai G2 F20 microscope (FEI Company, Hillsboro, OR) equipped with a Gatan CCD camera (Gatan Inc., Warrendale, PA). Dynamic Light Scattering (DLS). The hydrodynamic diameter of the PEG2000-DSPE micelles was measured by DLS at an angle of 90° and a temperature of 25 °C (DynaPro-MS/X, Protein Solutions Inc., Lakewood, NJ). For analysis, the samples were diluted using double distilled water. The mean diameter and the size distribution of the micelles were obtained from the Regularization algorithm (DYNAMICS V6 version 6.7.1, Whatt Technology Corp.). Differential Scanning Calorimetry (DSC). Thermal analysis was performed using a Q100 differential scanning calorimeter (DSC) (TA Instruments, New Castle, DE) on lyophilized samples of micelles (i.e., 3-5 mg) that were prepared as outlined above. Samples were cooled to -70 °C and then heated to 100 °C at a temperature ramp speed of 10 °C/min under nitrogen purge. The data was analyzed using TA universal software (TA Instruments, Inc., New Castle, DE) and the melting temperature (Tm) was taken to be the peak of the endotherm.
Results and Discussion Synthesis and Characterization of PEG-b-PCL Copolymer. PEG5000-b-PCL4100 was synthesized by metal-free cationic ring-opening polymerization of ε-CL by an activated monomer
Morphological Control of Copolymer Aggregates
mechanism, with HCl in ether as the catalyst.20,21,26 Conventional ring opening polymerization strategies involving organometallic catalysts have typically resulted in metal contamination of the polymer as a function of incomplete catalyst removal. As previously established, this metal free cationic synthetic strategy allows for low temperature polymerization of ε-CL with the formation of HCl salts that are easily removed by filtration and precipitation.27 The molecular weight of the PEG-b-PCL was determined to be 15761 and 9100 g/mol from GPC and 1H NMR analysis, respectively (Figures S1 and S2, Supporting Information). The discrepancy between the values obtained for the molecular weight of the copolymer from 1H NMR and GPC analyses can be attributed to differences in the hydrodynamic volume of the block copolymer and the polystyrene standards used for calibration of the GPC. The diblock copolymer was found to conform to the predicted theoretical composition (Mn ) 9000), defined as the calculated molecular weight of the polymer based upon the feed ratio of ε-CL to MePEG. The polydispersity index of the PEG-b-PCL diblock copolymer indicated a relatively narrow molecular weight distribution (Mw/Mn ) 1.15). Critical Micelle Concentration. The use of block copolymer micelles as long circulating drug delivery vehicles relies heavily on their thermodynamic stability which is determined by the CMC of the copolymer material. A low CMC is important to retain micelle integrity following the dilution that occurs in vivo after intravenous administration.28,29 The CMC of amphiphilic copolymers is largely determined by the nature of the hydrophobic and hydrophilic blocks as well as their relative lengths. Typically, an increase in the length of the hydrophobic block results in a decrease in the CMC when the length of the hydrophilic block is held constant.17,30,31 For example, Yoo et al. reported the CMC of PEG5000-b-PCL2500, PEG5000-bPCL5000, and PEG5000-b-PCL7500 copolymers to be 1.3-10-5, 5.0-10-6, and 2.0-10-6 M, respectively, as measured by surface tensiometry.32 In contrast, an increase in the length of the hydrophilic block results in an increase in the CMC when the length of the hydrophobic block is held constant.17 PEG-conjugated phospholipids are also reported to have CMC values in the low micromolar range.33 Ashok et al. determined the CMC values for PEG-DSPE with different PEG chain lengths (i.e., 2000, 3000, and 5000 g/mol) using the fluorescent probe DPH. The CMC values increased with increasing PEG chain length and ranged from 0.5 to 1.5-10-6 M.34 Vakil and Kwon found the CMC of PEG5000-DSPE to be 1.4-10-6 M using pyrene as the probe.35 In this way, PEG2000-DSPE employed in this research was expected to have a lower CMC than the PEG5000-b-PCL4100 copolymer. In this study, the CMC values of the PEG5000-b-PCL4100 copolymer, PEG2000-DSPE lipid, and a mixture of the two materials (PEG5000-b-PCL4100/PEG2000-DSPE) were investigated to gain insight into the thermodynamic stability of the pure and mixed micelle systems. The mixed micelle system that was investigated had a copolymer-to-PEG-lipid ratio of 80:20 mol %. The 80:20 mol % mixed system was the only mixed sample that remained transparent, with no precipitation, following hydration. The PEG5000-b-PCL4100/PEG2000-DSPE 70:30 mol % system was stable for short periods (i.e.,
