1490
Bioconjugate Chem. 2009, 20, 1490–1496
PEGylated Degradable Composite Nanoparticles Based on Mixtures of PEG-b-Poly(γ-benzyl L-glutamate) and Poly(γ-benzyl L-glutamate) Ma. Elisa Martı´nez-Barbosa,†,‡ Sandrine Cammas-Marion,§ Laurent Bouteiller,| Christine Vauthier,‡ and Gilles Ponchel*,‡ UMR CNRS 8612, Laboratoire de Pharmacie et Biopharmacie, Faculte´ de Pharmacie, Universite´ Paris-Sud, 5 rue J.B. Cle´ment, 92290 Chaˆtenay-Malabry, France, UMR 6226 CNRS-ENSCR, Equipe Chimie Organique et Supramole´culaire, Avenue du Ge´ne´ral Leclerc, 35700 Rennes, France, and UMR CNRS 7610, Chimie des Polyme`res, Universite´ Pierre et Marie Curie, 4 Place Jussieu, 75252, Paris cedex 05, France. Received January 15, 2009; Revised Manuscript Received May 18, 2009
In the present work, the possibility to obtain PEGylated nanoparticles from two PBLG derivatives, PEG-b-poly(γbenzyl L-glutamate), PBLG-PEG-60, and poly(γ-benzyl L-glutamate), PBLG-Bnz-50, by nanoprecipitation has been investigated. Particles were prepared not only from one polymer (PBLG-PEG-60 or PBLG-Bnz-50), but also from mixtures of two PBLG derivatives, PBLG-PEG-60 and PBLG-Bnz-50, in different ratios (90/10, 77/ 23, and 40/60 wt %). Because of the presence of PEG chains, hydrophilic particles were obtained, which was confirmed by ζ potential measurements (ζ from -13 mV and -21 mV) and by isothermal titration microcalorimetry (ITC). This last technique has shown no heat exchange when BSA was added to PEGylated nanoparticles. Further, complement activation has been evaluated by crossed immuno-electrophoresis demonstrating that the introduction of 77 wt % of PEGylated PBLG chains in the particles was enough to ensure a low complement activation activity. This effect was strongly correlated to the ζ potential of the particles, which decreased with an increase of PEG chains content. Interestingly, such properties are of interest for the preparation of degradable stealth nanocarriers. Moreover, it is suggested that the introduction of a reasonable amount (up to 20 wt %) of a second copolymer in the particle composition can be possible without modifying their stealth character. Moreover, the presence of this second copolymer would help to introduce a second functionality to the particles.
INTRODUCTION Over the past years, nanoparticulate carriers have been developed in order to improve efficiency and site-specificity of drug delivery (1, 2). As a result of their small size, this first generation of nanocarriers can be administrated intravenously. However, such small hydrophobic particles are rapidly (halftime of clearance about 1 min) and efficiently (about 90% uptake) cleared from the circulation due to their recognition by opsonins (3-5). Therefore, to address these limitations, a second generation of nanocarriers is under development taking into account the various parameters which influence carrier recognition by reticuloendothelial system (3-5). It has been shown that the coating of hydrophobic nanocarrier surfaces by hydrophilic polymers, such as poly(ethylene glycol) (PEG), induces a significant decrease of the opsonisation process (6). The coating has to be dense and flexible in order to reduce all types of interactions with hydrophobic nanocarrier surfaces. Moreover, hydrophilic polymer chains, such as PEG, have to be well anchored on nanoparticles to avoid desorption of chains (4). Surface modification on hydrophobic nanoparticulate systems has been carried out. Coating or covalent coupling methods (7, 8) have been proposed to introduce hydrophilic chains on nanoparticle surface. Hydrophilic polymers such as poly(ethylene glycol), poly(vinyl alcohol), poloxamers, or dextran have been * Corresponding author. Gilles Ponchel, e-mail: gilles.ponchel@ u-psud.fr, Te´l: +33 (0)1 46 83 59 19, Fax: +33 (0)1 46 61 93 34. † Present address: Departamento de Investigacio´n en Polı´meros y Materiales (DIPM), Universidad de Sonora. Rosales y Blvd. Luis Encinas s/n. C.P. 83000, Hermosillo, Sonora, Me´xico. ‡ UMR CNRS 8612. § UMR 6226 CNRS-ENSCR. | UMR CNRS 7610.
shown to decrease significantly protein adsorption (4). In order to obtain a high density of hydrophilic chains on the surface and an efficient anchoring of these chains in the carrier, core-shell type nanoparticles and macromolecular micelles formed in aqueous media by self-assembly of amphiphilic block copolymers have attracted considerable attention for application as drug delivery systems (9-14). Besides overcoming the rapid recognition of nanoparticles by the reticuloendothelial system, research has been also focused on overcoming the many-faceted problem of biocompatibility. In this context, efforts have been directed toward the use of natural-like materials such as poly(Ramino acids) (15). Poly(γ-benzyl L-glutamate), PBLG, a synthetic polypeptide, has attracted attention for biomedical applications (15-21) because of the presence of a degradable amide bound in the polymer backbone (21) and a carboxylic acid function in the side chain, available after hydrolysis of the lateral benzyl ester. Poly(ethylene oxide)-block-poly(γ-benzyl L-glutamate) (22, 23) allowed macromolecular micelle formation as a result of their amphiphilic nature (24, 25). Recently, in response to the need for targeting systems, poly(γ-benzyl L-glutamate) and poly(ethylene glycol) diblock copolymers endcapped with a galactose moiety have been proposed for sitespecific drug delivery (26). In this context, we focused our interest on the synthesis of PBLG derivatives using initiators allowing the preparation of biocompatible nanoparticulate systems possessing different specific functionalities. For that, PBLG derivatives were synthesized by ring-opening polymerization (ROP) of γ-benzyl L-glutamate N-carboxyanhydride (BLG-NCA) using selected amine-terminated initiators (27). In the present work, we report the possibility to obtain PEGylated nanoparticles from two PBLG derivatives, PBLGPEG-60 and PBLG-Bnz-50, by nanoprecipitation. Composite nanoparticles with different PBLG-PEG-60/PBLG-Bnz-50 ratio
10.1021/bc900017c CCC: $40.75 2009 American Chemical Society Published on Web 07/13/2009
PEGylated Degradable Composite NPs
have been prepared and characterized. Interactions of the different nanoparticles with bovine serum albumin (BSA) have been measured by ITC. Moreover, complement activation has been evaluated by crossed immuno-electrophoresis.
MATERIALS AND METHODS Materials. N,N-Dimethylformamide (DMF, Acros, 99%) and benzylamine (Janssen Chimica) were distilled under reduced pressure over BaO and KOH, respectively, and stored under argon atmosphere. γ-Benzyl-L-glutamate N-carboxyanhydride (γ-BLG-NCA), from ISOCHEM-SNPE, was used as received. j w ) 5000 Methoxy poly(ethylene glycol) amine (mPEG-NH2), M Da from Shearwater Corporation was dried under vacuum over P2O5 at 30 °C for 24 h. Bovine serum albumin (BSA) was purchased from Sigma. Water was purified by reverse osmosis (Milli-Q, Millipore). Poloxamer 188 (Lutrol F68), diethyl ether, and other solvents used were analytical grade, and all other chemicals were commercially available reagent grade. Synthesis and Characterization of PBLG-Derivatives. PBLG-derivatives, PBLG-Bnz-50, and PBLG-PEG-60 were prepared by anionic ring-opening polymerization as described previously (27). Briefly, PBLG-Bnz-50 was obtained by anionic ring-opening polymerization of γ-benzyl-L-glutamate N-carboxyanhydride initiated by benzylamine in DMF at 30 °C. The reaction mixture was stirred at 30 °C until the characteristic NCA bands disappeared from FT-IR spectrum and then poured into a large volume of cold diethyl ether. The precipitate was dried under vacuum at 35 °C for at least 12 h. This precipitation, purification, and drying procedure was performed twice. The molecular weight of PBLG-Bnz-50 was estimated from intrinsic viscosity measured in DMF at 25 °C using the Mark-Houwink equation for PBLG in DMF at 25 °C (28).The molecular weight j w ) 45 000 g · mol-1. estimated for the PBLG-Bnz-50 was M Similarly, PBLG-PEG-60 was obtained by anionic ring-opening polymerization of γ-benzyl-L-glutamate N-carboxyanhydride initiated by methoxy poly(ethylene glycol) amine (mPEG-NH2), j w ) 5000 g · mol-1, in DMF at 30 °C. The reaction average M mixture was stirred at 30 °C until the characteristic NCA bands disappeared from FT-IR spectrum and poured into a large volume of cold diethyl ether. The precipitate was filtered, washed with water and diethyl ether, and then dried under vacuum at 35 °C for at least 12 h. A second purification j Pn of the PBLG segment in PBLGprocedure was performed. D PEG-60 was determined from 1H NMR spectra by the ratio between the peak intensities of methylene protons of the PEG chain, OCH2CH2, and the benzyl protons of PBLG chain, j Pn of 231 was found. COOCH2C6H5. A D Nanoparticles Preparation. Nanoparticles were prepared using a slightly modified nanoprecipitation technique described elsewhere (29), in the absence of any surfactant agent, either from one polymer (PBLG-PEG-60 or PBLG-Bnz-50) or from mixtures of these two polymers in different ratios (PBLG-PEG60/PBLG-Bnz-50 at 90/10, 77/23, and 40/60 wt % ratio and named composite nanoparticles). Briefly, 15 mg of one polymer or a mixture of polymers were dissolved in 5 mL of THF at 30 °C during 18 h, without stirring. The solution was then homogenized under gentle agitation for a few seconds and added, by dripping to 10 mL of Milli-Q water, under magnetic stirring. The mixture was left under magnetic stirring for 15 min and then transferred in a Teflon-coated recipient. The solvent was evaporated, at 30 °C, under a light air flux. In order to eliminate residual solvent, nanoparticles were further diluted with 5 mL of Milli-Q water, and evaporation was carried out to adjust the volume of the nanoparticle suspension to 10 mL. Preparations were kept at 4 °C until characterization. Nanoparticles Characterization. Size Distribution. Mean diameter (n ) 3) was determined, after suitable dilution of bulk
Bioconjugate Chem., Vol. 20, No. 8, 2009 1491
suspensions in Milli-Q water, by dynamic laser light scattering (Nanosizer Coulter N4 Plus, Margency, France). ζ Potential. ζ potential measurements (n ) 5) were carried out using a Zetasizer 4, Malvern Instrument after adequate dilution of 200 µL of nanoparticles suspensions into 10 mL of 1 mM KCl filtered through a 0.22 µm filter. Transmission Electron Microscopy. Nanoparticles were observed by TEM (transmission electron microscope, Philips EM 208). 7 µL of nanoparticles suspension were half-diluted in a 0.125% (w/v) poloxamer solution and placed on a Formvarcarbon film previously coated on a copper grid (400 mesh). After 6 min of deposition at room temperature, nonadherent nanoparticles were eliminated and the sample was stained by phosphotungstic acid (1%) during 30 s. TEM bright field imaging was performed under a 80 kV accelerating voltage. Isothermal Titration Microcalorimetry (ITC) Studies. ITC experiments were carried out with an VP-ITC, MicrolCal. Interactions between bovine serum albumin (BSA), a model globular protein, and nanoparticles constituted by PBLG-Bnz50 or PBLG-PEG-60 were evaluated. Solution of BSA at 5.4 × 10-2 mM was placed in a 295 µL continuously rotating (270 rpm) syringe. Nanoparticle suspension at 2.7 × 10-2 mM of polymer was placed in the sample cell of 1.43 mL. A first 2 µL aliquot was injected, without taking into account the observed heat, to remove the effect of solute diffusion across the syringe tip during the equilibration period. Then, injections of 10 µL of the BSA solution were made at intervals of 10 min. Experiments were carried out at 25 °C. EValuation of Complement ActiVation by Crossed ImmunoElectrophoresis. The specific activation of the C3 component of the complement from human serum and induced by the nanoparticle surface was evaluated by comparative measurements of C3 cleavage into C3b. Nanoparticle suspensions (1.5 mg mL-1) were first concentrated by evaporation of part of the water to reach a nanoparticle concentration of 9 mg mL-1 or 6 mg mL-1 in order to obtain a surface area of 500 cm2 in a volume V e 100 µL. The amount of C3 and C3b was measured by crossed immunoelectrophoresis by an adaptation of the method described elsewhere (7, 30) after incubation of the nanoparticles with serum. Briefly, human serum obtained from healthy donors was stored at -80 °C until use. Equivalent volumes of nanoparticle suspensions, corresponding to a surface area of 500 cm2 in 100 µL, were incubated with 50 µL of human serum and 50 µL of veronal-buffered saline containing 0.15 mM Ca2+ and 0.5 mM Mg2+ in ions (VBS2+), under gentle agitation, for 1 h at 37 °C. Nanoparticle surface area was calculated, assuming that particles were spherical and using the following equations: S ) n4πr2
(1)
V ) n4/3πr3
(2)
leading to S(sp) ) S/V ) 6/dF
(3)
where S and V are the surface area and volume, respectively, of n spherical nanoparticles of average radius r. S(sp) is the specific surface area of 1 g of spherical nanoparticles of average diameter d and density F (taken as F ) 1.2 g/cm3 relative to F for PBLG). After incubation with human serum, samples (7 µL) were subjected to a first electrophoresis in 1% agarose gel for about 45 min. The second-dimension electrophoresis was carried out during 18 h with 1% agarose gel on Gelbond films in the presence of a polyclonal antibody to human C3, recognizing both C3 and C3b (complement C3 antiserum developed in goat, Sigma). Buffer used for electrophoresis migration was Tris veronal buffer diluted 1:5 in water and prepared as follow: 44.3 g
1492 Bioconjugate Chem., Vol. 20, No. 8, 2009
Martı´nez-Barbosa et al.
Table 1. Physicochemical Characteristics of Nanoparticles Constituted by PBLG-PEG-60 and PBLG-Bnz-50 polymer ratio (wt %)
mean size (nm)b
ζ potential (mV)c
100/0a 90/10a 77/23a 40/60a 0/100a
58 ( 20 62 ( 20 60 ( 21 54 ( 17 85 ( 30
-13.3 ( 0.6 -12.6 ( 0.3 -13.6 ( 0.3 -16.1 ( 0.6 -21.1 ( 0.2
a
PBLG-PEG-60/PBLG-Bnz-50 ratio (in wt %) in the formulation. Measured using dynamic laser light scattering, n ) 3. c Measured using a Zetasizer 4 Instrument, n ) 5. b
of Tris(hydroxy-methyl-amino-methane), 22.4 g of phenobarbital, and 0.53 g of lactate calcium were dissolved in 1 L of water at 50 °C, sodium azide was added to avoid bacterial development, pH was adjusted to pH ) 8.6, if necessary. C3 and C3b migration was followed by Coomasie blue staining. The height of the peaks given by the immunoprecipitation of C3 and C3b gave a semiquantitative estimation of the amount of each component in the serum. The complement activation factor (CAF) was expressed as the ratio of the peaks height as follows: CAF ) C3b/(C3 + C3b)
(4)
RESULTS Nanoparticle Preparation and Characterization. Possibilities to prepare, by nanoprecipitation, nanoparticles, either from one polymer (PBLG-Bnz-50 or PBLG-PEG-60) or from mixtures of these two polymers (composite nanoparticles) in different ratios, have been evaluated. As shown in Table 1, nanoparticles could be easily and reproducibly obtained without any surfactant agents by a slightly modified method described elsewhere (29). THF was selected, being both a good solvent of the polymer and miscible to water. Nanoparticle diameter, determined by dynamic laser light scattering, was typically lower than 100 nm. From transmission electron microscopy (Figure 1), it could be seen that nanoparticles prepared from both polymers exhibited ellipsoidal geometries, whose dimensions were compatible with dynamic light scattering measurements. Moreover, it can be seen that particles issued from mixtures of the two polymers were quite homogeneous, suggesting that they formed a single population. The ζ potential (Table 1) was similar for nanoparticles at 100/0, 90/10, and 77/23 wt % ratio. A weak decrease in the ζ potential was observed for a 40/60 wt % ratio. As expected, the stronger ζ potential value was found for non-PEGylated nanoparticles (0/100 wt % ratio). Isothermal Titration Microcalorimetry (ITC) Studies. In order to evaluate the capacity of nanoparticles to avoid protein adsorption and, therefore, to confirm the presence of PEG chains at the surface of nanoparticles, isothermal titration microcalorimetry (ITC) has been applied. PEGylated (PBLG-PEG-60) and non-PEGylated (PBLG-Bnz-50) nanoparticle-protein interactions were measured using bovine serum albumin (BSA) as model globular protein. As shown by Figure 2, the interaction of BSA with the nanoparticle surface is an exothermic phenomenon. The area underneath each injection peak is equal to the total heat released for that injection. In the case of nanoparticles constituted by PBLG-Bnz-50 (Figure 2a), signals after each BSA injection were much stronger than signals of nanoparticles constituted by PBLG-PEG-60 (Figure 2b). The corresponding control dilution experiment consisted of BSA injected in water and is shown by Figure 2c. Evaluation of Complement Activation by Crossed Immuno-Electrophoresis. Figure 3 presents the immuno-electrophoresis patterns of the complement proteins C3 and C3b in serum incubated with the different nanoparticles prepared and
of control experiments. The control experiments performed under experimental conditions in the absence of nanoparticles (Figure 3a) showed a major peak of protein C3 and a small peak of protein C3b. It gives the complement activation background under experimental conditions. The CAF value determined for the background activation was 0.16. In the absence of divalent ions (Figure 3b), no C3b peak can be seen on the immuno-electrophoregram in agreement with the fact that activation of the complement system requires divalent cations. The immuno-electrophoregram of serum incubated with nanoparticles presented both C3 and C3b peaks. The height of the C3b peak was less pronounced for the serum incubated with PEGylated PBLG-PEG-60 nanoparticles (Figure 3d) than the one observed for serum incubated with non-PEGylated PBLGBnz-50 nanoparticles (Figure 3c). As reported in Table 2, the CAF was higher for serum incubated with the non-PEGylated nanoparticles (0.40) than the CAF measured for serum incubated with the PEGylated nanoparticles (0.21), indicating that nonPEGylated particles could be rapidly opsonized by serum proteins. The CAF value of the serum incubated with the PEGylated nanoparticles was close to the CAF value of the background activation of the complement under experimental conditions. The immuno-electrophoresis patterns of the serum incubated with the different composite nanoparticles (Figure 3e,f,g) were very similar to the immuno-electrophoresis pattern shown for PBLG-PEG-60 nanoparticles (Figure 3d). The CAF values reported in Table 2 only showed a slight shift toward a higher value for the serum incubated with the composite nanoparticles containing the lower percentage (40 wt %) of PBLG-PEG-60 polymer. The CAF values obtained for serum incubated with the two other composite nanoparticles were the same than the CAF of serum incubated with nanoparticles prepared with 100 wt % PBLG-PEG-60 polymer.
DISCUSSION The development of efficient targeting systems necessitates the simultaneous introduction of a set of functionalities within the system itself, regardless of its physical structure. Various polymers have been used to prepare nanoparticles for pharmaceutical applications, and different strategies have been proposed to introduce specific functionalities, e.g., optimized drug encapsulation, controlled release properties, stealthiness in the bloodstream, organ or cell targeting by ligands. It has been shown previously (27) that copolymers of poly(γ-benzyl Lglutamate), PBLG, bearing pharmaceutically interesting moieties could be quite easily prepared. Interestingly, mixtures of these copolymers can easily self-associate, which offers a convenient means for introducing by a soft technique desired functionalities within the particles. In the present work, we investigated the possibility to precisely tune nanoparticle surfaces by introducing an optimal amount of a PEGylated copolymer of PBLG within the structure of nanoparticles. The nanoprecipitation method could be used conveniently to prepare nanoparticles either from one polymer (PBLG-Bnz-50 or PBLG- PEG-60) or from mixtures of these two polymers (composite nanoparticles) in different ratios. Nanoparticles could be easily and reproducibly obtained without any surfactants by a slightly modified method described elsewhere (29) as a result of the hydrophobic nature of PBLG. Moreover, DLS and TEM measurements showed that very small particles (lower than 100 nm) could be prepared. In addition, the homogeneous aspect of the particle shapes for the different preparations made from the polymer mixtures suggested that in this case the suspensions were composed of a single population of composite nanoparticles.
PEGylated Degradable Composite NPs
Bioconjugate Chem., Vol. 20, No. 8, 2009 1493
Figure 1. Transmission electron microscope micrographs of the nanoparticles. Simple nanoparticles constituted by (a) PBLG-PEG-60 (100/0 wt %) and (b) PBLG-Bnz-50 (0/100 wt %); Composite nanoparticles constituted by PBLG-PEG-60/PBLG-Bnz-50 at (c) 90/10, (d) 77/23, and (e) 40/60 (wt %), respectively.
Figure 2. Plots of heat flow vs time of microcalorimetric titration of (a) PBLG-Bnz-50 nanoparticles and (b) PBLG-PEG-60 nanoparticles, by BSA solution. Injections of 10 µL, at 25 °C, of 5.4 × 10-2 mM of BSA solution were made on nanoparticle suspensions (2.7 × 10-2 mM). Control of BSA dilution (c) was carried out by injections of BSA solution in water.
Characterization by ITC of PBLG-PEG-60 or PBLG-Bnz50 nanoparticle-protein interactions using bovin serum albumin (BSA), a model globular protein, permitted confirmation of the difference in surface properties of PEGylated and non-PEGylated nanoparticle suspensions. Theoretically, PEG chains of PEGylated nanoparticles should avoid or, at least, reduce protein adsorption on the nanoparticle surface as a result of steric repulsions and hydrophilization induced by PEG chains. In the case of nanoparticles constituted by PBLG-Bnz-50 (Figure 2a), signals after each BSA injection were higher than signals observed for nanoparticles constituted by PBLG-PEG-60 (Figure 2b). From this result, it could be concluded that interactions of BSA with nanoparticle surfaces were stronger in the absence of PEG chains. Quantitative interpretation of observed enthalpy effect after an adsorption process (∆Ηads) could be difficult because various
factors can contribute to this effect. For example, Norde and Lyklema (8, 31) took into account several factors while studying adsorption of human plasma albumin and bovin pancreas ribonuclease on negatively charged polystyrene surfaces. Some of these factors were as follows: the structural changes that the protein molecule and other involved molecules (for example, water and electrolyte) can undergo during the adsorption process; the alterations in the hydration state, disruption of the surface/solvent contacts, and simultaneous formation of protein/ surface contacts; the modification of protein/solvent interactions due to differences of the chain density between dissolved and adsorbed protein molecules; and the effect of partners dilution. Because such factors have not yet been elucidated for PBLGBnz-50 and PBLG-PEG-60 systems titrated with BSA, interpretation of ∆Ηads remains difficult. However, as shown in Figure 2, these studies made possible a qualitative surface characterization of tested nanoparticles. From these results, enthalpies of interactions could be obtained after integration of injection signals and subtraction of control (Figure 2c). Values of ∆Ηads ) 414 kCal/mol and 83 kCal/mol were obtained for the interaction of BSA with PBLG-Bnz-50 and PBLG-PEG-60 nanoparticles, respectively. Thus, it could be concluded from the existence of this 5-fold factor between adsorption enthalpies that interactions of BSA with non-PEGylated nanoparticles were much important than those observed for PEGylated nanoparticles. In Vitro evaluation of complement activation induced by nanoparticles is a very convenient method that can be used to predict their fate after intravenous injection in ViVo (4, 32). Indeed, most of the nanoparticles inducing a strong activation of the complement systems are generally rapidly taken up by macrophages of the mononuclear phagocyte systems, while nanoparticles appearing as low or non-activators of the complement systems are not recognized by the macrophages and remain in the bloodstream. In the present work, we have investigated the activation of the complement system induced by both PBLGPEG-60 nanoparticles and PBLG-Bnz-50 nanoparticles, as well as by composite nanoparticles made from blends of both types of polymers at different composition. The CAF value of PBLGBnz-50 nanoparticles was much higher than the CAF value of
1494 Bioconjugate Chem., Vol. 20, No. 8, 2009
Martı´nez-Barbosa et al.
Figure 3. Crossed immunoelectrophoresis of C3 antigens in human serum diluted 1:4 after incubation in the presence of nanoparticles (500 cm2). (A) Serum/VBS2+ (control of complement activation), background. (B) Serum/VBS2+-EDTA (negative control). (C) Serum/VBS2+/PBLG-Bnz-50 nanoparticles. (D) Serum/VBS2+/PBLG-PEG-60 nanoparticles. (E) Serum/VBS2+/(PBLG-PEG-60/PBLG-Bnz-50, 90/10 wt %) nanoparticles. (F) Serum/VBS2+/(PBLG-PEG-60/PBLG-Bnz-50, 77/23 wt %) nanoparticles. (G) Serum/VBS2+/(PBLG-PEG-60/PBLG-Bnz-50, 40/60 wt %) nanoparticles. Table 2. Relationship between PBLG-PEG-60 Content into the PBLG-PEG-60/PBLG-Bnz-50 Nanoparticles and the Complement Activation Factor controls nanoparticles polymer ratio (wt%)
sample
CAF
background activation serum + EDTA 100/0 90/10 77/23 40/60 0/100
0.16 0.05 0.21 0.21 0.21 0.25 0.40
the PBLG-PEG-60 nanoparticles. This result indicated the tendency of uncoated nanoparticles to activate the complement system contrarily to nanoparticles bearing PEG chains. These nanoparticles which gave a CAF value close to the background activation of the serum under experimental conditions appeared as low complement activating nanoparticles. The results observed here are in good agreement with the literature, which reports that nanoparticles prepared with PEG-containing polymers are generally low complement activators due to the presence of PEG chains at the nanoparticle surface (4, 5). Thus, the results also suggested that the organization of PBLG-PEG60 polymer chains to form the particle was highly efficient to expose PEG moieties at the surface of nanoparticles, ensuring a coating layer able to avoid complement activation. All these
results were also in good agreement with the results obtained from the evaluation of PEGylated and non-PEGylated nanoparticle interactions with BSA. Indeed, the nanoparticles showing the lowest adsorption of BSA were also the lowest complement activators in agreement with the fact that surfaces of such nanoparticles are generally less opsonized as a result of the presence of hydrophilic PEG chains. In contrast, nonPEGylated nanoparticles, which adsorbed higher amounts of BSA, appeared as the highest complement activators in agreement with the fact that they are generally strongly opsonized by seric proteins. Composite nanoparticles also displayed a low capacity to induce activation of the complement system. It is noteworthy that composite nanoparticles containing 90 and 77 wt % of PBLG-PEG-60 showed identical behavior to nanoparticles composed of 100 wt % of this polymer. The PEG coating formed at the surface of these nanoparticles seemed to have identical properties to the one formed on the surfaces of 100 wt % of PBLG-PEG-60 based nanoparticles. Interestingly, the three types of nanoparticles giving the lowest CAF value also showed identical ζ potential suggesting that they showed similar properties (Figure 4). The composite nanoparticles containing the lowest amount of PBLG-PEG-60 (40 wt %) showed a slightly higher capacity to induce complement activation as suggested by their CAF,
PEGylated Degradable Composite NPs
Bioconjugate Chem., Vol. 20, No. 8, 2009 1495
CONACyT (Mexico) for the financial support, which enabled her to conduct this study. TEM experiments where performed at the CCME (U. Paris-Sud) facility and authors acknowledge Jeril Degrouard for technical support and fruitful help. The authors would like to acknowledge Laurent Fontaine and Ve´ronique Montembault for their advice and support in the synthesis of the polymers used in the formulation of the nanoparticles.
LITERATURE CITED
Figure 4. Evolution of ζ potential and complement activation factor as a function of PBLG-PEG-60 ratio (wt %) in nanoparticles constituted of PBLG-PEG-60 and PBLG-Bnz-50. Circle: Complement activation factor. Square symbols: ζ potential.
which was shifted toward higher values. When combining this result with the fact that the ζ potential was also shifted toward the value of the ζ potential measured for nanoparticles composed of 100 wt % non-PEG polymer, it can be suggested that the PEG coating characteristics of the surface of these nanoparticles was modified in such a way that it reduced the effectiveness to prevent complement activation. The lower value of the ζ potential shown by these nanoparticles was in favor of a modification of the surface composition of the nanoparticles, which can be attributed to a decrease in the density of the PEG chains at the nanoparticle surface. Thus, a higher contribution of the core polymer can be associated with the measured ζ potential. This may also contribute to increasing the capacity of inducing complement activation, hence increasing the chance to be taken up by macrophages after intravenous administration in animals. Finally, from a practical point of view, it is interesting to note that low complement activation characteristics could be obtained with only 77% of PBLG-PEG copolymer chains in the particles, making possible to imagine composite particles containing not only PBLG-PEG copolymer (for ensuring stealth properties in the body), but also other copolymers bearing other chemical moieties, such as recognition ligands, in order to confer simultaneously other properties to the particles.
CONCLUSIONS The results reported in this contribution permit one to demonstrate that nanoparticles based on PEGylated derivatives of PBLG with a low complement activation pattern can be obtained by nanoprecipitation, an easy and reproducible method. Because of this property, it is strongly expected that these nanoparticles will behave as stealth nanoparticles after in ViVo administration by the intravenous route. Isothermal titration microcalorimetry and complement activation measurements allowed us to conclude that recognition of PEGylated PBLGbased nanoparticles by proteins is significantly lower than recognition of non-PEGylated nanoparticles. It is worth noting that composite nanoparticles based on a mixture of PBLG-PEG60 and PBLG polymers have been prepared by nanoprecipitation and present similar properties to PBLG-PEG-60 based nanoparticles, suggesting that it may prove possible to introduce in the particles reasonable amounts of other structurally similar copolymers chains bearing other ligands, without modifying their complement activation properties. This would help to adjunct on demand additional functionalities to the particles.
ACKNOWLEDGMENT Pr. Denis Labarre is warmly acknowledged for fruitful discussions concerning the discussion of the complement activation experiments. M.EMB would like to thank the
(1) Lee, V. H. L. (2004) Nanotechnology: challenging the limit of creativity in targeted drug delivery. AdV. Drug DeliVery ReV. 56, 1527–1528. (2) Brannon-Peppas, L., and Blanchette, J. O. (2004) Nanoparticle and targeted systems for cancer therapy. AdV. Drug DeliVery ReV. 56, 1649–1659. (3) Illum, L., Hunneyball, I. M., and Davis, S. S. (1986) The effect of hydrophilic coatings on the uptake of colloidal particles by the liver and the peritoneal macrophages. Int. J. Pharm. 29, 53– 65. (4) Vonarbourg, A., Passirani, C., Saulnier, P., and Benoit, J.-P. (2006) Parameters influencing the stealthness of colloidal drug delivery systems. Biomaterials 27, 4356–4373. (5) Owens, III, D. E., and Peppas, N. A. (2006) Opsonization, biodistribution, and pharmacokinetics of polymeric nanoparticles. Int. J. Pharm. 307, 93–102. (6) Gref, R., Luck, M., Quellec, P., Marchand, M., Dellacherie, E., Harnisch, S., Blunk, T., and Muller, R. H. (2000) ‘Stealth’ corona-core nanoparticles surface modified by polyethylene glycol (PEG): influences of the corona (PEG chain length and surface density) and of the core composition on phagocytic uptake and plasma protein adsorption. Colloids Surf., B 18, 301– 313. (7) Passirani, C., Barratt, G., Devissaguet, J. P., and Labarre, D. (1998) Interactions of nanoparticles bearing heparin or dextran covalently bound to poly(methyl methacrylate) with the complement system. Life Sci. 62, 775–785. (8) Norde, W., and Lyklema, J. (1978) The adsorption of human plasma albumin and bovine pancreas ribonuclease at negatively charged polystyrene surfaces. V. Microcalorimetry. J. Colloid Interface Sci. 66, 295–302. (9) Gref, R., Minamitake, Y., Peracchia, M. T., Trubetskoy, V., Torchilin, V., and Langer, R. (1994) Biodegradable longcirculating polymeric nanospheres. Science 263, 1600–1603. (10) Kataoka, K., Kwon, G. S., Yokoyama, M., Okano, T., and Sakurai, Y. (1993) Block copolymer micelles as vehicles for drug delivery. J. Controlled Release 24, 119–132. (11) Nah, J. W., Jeong, Y. I., and Cho, C. S. (1998) Clonazepam release from core-shell type nanoparticules composed of poly (γ-benzyl L-glutamate) as the hydrophobic part and poly(ethylene oxide) as the hydrophilic part. J. Polym. Sci., Part B. 36, 415– 423. (12) Peracchia, M. T., Gref, R., Minamitake, Y., Domb, A., Lotan, N., and Langer, R. (1997) PEG-coated nanospheres from amphiphilic diblock and multiblock copolymers: investigation of their drug encapsulation and release characteristics. J. Controlled Release 46, 223–231. (13) Kwon, G. S., Yokoyama, M., Okano, T., Sakurai, Y., and Kataoka, K. (1993) Biodistribution of micelle-forming polymerdrug conjugates. Pharm. Res. 10, 970–974. (14) Kwon, G., Naito, M., Yokoyama, M., Okano, T., Sakurai, Y., and Kataoka, K. (1993) Micelles based on AB block copolymers of poly(ethylene oxide) and poly(β-benzyl L-aspartate). Langmuir 9, 945–949. (15) Anderson, J. M., Gibbons, D. F., Martin, R. L., Hiltner, A., and Woods, R. (1974) The potential for poly-R-amino acids as biomaterials. J. Biomed. Mater. Res. 5, 197–207. (16) Yang, J. Z., Antoun, S., Ottembrite, R. M., and Milstein, S. (1996) Amino acid oligomer microspheres as drug delivery systems. I. Synthesis of oligo(amino acids) via NCAs and their
1496 Bioconjugate Chem., Vol. 20, No. 8, 2009 microsphere formation. J. BioactiVe Compat. Polym. 11, 219– 235. (17) Yang, J. Z., Wang, M., Ottembrite, R. M., and Milstein, S. (1996) Amino acid oligomer microspheres as drug delivery systems. II. Oligomerization of NCAs initiated with L-tyrosine methyl ester. J. BioactiVe Compat. Polym. 11, 236–247. (18) Anderson, J. M., Spilizewski, K. L., and Hiltner, A. (1985) Biocompatibility of Tissue Analogs, Volume I (Williams, D. F., Ed.) pp 68-88, Chapter 4, CRC Press, Boca Raton, FL. (19) Hirayama, C., Ihara, H., Ishimara, S. I., and Imanishi, Y. (1994) Preparation of porous spherical poly(γ-methyl L-glutamate) particles containing β-structure. Polym. J. 26, 61–66. (20) McCormick, T. L. A., Sgouras, D., and Duncan, R. (1989) Poly(amino acid) copolymers as a potential soluble drug delivery system. 2. Body distribution and preliminary biocompatibility testing in vitro and in vivo. J. BioactiVe Compat. Polym. 4, 252–268. (21) Hayashi, T., and Iwatsuki, M. (1990) Biodegradation of copoly(L-aspartic acid/L-glutamic acid) in vitro. Biopolymers 29, 549–557. (22) Oh, I., Leek, K., Kwon, H. Y., Lee, Y. B., Shin, S. C., Cho, C. S., and Kim, C. K. (1999) Release of adriamycin from poly(γbenzyl glutamate)/poly(ethylene oxide) nanoparticles. Int. J. Pharm. 181, 107–115. (23) Cho, C. S., Nah, J. W., Jeong, Y. I., Cheon, J. B., Asayama, S., Ise, H., and Akaike, T. (1999) Conformational transition of nanoparticles composed of poly(γ-benzyl glutamate) as the core and poly(ethylene oxide) as the shell. Polymer 40, 6769–6775. (24) Kwon, G. S., Naito, M., Yokoyama, M., Okano, T., Sakurai, Y., and Kataoka, K. (1995) Physical entrapment of adriamycin in AB block copolymer micelles. Pharm. Res. 12, 192–195. (25) Nah, J. W., Jeong, Y. I., Cho, C. S., and Kim, S. I. (2000) Drug-delivery system based on core-shell-type nanoparticles composed of poly (γ-benzyl-L-glutamate) and poly(ethylene oxide). J. Appl. Polym. Sci. 75, 1115–1126.
Martı´nez-Barbosa et al. (26) Jeong, Y. I., Seo, S. J., Park, I. K., Lee, H. C., Kang, I. C., Akaike, T., and Cho, C. S. (2005) Cellular recognition of paclitaxel-loaded polymeric nanoparticles composed of poly(γbenzyl L-glutamate) and poly(ethylene glycol) diblock copolymer endcapped with galactose moiety. Int. J. Pharm. 296, 151–161. (27) Martı´nez Barbosa, M. E., Montembault, V., Cammas-Marion, S., Ponchel, G., and, and Fontaine, L. (2007) Synthesis and characterization of novel poly(γ-benzyl L-glutamate) derivatives tailored for the preparation of nanoparticles of pharmaceutical interest. Polym. Int. 56, 317–324. (28) Temyanko, E., Russo, P. S., and Ricks, H. (2001) Study of rodlike homopolypeptides by gel permeation chromatography with light scattering detection: validity of universal calibration and stiffness assessment. Macromolecules 34, 582–586. (29) Thioune, O., Fessi, H., Devissaguet, J. P., and Puisieux, F. (1997) Preparation of pseudolatex by nanoprecipitation: influence of the solvent nature on intrinsic viscosity and interaction constant. Int. J. Pharm. 146, 233–238. (30) Laurell, C. B. (1966) Quantitative estimation of proteins by electrophoresis in agarose gels containing antibodies. Anal. Biochem. 15, 45–52. (31) Norde, W., and Lyklema, J. (1978) The adsorption of human plasma albumin and bovine pancreas ribonuclease at negatively charged polystyrene surfaces. I. Adsorption isotherms. Effects of charge, ionic strength, and temperature. J. Colloid Interface Sci. 66, 257–265. (32) Labarre, D., Vauthier, C., Chauvierre, C., Petri, B., Muller, R., and Chehimi, M. M. (2005) Interactions of blood proteins with poly(isobutylcyanoacrylate) nanoparticles decorated with a polysaccharidic brush. Biomaterials 26, 5075–5084. BC900017C
