Block Copolymer Micelle Complexes

Langmuir 2008, 24, 12189-12195

12189

Formation of Annexin-A5 Protein/Block Copolymer Micelle Complexes: QCM-D and PAGE Experiments Vanessa Schmidt,† Cristiano Giacomelli,† Celine Gounou,‡ Josephine Lai-Kee-Him,‡ Alain R. Brisson,*,‡ and Redouane Borsali*,§ Laboratoire de Chimie des Polyme`res Organiques (LCPO)-ENSCPB-UniVersite´ Bordeaux 1, 16 AVenue Pey Berland, 33607 Pessac Cedex, France, Laboratoire de Chimie et Biologie des Membranes et Nanoobjets (CBMN, UMR-5248) and Institut Europe´en de Chimie et Biologie, AVenue des Faculte´s, 33405 Talence, France, and Centre de Recherche sur les Macromole´cules Ve´ge´tales (CERMAV, UPR 5301) and UniVersite´ Joseph Fourier, BP53, 38041, Grenoble Cedex 9, France ReceiVed July 31, 2008. ReVised Manuscript ReceiVed September 1, 2008 The Annexin-A5 (Anx5) protein is a specific marker of the exposure of phosphatidylserine molecules at the surface of cells, which occurs in processes such as apoptosis and platelet activation. Decoration of self-assembled block copolymer nanostructures by Anx5 is of particular interest in micelle-mediated target drug delivery or in vivo magnetic resonance imaging, the Anx5 imparting (bio)functionality to the system. In this work, the reversible binding of the Anx5 onto polystyrene-b-poly(2-phosphatethyl methacrylate-co-2-hydroxyethyl methacrylate) (PS-b-P(PEMA-coHEMA)) block copolymer micelles in the presence of Ca2+ ions is described using Quartz crystal microbalance with dissipation monitoring (QCM-D) and polyacrylamide gel electrophoresis (PAGE) analysis. QCM-D experiments confirmed the binding process as well as its reversibility and dependence on the characteristics of macromolecular assemblies, such as the number of phosphonic diacid groups (Pmic) and hydrodynamic diameter (2RH). A linear relationship between the amount of micelles and the amount of protein bound onto the micelle surface until a saturation point was established by QCM-D. The amount of Anx5 bound to PS-b-P(PEMA-co-HEMA) micelles was successfully quantified by PAGE experiments in nondenaturing conditions, which also corroborated that the binding process is mediated by Ca2+ ions. The ability of such surface (bio)-functionalized nanoparticle systems to stabilize and transport hydrophobic loads was highlighted by transmission electron microscopy (TEM) of assemblies with entrapped iron oxide particles.

Introduction Eukaryotic cell membranes present an asymmetric distribution of lipid molecules, the majority of phosphatidylserine molecules being located in the inner leaflet of the membrane bilayer.1,2 The regulated exposure of phosphatidylserine on the outer leaflet of cell membranes is a hallmark of two basic processes, apoptosis or programmed cell death3 and platelet activation.4 The AnnexinA5 protein (Anx5), a member of the annexin protein family, is a 35 kDa hydrophilic molecule with an isoelectric point (pI) of 4.5. The annexins usually contain 4- or 8-fold repeat of domains with about 70 amino acid residues each. In the Anx5, such domains are arranged in a cyclic way, giving the molecule an overall flat, slightly curved shape with the Ca2+ binding sites located on the concave membrane-binding face.5,6 The ability of Anx5 to bind with high affinity to membranes exposing negatively charged lipids in a Ca2+-dependent manner7 explains the use of Anx5* To whom correspondence should be addressed. E-mail: A.R.B. a.brisson@ iecb.u-bordeaux.fr; R.B. [email protected]. † LCPO-ENSCPB-Universite´ Bordeaux. ‡ CBMN-IECB-Universite´ Bordeaux. § CERMAV-Universite´ Joseph Fourier.

(1) Devaux, P. F. Biochemistry 1991, 30, 1163–1173. (2) Williamson, P.; Schlegel, R. A. Mol. Membr. Biol. 1994, 11, 199–216. (3) Fadok, V. A.; Voelker, D. R.; Campbell, P. A.; Cohen, J. J.; Bratton, D. L.; Henson, P. M. J. Immunol. 1992, 148, 2207–2216. (4) Bevers, E. M.; Comfurius, P.; Zwaal, R. F. A. Biochim. Biophys. Acta 1983, 736, 57–66. (5) Sopkova-de Oliveira Santos, J.; Fischer, S.; Guilbert, C.; Lewit-Bentley, A.; Smith, J. C. Biochemistry 2000, 39, 14065–14074. (6) Ko¨hler, G.; Hering, U.; Zscho¨rnig, O.; Arnold, K. Biochemistry 1997, 36, 8189–8194. (7) Gerke, V.; Creutz, C. E.; Moss, S. E. Nat. ReV. Mol. Cell Biol. 2005, 6, 449–461.

labeled proteins as research tools8,9 and in vivo diagnostic agents to detect cell death in cancer chemotherapy, organ transplant rejection, and myocardial infarction.10-13 Development of chimerical Anx5 entities obtained by coupling Anx5 to various types of bioactive moieties opens exciting avenues for the targeted delivery of therapeutic compounds.14-17 Consequently, great potential for simultaneous in vivo monitoring and treatment is expected for Anx5 protein/drug carrier assemblies. The ability of block copolymers to self-assemble into ordered nano-objects (micelles, vesicles, lamellae, etc.) if dissolved in a selective solvent has been largely explored in modern (8) Dachary-Prigent, J.; Freyssinet, J. M.; Pasquet, J. M.; Carron, J. C.; Nurden, A. T. Blood 1993, 81, 2554–2565. (9) Vanoers, M. H. J.; Reutelingsperger, C. P. M.; Kuyten, G. A. M.; Vondemborne, A.; Koopman, G. Blood 1994, 84, A291-A291. (10) Blankenberg, F. G.; Katsikis, P. D.; Tait, J. F.; Davis, R. E.; Naumovski, L.; Ohtsuki, K.; Kopiwoda, S.; Abrams, M. J.; Darkes, M.; Robbins, R. C.; Maecker, H. T.; Strauss, H. W. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 6349–6354. (11) Narula, J.; Acio, E. R.; Narula, N.; Samuels, L. E.; Fyfe, B.; Wood, D.; Fitzpatrick, J. M.; Raghunath, P. M.; Tomaszewski, J. E.; Kelly, C.; Steinmetz, N.; Green, A.; Tait, J. F.; Leppo, J.; Blankenberg, F. G.; Jain, D.; Strauss, H. W. Nat. Med. 2001, 7, 1347–1352. (12) Thimister, P. W.; Hofstra, L.; Liem, I. H.; Boersma, H. H.; Kemerink, G.; Reutelingsperger, C. P.; Heidendal, G. A. J. Nucl. Med. 2003, 44, 391–396. (13) Tait, J. F.; Gibson, D. F.; Smith, C. Anal. Biochem. 2004, 329, 112–119. (14) Tanaka, K.; Einaga, K.; Tsuchiyama, H.; Tait, J. F.; Fujikawa, K. Biochemistry 1996, 35, 922–929. (15) Kenis, H.; van Genderen, H.; Bennaghmouch, A.; Rinia, H. A.; Frederik, P.; Narula, J.; Hofstra, L.; Reutelingsperger, C. P. M. J. Biol. Chem. 2004, 279, 52623–52629. (16) Berat, R.; Remy-Zolghadry, M.; Gounou, C.; Manigand, C.; Tan, S.; Salto, C.; Arenas, E.; Bordenave, L.; Brisson, A. R. Biointerphases 2007, 2, 165–172. (17) Brisson, A. R. A deVice for binding a targeting entity to a bait entity and detection methods using the same. WO2005/114192, 2005.

10.1021/la8024815 CCC: $40.75  2008 American Chemical Society Published on Web 10/02/2008

12190 Langmuir, Vol. 24, No. 21, 2008

Schmidt et al.

Table 1. Physico-Chemical Characteristics of the Polymeric Structures and QCM-D Data (frequency (∆F), Adsorbed Mass (∆m), and Dissipation (∆D)) Related to Formation of Anx5/PS-b-P(PEMA-co-HEMA) Assemblies entry PSx-b-P(PEMAa-co-HEMAb)yx-(a-b)y Mw,mica (g/mol) Naggb 2RHc(nm) Pmicd (at./mic) ∆F (Hz) ∆me (ng/cm2) ∆D (×1 × 10-6) 1 2 c

30-(15-55)70 30-(48-22)70

3.4 × 106 2.3 × 106

253 143

14 18

3.8 × 103 6.9 × 103

55 76

974 1350

2.2 2.0

a Determined by static light scattering (SLS). b Calculated using the data obtained by static light scattering (SLS) and gel permeation chromatography (GPC). Determined by dynamic light scattering (DLS). d Number of binding sites (phosphonic diacid groups) per micelle. e Calculated using the Saurbrey equation.

science.18-22 These structures are potential delivery vehicles for pharmaceuticals and gene therapy agents due to the fact that they can stabilize, within their core, hydrophobic molecules with otherwise limited water solubility and controllably release them.21,23,24 In some cases, the high toxicity of potent antitumor drugs to healthy cells is decreased through encapsulation. Currently, the challenge is to give (bio)functionality and guide such structures toward specific sites. Using controlled polymer synthesis, we recently synthesized amphiphilic block copolymer systems carrying phosphonic diacid groups at the corona, harboring the ability to bind the Anx5 protein in a similar way to natural phospholipid structures.25,26 Herein, we describe the reversible binding of Anx5 proteins onto polystyrene-b-poly(2-phosphatethyl methacrylate-co-2hydroxyethyl methacrylate) (PS-b-P(PEMA-co-HEMA)) block copolymer micelles in the presence of Ca2+ ions. Using quartz crystal microbalance with dissipation monitoring (QCM-D) and polyacrylamide gel electrophoresis (PAGE) in nodenaturing conditions, the Anx5 binding ability and efficiency is described as a function of the characteristics of macromolecular assemblies, namely, the number of phosphonic diacid groups (binding sites) located at the micelle corona, aggregation number, and hydrodynamic micellar size.

Experimental Section Materials. 1,2-Dioleoyl-sn-glycero-3-phosphocholine (DOPC) and 1,2-dioleoyl-sn-glycero-3-[phospho-L-serine] (DOPS) were purchased from Avanti Polar Lipids (Alabama). All other chemicals were of ultrapure grade. Water was purified with a RiOs system (Millipore, France). QCM-D sensor crystals, covered with 100-nm evaporated gold and reactively sputter coated with 50-nm silicon oxide, were purchased from Q-Sense (Gothenburg, Sweden). Recombinant Anx5 was produced as described in Richter et al.27 Nanoparticles Preparation. The synthesis and characterization of polystyrene-b-poly(2-phosphatethyl methacrylate-co-2-hydroxyethyl methacrylate) (PSx-b-P(PEMAa-co-HEMAb)y (here and throughout the text subscripts x, a, b, and y (with y ) a + b) refer to the mean degree of polymerization) amphiphilic macromolecules (Table 1), and their respective self-assembly behavior in aqueous solution have been recently described.25 In short, dilute nanoparticle aqueous solutions (Cp ) 1.0 mg/mL) were prepared using the indirect dissolution method. Typically, 3.5 mg of polymer was dissolved in 0.5 mL of acetone in a closed vial. The solution was allowed to stir for at least 3 h. Micellization was subsequently induced by slow (6.0

mL/h) addition of 7.0 mL of Tris buffer at pH ) 8.0. Next, the solution was stirred for at least 2 days, thus allowing evaporation of the organic solvent. Small unilamellar vesicles (SUV) made from DOPC/DOPS 4/1 w/w were prepared by sonication as reported by Richter at al.28 Polyacrylamide Gel Electrophoresis (PAGE). PAGE was performed in 1.0 mm thick slab gels. Separating gels contained 8% acrylamide and 0.05% bis-acrylamide in 400 mM Tris (pH ) 8.8). Before deposition on the gel, samples (20-40 µL maximum volume) were mixed with 5 µL of a solution containing bromophenol blue in 30% glycerol. The migration buffer was 25 mM Tris, 192 mM glycine (pH ) 8.3). Experiments were carried out in the absence of SDS to prevent protein denaturation and possible interaction with block copolymer micelles. In order to better understand the calcium dependency on the binding process, experiments were performed both in the absence and in the presence of 2 mM Ca2+. Gels were stained with Coomassie brilliant blue according to standard protocols. Quartz Crystal Microbalance with Dissipation Monitoring (QCM-D). QCM-D measurements were performed with the Q-SENSE D300 system equipped with an Axial Flow Chamber (QAFC 301) (Q-SENSE AB, Gothenburg, Sweden).28,29 Briefly, upon interaction of (soft) matter with the surface of a sensor crystal, changes in the resonance frequency (F) related to attached mass (including coupled water) and in the dissipation (D) related to frictional (viscous) losses in the adlayer are measured with a time resolution of better than 1 s. Measurements in liquid environment were performed at a working temperature of 24 °C in exchange mode. Resonance frequency and dissipation were measured at several harmonics (15, 25, 35 MHz) simultaneously. If not stated otherwise, (i) changes in dissipation and normalized frequency (∆F norm ) ∆fn/n, with n being the overtone number) are presented and (ii) adsorbed masses (∆m) are calculated according to the Sauerbrey equation, ∆m ) -C∆F norm, with -C ) 17.7 ng cm-2 Hz-1.30 Unless otherwise specified, polymer concentrations (Cp) of 1.0 mg/ mL were used. Transmission Electron Microscopy (TEM). TEM images were recorded using a CM 120 Philips microscope operating at 120 kV and equipped with a USC1000-SSCCD 2k × 2k Gatan camera. To prepare the TEM samples, 5 µL of an aqueous solution of block copolymer micelles was deposited onto a carbon-coated copper grid, which was rendered hydrophilic by UV/ozone treatment. Excess micelle solution was gently removed using absorbent paper. Samples were then negatively stained by adding a 5 µL droplet of 2% sodium phosphotungstate solution at pH 7.5, and the excess solution was again removed prior to drying under ambient conditions.

Results and Discussion (18) Lodge, T. P.; Pudil, B.; Hanley, K. J. Macromolecules 2002, 35, 4707– 4717. (19) Discher, B. M.; Won, Y.-Y.; Ege, D. S.; Lee, J. C.-M.; Bates, F. S.; Discher, D. E.; Hammer, D. A. Science 1999, 284, 1143–1146. (20) Discher, B. M.; Hammer, D. A.; Bates, F. S.; Discher, D. E. Curr. Opin. Colloid Interface Sci. 2000, 5, 125–131. (21) Allen, C.; Maysinger, D.; Eisenberg, A. Colloid Surf. B: Biointerfaces 1999, 16, 3–27. (22) Riess, G. Prog. Polym. Sci. 2003, 28, 1107–1170. (23) Kataoka, K.; Harada, A.; Nagasaki, Y. AdV. Drug DeliV. ReV. 2001, 47, 113–131. (24) Duncan, R. Nat. ReV. Drug DiscoV. 2003, 2, 347–360. (25) Schmidt, V.; Giacomelli, C.; Brisson, A.; Borsali, R. Mater. Sci. Eng., C 2008, 28, 479–488. (26) Schmidt, V.; Giacomelli, C.; Lecolley, F.; Lai-Kee-Him, J.; Brisson, A. R.; Borsali, R. J. Am. Chem. Soc. 2006, 128, 9010–9011. (27) Richter, R. P.; Lai Kee Him, J.; Tessier, L.; Tessier, C.; Brisson, A. R. Biophys. J. 2005, 89, 3372–3385.

In order to broaden the range of applications that exploit the distinguished binding ability of Anx5, self-assembled nanostructures made of block copolymer systems carrying phosphonic diacid groups were prepared, and their interaction with the protein was investigated in detail. PS-b-P(PEMA-co-HEMA) copolymers are amphiphilic macromolecules exhibiting self-organizing ability when adequately manipulated in the presence of selective solvents. In aqueous media, these block copolymers self-assemble into (28) Richter, R.; Mukhopadhyay, A.; Brisson, A. Biophys. J. 2003, 85, 3035– 3047. (29) Rodahl, M.; Hook, F.; Krozer, A.; Brzezinski, P.; Kasemo, B. ReV. Sci. Instrum. 1995, 66, 3924–3930. (30) Sauerbrey, G. Z. J. Phys. 1959, 155, 206–222.

Decoration of Block Copolymer Assemblies with Annexin-A5

Langmuir, Vol. 24, No. 21, 2008 12191

Scheme 1. Structure and Aqueous Solution Behavior of PS-b-P(PEMA-co-HEMA) Copolymers (A) and Formation of Anx5/ PS-b-P(PEMA-co-HEMA) Complexes (B)

micellar structures with a hydrophobic PS core and a hydrophilic P(PEMA-co-HEMA) corona bearing negatively charged phosphate moieties (Scheme 1A).25 The interaction of such micellar structures with Anx5 led to the reversible formation of complexes. The approach undertaken in this endeavor is depicted in Scheme 1B, and the corresponding response recorded by QCM-D is shown in Figure 1. The arrows

indicate the start of the exposure to solutions containing different entities (lipids, proteins, micelles, rinsing solutions, etc.) as given in the legends. Before addition of copolymer micelles, two other preliminary steps (steps i and ii) are necessary in order to build the assemblies, as illustrated in Scheme 1B. The first step corresponds to addition of a solution containing DOPC/DOPS (4:1) small unilamellar vesicles (SUV) forming a supported lipid

12192 Langmuir, Vol. 24, No. 21, 2008

Schmidt et al.

Figure 1. QCM-D data showing formation of Anx5/PS-b-P(PEMA-co-HEMA) assemblies onto lipid bilayers.

bilayer (SLB) onto the quartz crystal (typical ∆F ≈ 25 Hz for maximum surface coverage).28 Formation of SLBs from lipids and their characterization by QCM-D has been reviewed by Keller and Kasemo31 and Richter et al.32 The second step consists of deposition of a monolayer of chemically engineered “double” Anx5 molecules, here referred to as Anx5-dimer (i.e., one side is bound to the lipid bilayer while the other remains exposed to the surface). After addition of PSx-b-P(PEMAa-co-HEMAb)y micelles (step iii) with different physical-chemical characteristics, changes in frequency and dissipation were observed as a consequence of micelles binding onto the protein layer. After rinsing with buffer solution, a slight decrease in ∆F took place due to release of nonspecifically bound micelles (less than 5% for systems corresponding to entries 1 and 2 in Table 1), thus indicating formation of stable and specific interactions. Addition of EGTA (a chelating agent for Ca2+ ions) to these hierarchically assembled complexes provokes release of all species bound to the surface via Anx5 (i.e., polymeric micelles and the protein layer), as judged by the residual frequency value (∼25 Hz) which correspond to the initial value measured for formation of a lipid bilayer. Addition of such a chelating agent confirms, therefore, the reversibility of the process that eventually leads to recovery of initial supported lipid bilayer (Scheme 1B), allowing for subsequent depositions to be carried out for other block copolymer systems as well. The changes in frequency and dissipation depended on the micellar characteristics, as summarized in Table 1. The frequency shift (∆F), which is related to the adsorbed mass (∆m), varied as a function of the hydrodynamic micellar size (2RH) and number of phosphonic diacid moieties present at the micelle corona (∆F ) 55 Hz for micelles with 2RH ) 14 nm and Pmic ) 3.8 × 103 atoms/micelle, and ∆F ) 76 Hz for micelles with 2RH ) 18 nm and Pmic ) 6.9 × 103 atoms/micelle). Pmic is estimated from physical-chemical properties of micelles from Pmic ) Nagg × DP(PEMA), where Nagg is the aggregation number determined by static light scattering (SLS) in the presence of salt and DP(PEMA) is the degree of polymerization of PEMA determined by nuclear magnetic resonance (NMR). In order to determine the amount of polymeric micelles needed to saturate the protein layer (i.e., to exploit at the maximum the (31) Keller, C. A.; Kasemo, B. Biophys. J. 1998, 75, 1397–1402. (32) Richter, R. P.; Be´rat, R.; Brisson, A. R. Langmuir 2006, 22, 3497–3505.

Figure 2. Variation of frequency (∆F) measured by QCM-D as a function of the polymer concentration (Cp) for the PS30-b-P(PEMA15-coHEMA55)70 system.

binding ability of the protein-based surface), the influence of polymer concentration at constant volume of added solution on the QCM-D data was investigated. Figure 2 shows the variation of ∆F as a function of the Cp for a selected system (PS30-bP(PEMA48-co-HEMA22)70; entry 1 in Table 1), where a steep increase in ∆F was observed as Cp increased to 20 µg/mL. For more concentrated solutions (Cp g 20 µg/mL), a plateau was reached, ∆F remaining approximately constant at 55 Hz. These results suggested, therefore, that the concentration used in QCM-D experiments (Cp ) 1.0 mg/mL) was well above the amount required to saturate the protein layer (binding saturation point). The amount of block copolymer micelles (adsorbed mass) deposited onto the protein layer was estimated using the Sauerbrey equation, neglecting contributions from coupled water.30 The data was correlated with micellar properties (Mw,mic, Nagg, and 2RH) determined by light scattering techniques in order to estimate the extent of surface coverage (θ) through the relation

Decoration of Block Copolymer Assemblies with Annexin-A5

θ ) (∆mNARH2) ⁄ (Mw,micRcrl2)

Langmuir, Vol. 24, No. 21, 2008 12193

(1)

where ∆m is the adsorbed mass, NA is Avogadro’s number, RH is the micellar hydrodynamic radius, Mw,mic is the micellar molar mass, and Rcrl is the radius of the quartz crystal. In this calculation, it is assumed that micelles behave as hard spheres and do not flatten onto the surface. This is in fact reasonable as long as the PS-based micelle core is presumably in a glassy state at room temperature, implying that the macromolecular chains have highly reduced mobility to escape from these so-called frozen aggregates. Diassembly of micelles is likewise not a favorable process. Using this approach, θ was estimated to be on the order of 0.40 (i.e., formation partially micelle-covered surface). Such an observation can be explained by the repulsion between adjacent micelles as a consequence of the high density of negative charges present at their corona.33 Indeed, partial coverage of the Anx5-dimer surface (structure ii in Scheme 1B) with micelles (structure iii

in Scheme 1B) was corroborated by further addition of a liposome solution. In this experiment (Figure 3), liposomes were able to form complexes with the underlying layer of Anx5-dimer adsorbing to the surface (formation of structure iv in Scheme 1B), indicating that the uppermost micelle layer does not cover the surface entirely. This process is accompanied by an increase in ∆F after addition of lipid solution, as indicated in Figure 3 (arrow at ∼65 min). Very importantly, no interaction between lipid- and polymer-based aggregates is detected by QCM-D (see Figure S1, see Supporting Information), thus confirming that the response recorded after addition of lipids is only related to their binding to Anx5-dimer. Deposition of block copolymer micelles bearing Anx5 binding sites produces a micelle-covered surface onto which further hierarchical assembly can be induced by addition of Anx5 (formation of structure v in Scheme 1B). The results corresponding

Figure 3. Effect of lipid addition on the QCM-D response evidencing the partial surface coverage by Anx5/PS30-b-P(PEMA15-co-HEMA55)70 assemblies.

Figure 4. QCM-D data showing the binding process of Anx5 protein onto a micelle-covered surface leading to formation of hierarchically assembled structure (v).

12194 Langmuir, Vol. 24, No. 21, 2008

Schmidt et al.

only protein (left side of gels) or micelles (right side of gels) as a control of object displacements. As judged from the results presented in Figure 6A, block copolymer micelles and polymer-protein assemblies were stopped at the top of the gel, while the Anx5 migrates through the gel, moving toward the bottom. The effect of increasing the concentration of micelles, that is, the concentration of phosphonic diacid binding sites, is clearly evidenced in this experiment through the progressive decrease in the amount of free Anx5 capable of moving toward the bottom of the gel as the mass of micelles increases from 0.5 to 2.5 µg. For 2 µg of Anx5, one observes that with 3 µg of block copolymer no protein remains free in solution (no migration toward the bottom) due to quantitative binding onto the block copolymer micelles. As expected, binding of Anx5 onto polymer micelles did not occur in the absence of Ca2+ no matter how much polymer micelles were used (Figure 6B). Essentially the same comments also apply for the other block copolymer system listed in Table 1. Figure 5. Frequency shifts due to Anx5 binding (∆FAnx5) onto block copolymer micelles as a function of the frequency shifts corresponding to the added block copolymer micelles (∆FMicelles).

to such experiments are shown in Figure 4. After addition of Anx5 (arrow at ∼200 min) the absolute ∆F value increased due to the protein binding to the surface of block copolymer micelles. Upon rinsing with buffer solution, a slight decrease in ∆F took place, suggesting release of nonspecifically bound proteins. Very interestingly, the amount of protein bound to the micelle surface in this protein deposition step, as reflected by variation in ∆F, depended on the number of phosphate groups (available binding moieties) at the micellar corona (see also discussion below). The selectivity of Anx5 binding to polymeric micelles was proven by the linear variation of ∆F associated with the protein deposition (∆FAnx5) as a function of ∆F due to micelle deposition (∆Fmicelles) until a binding saturation point is reached (Figure 5). In parallel with QCM-D, PAGE experiments were carried out in nondenaturizing conditions in order to quantify the amount of Anx5 bound to PS-b-P(PEMA-co-HEMA) micelles. The results are illustrated in Figure 6 for PS30-b-P(PEMA15-co-HEMA55)70 micelles in the presence (Figure 6A) and absence (Figure 6B) of calcium ions. In these experiments, different amounts of polymeric micelles (from 0.5 to 3.0 µg) at a constant amount of Anx5 (2.0 µg) were investigated as well as solutions containing

The (bio)functionalization of polymeric micelles with Anx5 and its derivatives (i.e., modified Anx5 molecules that can be used to target various types of cell receptors)17,34 exhibits, therefore, great potential in target delivery of therapeutic agents and in vivo imaging and monitoring. For example, PS30-bP(PEMA48-co-HEMA22)70/Anx5 complexes carrying hydrophobically modified Fe2O3 nanoparticles within the PS-based core can be easily prepared following the experimental protocol described above. Effectively, encapsulation of iron oxide nanoparticles was successfully performed during the micellization step. Exploring the fact that the block copolymer micelles are prepared using the cosolvent method (dissolution in organic solvent followed by slow addition of aqueous solution and dialysis), a solution of iron oxide particles in organic solvent was mixed with the molecularly dissolved block copolymer unimers. Upon water addition, the hydrophobic iron particles were partially trapped inside the micelle core during self-assembly, as corroborated by the high contrast areas (dark spots) observed within selected nanoparticles in the TEM image displayed in Figure 7. Entrapment of iron oxide nanoparticles inside delivery vehicles is important in the fields of nanobiotechnology and biomedicine since such systems may be used, for instance, in magnetic resonance imaging related applications, magnetic field-induced release of micellar payloads.35,36

Figure 6. PAGE results obtained for the Anx5 protein, PS30-b-P(PEMA15-co-HEMA55)70 micelles, and mixtures in the presence (A) and absence (B) of CaCl2. Mixtures contained different amounts of block copolymer micelles and a constant amount of protein, as indicated.

Decoration of Block Copolymer Assemblies with Annexin-A5

Langmuir, Vol. 24, No. 21, 2008 12195

presence of Ca2+ ions, as characterized by QCM-D and PAGE experiments. The binding process depends on the characteristics of macromolecular assemblies (number of phosphonic diacid groups (Pmic), aggregation number (Nagg), and hydrodynamic diameter (2RH). A linear relationship between the amounts of micelles and protein bound onto the micelle surface until a saturation point was established by QCM-D. Therefore, target delivery of therapeutic agents and in vivo imaging and monitoring can be achieved through decoration of polymeric micelles with modified Anx5 molecules (i.e., chemically engineered Anx5 containing piloting/targeting functions that target various types of cell receptors). Such nanostructures have a cargo space for hydrophobic payloads (e.g., drugs, iron oxide nanoparticles, etc.) which is customizable via proper choice of core-forming block. Acknowledgment. R.B. and A.R.B. acknowledge financial support from CNRS, Universite´ Bordeaux 1, Re´gion Aquitaine, EuropeanUnion(ECgrantFP6-NMP4-CT2003-505868“Nanocues” to A.R.B.). V.S. and C.G. thank, respectively, CNPq and CAPES. Supporting Information Available: Quartz crystal microbalance with dissipation monitoring (QCM-D) analysis. This material is available free of charge via the Internet at http://pubs.acs.org. LA8024815 Figure 7. Digital picture of hydrophobically modified Fe2O3 nanoparticles in pure water and encapsulated within the PS-based core of PS30-bP(PEMA48-stat-HEMA22)70 micelles (top), and TEM image of the latter (bottom). Scale bar: 50 nm.

Conclusions The Anx5 reversibly binds to phosphonic diacid functional groups located at the surface of block copolymer micelles in the

(33) Schmidt, V.; DiCola, E.; Giacomelli, C.; Brisson, A. R.; Narayaman, T.; Borsali, R. Macromolecules 2008, 41, 2195–2202. (34) Be´rat, R.; Re´my-Zolghadry, M.; Gounou, C.; Manigand, C.; Tan, S.; Salto´, C.; Arenas, E.; L., B.; Brisson, A. R. Biointerphases 2007, 2, 165–172. (35) Pridgen, E. M.; Langer, R.; Farokhzad, O. C. Nanomedicine 2007, 2, 669–680. (36) Nasongkla, N.; Bey, E.; Ren, J.; Ai, H.; Khemtong, C.; Guthi, J. S.; Chin, S. F.; Sherry, A. D.; Boothman, D. A.; Gao, J. Nano Lett. 2006, 6, 2427–2430.