Article pubs.acs.org/Langmuir
Magnetic Nanocarriers of Doxorubicin Coated with Poly(ethylene glycol) and Folic Acid: Relation between Coating Structure, Surface Properties, Colloidal Stability, and Cancer Cell Targeting Karine Kaaki,†,§ Katel Hervé-Aubert,† Manuela Chiper,† Andriy Shkilnyy,‡ Martin Soucé,† Roland Benoit,§ Archibald Paillard,† Pierre Dubois,† Marie-Louise Saboungi,§ and Igor Chourpa*,† †
EA 4244, Physicochimie des matériaux et des biomolécules, équipe Nanovecteurs magnétiques pour la chimiothérapie, Université F. Rabelais, Faculté de Pharmacie, 31 avenue Monge, F-37200 Tours, France; IFR-135 Imagerie Fonctionnelle, F-37200 Tours, France ‡ Department of Chemical and Biotechnological Engineering, Université de Sherbrooke, Sherbrooke QC J1H 5N4, Canada § UMR 6619, Centre de Recherche sur la Matière Divisée, CNRS, Université d’Orléans, 1B rue de la Férollerie, F-45071 Orléans cedex 2, France ABSTRACT: We report the efficient one-step synthesis and detailed physicochemical evaluation of novel biocompatible nanosystems useful for cancer therapeutics and diagnostics (theranostics). These systems are the superparamagnetic iron oxide nanoparticles (SPIONs) carrying the anticancer drug doxorubicin and coated with the covalently bonded biocompatible polymer poly(ethylene glycol) (PEG), native and modified with the biological cancer targeting ligand folic acid (PEG-FA). These multifunctional nanoparticles (SPIONDOX-PEG-FA) are designed to rationally combine multilevel mechanisms of cancer cell targeting (magnetic and biological), bimodal cancer cell imaging (by means of MRI and fluorescence), and bimodal cancer treatment (by targeted drug delivery and by hyperthermia effect). Nevertheless, for these concepts to work together, the choice of ingredients and particle structure are critically important. Therefore, in the present work, a detailed physicochemical characterization of the organic coating of the hybrid nanoparticles is performed by several surface-specific instrumental methods, including surface-enhanced Raman scattering (SERS) spectroscopy, X-ray photoelectron spectrometry (XPS), and time-of-flight secondary ion mass spectrometry (ToFSIMS). We demonstrate that the anticancer drug doxorubicin is attached to the iron oxide surface and buried under the polymer layers, while folic acid is located on the extreme surface of the organic coating. Interestingly, the moderate presence of folic acid on the particle surface does not increase the particle surface potential, while it is sufficient to increase the particle uptake by MCF-7 cancer cells. All of these original results contribute to the better understanding of the structure−activity relationship for hybrid biocompatible nanosystems and are encouraging for the applications in cancer theranostics. blood circulation by the immune system.11 These properties (stability and stealthiness) can be modulated by coating nanoparticles with a biocompatible polymer.12,13 In the past decade, numerous papers have reported on the stability of SPION suspensions under physiological conditions after modification of their surface with biocompatible polymers such as poly(amino acid),14 dextran,15 and poly(ethylene glycol) (PEG).16,17 These studies showed that the uptake of SPIONs modified with biocompatible polymers by the macrophage cells is lower than for nonmodified ones.18 This is concomitant with longer blood circulating lifetimes of nanocarriers, which favors their accumulation in tumor sites by both magnetic retention and the EPR effect.
1. INTRODUCTION Superparamagnetic iron oxide nanoparticles (SPIONs) are extensively studied as platforms for numerous biomedical applications including cancer diagnostics by magnetic resonance imaging (MRI)1,2 and cancer therapy by hyperthermia3,4 and/or magnetically targeted drug delivery.5,6 After intravenous administration, drug-loaded SPIONs can be specifically cumulated in a tumor site due to synergistic mechanisms such as (i) externally applied localized magnetic field,7 (ii) enhanced permeability and retention (EPR)8 effect related to the permeability of the leaky tumor neovasculature to small size nanoparticles (below ∼100 nm), and (iii) internalization of nanoparticles within cancer cell compartments.9 The intracellular release of anticancer drugs leads to an increase in antitumor efficacy and reduces systemic side effects.10 For in vivo application, two conditions are essential: stability of suspensions at physiological conditions and immunological stealthiness, i.e., reduced elimination of nanoparticles from the © 2011 American Chemical Society
Received: September 27, 2011 Revised: December 12, 2011 Published: December 15, 2011 1496
dx.doi.org/10.1021/la2037845 | Langmuir 2012, 28, 1496−1505
Langmuir
Article
Figure 1. Schematic representation of surface functionalization of superparamagnetic iron oxide nanoparticles, SPIONs.
proteins,23 peptides,24 aptamers,25 or smaller molecules such as folic acid (FA).26,29,30 Indeed, FA has the ability to preferentially target cancer cells because its receptor is frequently overexpressed on the surface of cancer cells which actively consume these molecules. Furthermore, FA has several advantages: it is nontoxic, nonimmunogenic, inexpensive, and stable. Several protocols have been reported for chemical conjugation of FA to a variety of therapeutic drugs31 and imaging agents.32 We have recently reported a one-pot synthesis protocol of SPION-DOX-PEG,17 i.e., of SPIONs modified with the fluorescent anticancer drug doxorubicin (DOX) and with PEG polymer, both covalently linked to the iron oxide surface
However, these two targeting concepts are not enough to achieve high intracellular drug concentrations if nanoparticles mainly release their drug load outside of the tumor cells. In order to enhance the intracellular antitumor action, the drug has to be released by nanoparticles internalized within tumor cells. The drug delivery within cells can then be accelerated by various stimuli, for example pH,19 temperature,20 or enzymatic activity.21 One of the common strategies to improve drug delivery into cancer cells involves attaching a specific ligand to the nanoparticle surface that is recognized by receptors expressed on the external membrane of cancer cells. A variety of potential ligands have been conjugated to the nanoparticle surface including macromolecules such as antibodies,22 1497
dx.doi.org/10.1021/la2037845 | Langmuir 2012, 28, 1496−1505
Langmuir
Article
aqueous mixture of Fe3+ (0.032 mol, 350 mL) and Fe2+ (0.016 mol, 20 mL, HCl 1.5 M) salts. The black precipitate was isolated from the solution by magnetic decantation and washed thrice with water. Nanoparticles were treated with nitric acid (30 mL, 2 M) prior to the oxidation of the outer shell by iron(III) nitrate (60 mL, 0.33 M) at 100 °C and peptized again with nitric acid (30 mL, 2 M). Finally, nanoparticles were washed thrice with acetone and redispersed in water at pH 3. The prepared nanoparticles were characterized by TEM, XRD, dynamic light scattering (DLS), ζ-potential measurements, and FTIR spectroscopy. Functionalized SPIONs (SPION-DOX-PEG-FA and SPION-DOXPEG). Functionalized SPIONs were prepared according to a one-pot synthesis protocol similar to that previously published by our group.17 100 μL aqueous suspension of SPIONs at an iron concentration of 20 g/L was suspended in 5 mL of DMF, and water was removed by evaporation. The suspension of SPIONs in DMF was sonicated for 30 min prior to functionalization. NH2-PEG-OCH3, NH2-PEG-FA, and DOX were freeze-dried. An approximately 4:1 molar mixture of NH2PEG-OCH3 (216 mg, 0.043 mmol) and NH2-PEG-FA (70 mg, 0.011 mmol) dissolved in 5 mL of anhydrous DMF was allowed to react with GOPTS (12 μL, 0.054 mmol), dissolved in 1.3 mL of anhydrous DMF, for 4 h at 65 °C (Figure 1, step I). DOX (1.2 mg, 2 μmol), dissolved in 1 mL of anhydrous DMF, was added to the reaction mixture and left to react with GOPTS for 1 h (Figure 1, step II). Then, SPIONs in DMF were added, and the reaction was kept under stirring for 48 h at 65 °C in the dark under nitrogen (Figure 1, step III). Afterward, 2 mL of glycerol was added to the reaction mixture, and DMF was evaporated under vacuum. The glycerol suspension was dispersed in 10 mL of ultrapure water and purified by dialysis against deionized water on MWCO 10 000 Da. Dialysis was carried out for 48 h, while water in the acceptor volume was changed every 2 h for the first 12 h and then every 12 h. The donor/acceptor volume ratio was 10 mL/5 L. This purification eliminates molecules not bonded to the SPION surface (NH2-PEG-OCH3, NH2-PEG-FA, and DOX). A small amount of nonfunctionalized nanoparticles was removed by magnetic decantation at pH 7. To be used as a model system, SPION-DOXPEG nanoparticles were prepared by the same protocol, using only NH2-PEG-OCH3, without NH2-PEG-FA. The resulting nanoparticles were characterized by TEM, DLS, ζ-potential measurements, FTIR, fluorescence spectroscopy, SERS, XPS, and ToF-SIMS. 2.2. Nanoparticle Characterization. 1H NMR Spectrometry. NH2-PEG-FA structure was determined by 1H NMR spectrometry on a Bruker 200 MHz spectrometer, using tetramethylsilane as reference. Samples were prepared in deuterated DMSO. Transmission Electron Microscopy. The morphology and size of nanoparticles were examined by transmission electron microscopy using a JEOL JEM-1230 microscope with 0.2 nm resolution, operating at 120 kV. Samples were diluted with ultrapure water to an iron concentration of 20 mg/L before being deposited on a carbon-coated copper grid and left to air-dry. Photon Correlation Spectroscopy. The hydrodynamic diameter and size distribution of nanoparticles were studied by DLS on a Malvern HPPS 5001 instrument. The ζ-potential measurements were performed on a Malvern Zetasizer 2600 instrument. Both instruments were operating with a 633 nm red laser. All samples were diluted with ultrapure water to an iron concentration of 20 mg/L. Titrations were carried out manually at an ionic strength of 10−2 mol/L NaNO3, using a series of HNO3 and NaOH solutions at 0.1, 0.01, and 0.001 mol/L. Fourier Transform Infrared Spectroscopy. The chemical composition of nanoparticles was studied by FTIR on a Bruker Vector 22 FTIR spectrometer. All dried samples were prepared as KBr pellets. FTIR spectra were recorded from 400 to 4000 cm−1 with a resolution of 4 cm−1, using a pure KBr pellet as zero absorbance reference. Atomic Absorption Spectroscopy. The total iron concentration was determined by atomic absorption spectrophotometry at 248 nm. The measurements were performed on an atomic absorption spectrometer (AAS) iCE 3000 Series Thermo Scientific. Prior to AAS measurements, samples were digested by concentrated hydrochloric acid (37%) and then diluted with hydrochloric acid (1%). The calibration procedure for the iron absorbance at 248 nm was
through silane and epoxide chemistry. In the present work, in order to increase both the efficiency and the specificity of nanoparticle uptake by cancer cells, we synthesized and studied SPION-DOX-PEG-FA nanoparticles functionalized with FA. We focus on investigating the chemical composition/ organization of the organic coating of SPIONs, composed of DOX and of two polymers (PEG5000 and FA-PEG6000). Surfaceenhanced Raman scattering (SERS) spectroscopy, X-ray photoelectron spectroscopy (XPS), and time-of-flight secondary ion mass spectrometry (ToF-SIMS) were used to determine the localization of DOX and FA relative to the inorganic/organic surface. Furthermore, the effects of FA on nanoparticle size, ζ-potential, and uptake by cancer cells in vitro have been studied and discussed in terms of the relation between physicochemical and biological properties of these novel biocompatible nanoparticles useful for cancer theranostics.
2. MATERIALS AND METHODS 2.1. Nanoparticle Synthesis. Chemicals. Iron(II) chloride tetrahydrate (FeCl2·4H2O, 99%), folic acid (96−102%), and N,N′dicyclohexylcarbodiimide (DCC) were purchased from Acros Organics (Noisy le Grand, France). Anhydrous iron(III) chloride (FeCl3), hydrochloric acid (HCl, 37%), and ammonium hydroxide (NH4OH, 35%) were purchased from Fisher Scientific (Illkirch, France). Ferric nitrate nonahydrate (Fe(NO3)3·9H2O), nitric acid (HNO3, 65%), acetone, and diethyl ether (Et2O) were purchased from Carlo Erba (Val de Reuil, France). N-Hydroxysuccinimide (NHS) was purchased from Fluka Chemika (France). Poly(ethylene glycol)bisamine (NH2-PEG-NH2, Mw 6000 Da) and methoxy poly(ethylene glycol)amine (NH2-PEG-OCH3, Mw 5000 Da) were obtained from Rapp Polymer GmbH (Tübingen, Germany). 3-Glycidoxypropyltrimethoxysilane (GOPTS), dimethyl sulfoxide (DMSO), dimethylformamide (DMF), and triethylamine (Et3N) were purchased from Sigma-Aldrich (Saint-Quentin-Fallavier, France). Dialysis membranes (MWCO 3500 and 10 000 Da) were purchased from BioValley (Marne La Vallée, France). Doxorubicin hydrochloride (DOX) was obtained from the EDQM (Strasbourg, France). Ultrapure water was produced using a Barnstead EASYpure RoDi system (Thermo Fisher Scientific, France). All reagents were used as received unless specified. PEG-FA Conjugates. FA was conjugated to PEG using the NHS/ DCC coupling system as described elsewhere33,34 with some modifications. NHS-FA ester was prepared as follows: FA (441.4 mg, 1 mmol) was dissolved in 10 mL of anhydrous DMSO, followed by the addition of DCC (248 mg, 1.2 mmol) and NHS (173 mg, 1.5 mmol). The reaction mixture was stirred overnight at room temperature in the dark under nitrogen. The white dicyclohexylurea precipitate (DCU), a side product of the reaction, was filtered off, and NHS-FA was precipitated with a mixture of acetone−ether (30−70). The yellow precipitate was recovered by filtration on sintered glass, triturated several times with anhydrous ether, and dried under vacuum to yield a yellow powder. NHS-FA ester was characterized by FTIR and used immediately for the following reaction. NHS-FA was coupled to PEG as follows: freeze-dried NH2-PEGNH2 (600 mg, 0.1 mmol) was dissolved in 10 mL of anhydrous DMSO and 350 μL of Et3N, and then NHS-FA (54 mg, 0.1 mmol) in 2 mL of DMSO was added dropwise. The reaction mixture was left under stirring for 24 h at room temperature in the dark under nitrogen. DMSO was evaporated under reduced pressure to yield a viscous yellow solid. The product was purified by ion-exchange chromatography on a Sephadex CM25 column (350 × 24 mm) equilibrated with phosphate buffer pH 7.4, 0.007 mol/L, followed by dialysis against deionized water on MWCO 3500 Da. The yield after purification was 40%. The product, NH2-PEG-FA, was characterized by FTIR spectroscopy and 1H NMR. Initial SPIONs. Initial SPIONs were prepared by the coprecipitation method as previously described.35 Magnetite nanoparticles were precipitated by adding ammonia solution (30 mL, 35%) to an 1498
dx.doi.org/10.1021/la2037845 | Langmuir 2012, 28, 1496−1505
Langmuir
Article
Figure 2. FTIR spectra of initial SPIONs, polymers, and functionalized SPIONs. performed with a series of Fe solutions at concentrations of 0.5, 1, 1.5, 2, 2.5, and 3 mg/L (prepared from a 1 g/L Fe standard solution). Fluorescence Spectroscopy. The presence of DOX was investigated by fluorescence spectroscopy. The measurements were performed on a LabRAM laser scanning confocal microspectrometer (Horiba Jobin-Yvon, France) equipped with an argon laser (MellesGriot, model 36-LAP-431-230, France). Samples were diluted in PBS buffer pH 7.4 to an iron concentration of 2 mg/L. Fluorescence emission spectra were recorded from 500 to 800 nm with the excitation wavelength at 488 nm of the argon laser. Surface-Enhanced Raman Spectroscopy. The presence of FA was investigated by SERS on silver colloids. The measurements were performed on a LabRAM laser scanning confocal microspectrometer (Horiba Jobin-Yvon, France) equipped with a built-in He−Ne laser. Citrate-stabilized Ag colloids were prepared by reducing silver nitrate in the presence of an excess of trisodium citrate, according to the Lee− Meisel protocol.36 Silver colloids were preaggregated by increasing the ionic strength of the medium upon addition of phosphate buffered saline (PBS pH 7.4). The preaggregation involves destabilizing the particle solvation layers and, consequently, can stimulate interactions between the metal surface and the analyte. SERS samples were prepared by mixing preaggregated Ag colloids with either FA solution or with suspensions of FA-functionalized nanoparticles. SERS spectra were recorded from 1800 to 900 cm−1 with the excitation wavelength at 632.8 nm of the He/Ne laser. No sample photodegradation was observed under the conditions used. Each spectrum was recorded as an average of 36 scans (2 s per scan). The spectra presented in Figure 5 are averages of at least three independent measurements. X-ray Photoelectron Spectroscopy. The surface composition was studied by XPS. The measurements were performed on a Thermo Scientific ESCALAB 250 system fitted with a monochromatic Al Kα X-ray source operating at 100 W. A drop of a suspension of nanoparticles was deposited onto an Au substrate and left to dry at room temperature. Samples were cooled with liquid nitrogen to minimize thermal effects of irradiation by X-rays. Wide-scan spectra were acquired for each sample, with high-resolution C 1s, N 1s, Si 2p, O 1s, and Fe 2p spectra at respectively pass energies of 50 and 20 eV. The background was subtracted using a Shirley function calculated from a numerical iterative method. Spectral calibration was carried out by setting the main C 1s peak at 284.6 eV. The atomic fractions of elements were calculated using the peak intensity and Scofield X-ray photoelectron spectrometry (XPS) cross sections. For each reference,
two points of analysis were carried out in order to check the homogeneity of the surface composition. Time-of-Flight Secondary Ion Mass Spectrometry. ToF-SIMS experiments were performed on an IONTOF, TOF.SIMS 5 equipped with a 25 kV bismuth source and a 10 kV C60 source operating at an incident angle of 45°. A primary ion dose of less than 1013 ions/cm2 was used to ensure static conditions. Both positive and negative spectra were collected. The analyzed samples were prepared according to the same method used for the XPS studies. 2.3. Nanoparticle Uptake by Cancer Cells. Cell Culture. Human breast carcinoma cells (MCF-7) were obtained from the American Type Culture Collection (LGC Promochem, Molsheim, France). The cells were grown at 37 °C/5% CO2 in Dulbecco’s Modified Eagle Medium (DMEM) with glucose and L-glutamine (Fisher Bioblock Scientific, Illkirch, France) containing 5% fetal calf serum (Fisher Bioblock Scientific, Illkirch, France) and 50 IU/mL penicillin and 50 μg/mL streptomycin (Sigma-Aldrich). MCF-7 cells were plated on 24-well plates for 48 h at 40 000 cells/well in DMEM containing 5% FCS. After washing the cells with Hank’s Balanced Salt Solution (HBSS, Fisher Bioblock Scientific, Illkirch, France), MCF-7 cells were incubated with either control culture medium, SPIONDOX-PEG or SPION-DOX-PEG-FA for 2 h at 37 °C/5% CO2. The nanoparticle concentrations were adjusted to give the same DOX fluorescence intensity for both preparations, corresponding to a total iron concentration of 1.7 and 1.4 μg/mL for SPION-DOX-PEG and SPION-DOX-PEG-FA, respectively. After the incubation with NPs, the medium was removed and the cells were washed with fresh HBSS. Quantitative NPs Uptake Study by Flow Cytometry. The cells were detached by trypsinization. After centrifugation, they were resuspended in a 0.4% (w/v) trypan blue solution in HBSS to quench the extracellular fluorescence,37,38 thus enabling determination of the fraction that was actually internalized. The trypan blue-treated cells were subsequently washed twice, fixed with formaldehyde, and analyzed by flow cytometry, with ∼10 000 cells measured in each sample (n ≥ 3). Cells were analyzed using a MoFlo (Beckman Coulter, Fort Collins, CO) high-speed cell sorter equipped with a solid state laser operating at 488 nm and 100 mW. Summit software (Beckman Coulter, Fort Collins, CO) was used to perform flow cytometry data analysis and remove debris depending on morphological criteria, as previously described.37 Qualitative NPs Uptake Study by Laser-Scanning Confocal Microscopy (LSCM). The analysis of NPs distribution was carried out on the cells adherent on coverslips. The cells treated and washed 1499
dx.doi.org/10.1021/la2037845 | Langmuir 2012, 28, 1496−1505
Langmuir
Article
Figure 3. TEM images of initial SPIONs (a), SPION-DOX-PEG (b), and SPION-DOX-PEG-FA (c).
synthesis route compared to thermal decomposition,43 the distribution width remains acceptable and the avoidance of organic solvents turns out to be beneficial for biomedical use.44 It is important to underline that this phenomenon did not affect the superparamagnetic properties described above nor the colloidal stability, since no sedimentation was observed over several months. The ζ-potential of initial SPIONs varies from +35 to −35 mV between pH 3 and pH 11 in 10−2 mol/L NaNO3 (Figure 7). The isoelectric point (IEP) was found to be at pH 7.9, which means that at this pH the SPIONs have no electrical charge on their surface and tend to flocculate. These results agree with those reported for iron oxide nanoparticle suspensions (ferrofluids) in the literature.45,46 The total iron concentration of the ferrofluids determined by atomic absorption at 248 nm was about 20 g/L. 3.2. NH2-PEG-FA. NH2-PEG-FA was prepared by a NHS/ DCC mediated coupling reaction between NH2-PEG-NH2 and NHS-FA (N-hydroxysuccinimide ester of FA). NHS-FA was prepared by the reaction of FA with N-hydroxysuccinimide (NHS) in the presence of dicyclohexylcarbodiimide (DCC). FA has two carboxylic groups that can be activated. In order to activate only one of them, the coupling reagents (DCC and NHS) were introduced in a small molar excess. The NHS-FA reacts with the primary amine of PEG to form an amide linkage (Figure 1, step I). In order to favor coupling of one FA to one amino group of PEG, NH2-PEG-NH2 and NHS-FA were introduced in the molar ratio 1:1. Reactions were done in anhydrous solvents since NHS esters hydrolyze in water. Excess of FA was eliminated by dialysis against deionized water. NH2PEG-FA was purified from unreacted NH2-PEG-NH2 and from possibly formed FA-PEG-FA by ion exchange chromatography on Sephadex CM25. The FTIR spectrum of NH2-PEG-FA (Figure 2) exhibits all the characteristic bands of PEG.47 The amide bond formed between FA and the polymer was revealed by the relatively weak IR band at 1640 cm−1 assigned to the CO stretching vibration. The weak band observed at 1606 cm−1 corresponds to the aromatic ring stretching vibration of the FA. In the 1H NMR spectrum, we observed the signals of protons from both aromatic cycles of FA and from methylenelike structures (−CH2CH2O−) of PEG, thus confirming the presence of both moieties. 3.3. Functionalization of SPIONs with DOX and PEGFA. Functionalized nanoparticles SPION-DOX-PEG and SPION-DOX-PEG-FA were prepared in our lab by a one-pot synthesis method similar to that previously described.17 Briefly, DOX, NH2-PEG-OCH3, and/or NH2-PEG-FA were coupled to epoxysilane to form silylated conjugates (Figure 1, steps I and II). The reaction consists in the opening of the epoxide ring by the amino group to form an amine linkage. These silylated conjugates were bonded to the SPIONs surface by reaction of the silanol groups with the hydroxyl groups of iron oxide
as described above were fixed with 4% paraformaldehyde for 15 min and then mounted on microscope glass slide in a Dacko medium (Dako North America, Inc., Carpinteria, CA) diluted with 50% of PBS. The image acquisitions (LSCM optical sections) were carried out with an Olympus Fluoview 500 microscope through a 60× PlanApo Oil objective (NA = 1.4). The excitation and emission wavelengths employed were 488 and 505−525 nm, respectively.
3. RESULTS AND DISCUSSION 3.1. Initial SPIONs. Iron oxide nanoparticles were prepared by coprecipitating iron(II) and (III) salts in alkaline medium according to Massart’s protocol.35 In order to prevent uncontrolled oxidation, the surface of the nanoparticles was partially oxidized.39 The resulting nanoparticles consist of a magnetite core surrounded by a layer of maghemite.40 The FTIR spectrum (Figure 2) exhibits the characteristic band of maghemite at 630 cm−1, while the band at 590 cm−1 is characteristic of both maghemite and magnetite.41 The strong narrow band at 1380 cm−1 is assigned to nitrate ions on the surface of initial SPIONs, which had been dispersed in the presence of nitric acid. TEM micrographs of initial SPIONs (Figure 3a) show mostly spherically shaped nanoparticles with a narrow size distribution. The mean diameter dm and the standard deviation σ calculated from the TEM images by averaging over 150 nanoparticles (NPs) were found to be dm = 9.6 nm and σ = 1.8 nm. The X-ray diffraction patterns (data not shown) of the same SPION crystallites in powder exhibited a good fit with the spinel phase. The mean crystallite size determined from the X-ray data using the Debye−Scherrer formula was 10.86 ± 0.04 nm. The agreement between these two techniques shows that the diffracting volume extends over the whole particle and that the particles are thus monocrystalline.42 As reported in our previous studies,16,17 at room temperature aqueous suspensions of our SPIONs have a superparamagnetic behavior that also corresponds to ca. 10 nm diameter monocrystallites. Indeed, the size limit between superparamagnetism and blocked monodomain behavior depends on the temperature and is about 15−20 nm at room temperature (Salazar et al.42 and references therein). The mean hydrodynamic diameter DH of the initial SPIONs dispersed in acidic aqueous medium in 10−2 mol/L NaNO3 (pH 3) was about 50 nm (intensity mean or Z-average), with a polydispersity index of ca. 0.2. These data are close to the values previously reported for similar nanoparticles16,17,30,39 and correspond to a moderately polydispersed population. The DH values are considerably larger than dm seen in TEM because (i) the DLS analysis comprises solvation layers (water molecules and ions) of colloids and (ii) the intensity-weighted mode of DH calculation exaggerates contributions due to small aggregates that cannot be totally excluded with the coprecipitation protocol.35,39 In spite of this minor drawback of partial agglomeration generally observed with the aqueous 1500
dx.doi.org/10.1021/la2037845 | Langmuir 2012, 28, 1496−1505
Langmuir
Article
the sample surface and tilted at a 60° angle, were used for the analysis of two different depths, up to 10 and about 2 nm, respectively. For both depths, the XPS spectra of SPION-DOX-PEG and SPION-DOX-PEG-FA surfaces showed mostly carbon, oxygen, and small amounts of nitrogen and silicon, in agreement with the composition of the nanoparticles. The carbon and oxygen reside mainly in the poly(ethylene glycol), which constitutes the major fraction in the superficial layers, whereas nitrogen is supplied by FA (when present). For the larger depth analysis, the atomic fraction of nitrogen was higher for SPION-DOXPEG-FA (0.5%) than for SPION-DOX-PEG (0.3%). This result is consistent with the fact that nitrogen is mainly provided by the FA. Moreover, when the analyzed depth decreased to ∼2 nm, the atomic fraction of nitrogen increased from 0.5% to 1.0% for SPION-DOX-PEG-FA, whereas no nitrogen was observed for SPION-DOX-PEG. The latter result indicates that nitrogen is more concentrated at the superficial layers of the organic coating of SPION-DOX-PEG-FA. Consequently, the XPS data on atomic fractions indicate that FA is located in the outside layer of the organic coating. To confirm this conclusion by analysis at the molecular level, we performed SERS spectroscopy measurements using aqueous silver nanoparticle colloids. With SERS, the spectroscopic detection of analytes is possible only if the analyte molecules are in the vicinity (less than 5 nm) of the metal surface responsible for the enhancement of the scattering. SERS spectra of SPION-DOX-PEG-FA mixed with preaggregated silver colloids (Figure 5) exhibit all the main bands of
(Figure 1, step III). We have chosen a ratio between PEG and PEG-FA equal to 4 in order to avoid too large a presence of FA on the particle surface, as discussed below. The FTIR spectra of the freeze-dried functionalized SPIONs (Figure 2) contain all the main bands of the PEG polymer. The characteristic absorption bands of PEG mainly include C−O− C antisymmetric and symmetric stretching vibrations at 1150− 1085 and 1055−870 cm−1 and ethylenic CH2 group stretching and bending vibrations around 2880 and 1467 cm−1, respectively.47 As expected, the characteristic bands of FA at 1640 and 1606 cm−1 described above were only observed in the spectra of SPION-DOX-PEG-FA. Neither the absorption bands of DOX nor those of magnetite and maghemite were detectable due to their relatively low concentration in the sample, below the typical sensitivity threshold in IR spectroscopy of about 5%. However, the presence of DOX was attested by the strong fluorescence signal identified in functionalized SPIONs suspended in a PBS buffer, pH 7.4 (Figure 4). The attachment
Figure 4. Fluorescence spectra of free DOX in solution (a) and DOX covalently bound to SPION-DOX-PEG (b) and SPION-DOX-PEGFA (c). The spectra were recorded from aqueous solutions (a) or suspensions (b, c) in PBS (pH 7.4).
of DOX through the amine group preserves the fluorescence activity of the chromophore. Although recognizable, the fluorescence of the drug linked to the surface of modified SPIONs exhibits a modified spectral shape in comparison with the fluorescence of free DOX in the PBS buffer. As discussed previously,17 these spectral modifications result from the fact that the drug molecule is immobilized on the SPIONs and buried within a polymeric layer. The iron oxide fraction was determined as total iron concentration measured by AAS. For SPION-DOX-PEG and SPION-DOX-PEG-FA it was 0.20 and 0.17 g/L, respectively. As expected, superparamagnetic properties were observed for both SPION-DOX-PEG and SPION-DOX-PEG-FA, the magnetization curves versus magnetic field showing no hysteresis (data not shown). These data confirm the chemical composition of the functionalized nanoparticles, in terms of both inorganic and organic components. 3.4. Structure of the Organic Coating of SPION-DOXPEG-FA. In order to investigate the organization of the organic coating layers around the nanoparticles, we used XPS, SERS, and ToF-SIMS techniques, all three specific to surface phenomena. XPS provides quantitative or semiquantitative analysis of the surface layers by providing information on the elements and their chemical state. SPION-DOX-PEG and SPION-DOX-PEG-FA were analyzed by angle-resolved XPS (ARXPS). In this work, two photoemission angles, normal to
Figure 5. SERS spectra of DOX (a), free FA (b), and SPION-DOXPEG-FA (c).
FA, although with some differences in the band ratios. These changes may be assigned to the different adsorption geometries of FA attached to the PEG polymer of functionalized SPIONs compared with free FA molecules. Because of these changes, one can exclude that the FA SERS signal from the functionalized SPIONs results from free FA. The SERS data confirmed that SPION-DOX-PEG-FA nanoparticles are functionalized with FA in the outside layer of their organic coating. Remarkably, no signal from DOX was observed in the spectra of the particles, in spite of the fact that the drug chromophore has a good SERS activity.48 This implies that DOX is buried inside the organic coating, as expected for its linking to the SPION surface (Figure 1). 1501
dx.doi.org/10.1021/la2037845 | Langmuir 2012, 28, 1496−1505
Langmuir
Article
Figure 6. ToF-SIMS spectra of SPION-DOX-PEG-FA and SPION-DOX-PEG.
Figure 7. Variation of zeta potential (ζ) and hydrodynamic diameter (DH) versus pH of SPIONs and functionalized SPIONs in 10−2 M NaNO3.
common for both types of particles, mainly correspond to CxHyOz-like components. For SPION-DOX-PEG samples, mainly the PEG signal was detected (see inset), as characterized by mass increments of 14 or 16 mass units. The absence of a PEG signal for SPION-DOX-PEG-FA can be explained by the following reasons: (i) the FA signal hides that of the underlying polymer chain, or (ii) the FA-free PEG chains are shorter than those of PEG-FA (MW 5000 Da vs 6000 Da). 3.5. Colloidal Properties of SPION-DOX-PEG-FA. As seen in TEM (Figure 3b,c), the functionalized SPIONs have nearly the same shape and size as the initial SPIONs. Some aggregation of nanoparticles observed in TEM images could be due to the protocol of the sample preparation, i.e., adsorption of the particles on the carbon-coated copper grid, and their
Since SERS exaltation detects analyte molecules up to a few nanometers from the silver surface, the FA might have been buried under a thin outer layer of PEG. By using ToF-SIMS, we are able to analyze a surface thickness of only a few angstroms. Both SPION-DOX-PEG-FA and SPION-DOX-PEG were analyzed by ToF-SIMS under similar experimental conditions (Bi+, 25 kV, pulse 20 ns; mass range: 0−1200 mass units). The data shown in Figure 6 are normalized to the total received dose of primary ions. The presence of FA in the SPION-DOXPEG-FA sample was revealed by the characteristic fragments at 295.1 and 309.1 mass units (see insets a and b), while the fraction of the intact FA molecule was minor (peak at 441.3 mass units, close to the detection limit of the instrument,
