CH V 97

CHAPTER V

Magnetic nanoparticles coated by temperature responsive based polymers for hyperthermia

Abstract.

Various temperature-responsive N-isopropylacrylamide-based functional copolymers were

prepared and used for the stabilization of iron oxide nanoparticles. The copolymers

investigated are poly(acrylic acid)-b-poly(N-isopropylacrylamide) (PAA-PNIPAM) and

poly(acrylic acid)-b-poly(N-isopropylacrylamide)-b-poly(acrylate methoxy poly(ethylene

oxide)) (PAA-PNIPAM-PAMPEO), with different molecular weights. The coated

nanoparticles were characterized in term of size by a combination of dynamic light scattering

(DLS) and transmission electron microscopy (TEM). A sharp temperature transition was

confirmed by particle size measurements vs. temperature. In addition, the stealthiness of the

coated nanoparticles has been assessed in vitro by the haemolytic CH50 test. These

measurements evidenced the crucial role of the PEO segments on the stealthiness of the

nanoparticles and thus that such copolymers are particularly suitable for biomedical

applications. Preliminary experiments of alternating magnetic field induced heating were

performed and specific absorption rate of the various samples were recorded.

Magnetic nanoparticles coated by temperature responsive based polymers for hyperthermia

1. Introduction.

Magnetic nanoparticles are emerging materials particularly promising in the biomedical domain [1]. Indeed, by applying an appropriate alternating magnetic field to a suspension of these nanoobjets, a locale increase of the temperature can be triggered. Such temperature rising is able to kill living cells of the body, and is usually referred as hyperthermia [2]. For such purpose, magnetic nanoparticles commonly prepared by Massart process [3], i.e. alkaline nanoprecipitation from Fe ²⁺ and Fe ³⁺ aqueous solution, exhibit magnetic properties (typically their specific absorption rates (S.A.R.) is about 20 W/g (Fe) ) well-suited to be applied in cancer treatment by hyperthermia.[2] However iron oxide nanoparticles suspensions produced by Massart process are not stable enough in physiological conditions to be used as such.[1] A stabilizing coating is needed to avoid aggregation and consequent precipitation of the colloids in body fluids. Such coating should also confer stealthiness to the nanoparticles in order to avoid their rapid removal from the body.[4] For this purpose, the high flexibility and hydrophilicity of poly(ethylene oxide) chains make it a candidate of choice to coat nanoparticles surface. In order to strongly link the coating to the magnetic nanoparticles, a poly(acrylic acid) block is used as anchoring block [5]. Finally, in order to improve the tumor treatment, it is expected that the release of a drug simultaneously to hyperthermia would act synergistically. Therefore, the use of a thermoresponsive polymer having a thermal transition close to 37°C, i.e. poly(N-isopropyl acrylamide) (PNIPAM) [6] for the coating of the magnetic nanoparticles is the purpose of the present chapter.

Poly(N-isopropylacrylamide) (PNIPAM) is one of the thermo-responsive polymers that

produces a coil-to-globule transition around 32°C called the lower critical solution

temperature (LCST). This phenomenon is based on a reversible hydration–dehydration of

amide groups in the molecules [7]. At a temperature lower than the LCST, the polymer is

completely soluble in aqueous media. However, increasing temperature, the polymer solution

becomes opaque due the polymer precipitation above the LCST. Such a phase transition

occurs in a narrow temperature range, that can be adjusted to 37°C by random

copolymerization with acrylic acid [8].

Only limited works have been published on the stabilisation of magnetic nanoparticles by thermosensitive PNIPAM based derivatives. Entrapped nanoparticles in the PNIPAM matrix were obtained by Kondo et. al. using a two step polymerization method, emulsion polymerization of styrene in the presence of iron oxide nanoparticles was followed by precipitation polymerization of NIPAM and methacrylic acid [9]. They used thermoresponsive magnetite nanoparticles [10] for easy separation and recovery of proteins by the synergism of magnetic force and hydrophobic interaction. H. Wakamatsu et al.

succeeded in preparing PNIPAM-based copolymers with both sensitive temperature-response and chemical reactivity, by dispersing 3-aminopropyltrimethoxysilane modified magnetite nanoparticles in an aqueous solution of Poly(NIPAM-co-2-carboxyisopropyl acrylamide) in the presence of coupling agent, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride.[11] It is known that acid function can interact easily with iron oxide surface.

Surfactants presenting such function, oleic acid, citrate or lauric acid [12] were commonly used as dispersing agent to prepare ferrofluids. Recently, poly(acrylic acid) was used as stabilizer, and can provide both electrostatic and steric repulsion against nanoparticles aggregation. [13]

In the present chapter, we report on a new approach for the preparation of thermoresponsive iron oxide nanoparticles using N-isopropylacrylamide (NIPAM)-based copolymers prepared by RAFT polymerization. Sensitive temperature-response and high stealthy behaviour were targeted for in vivo applications, and thus two block copolymers, i.e. poly(acrylic acid)-b- poly(N-isopropylacrylamide) (PAA-PNIPAM) and poly(acrylic acid)-b-poly(N- isopropylacrylamide)-b-poly(acrylate methoxy poly(ethylene oxide)) (PAA-PNIPAM- PAMPEO), with different molecular weights were prepared and compared as stabilizing ligands for magnetic nanoparticles. The size, structure and magnetic properties of the resulting stabilized NPs were characterized by TEM, DLS, XRD and magnetometry. Their stealth behaviour has been assessed in vitro by the haemolytic CH50 test,[14] and their specific absorption rate has been recorded by calorimetry when submitted to an alternating magnetic field, which is the basic concept of magnetic hyperthermia.

2. Experimental Section

Materials.

N-isopropylacrylamide (Aldrich; 97%) was recrystallized twice from benzene/hexane 3:2 (v/v) and dried under vacuum prior to use. 2,2’-Azobis(isobutyronitrile) (AIBN, Fluka) was recrystallized from methanol.Toluene was dried by refluxing over the sodium/benzophenone complex and distilled under nitrogen before use. Poly(ethylene oxide) monomethyl ether (MPEO-OH), (Mn = 2000 g/mol), was purchased from SIGMA. Acrylic acid (AA) was purified by distillation under reduced pressure. α-acrylate ω-methoxy poly(ethylene oxide) (AMPEO), dimethylformamide (DMF), 2,2-azobis(4-methoxy-2,4-dimethylvaleronitrile) (V70) were used as received. 2-dodecylsulfanylthiocarbonylsulfanyl-2-methyl propionic acid (DMP) was synthesized according to J.T. Lai et al. [15]. FeCl 3 .6H 2 O and FeCl 2 .4H 2 O (Aldrich) were used without further purification. Water MiliQ was deoxygenated for at least 30 min with ultra-high-purity nitrogen (99.9+%). Hydrochloric acid (Aldrich) was used as a 25% v/v aqueous solution.

Synthesis of block copolymers.

i. Synthesis of PAA first block. 0.012 g azo-bis-isobutyronitrile (7.31×10 ^-5 mol), 1.09 g DMP (3×10 ^-3 mol), 15 mL of AA (1.98×10 ^-1 mol) and 15 mL of DMF were mixed together in a 250 mL Schlenk flask. The mixture was degassed by four freeze-pump-thaw cycles. This reaction mixture was heated in an oil bath at 70ºC for 4 h. The polymer was precipitated by addition of the solution to ether, and dried under vacuum up to constant weight. The molecular weight was determined by ¹ H NMR in DMSO-d6 (M n =3xI 2.44 /I 0.8 +364), where I 0.8 and I 2.44 are the intensity of the proton resonances at 0.8 ppm (CH 3 -C 11 H 22 , t) and 2.44 ppm (CH-COOH, m), respectively. Polydispersity was measured by SEC in DMF.

ii. Synthesis of PAA-b-PNIPAM. Typically, 1g of trithiocarbonate-capped PAA (3.33 10 ^-4 mol;

M _n (NMR)= 3000 and M _w /M _n =1.10), 17g of NIPAM (0.15 mol), 10 ^-2 g V70 (3.33 10 ^-5 mol)

and 55 mL of DMF were mixed together. The reactive mixture was degassed by four freeze-

pump-thaw cycles and the reaction solution was kept stirring at room temperature, to obtain

different molecular weights for the PNIPAM block, 10ml were syringed from the reaction

solution each 2h. The copolymer was precipitated into ether and dried in vacuum up to

constant weight. The molecular weight of the second block was determined by ¹ H NMR in

DMSO-d6 by comparing the peak at 4.01 ppm (N-CH<) for the PNIPAM block to the peak at

2.44 ppm (CH-COOH) for PAA block. Polydispersity was determined by SEC in DMF. (Fig

1)

Figure 1: Evolution of the SEC traces during the copolymerization of NIPAM from PAA:

samples pick-out after 2h, 4h, 6h and 22h.

iii. Synthesis of PAA-b-PNIPAM-b-PAMPEO.

Typically, 1g of trithiocarbonate-capped PAA 41 -NIPAM 150 (5.10 ^-5 mol), 3g of AMPEO (6.61mmol), 0.8×10 ^-3 g azo-bis-isobutyronitrile (5.10 ^-6 mol) and 10 mL of DMF were mixed together. The reactive mixture was degassed by four freeze-pump-thaw cycles and heated in an oil bath at 80ºC for 3 h. The copolymer was precipitated into ether and dried in vacuo up to constant weight. The molecular weight of the third block was determined by ¹ H NMR in DMSO-d6 by comparing the peak at 3.54 ppm (CH 2 -CH 2 -O) for PAMPEO block to the peak at 4.01 ppm (N-CH<) for the PNIPAM block. polydispersity was determined by SEC in DMF.

Synthesis of iron oxide nanoparticles.

The iron based magnetic nanoparticles were prepared by the Massart process [3]. All the solutions were deoxygenated just prior to use in order to minimize parasitic oxidation.

Required amounts of FeCl 3 .6H 2 O (40 mL, 1M in HCl solution 2M) and FeCl 2 .4H 2 O (10 mL, 2M in HCl solution 2M) were mixed in an additional funnel and added dropwise within 15 min to an alkaline solution (400 mL, 0.75 M) at 100°C under magnetic stirring in order to target magnetite particles. The solution quickly turned black as result of magnetite formation.

The magnetite particles were let to grow for 1 h under stirring and nitrogen. After cooling down to room temperature, they were collected with a permanent magnet, and the supernatant

11 12 13 14 15 16 17 18 19 20 21 22 23

Elution time (min.)

was discarded by decantation. Salt excess and NaCl byproduct were eliminated by suspending the particles within 100 mL of nitric acid for 10 min. This purification procedure was 3 times repeated. Finally, the purified uncoated particles were dispersed within deionized water and dialyzed (Spectra pore 7, MWCO 8000) against water (pH~4) for 2 days, the water being replaced twice a day. Particle aggregates were removed by centrifugation for 30 min, which was repeated until no insoluble was deposited at the bottom of the tube. Approximately 6 mg Fe 3 O 4 /mL was collected (checked by volumic titration).

Coating of the Fe 3 O 4 nanoparticles by a block copolymer.

A representative recipe (thus, whatever the block copolymer) was as follows. 50mg of block copolymer was dissolved in 3 mL double-distilled water in a 20 mL round-bottom flask equipped with a magnetic stirring bar. The pH was adjusted to pH~6.5. 3 mL of the uncoated Fe 3 O 4 suspension (6 mg/mL, pH~4) were then added dropwise to the copolymer solution.

Complement consumption studies.

Complement activation was measured as the lytic capacity of a normal human serum (NHS) towards antibody-sensitized sheep erythrocytes after exposure to the nanoparticles. Aliquots of NHS were incubated with increasing amounts of nanoparticles. The amount of serum, able to haemolyse 50% of a fixed number of the sheep erythrocytes after exposure to the nanoparticles, was determined (“CH50 units”) for each sample. NHS was provided by the

“Etablissement Français du Sang” (Angers, France) and stored as aliquots at – 80°C until use.

Veronal-buffered saline containing 0.15 mM Ca ²⁺ and 0.5 mM Mg ²⁺ (VBS++) was prepared

as reported elsewhere.[4] Firstly, sheep erythrocytes were sensitized by rabbit anti-sheep

erythrocytes antibodies (Sérum hémolytique, Biomérieux, Marcy-l’Etoile, France) and diluted

by the veronal-buffered saline at a final concentration of 2.109 cells/ml in VBS++. Increasing

amounts of the particle suspension were added to NHS diluted in VBS++ such that the final

dilution of NHS in the mixture was 1/4 (v/v) in a final volume of 1 mL. After 1 h of

incubation at 37°C under gentle agitation, the suspension was diluted 1/25 (v/v) in VBS++,

and aliquots of 8 different dilutions were added to a given volume of sensitized sheep

erythrocytes. After 45 min of incubation at 37°C, the reaction mixture was slightly

centrifuged at 2000 rpm for 10 min. The absorption of the supernatant was determined at 414

nm with a microplate reader (Multiskan Anscent, Labsystems SA, Cergy-Pontoise, France)

and compared to the results obtained with control serum in order to evaluate the amount of

haemolysed erythrocytes. Positive and negative controls were made in each series of

experiments in order to account for any difference in the hemoglobin response from a given

erythrocyte preparation. Furthermore, corrections for particle light-scattering and spontaneous

erythrocyte haemolysis were estimated by UV/VIS measurements using blanks containing only particles and only erythrocytes, respectively. In order to compare nanoparticles of different average radius, their surface area was calculated as follows: S = 3 m/rρ, where S is the surface area (cm ² ), m the weight (µg) in 1 mL of suspension, r the average radius (cm) determined by DLS, and ρ the volumic mass (µg/cm ³ ) of the nanoparticles estimated at 106 µg/cm ³ . The experimental data were the average of 3 independent experiments with a 10%

standard deviation.

Calorimetric Determination of Specific Absorption Rate (SAR).

For the calorimetric determination of SAR, the iron oxide suspensions were thermally isolated in a vessel and placed into a coil. Temperature changes vs. time of exposure to an alternating magnetic field (amplitude, 10 kA/m; frequency, 108 kHz produced by a Celes inductor C97104) were automatically recorded with an optical fiber connected to a multimeter.

Characterizations.

1 H NMR. Samples were analyzed by 1H NMR spectroscopy with a Bruker AM 400 apparatus at 25°C, in deuterated chloroform (CDCl3) added with tetramethylsilane as an internal reference.

Gel permeation chromatography (GPC). Molecular weight and polydispersity index (Mw/Mn) were determined by size exclusion chromatography (SEC), using a 25mM solution of LiBr in DMF as the eluent at 50°C. The columns were calibrated with polystyrene standards.

Dynamic Light Scattering (DLS). The radius of the micelles was measured by DLS with a Malvern Instrument Model ZetaSizer Nano ZS.

X-Ray diffraction (XRD) measurement. The crystal structure of the uncoated iron oxide nanoparticles was obtained by the powder X-ray diffraction (XRD) pattern of sample recorded with a difractometer PHILIPS PW1700 with CuKα (λ = 1,5418 Ǻ).

Transmission electron microscopy (TEM) measurement. The average particle size, size distribution and morphology of the samples were studied using a Philips CM-100 microscope, at an accelerating voltage of 100 kV. The images were recorded by a camera. Samples were prepared by deposition of one drop of an appropriately diluted solution onto the copper grid coated with Formvar and drying it in air before it was loaded into the microscope.

Magnetic measurements. The magnetization of iron oxide nanoparticles were measured as a

function of the applied magnetic field H with a SQUID MPMS-5S magnetometer from

Quantum design. The hysteresis curve was obtained by changing H between -6000 and 6000Oe, these measurements were carried out at 290 K.

3. Results and discussion.

In a previous work, some of us reported on the controlled radical block copolymerization of acrylic acid [16a] and α-methoxy,ω-acrylate (polyethylene oxide) (AMPEO) by RAFT by using DMP with formation of diblock copolymers with a narrow molecular weight distribution and a predictable molecular weight at high conversion.[16] This control has been extended to the block copolymerization of AA, NIPAM and AMPEO. (Scheme 1)

Scheme1 : Sequential RAFT copolymerization of acrylic acid, N-isopropyl acrylamide and acrylate methoxy PEO for the preparation of PAA-PNIPAM and PAA-PNIPAM-PAMPEO.

As reported by Müller et al.,[17] the synthesis of PNIPAM-b-PAA with M ⁿ around (22.9-

29.2).10 ⁴ g/mol and low polydispersities (around 1.11-1.03), by RAFT polymerization can be

performed in methanol. Whenever the AA monomer is first polymerized, the PAA polymer

has the expected molecular weight and a low polydispersity (Table 1, entry 1 and 2). This

polymer was used as a macroinitiator for the polymerization of NIPAM in DMF at 25°C

using V70 as radical source. (Table 1, entry 3-4) Figure 1 shows SEC traces of the

copolymerization of NIPAM from the PAA macroinitiator. The traces were monomodal,

symmetrical and clearly shifted to lower elution time with increasing conversion, one

indicator of a controlled polymerization. Moreover, the chromatograms show no evidence of

higher molecular weight impurities (normally visualised as shoulder), even at extended

polymerization times (22h), the same observation was reported by McCormick.[18] For the

triblock-copolymer, PAA-b-PNIPAM was used as a macroinitiator for the polymerization of

AMPEO in DMF at 75°C, i.e. conditions similar to already published AMPEO

polymerization [16b].(Scheme 1) Table 1 entry 5-6 shows different copolymers compositions used for the coating of the iron oxides nanoparticles.

Table 1: Degree of polymerization determined by NMR (DP), polydispersity index of copolymers determined by SEC in DMF (IP) characterizing the copolymers. Hydrodynamic radius (Rh) and polydispersity index determined by DLS (IP), zeta potential (ζ), and conductivity for naked NPs (0) and magnetic nanoparticles stabilized by the different copolymers measured at room temperature (~20°C) (1-4).

Scheme 2: different steps to obtain stable coated nanoparticles, coprecipitation of Fe ²⁺ and Fe ³⁺ , modification of surface charge and chemisorption of the ligand.

Magnetic nanoparticles were synthesized by co-precipitating aqueous FeCl 2 and FeCl 3 salt solutions at room temperature under N 2 with a sodium hydroxide base (pH~13). The stoechiometric molar ratio of Fe ²⁺ /Fe ³⁺ was 0.5 to achieve quantitative conversion. After washing the precipitate by nitric acid, a dispersion of iron oxide nanoparticles in water is obtained, stabilized through electrostatic repulsions. The nitric-acid presents a complex solution system in which the ions are either in solution or surround the particles as counterions, forming an electric double layer, the free nitric acid was removed by dialysis.

(Scheme 2) The magnetic nanoparticle radius was 3±2nm as checked by TEM.(Fig 2a) As shown in figure 3, the magnetization of these magnetic nanoparticles disappears when the external field is removed (i.e., they have near zero magnetic remanence and coercivity) in short times relative to the experiment time. This behaviour is characteristic for

Entry Ligand DP

(NMR)

IP (SEC)

Rh(nm) (DLS)

IP (DLS)

Zeta ζ (mV)

conductivity (mS/cm)

SAR (W/g

) 0 ___ ___ ___ 25 0.20 45.7 0.07 19.6

1 PAA 61 1.17 ___ ___ ___ ___ ___

2 PAA 41 1.10 ___ ___ ___ ___ ___

3 PAA-PNIPAM 61-97 1.09 80 0.17 -39.7 0.31 10.8 4 PAA-PNIPAM 61-239 1.08 90 0.18 -34.5 0.26 13.5 5 PAA-PNIPAM-

PAMPEO

41-290-90 1.06 60 0.22 -21.0 0.91 12.7

6 PAA-PNIPAM- PAMPEO

41-150-90 1.07 70 0.20 -22.8 0.21 8.2

superparamagnetic property, which suggests that the nanoparticles may be ideal for hyperthermia.

Figure 2: TEM images of iron oxide nanoparticles (a) after Massart synthesis, and coated with: (b-c) PAA 61 -PNIPAM 239 (Table 1, entry 4), (d-e) PAA 41 -PNIPAM 290 -PAMPEO 90

(Table 1 entry 5)

14 nm

a

200nm 20nm

b c

200nm 20nm

d e

The X-ray powder diffraction spectrum of the magnetic nanoparticles is shown in Figure 3.

Five characteristics peaks for Fe 3 O 4 corresponding to indices (2 2 0), (3 1 1), (4 0 0), (5 1 1), and (4 4 0) are observed, but the entire peaks of the diffractogram are not very well-defined;

they can be indexed either to maghemite or magnetite. It indicates that probably both structures coexist in the ferrofluid. In order to discern if these iron oxide particles are built up of magnetite (Fe 3 O 4 =2Fe ³⁺ Fe ²⁺ 4O ^2- ) or maghemite (Fe 2 O 3 = 2Fe ³⁺ 3O ^2- ), the existence of Fe ²⁺

ions was determined by the titration with potassium dichromate K 2 Cr 2 O 7 . No significant Fe ²⁺

ion concentration was determined (~5%), suggesting the maghemite phase as the dominant crystalline phase in the samples. Moreover, the solution presents a red colour, as it is expected for maghemite solutions. This oxidation probably occurred during washing steps by nitric acid solution and water dialysis. However, the two structures present a similar magnetic property and both of them can be used for hyperthermia.

a b

Figure 3: (a) magnetization curve of iron oxide nanoparticles before coating. (b) XRD patterns of iron oxide nanoparticles before coating.

The ferrofluids were also characterized by dynamic light scattering (DLS). The average radius obtained by this technique for uncoated NPs are much larger than those by TEM (25nm). This is because even in the absence of any external magnetic field, the magnetostatic (magnetic dipole–dipole) interactions between the particles can cause their agglomeration. It has been determined, experimentally [19] as well as by Monte Carlo simulations that the particles form closed rings and long open loops with no particular spatial orientation due to magnetostatic interaction in the absence of any external magnetic field. This aggregation have a lower diffusion coefficient than the single particle and the equivalent sphere radius measured by light scattering is higher than the elementary particle size as revealed by TEM.(Fig.2a) The agglomerated ring and loop structures are not seen in TEM, possibly because they are disturbed due to the drying forces present during the sample preparation for TEM. [19]

In this chapter a simple strategy has been tested which consists in the mixing of the preformed well-defined PAA based block copolymers with the magnetic nanoparticles produced by the Massart process (Scheme 2), PAA being used as chemisorption agent to anchor the polymers to the surface of the nanoparticles. In order to prevent complexation between the amide functions of PNIPAM and protonated polyacids,[20] deprotonation of PAA in carboxylate functions are performed before mixing by increasing pH of the copolymers solution up to pH=6.5. In addition, this ionisation of PAA enhances interactions with the surface of NPs.

(scheme 2) Indeed, the uncoated NPs in suspension at pH=4 due to slight nitric acid excess are positively charged as confirmed by Zeta potential (ξ= +45.7 mV),(Table 1, entry 0). After mixing the copolymer and nanoparticles solutions, the nanoparticles are complexed with asymmetric block copolymers and become negatively charged (zeta potential, entry 3-6) as result of the excess of deprotonated acrylic acid functions. Zeta potential of stabilized nanoparticles appears to depend on the molecular weight of the only negatively charged PAA block and not on the total molecular weight of the copolymer. Table 1 shows that the zeta potential is about 15 mV less negative when the DP of the PAA block is decreased from 61 to 41. Close Zeta potential values are measured keeping constant the PAA block and varying the other block length (Table 1 entry 3-4 and 5-6). For the ease of reading, PAA-PNIPAM copolymer (Table 1, entry 3 and 4) stabilized NPs and PAA-PNIPAM-PAMPEO triblock- copolymer (Table 1, entry 5 and 6) stabilized NPs will be respectively abbreviated as Di-NPs and Tri-NPs in the following.

0 0.2 0.4 0.6 0.8 1 1.2

0 50 100 150 200 250 300

Rh F(Rh)

39°C 25°C a

0 0.4 0.8 1.2

0 50 100 150 200 250

Rh F(Rh)

39°C 25°C

Figure 4: size distribution above and below LCST determined by DLS measurements of magnetic NPs stabilized by the triblock copolymer of different molecular weights:(a) PAA 41 - PNIPAM 290 -PAMPEO 90 (Table 1, entry 5) and (b) PAA 41 -PNIPAM 150 -PAMPEO 90 (Table 1, entry 6).

TEM micrographs clearly evidence the difference between naked and stabilised NPs. In the first case some small aggregates were observed due probably to dipole-dipole interaction between magnetic crystals (Fig 2a), and after stabilisation higher dispersion was observed proving the presence of the ligand around the NPs. (Fig 2b-c) As reported, such double responsive block copolymers PAA-PNIPAM may form micelles or other aggregates depending on solvent, temperature, pH and block lengths.[20] Dynamic light scattering (DLS) of the resulting polymer-coated magnetic nanoparticles show the presence of stable micelles with narrow size distribution, (Table 1, entry 3 to 6) (Fig 4) the new hydrodynamic radius observed (from 60 to 90nm) is higher than the one observed for uncoated NPs (25nm), this observation provides an additional evidence for the nanoparticles coating. Even if the magnetic cores are embedded within micelles of the block copolymers, the diameter of these colloidal hybrid particles is still below 200 nm and thus suitable for body injections. Thanks to the copolymer coating, magnetic suspensions in buffer or physiological fluids are now stable for months while ligand-free colloids flocculate immediately in such media.

When the temperature-responsive polymer is immobilized on solid surfaces, the modified surface can show a reversible hydrophilic–hydrophobic transition because of the reversible hydration–dehydration of the polymer chains. This transition has been observed for both DiNPs and TriNPs above and below the LCST at 32°C by means of DLS.[21] The

b

temperature-dependent hydrodynamic radius (Rh) of Tri-NPs in aqueous solutions as measured by dynamic light scattering is shown in Fig. 4-b. For both Tri-NPs, the Rh decreases when the temperature is increased above the LCST, this effect being larger, i.e.

50% (from 60 to 30 nm) when the copolymer PAA 41 -PNIPAM 290 -PAMPEO 90 (entry 5 table 1) was used and 35% (from 70 to 45nm) for the copolymer PAA 41 -PNIPAM 150 -PAMPEO 90

(entry 6 table 1), when the molecular weight of the PNIPAM inner block is higher, the other blocks length being kept constant. In addition, the polydispersity of the Tri-NPs with the higher PNIPAM block length (Figure 4a), was increasing from 0.29 to 0.36 at 39°C which could indicate that such big hydrophobic core cannot be efficiently stabilized by the hydrophilic corona constituted by the only PEO chains which are unable to prevent aggregation. In contrast, when the PNIPAM chain length is smaller, the PEO coating appears to remain very efficient against aggregation and the particles remains highly monodisperse (polydispersity 0.17). (Figure 4b)

The surface of proteins is heterogeneous, and exposes positively and negatively charged groups with hydrogen bonding abilities as well as non-polar regions [19, 22]. The interactions with NPs can be of a different nature, repulsive and attractive ionic interactions as well as hydrogen bondings, hydrophobic interactions, Van der Waals interactions and steric repulsion resulting from the presence of the polymer in the surface of NPs.[19–22] The influence of charge presents on the surface of nanoparticles has been well described in many studies [4a].

Indeed, from different surface structures presenting negative and/or positive charges and

hydroxyl groups at different depths in the coating, Luck et al. [4b] showed that it exerted

different operating forces, and that charge density, accessibility and hydrophobicity defined

the protein adsorption pattern. In the present work, Zeta potential measurements of the

different solutions (table 1) evidenced similar charge densities on the nanoparticles surface,

solution conductivities being in the same order of magnitute. This would suggest that the

stealthy behaviour of stabilized nanoparticles should be governed by the nature of the neutral

block in the shell structure.

0 20 40 60 80 100

0 100 200 300 400 500 600

NPs surface area (cm2/ml) CH50 units

consumption (%)

PAA61-PNIPAM97 PAA61-PNIPAM239

a

-5 15 35 55 75 95

0 100 200 300 400 500 600

NPs surface area (cm2/ml) CH50 units

consumption (%)

PAA41-PNIPAM290-PAMPEO90

PAA41-PNIPAM150-PAMPEO90

b

Figure 5: Consumption of CH50 units vs. surface area of coated magnetic nanoparticles stabilized by (a) PAA-PNIPAM and (b) PAA-PNIPAM-PAMPEO block copolymers.

Protein adsorption test (CH50) was performed on the various coated nanoparticles in order to compare their propensity to resist proteins adsorption, a first requirement for i.v.

injection.[23] The CH50 test is based on the activation of the complement system by the

nanoparticles in normal human serum (diluted 1/4 (v/v)). The amount of serum proteins

adsorbed on the NPs surface decreases with increasing stealthiness. Basically, after exposure

of the human normal serum to increasing amounts of coated nanoparticles, the amount of

serum needed to haemolyse 50% of a fixed number of sensitized sheep erythrocytes is

determined. By this way, the complement consumption was evaluated after incubation with

Di-NPs and Tri-NPs (Figure 5). It is known that the hydrophobic interaction has a major role in protein adsorption phenomena [24]. Since the nanoparticles were incubated with the proteins at 37°C, the thermosensitive polymer PNIPAM offers a hydrophobic surface above the lower critical solution temperature (LCST) of 32°C.[25] Hydrophobic particles are susceptible to adsorb larger amounts of protein. As expected, Di-NPs adsorbed large amounts of serum proteins, i.e., they are strong activators of the complement system (Figure 5a). 100%

of CH50 units are indeed consumed when the serum protein solution is exposed to 600cm ² of colloids surface. This adsorption behaviour is increase with the content of PNIPAM in the copolymer (Figure 5a). In contrast, Tri-NPs did not induce strong activation of the complement system which remains lower than 35% even at high surface contact (600cm ² /ml).

(Figure 5b) This observation confirms the unique capacity of PEO chains to prevent protein adsorption as a result of specific solution properties and chain conformation in water.[4, 16, 23] In accordance with the DLS above LCST results, the less efficient protection of the hydrophobic core by the PEO when high molecular weight PNIPAM (DP=290) is used is confirmed by the protein adsorption test (Figure 5b). The PEO shell is highly efficient in preventing protein adsorption for the PAA 41 -PNIPAM 150 -PAMPEO 90 coated NPs even above the PNIPAM LCST. The uncoated iron oxide NPs could not be tested by CH50 due to the precipitation after addition of VBS ²⁺ solution, showing the determinant role of the ligands in nanoparticles stabilization.

25 30 35 40 45 50

0 200 400 600 800 1000 1200

entry 0

entry 4 entry 5 T(°C)

time (s)

Figure 6: Temperature increase triggered by magnetite nanoparticles uncoated (entry 0, Table

1) and coated by various polymers (entry 4 and 5, Table 1) in a magnetic field of 10 kA/m

amplitude and a frequency of 108 kHz.

Because superparamagnetic iron oxide NPs generate heat by Néel and/or Brown relaxation losses in an alternating magnetic field [26], their use was reported for the heat ablation of tumors [27]. Therefore, the temperature profile of the magnetic nanoparticles used in this work has been recorded (Figure 6). Inductive heating experiments show that a magnetic field of 100 kHz is able to produce enough energy for temperature to increase by approximately by 8°C within a short period of time (~10min).(Figure 6) The specific absorption rates (SAR) are reported in Table 1. They were calculated by the expression SAR= C ∆T/∆t, where C is the sample specific heat capacity and ∆T/∆t is the slope of temperature (from 36°C to 38°C) versus time curve. Since C strongly depends on the weight of iron in the sample, the iron mass was determined by volumetric titration of the starting solution considering, in first approximation, that the further sampling of the starting solution distributes constant amount of iron in each sample. The higher SAR value (~20W/g (Fe) ) was measured for the uncoated nanoparticles in good agreement with reported data [28]. The SAR value were found systematically lower (between 8 and 13 W/g (Fe) ) for the coated samples. However, the slight SAR decrease due to the organic coating does not preclude the use of such colloids for hyperthermia.

4. Conclusion.

PAA-PNIPAM block copolymers have been synthesized by RAFT by sequential RAFT

copolymerization of acrylic acid and N-isopropyl acrylamide and used as macroinitiator to

obtain triblock copolymer PAA-PNIPAM-PAMPEO. The ability of these double responsive

block copolymers to stabilize magnetic nanoparticles prepared by coprecipitation of Fe ²⁺ and

Fe ³⁺ in aqueous media was studied by TEM, DLS, and zeta potential. When the temperature-

responsive polymer is immobilized on solid surfaces, the modified surface can show a

reversible hydrophilic–hydrophobic change because of the reversible hydration–dehydration

of the polymer chains. The outer PEO block plays a crucial role, preventing aggregation of the

NPs above the LCST and enhancing the stealthiness of the nanoparticles. The block length

ratio PNIPAM/PEO is also crucial; the contraction of the coated NPs going above the LCST

is higher as far as the molecular weight of the PNIPAM is increased. However, too high

PNIPAM block prevent homogeneous PEO shell at the outer periphery of the NPs, which

reduces the stealth behaviour and stability of the NPs. Preliminary experiments of magnetic

heating proved that magnetic nanoparticles stabilized by these double responsive block

copolymers are particularly promising in the hyperthermia therapies. In the light of these

reported results, the adjustment of the PNIPAM LCST to 38°C is under current investigation, together with thermal triggered release of drugs.

5. Acknowledgments.

A.A., C. J. and R. J. are grateful to the ‘Région Wallonne’ for support in the frame of the

“NOMADE” program. CERM is much indebted to the “Belgium Science policy” for general support in the frame of the IAP VI/27 program “Functional Supramolecular Systems”. A.A. is much indebted to the European NoE “FAME” and to CGRI-FNRS-Inserm cooperation for grant supporting research stays in Bordeaux and Angers, respectively.

5. References.

(1) a) Cheng, F. Y.; Su, C. H.; Yang, Y. S.; Yeh, C. S.; Tsai, C. Y.; Wu, C. L.; Wu, M. T.;

Shieh, D. B. Biomaterials 26 (2005) 729–738 b) Bergemann C, Muller-Schulte D, Oster J, BrassardL, Lubbe AS. J Magn Magn Mater 1999,194, 45–52.

(2) a) Mornet, S.; Vasseur, S.; Grasset, F.; Duguet, E.; J. Mater. Chem. 2004, 14, 2161-2175.

b) Duguet, E. ; Mornet, S. ; Vasseur, S. ; Devoisselle, J. M. Nanomedicine, 2006, 1, 257. c) Chan DCF, Kirpotin DB, Bunn P. J Magn Magn Mater 1993; 122:374–8.

(3) a) Massart, R. IEEE Trans. Magn. 17 (1999) 1247. b) Bacri, J.; Perzynski, R.; Salin, D.;

Cabuil, V.; Massart, R. J. Magn. Magn. Mater. 1990, 85, 27-32.

(4) : a) Vonarbourg, A.; Passirani, C.; Saulnier, P.; Benoit, J. P. Biomaterials 27 (2006) 4356–

4373. b) Luck, M.; Paulke, B.R.; Schroder, W.; Blunk, T.; Muller, R.H. J Biomed Mater Res 1998; 39: 478–85.

(5) Harris, L. A.; Goff, J. D.; Carmichael, A. Y.; Riffle, J. S.; Harburn, J. J.; St. Pierre, T. G.;

Saunders M.; Chem. Mater. 2003, 15, 1367-1377

(6) a) H. Furukawa, R. Shimojyo, N. Ohnishi, H. Fukada, A. Kondo, Appl. Microbiol.

Biotechnol. 62 (2003) 478. b) F. Takahashi, Y. Sakai, Y. Mizutani, J. Ferm. Bioeng. 83 (1997) 152. c) N. Kato, Y. Takizawa, F. Takahashi, J. Intellegent Mater. Sys. Struct. 8 (1997) 588. d) F. Eeckman, A.J. Moes, K. Amighi, J. Contr. Release 88 (2003) 105. b) D.C.

Coughlan, F.P. Quility, O.I. Corrigan, J. Contr. Release 98 (2004) 97.

(7) a) M. Heskins, J.E. Guillet, J. Macromol. Sci. Chem. A2 (1968) 1441. b) R. Yoshida, K.

Uchida, Y. Kaneko, K. Sakai, A. Kikuchi, Y. Sakurai, T. Okano, Nature 374 (1995) 240.

(8) Kayi Kayi, H.; Tuncel, S. A.; Elkamel, A.; Alper, E. J. Mol. Model 2005;11:55-60.

(9) Kondo A, Kamura H, Higashitanin K, 1994 Appl Microbiol Biotechnol 41:99.

(10) H. Furukawa, R. Shimojyo, N. Ohnishi, H. Fukuda, A. Kondo, Appl. Microbiol.

Biotechnol. 62 (2003) 478.

(11) Wakamatsu H.; Yamamoto K.; Nakao A.; Aoyagi T.; Journal of Magnetism and Magnetic Materials 302 (2006) 327–333.

(12) a) M. Shinkai, H. Honda, T. Kobayashi, Biocatalysis 5 (1991) 61. b) S.E. Khalafalla, G.W. Reimers, IEEE Trans. Magn. 16 (1980) 178. c) D.K. Kim, Y. Zhang, W. Voit, K. Rao, M. Muhammed, J. Magn. Magn. Mater. 225 (2001) 30. d) L. Shen, P.E. Laibinis, T.A. Hatton, Langmuir 15 (1999) 447.

(13) Mak S. Y.; Chen D. H.; Macromol. Rapid. Commun. 2005, (26) 1567-1571 (14) Mayer, M. M. Eds: Kabat E.A.; Mayer, M.M., Springfield, IL, USA 1961, 133 (15) J. T. Lai, D. Filla and R. Shea, Macromolecules 2002, 35, 6754.

(16) a) Aqil A. ; Detrembleur C.; Gilbert B.; Jérôme R. and Jérôme C. Chemistry of materials Chem. Mater. 2007, 19, 2150-2154 b) Aqil A. ; Vasseur S. ; Duguet E. ; Passirani A. ; Jérôme R. and Jérôme C. (submitted biomacromolecules)

(17) Schilli CM, Zhang M, Rizzardo E, Thang SH, Chong YK, Edwards K, Karlsson G, Müller AHE. Macromolecules 2004; (37) 7861-6.

(18) A. J. Convertine; N. Ayres; C. W. Scales; A. B. Lowe; C. L. McCormick Biomacromolecules 2004, 5, 1177-1180

(19) Maity, D.; Agrawal, D. C. Journal of Magnetism and Magnetic Materials 308 (2007) 46–55

(20) a) Wang YC, Morawetz H. Macromolecules 1989 (22); 164-67. b) Shibanuma T, Aoki T, Sanui K, Ogata N, Kikuchi A, Sakurai Y, Okano T. Macromolecules 2000; (33) 444-50. c) Ramon O, Kesselman E, Berkovici R, Cohen Y, Paz Y. J Polym Sci Part B: Polym Phys 2001; (39)1665-77. d) Blatchford MA, Raveendran P, Wallen SL. J Am Chem Soc 2002;

(124)14818-9.

(21) Suzuki, D.; Kawaguchi, H. Langmuir 2005, 21, 8175-8179

(22) Claesson, P.M.; Blomberg, E.; Froberg, J.C.; Nylander, T.; Arnebrant, T. Adv Colloid Interface Sci. 1995;57:161–227.

(23) Rieger, J.; Passirani, C.; Benoit, J.P.; Van Butsele, K.; Jerome, R.; Jerome, C. Advanced Functional Materials (2006), 16(11), 1506-1514.

(24) Xua, L. C.; Siedlecki, C. A. Biomaterials 28 (2007) 3273–3283

(25) N. Shamim, L. Hong, K. Hidajat, M.S. Uddin Journal of Colloid and Interface Science

304 (2006) 1–8

(26) Ma, M. ; Wu, Y. ; Zhou, J. ; Sun, Y. ; Zhang, Y. ; Gu, N. J. Magn. Magn. Mater. 268 (2004) 33.

(27) Hilger, I.; Hiergeist, R.; Winnefeld, K.; Schubert, H.; Kaiser, W.A. Invest. Radiol. 37 (2002) 580.

(28) Hergt, R.; Andra,W.; d’Ambly, C.G.; Hilger, I.; Kaiser, W. A.; Richter, U.; Schmidt,

H. G. IEEE Trans. Magn., 1998, 34, 3745–3754.