Evaluation of multilayer coated magnetic nanoparticles as biocompatible curcumin delivery platforms for breast cancer treatment

Abstract

Biocompatible multi-layer iron oxide magnetic nanoparticles (MNPs) for drug delivery applications with increased loading capacity, sustained sensitive release profile, and high inherent magnetic properties as well as improved cellular uptake were prepared. In this approach, Fe₃O₄ MNPs were obtained by a co-precipitation method and functionalized using hydroxyapatite (HAP) and/or polyethyleneimine (PEI). They were then modified with β-cyclodextrin (CD) to increase their loading capacity. These MNPs allowed suitable encapsulation of hydrophobic curcumin (CUR) in the CD shell and CUR adsorption into the polymeric layers. The dissolution profile of CUR showed pH sensitive release of CUR. The protein corona pattern of the MNPs by electrophoresis showed lower protein adsorption for CD modified MNPs than for other MNPs. No significant toxicity was observed for the target MNPs, whereas the CUR loaded MNPs inhibited MCF-7 breast cancer cells more efficiently than free CUR. Moreover, the negligible hemolytic activity of the target MNPs showed their excellent haemocompatibility for cancer treatment. The preferential uptake of the drug by MCF-7 cells was observed for CUR loaded MNPs in comparison with free CUR using flow cytometric analysis. As a result, the designed and prepared MNPs can be considered to be a promising CUR delivery platform.

---

1. Introduction

The global incidence of cancer as the second leading cause of death is increasing in populations, and many new cases of cancer diseases are being diagnosed according to annual reports.¹ Different chemotherapy agents have been administered for cancer treatment, but unlike traditional medicines, most of them are toxic, cause adverse reactions for patients and can be very expensive.^2,3 Curcumin (CUR) is an inexpensive drug and a natural polyphenol compound, which is extracted from the herb Curcuma longa (turmeric). Curcumin has diverse therapeutic properties, including wound healing, anti-inflammatory and anti-microbial properties, and especially anti-cancer effects.^4,5 The low toxicity, high dose tolerance and safe profile of CUR in addition to its anti-proliferative properties make it a good candidate for cancer therapy.⁶ However, the pharmaceutical preparations of CUR are restricted due to its low water solubility and poor bioavailability; these are the main challenges that can be improved through a suitable drug delivery system (DDS).^7,8 An efficient anticancer DDS needs to have a number of desirable properties, including water solubility, biocompatibility, haemocompatibility, physicochemical properties, cell internalization and selective protein adsorption.⁹

These major obstacles can be overcome by anticancer nanomedicines in which an engineered and optimized nanoparticle combines multiple functions of DDS.^10,11 Iron oxide magnetic nanoparticles (MNPs) have received a great deal of attention from researchers in nanotechnology. The unique magnetic properties of these nanoparticles, especially Fe₃O₄ magnetic nanoparticles, including easy surface functionalization and modification, make them promising platforms for drug delivery applications.^12–17 It has been shown that the surface coating of MNPs through a layer-by-layer strategy with biocompatible and biodegradable polymers will overcome the tendency of DDS to aggregate.^18,19 In addition, this process results in better physiochemical properties, colloidal stability, high blood circulation half life, the possibility of further modification and a considerable amount of drug loading into the polymer shell.^20–22 Two approaches that have been applied for coating of the iron oxide are ligand addition and ligand exchange.²³

Hydroxyapatite (HAP), a natural inorganic and biocompatible polymer, is the most widely accepted biomaterial, which has been used in medicinal implants, artificial blood trachea and controlled release systems.^24,25 HAP can provide greater loading capacity and biocompatibility to MNPs.²⁶ It has been shown that nanohydroxyapatite has acceptable efficiency for the passive targeting of animal liver cancer.²⁷ Likewise, polyethyleneimine (PEI) has gained fame as a water soluble cationic polymer, which can increase drug release and nanoparticle cellular uptake into the cytosol through the proton sponge effect and enhance avidity towards anionic proteoglycans in tumor vasculature.^28–32 The clinical application of PEI was restricted due to its toxicity.^33,34 However, the chemical modification of PEI alleviates the toxicity of the polymer.³⁵

Cyclodextrins (CD) are also known as pharmaceutical molecular cages, which can improve the solubility, bioavailability and safety of hydrophobic drugs due to possessing a lipophilic central cavity and hydrophilic outer surface.^36–39 Some studies suggested that the solubility of CUR can significantly increase due to the formation of a CUR–CD inclusion complex. However, this complex has some limitations in its application in cell-based treatments due to its negatively charged nature.^40–43

Recent studies showed that CD based CUR delivery systems have improved uptake and showed more potent therapeutic efficacy for cancer cells compared to free CUR.^44–46 Furthermore, the application of magnetic nanoparticles with multi-layer coatings can be useful for drug targeting, hyperthermia and imaging applications.^47–49

The incubation of surface modified nanoparticles with plasma results in the adsorption of proteins so that a protein corona is formed in contact with a biological milieu. The surface coating of the nanoparticles and the bio-nano interface determine the stability and function of the nanoparticles. Therefore, for the in vivo application of such DDS, investigation of the interaction of nanoparticles with the environmental proteins in a biological medium is important.⁵⁰

Taking into account the advantages of the excellent properties of HAP, PEI and CD, in addition to layer-by-layer modification of MNPs, we first prepared superparamagnetic iron oxide nanoparticles as a magnetically responsive core, and then grafted them with PEI as a hydrophilic polymer with the ability to increase drug release at acidic pH as well as nanoparticle cellular uptake. Likewise, HAP was used followed by PEI coating to evaluate the ability of HAP to enhance the loading capacity and biocompatibility of the carrier. In the next step, CD was used to improve the solubility, bioavailability and loading capacity through its hydrophobic cavity and inclusion complex process. The advantages of our formulated MNPs were lower toxicity, haemocompatibility, relatively low protein adsorption, high water solubility, high drug loading capacity, and improved uptake in cancer cells as well as inherent magnetization characteristics.

2. Materials and methods

2.1. Materials

Curcumin (CUR), hexamethylene diisocyanate (HMDI), ethylenediaminetetraacetic acid (EDTA), β-cyclodextrin (CD) and polyethyleneimine (PEI, M_w = 60 000) were purchased from Sigma-Aldrich. Absolute ethanol (EtOH), dimethylformamide (DMF), dimethyl sulfoxide (DMSO), ammonia solution (NH₃·H₂O, 28% ammonia), ferric chloride hexahydrate (FeCl₃·6H₂O), ferrous chloride tetrahydrate (FeCl₂·4H₂O), calcium nitrate [Ca(NO₃)₂], and diammonium hydrogen phosphate ((NH₄)₂HPO₄) were obtained from Merck (Darmstadt, Germany). MCF-7 breast cancer cells were purchased from the Pastor Institute Cell Bank (Iran, Tehran). Millipore Milli-Q® (Burlington, MA, USA) highly purified water was used to make aqueous solutions.

2.2. Synthesis of bare and surface modified Fe₃O₄ MNPs

A modified and simple method was used to decorate MNPs with PEI and CD.⁵¹ The synthesis of Fe₃O₄ MNPs was initially carried out by a chemical co-precipitation method.

Hydroxyapatite coated MNPs were then prepared according to a previously reported method.⁵² Thus, solutions of Ca(NO₃)₂·4H₂O (50 mL, 33.7 mmol) and (NH₄)₂HPO₄ (50 mL, 20 mmol) at a Ca²⁺ : PO₄³⁻ molar ratio of 3 : 1 with pH adjusted to 10 were prepared. Both solutions were added dropwise to an aqueous solution of 100 mg MNPs under nitrogen atmosphere and vigorous stirring. After refluxing the brown solution at 90 °C for 2 h, the mixture was cooled at room temperature followed by storing overnight. Thereafter, the resulting precipitate (Fe@HAP) was collected magnetically and washed with deionized water several times.

PEI was grafted onto the surface of the prepared Fe₃O₄ MNPs or Fe@HAP according to a previously reported method.⁵³ 10 mL of an aqueous solution of PEI (10%, w/v) was added to 100 mg of precisely dispersed MNPs in 10 mL of deionized water. The mixture was vigorously stirred under nitrogen atmosphere and heated to 90 °C for 3 h. The resulting precipitate was magnetically isolated and washed with deionized water and ethanol to remove excess substrate.

For CD modification, HMDI was used as a cross linker to graft CD onto the surface of as-synthetized MNPs. Briefly, 100 mg of HMDI was dissolved in 5 mL dry DMF and added to 500 mg of Fe@PEI or Fe@HAP–PEI, which was previously dispersed in 30 mL dry DMF by sonication. The mixture was stirred at room temperature for 3 h under nitrogen atmosphere. Then, the mixture was added to a prepared solution of 0.88 mmol CD and 1 mmol NaH, which was stirred in dry DMF (10 mL) for 5 hours. The obtained mixture was heated to 70 °C for 3 h, and the precipitate was collected with the help of an external magnetic field and washed 3 times with deionized water and ethanol to remove excess reactants and by-products.

2.3. Characterization

Fourier transform infrared (FT-IR) spectra were obtained on a Nicolet FT-IR Magna 550 spectrophotometer in the range of 500–4000 cm⁻¹. X-ray diffraction (XRD) data were obtained on an XPert MPD advanced diffractometer with Cu (Kα) radiation (wavelength: 1.5406 Å) at a scanning rate of 0.02° s⁻¹ at room temperature in the range of 2θ from 4° to 120°. Thermogravimetric analysis (TGA) was conducted on the dried powder samples on the TGA Q50 thermogravimetric analyzer of a TA instrument from room temperature to 900 °C with a heating rate of 10 °C min⁻¹ under argon flow. A vibrating sample magnetometer (VSM, Meghnatis Kavir Kashan Co., Kashan, Iran) with a maximum magnetic field of 10 kOe at 298 K was used to study the magnetic behavior of the prepared MNPs. Ultraviolet-visible (UV-vis) spectra of all MNPs were obtained using a Jasco-530 UV-vis spectrophotometer in the range of 220–800 nm. The hydrodynamic size distribution and zeta potential of the as-synthesized MNPs were determined using a Zetasizer MS2000 (Nano ZS, Malvern Instruments, Malvern, UK) according to the dynamic light scattering principle technique. For these measurements, after diluting 50 μL of 1 mg mL⁻¹ nanoparticle suspension to 5 mL with distilled water, the mixture was sonicated for 1 min. The particle sizes were measured in 4 min at 25 °C. The results were reported from 5 runs of each MNP. The zeta potential of the MNPs was obtained from the average of 3 readings. The morphology of the samples was investigated by a transmission electron microscope (TEM, EM208 Philips Company). Surface images of the samples were acquired by a scanning electron microscope (SEM, VEGAII TESCAN) at an accelerating voltage of 15 kV. The SEM samples were prepared by dropping the diluted MNPs solution onto glass slides and drying the slides at room temperature.

2.4. CUR loading into the MNPs

200 μL of 10 mg CUR in acetone was added drop-wise to an aqueous dispersion of 10 mg of different MNPs in 3 mL water, followed by stirring at 400 rpm for 24 h to allow the CUR molecules to penetrate into the CD polymer layers around the nanoparticle core.⁴⁹ Then, MNPs containing CUR were separated from free CUR and washed with distilled water three times through interval re-suspension in water. The purified loaded MNPs were lyophilized.⁴⁴ To estimate the loading efficiency of CUR, 1.0 mg of lyophilized nanoparticles containing CUR were dissolved in DMSO followed by gentle shaking on an orbital shaker for 24 h at room temperature in the dark to completely extract the CUR in solution. For each MNP, DMSO solution containing extracted CUR was centrifuged at 11 000 rpm (Hettich, ROTINA 380R, rotor with maximum rate of 1000 rpm, Germany) to separate the MNPs. The yellow clear supernatants were collected and used for the estimations. The CUR concentration in each collected supernatant was quantified using the UV-vis spectrophotometer at 438 nm. A standard plot of CUR in DMSO (0–7 μg mL⁻¹) was prepared under the same conditions. The CUR loading percentage was defined as the weight of drug (micrograms) present in 1 mg of MNPs. The CUR loading content (CLC) and efficiency (CLE) were calculated as follows:

CLC (%) = weight of CUR in MNPs/weight of MNPs × 100 (1)

CLE (%) = C_E/C_T × 100 (2)

where C_E refers to the mass of loaded CUR in MNPs and C_T refers to the total mass of CUR added initially. The measurement was repeated 3 times.⁴⁸

2.5. In vitro drug release experiment

The release profile of CUR from the MNPs was determined by transferring 2 mg of CUR loaded MNPs into 3 mL phosphate buffer saline (PBS) of pH 5.5 or pH 7.4 containing 0.1% w/v Tween-80 in a 5 mL tube, followed by shaking the solution at 37 °C at 200 rpm in an orbital shaker. Tween-80 was used in the medium to maintain sink conditions and increase the solubility of CUR in the aqueous phase. At predetermined time points, the tubes were centrifuged at 11 000 rpm for 25 minutes, and the supernatant was stored at 4 °C in the dark. All medium solutions in the tube were removed and replaced with equivalent volumes of fresh medium. The concentrations of CUR were measured by ultraviolet-visible spectrophotometry based on the absorbance intensity at 450 nm. Data were expressed as the mean value of three repeated experiments. The standard curve of CUR (0.5–3.0 μg mL⁻¹) was prepared under the same conditions.⁵⁴

2.6. Protein corona study

The plasma protein adsorption on the surface of MNPs as a corona was evaluated by following a sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) protocol.⁵⁵

For the protein corona preparation, human plasma (∼100 mL each) was obtained from the Iran Blood Transfusion Institute, Tehran. The pooled units were liquated and stored in a −80 °C freezer until use. On the experiment day, any possible protein aggregate complex was separated as a pellet by centrifugation at 15 000 rpm for 4 min after bringing each sample to room temperature. A total of 0.1 mL of each MNP (1 mg mL⁻¹) was added to 0.9 mL of human blood plasma and incubated for 1 h at 37 °C. Then, the hard corona was isolated from the MNP protein corona complexes with centrifugation at 11 000 rpm for 10 min followed by resuspending the pellet in 0.5 mL of PBS. The washing process was repeated 3 times before resuspension of the final pellet. The SDS-PAGE analysis was performed within 24 h.

For protein corona identification by SDS-PAGE, the eluted hard corona proteins from all collections were mixed with sample buffer [Tris–HCl (125 mM); pH 6.8; SDS (4% w/v); glycerol (20% w/v); 2-mercaptoethanol (5% w/v) and bromophenol blue (0.06% w/v)] in PBS and boiled at 100 °C for 10 min. Equal volumes of samples were loaded on 12% acrylamide SDS-PAGE for separation and run for ∼90 min at 120 V. The acidic silver nitrate staining method was used for the detection of separated proteins. Gel densitometry was analyzed by ImageJ (1.410 version).

2.7. MTT cell viability assay

The in vitro cytotoxicity of the nanoparticles was assessed by a MTT (3-[4,5-dimethyl-thiazol-2-yl]-2,5-diphenyltetrazoliumbromide) cell viability assay after treating the MCF-7 breast adenocarcinoma cells with CUR loaded Fe@PEI–CD and Fe@HAP–PEI–CD as well as unloaded MNPs according to the manufacturer's instructions (Roche, Cell proliferation kit I, MTT). Briefly, the cancer cells were seeded into a 96-well plate at an initial density of 2 × 10³ cells per well in a final volume of 100 μL DMEM (Dulbecco's Modified Eagle's Medium-high glucose) followed by incubation overnight to bring the cells to confluence. Furthermore, the medium was replaced with 100 μL fresh medium containing pure PBS buffer (control), CUR and MNPs at different concentrations (5, 10, 20, 40 and 80 μM). After incubation (48 h) at 37 °C and 5% CO₂, 10 μL of the MTT labeling reagent (final concentration 0.5 mg mL⁻¹) was added to each well and incubated for 4 h to detect the metabolically active cells. Then, 100 μL DMSO was added to each well to replace the culture medium and dissolve the insoluble formazan crystals. After overnight incubation in a humidified atmosphere at 37 °C, the absorbance of each well was measured using an Asys Hitech ELISA reader spectrophotometer at 570 nm. The antiproliferation potential of the MNPs treatment was calculated as a percentage of cell growth with respect to the PBS controls using the following equation:

Cell viability (%) = A (in 570 nm of treated cells)/A (in 570 nm of control cells) × 100

Standard deviations were obtained from 3 replicates.

2.8. Hemolysis assay

Human blood was provided by the Clinic of Tehran University (Tehran, Iran). The red blood cells (RBCs) were prepared for hemolytic assay according to the reported procedure.⁵⁶ Briefly, the fresh male human blood was stabilized with EDTA and centrifuged at 3000 rpm for 5 min to remove the plasma as the supernatant. The resulting precipitation was purified 4 times through successive rinsing with PBS buffer. After diluting the suspension of RBC 10 times with PBS buffer, 100 μL of RBC suspension was separately added to 900 μL of each sample and control followed by gentle shaking and incubating for 2 h at room temperature. PBS and water were used as negative and positive controls in addition to different concentrations of MNPs (10–80 μM), respectively. Finally, after centrifugation of the samples and controls at 10 000 rpm for 3 min, absorbance of the hemoglobin in the supernatants was recorded by a UV-vis spectrophotometer at 541 nm. Images of the centrifuged samples were also taken with a camera. To determine the hemolytic percentages of the different samples, the difference in absorbance between the samples and the negative control were divided by the difference between the positive and negative controls.

2.9. Cellular uptake investigation

To investigate the cellular uptake of MNPs in MCF-7 human breast adenocarcinoma cells, 4 × 10⁵ cells were seeded in 6-well plates in 2 mL medium (DMEM) under standard culture conditions. After overnight incubation and cell attachment, media was replaced with 2 mL of serum-free medium containing 50 μg mL⁻¹ CUR or MNPs loaded with 50 μg mL⁻¹ CUR. After 12 hours, the cells were washed three times with 1× PBS (pH 7.4) and the fluorescence intensity (with excitation wavelength of 420 nm and emission wavelength of 470 nm) of CUR per cell was determined using a flow cytometer (FACScan, LYSIS II, Becton Dickinson). Standard deviations were calculated from 3 replicates.

2.10. Statistical analysis

The quantitative data are presented as the mean ± standard deviation (SD) for each experiment using Microsoft Excel 2013 software. All experiments were performed with three replicates, and the results presented were from representative experiments. Significant differences between the mean values were evaluated using a t-test. A value of P < 0.05 was considered to be statistically significant.

3. Results and discussion

Our primary goal was to evaluate biocompatible nanoparticles with suitable physiochemical properties for CUR delivery to breast cancer cells and to investigate the release patterns, protein adsorption and cell internalization of the nanoparticles. Therefore, we synthesized and modified iron oxide nanoparticles and determined their physiochemical properties through characterization by DLS, TGA, XRD, VSM and FTIR. Scheme 1 illustrates the sequence steps for the preparation of the target MNPs and drug (CUR) loading.

Scheme 1 Schematic showing MNP preparation and CUR loading.

3.1. Characterization of nanocomposites

a Morphology analysis by SEM and TEM. SEM images of the prepared nanoparticles are illustrated in Fig. 1A and B, and show the spherical shape and rough appearance of the CD modified MNPs. More accurate information on the morphology, polymeric shell and particle size of the MNPs was obtained by TEM micrograph (Fig. 1C and D). The spherical morphology of the prepared MNPs, with well-shaped particles, confirmed the SEM data in which a polymeric shell on the surface of the spherical core of the nanocomposites could be easily observed. All MNPs were found to be monodispersed. The estimated mean diameters for MNP@PEI–CD and MNP@HAP–PEI–CD were 195 nm and 240 nm, respectively.

Fig. 1 SEM images of Fe@PEI–CD (A) and Fe@HAP–PEI–CD (B); TEM micrographs of Fe@PEI–CD (C) and Fe@HAP–PEI–CD (D); hydrodynamic sizes of the prepared nanostructures (E).

b Dynamic light scattering (DLS) and zeta potential analysis. The particle size data of the MNPs by DLS was close to the TEM/SEM micrographs (Fig. 1E). The slight differences between the particle sizes can be attributed to the fact that the particles for the DLS measurement were water swollen, compared to the dried state of the particles used for TEM/SEM measurement. A difference in the measured sizes between DLS and TEM/SEM has also been reported with other nano-materials.⁵⁷ The modification of MNP@PEI with CD changed the zeta potential from 19.1 mV to −4.3 mV, and MNP@HAP–PEI modification with CD decreased the zeta potential from 4.9 mV to −11.9 mV.

c XRD analysis. The XRD patterns of the naked Fe₃O₄ MNPs and final nanocomposites are depicted in Fig. 2A. The peaks appearing at 2θ angles of 30.1°, 35.6°, 43.3°, 53.5°, 57°, 63° and 74° clearly show the cubic crystalline structure of the synthesized Fe₃O₄ MNPs. It is also clear that the surface modification of the MNPs is well matched with that of the bare Fe₃O₄ crystals and is in agreement with those reported in the literature.^58–60 In addition, the presence of HAP in the structure of Fe@HAP–PEI–CD could be confirmed by the appearance of new peaks at 2θ of 25.8° (002), 28.6° (210), 32.2° (112), 40° (310), 47° (222) and 49.5° (213), as shown in Fig. 2A with a star. These results are consistent with recent reports.^61,62

Fig. 2 XRD patterns of the naked Fe₃O₄ MNPs and final nanocomposites (A); TGA curves (B) and hysteresis loops (C) of all MNPs; UV-vis spectra of ethanolic dispersion of MNPs with or without CUR (D).

d TGA analysis. The TGA results showed the thermal decomposition behavior of all synthetic nanoparticles and could be useful to determine the percentage of ligand coated on the surface of the MNPs. As indicated in Fig. 2B, three steps of weight loss can be considered for the naked MNPs. These weight losses can be observed in the ranges of 25–150 °C, 150–580 °C and finally 580–800 °C, which could be ascribed to the hydrogen bonded water desorption, the removal of trapped water from the lattice and the phase transition of Fe₃O₄ to Fe–O, respectively.⁵⁹

The first weight loss for all modified MNPs in the range of about 25–150 °C could be attributed to the vaporization of water. The thermogram of Fe@PEI MNPs shows about 10% weight loss in the range of 150–400 °C in comparison to naked Fe₃O₄, which could be related to the oxidative decomposition of PEI. Furthermore, the grafting of CD on the surface of Fe@PEI using HMDI led to a weight loss percentage of about 35% in the range of about 150–400 °C. Therefore, approximately 20% of this weight loss could be related to the CD and HMDI. Similarly, a weight loss of about 37% in the range of about 150–400 °C in the thermogram of Fe@HAP–PEI could be ascribed to the oxidative decomposition of PEI as well as the dehydration of HAP. In addition, comparison between the thermograms of Fe@HAP–PEI and Fe@HAP–PEI–CD shows that approximately 18% of the grafting density is related to the PEI and HMDI moieties.

e Magnetization analysis. As illustrated in Fig. 2C, the zero coercivity and remanence of the hysteresis loops of the MNPs demonstrate the superparamagnetic behavior of all the MNPs. The saturation magnetization (MS) of bare Fe₃O₄ MNP was 90 emu g⁻¹, whereas those for Fe@PEI and Fe@HAP–PEI were 43.7 and 28.2 emu g⁻¹, respectively. The MS value decreased to 23.4 and 21.1 emu g⁻¹ after CD modification of Fe@PEI and Fe@HAP–PEI, respectively. The decreased magnetization value could be attributed to the increased mass of the particle on the surface of the MNPs when unmagnetized shells are used. This saturated magnetization for modified MNPs is sufficient for magnetic separation or direction.^63,64

f UV-vis analysis. Fig. 4B shows the UV-vis spectra of the CUR-loaded and unloaded MNPs. The UV-vis spectra of the MNPs containing CUR in ethanol showed a maximum absorbance at 420 nm. Another maximum absorbance was observed at about 280 nm for the CD containing MNPs.

Fig. 3 FTIR spectra of pure CUR (red dashed line), Fe@PEI–CD–CUR (blue line) and Fe@HAP–PEI–CD–CUR (black line).

Fig. 4 Dissolution release profiles of CUR from CUR loaded Fe@PEI–CD (A) and Fe@HAP–PEI–CD (B) nanoparticles.

g FTIR analysis. The stepwise chemical modification of the Fe₃O₄ MNPs was evaluated by FTIR analysis, as shown in Fig. S1 (see ESI†). The FTIR spectrum of the bare Fe₃O₄ MNPs shows main absorption peaks at around 600 and 3400 cm⁻¹, which are attributed to the stretching vibrations of the Fe–O and O–H groups of the Fe₃O₄ MNPs, respectively.^65,66 The FTIR spectrum of MNP@HAP exhibited a strong and broad absorption band at 1061 cm⁻¹, which is attributed to the PO₄³⁻ groups of HAP on the surface of the MNPs.⁶⁷ The presence of PEI on the surface of the MNPs could be assigned to the presence of C–H stretching vibrations at around 2800–2900 cm⁻¹ and the C–N stretching vibration at around 1340 cm⁻¹ (Fig. S1†).⁶⁸ HMDI was used to covalently attach CD on the surface of the MNPs. This aim can be affirmed by the formation of carbonyl bonds, which are attributed to the reaction of the PEI amine groups and CD hydroxyl groups with HMDI (Fig. S1†).^56,69

The CUR loaded MNPs could be also confirmed by the FTIR spectra (Fig. 3). A sharp band at 3511 cm⁻¹ and a broad band at about 3400 cm⁻¹ are assigned to the presence of the OH groups of CUR. The peak appearing at 1626 cm⁻¹ could be attributed to the vibration of C=O bonds, which have overlapped with C=C bonds. Furthermore, the band appearing at 1601 cm⁻¹ is due to the stretching vibrations of the C=C aromatic rings. All the other bands are in agreement with the FTIR spectra of CUR.^70,71 The FTIR spectra of the Fe@HAP–PEI–CD–CUR and MNP@PEI–CD–CUR show the combined characteristic peaks of CUR and MNPs.

3.2. CUR loading into the CD modified MNPs

Drug loading capacity and drug loading efficiency were two important factors to consider when investigating the performance of each carrier. As shown in Fig. 2D, CUR loading in the MNPs was estimated using a UV-vis spectrophotometer, according to a reported procedure.⁴⁸ A detailed calculation is provided in the ESI (Table S2 and Fig. S3†). The results showed that the CUR loading content and efficiency were 14.1% and 70.6% for Fe@PEI–CD, respectively. However, Fe@HAP–PEI–CD showed more potential for CUR loading (CLC = 18% and CLE = 76.6%) in comparison to Fe@PEI–CD, which could be related to the presence of the inner HAP layer as a porous material.⁷² Taking into account the advantages of CD, with its hydrophobic bucket structure and capability of forming an inclusion complex with CUR, the current MNPs showed proper CUR loading.⁵⁴

3.3. CUR release profile

Fig. 4 shows the dissolution release profiles of CUR from the MNPs at different pH values. This was calculated according to the CUR standard calibration curve, as shown in Fig. S4.† As depicted, both CD modified MNPs showed faster release profiles at pH 5.5 in comparison to pH 7.4. As can be observed from Fig. 4, release percentages of about 52% and 38.1% were obtained for Fe@HAP–PEI–CD and Fe@PEI–CD, respectively, after 5 days at pH 5.5. However, a significant diminution was observed in the release profile of MNPs at pH 7.4 at the same times (28% and 18.4% for Fe@HAP–PEI–CD and Fe@PEI–CD, respectively). The higher release percentage for the MNPs at pH 5.5 may be attributed to the sensitivity of PEI to acidic media. In comparison, the greater CUR release from Fe@HAP–PEI–CD than from Fe@PEI–CD nanoparticles at acidic pH could be related to the thicker PEI layer in the former MNP, as confirmed by TGA analysis. The release profile of CUR for Fe@HAP–PEI–CD at pH 7.4 showed a biphasic behavior in which a sustained release was observed for up to 5 days, followed by a delayed time release with a lag time of 4 days (Fig. 4B). This particular profile could be attributed to sequential CUR release from the CUR located in the inner HAP–PEI and outer CD layers. A more extended release of CUR was observed at pH 7.4 for CUR loaded Fe@PEI–CD than for Fe@HAP–PEI–CD nanoparticles.

3.4. Protein corona study

Different coatings of the MNP core can alter the composition and extent of the protein corona in a nanoparticle complex.^55,73 In this study, the effects of the PEI, HAP and CD coating layers on the pattern and amount of protein corona formation were investigated by 1-D SDS gel electrophoresis. According to the SDS-PAGE analysis shown in Fig. 5, although a very similar protein pattern was observed for the protein coronas of the different MNPs; however, the surface functional groups had different influences on plasma protein adsorption. It could be observed that the molecular weights of the proteins from the 45 to 75 kDa range were more enriched. In addition, the most intensive protein band was also observed for Fe@PEI. A decrease in the protein corona of Fe@PEI modified with CD in MNP@PEI–CD showed that the positive surface charge of PEI probably resulted in more protein adsorption. Because the protein bond intensity of Fe@HAP–PEI–CD also decreased to a lower extent than that of Fe@HAP–PEI, it could be concluded that modification with CD as a biocompatible agent can improve nanoparticles for in vivo application by recovering lower protein coronas.

Fig. 5 (A) SDS-PAGE gel (12%) of human plasma proteins obtained from Fe@PEI (2), Fe@PEI–CD (3), Fe@HAP–PEI (4) and Fe@HAP–PEI–CD (5). The M_w of the proteins in the standard ladder are reported in column 1 for reference. (B) Molecular weight histograms representing the total band intensity of proteins associated with nanoparticles after exposure with human plasma (bond intensity indicates amount or concentration of proteins that bind to nanoparticles).

3.5. Cytotoxicity

Cell viability assays in response to the MNPs were used to assess the cytotoxicity of the nanoparticles (Fig. 6). No significant effect for the control treatments was observed on cell growth. The IC50 values (concentrations of 50% MCF-7 cancer cells growth inhibition), as a quantitative measure for cytotoxicity, were 40 μM for free CUR and 22 and 13 μM for Fe@PEI–CD–CUR and Fe@HAP–PEI–CD–CUR, respectively. This data showed that the CUR loaded MNPs were more effective than free CUR in suppressing cell proliferation. After treating MCF-7 cells with free MNPs, all target carriers were found to be nearly nontoxic in the given concentration range, as shown in Fig. 6A and B. It has been well documented that PEI is toxic due to its enormous number of free amines and its positive surface chemistry.^74,75 However, it has been demonstrated that chemical modification can alleviate its toxicity.^76–78 In this study, CD played a role in biocompatibility by changing the surface charge of PEI functionalized MNPs from positive to negative and subsequently reducing the PEI toxicity. The lower zeta potential of the nanoparticles confirms the surface charge alteration (see Table S1†). In general, the cell viability data showed that after modification of the PEI containing MNPs, their cytotoxicity was significantly reduced (data not shown).

Fig. 6 Anti-proliferative effects (A & B) and hemolytic activity (C & D) of Fe@PEI–CD–CUR and MNP@HAP–PEI–CD–CUR. Cell viability of breast cancer cells (MCF-7) was measured using an MTT assay by UV-vis spectrophotometer at 492 nm, whereas the hemolysis percentages of the same NPs were investigated at 541 nm and in the 10–80 μM concentration range.

It was recently reported that MNP–CUR damages the mitochondrial membrane more effectively than free CUR, which thus increases the anticancer potential of CUR as a result of improved uptake.⁵⁴ It was suggested that Bcl-xL expression is strongly suppressed after treatment with CUR loaded MNPs, compared to equivalent amounts of CUR.⁴⁹

3.6. Hemolysis assay

Hemolytic tests were performed as a supplementary test for the evaluation of cytotoxicity and as a sensitive measure for the biosafety assessment of the effects of the MNPs on erythrocytes.⁷⁹ As shown in Fig. 6C and D, the treatment of diluted erythrocytes with Fe@PEI–CD, Fe@HAP–PEI–CD and PBS control solution resulted in sedimentation, whereas water as a positive control demonstrated apparent hemolysis behavior without any sedimentation. As can be observed in Fig. 6C and D, both target MNPs exhibited good hemocompatibility, similar to that of PBS. No significant difference in OD values was observed between the two groups (P > 0.05). The maximum hemolytic percentages of the Fe@PEI–CD and Fe@HAP–PEI–CD samples in the studied concentration range (10–80 mM) were 1.0% and 0.7%, respectively. The results suggest excellent hemocompatibility for these MNPs according to the standard acceptance limit (less than 5%).⁸⁰ Furthermore, the hemolytic results are in good agreement with the cytotoxicity results for each of the MNPs.

3.7. Cellular uptake of NPs

Because CUR inherently emits strong green fluorescence, flow cytometry analysis was utilized to quantify the cellular uptake of the CUR-loaded MNPs. Fig. 7 shows the histograms of cellular CUR fluorescence for MCF-7 cells incubated with various MNPs. Cells without CUR were used as a negative control, showing only autofluorescence (Fig. 7A). Strong CUR fluorescence was observed when MCF-7 cells were incubated with CUR loaded-nanoparticles. The mean CUR fluorescence intensities for CUR (Fig. 7B), CUR loaded Fe@PEI–CD (Fig. 7C) and CUR loaded Fe@HAP–PEI–CD (Fig. 7D) were approximately 1.3-fold, 1.5-fold and 4.3-fold higher than that for the negative control cells (Fig. 7A), respectively. The mixture of CUR loaded Fe@HAP–PEI–CD and Fe@PEI–CD resulted in 5.3-fold higher intensity (Fig. 7E). The uptake quantitative analyses of free CUR and CUR loaded MNP nanocomposites by flow cytometry have been provided in Fig. S5.† Higher uptake of CUR loaded MNPs in breast cancer cells could be an indication for an improved therapeutic index. It has been shown that CD can enhance the cell internalization of some hydrophobic drugs.⁸¹ It was suggested that the enhanced drug solubility and the decrease in the free fraction of the drug causes the enhancement of the cellular uptake of drugs by CD.⁸² It was recently noted that two distinct mechanisms, consisting of diffusion via the cell wall and phagocytosis, are implicated in the cellular uptake of PEG modified Fe₃O₄–PEI MNPs.⁵⁶ In addition, it has been reported that positive surface charge of the PEI layer on the surface of the MNPs can increase nanoparticle cellular internalization through high avidity towards the anionic proteoglycans on cancer cells following enhanced endocytosis.^83,84 According to some reports, complexation of the CUR with CDs results in improvement in CUR solubility.^38,85

Fig. 7 Cellular uptake of CUR and CUR loaded MNP formulations by flow cytometry. (A) Autofluorescence of MCF-7 cells; (B) MCF-7 cells incubated with CUR; (C) MCF-7 cells incubated with CUR loaded Fe@PEI–CD; (D) MCF-7 cells incubated with Fe@HAP–PEI–CD; and (E) MCF-7 cells incubated with both CUR loaded MNPs simultaneously.

According to the obtained results, this multilayer coated magnetic nanocomposite could be considered as a new nanomedicine platform for other hydrophobic dugs. The properties of relatively low protein corona, colloidal stability, higher drug loading content, pH sensitive and sustained release profile, hemocompatibility and enhanced uptake in MCF-7 breast cancer cells make these nanocomposites a convenient drug delivery system for cancer therapeutic applications. According to the level of the inherent magnetic properties of the target nanocomposites, we believe they could also be considered as novel theranostic agents and evaluated for magnetic resonance imaging (MRI) analysis in addition to therapeutic applications.

4. Conclusion

There has been a great interest in the design and preparation of formulations with safe, efficient and biocompatible properties that exhibit potent anticancer activity for drug delivery and other biomedical applications. To achieve this goal, it is necessary to overcome major obstacles for some MNPs such as opsonization by the reticuloendothelial system (RES), low loading capacity, high protein adsorption, low solubility, low cellular uptake and poor therapeutic efficacy. We constructed nanocarriers with a high loading capacity for CUR, a sustained release profile, haemocompatibility, biocompatibility and an appropriate protein corona pattern. In addition, the CUR loaded MNPs were more potent than free CUR for the inhibition of MCF-7 cell proliferation as a result of enhanced uptake in breast cancer cells. Due to their high inherent magnetic properties, these nanomedicine platforms could be used for theranostic applications after ensuring their MRI imaging abilities. These findings suggest that the developed multilayer MNPs, coated with polymers and modified with CD, are excellent candidates for efficient drug delivery applications.

Acknowledgements

This study was supported by a grant from the Research Council of Tehran University of Medical Sciences and from the Iran National Science Foundation (INSF).

References

1. A. Jemal, R. Siegel, E. Ward, Y. Hao, J. Xu and M. J. Thun, Ca-Cancer J. Clin., 2009, 59, 225.
2. A. Sudhakar, History of Cancer, Ancient and Modern Treatment Methods, J. Cancer Sci. Ther., 2009, 1, 1.
3. O. Tacar, J. Pharm. Pharmacol., 2010, 65, 157.
4. J. Ravindran, S. Prasad and B. Aggarwal, AAPS J., 2009, 11, 495.
5. Y. Sun, L. Dua, Y. Liu, X. Li, M. Li, Y. Jin and X. Qian, Int. J. Pharm., 2014, 469, 31.
6. A. B. Kunnumakkara, P. Anand and B. B. Aggarwal, Cancer Lett., 2008, 269, 199.
7. E. I. Paramera, S. J. Konteles and V. T. Karathanos, Food Chem., 2011, 125, 913.
8. S. S. Bansal, M. Goel, F. Aqil, M. V. Vadhanam and R. C. Gupta, Cancer Prev. Res., 2011, 4, 1158.
9. M. Salem, S. Rohani and E. R. Gillies, RSC Adv., 2014, 4, 10815.
10. R. Duncan, Nat. Rev. Cancer, 2006, 6, 688.
11. B. Sivakumar, R. G. Aswathy, Y. Nagaoka, S. Iwai, K. Venugopal, K. Kato, Y. Yoshida, T. Maekawa and D. N. S. Kumar, RSC Adv., 2013, 3, 20579.
12. W. Wu, Q. He and C. Jiang, Nanoscale Res. Lett., 2008, 3, 397.
13. T. K. Indira and P. K. Lakshmi, Int. J. Pharma Sci. Res., 2010, 3, 1035.
14. D. Chen, Q. Tang, X. Li, X. Zhou, J. Zang, W. Xue, J. Xiang and C. Guo, Int. J. Nanomed., 2012, 7, 4973.
15. B. Hu, J. Pan, H. L. Yu, J. W. Liu and J. H. Xu, Process Biochem., 2009, 44, 1019.
16. M. A. Faramarzi and A. Sadighi, Adv. Colloid Interface Sci., 2013, 189, 1.
17. K. Li, H. Zhu, Y. Zhou, J. Ding, J. Shen, Z. Li, Z. Xu and P. K. Chu, RSC Adv., 2013, 3, 10598.
18. A. Akbarzadeh, N. Zarghami, H. Mikaeili, D. Asgari, A. M. Goganian, H. K. Khiabani, M. Samiei and S. Davaran, Nanotechnol., Sci. Appl., 2012, 5, 13.
19. W. S. Arora, Int. J. Nanomed., 2012, 7, 3445.
20. X. Chen, J. Zhu, Z. Chen, C. Xu, Y. Wang and C. Yao, Sens. Actuators, B, 2011, 159, 220.
21. M. Mahmoudi, S. Sant, B. Wang, S. Laurent and T. Sen, Adv. Drug Delivery Rev., 2011, 63, 24.
22. A. K. Gupta and M. Gupta, Biomaterials, 2005, 26, 3995.
23. T. Ahmad, H. Bae, I. Rhee, Y. Chang, J. Lee and S. Hong, Curr. Appl. Phys., 2012, 12, 969.
24. V. S. Gshalaev and A. C. Demirchan, Hydroxyapatite: Synthesis, Properties and Applications (Biomaterials-Properties, Production and Devices), Nova Science Publisher, 2012.
25. M. Iafisco, E. Varoni, M. Di Foggia, S. Pietronave, M. Fini, N. Roveri, L. Rimondini and M. Prat, Colloids Surf., B, 2012, 90, 1.
26. H. Kim and J. C. Knowles, Biomaterials, 2004, 25, 1279.
27. K. Biswanath, D. Ghoshb, M. K. Sinhaa, P. S. Sena, V. K. Ballaa, N. Dasb and D. Basua, Ceram. Int., 2013, 39, 9557.
28. D. W. Dong, B. Xiang, W. Gao, Z. Z. Yang, J. Li and X. R. Qi, Biomaterials, 2013, 34, 4849.
29. B. Chertoketal, Biomaterials, 2010, 31, 6317.
30. S. M. Bain, H. Yiu, A. E. Haj and J. Dobson, J. Mater. Chem., 2007, 17(24), 2561.
31. F. M. Kievit, F. Y. Wang and C. Fang, J. Controlled Release, 2011, 152, 76.
32. O. Boussif, F. Lezoualc, M. A. Zanta, M. D. Mergny, D. Scherman, B. Demeneix and J. P. Behr, Proc. Natl. Acad. Sci. U. S. A., 1995, 92, 7297.
33. M. Chorny, B. Polyak and I. S. Alferiev, FASEB J., 2007, 21, 2510.
34. H. Mimi, K. M. Ho and Y. S. Siu, J. Controlled Release, 2012, 158, 123.
35. A. Zintchenko, A. Philipp and A. Dehshahri, Bioconjugate Chem., 2008, 19, 1448.
36. A. Rasheed, A. Kumar and V. Sravanthi, Sci. Pharm., 2008, 76, 567.
37. A. Vyas, S. Saraf and S. J. Saraf, J. Inclusion Phenom. Macrocyclic Chem., 2008, 62, 23.
38. L. Y. Qiu, Nanomedicine, 2010, 5, 193.
39. G. Horvath, T. Premkumar, A. Boztas, E. Lee, S. Jon and K. E. Geckeler, Mol. Pharm., 2008, 5, 358.
40. A. Popat, S. Karmakar, S. Jambhrunkar, C. Xu and C. Yu, Colloids Surf., B, 2014, 117, 520.
41. S. S. Dhule, P. Penfornis, T. Frazier, R. Walker, J. Feldman, G. Tan, J. He, A. Alb, V. John and R. Pochampally, Nanomed. Nanotechnol. Biol. Med., 2012, 8, 440.
42. R. S. Mulik, J. Mönkkönen, R. O. Juvonen, K. R. Mahadik and A. R. Paradkar, Int. J. Pharm., 2010, 398, 190.
43. A. Trapani, A. Lopedota, M. Franco, N. Cioffi, E. Ieva, M. G. Fuentes and M. J. Alonso, Eur. J. Pharm. Biopharm., 2010, 75, 26.
44. M. M. Yallapua, M. Jaggi and S. C. Chauhana, Colloids Surf., B, 2010, 79, 113.
45. M. M. Yallapu, M. Jaggi and S. C. Chauhan, Macromol. Biosci., 2010, 10, 1141.
46. B. Hegge, R. B. Schuller, S. Kristensen and H. H. Tonnesen, Pharmazie, 2008, 63, 585.
47. M. M. Lin, H. H. Kim, H. Kim, H. J. Dobson and D. K. Kim, Nanomedicine, 2010, 5, 109.
48. M. M. Yallapu, S. F. Othman, E. T. Curtis, B. K. Gupta, M. Jaggi and S. C. Chauhan, Biomaterials, 2011, 32, 1890.
49. S. S. Banerjee and D. H. Chen, Chem. Mater., 2007, 19, 6345.
50. I. Lynch and K. A. Dawson, Nano Today, 2008, 3, 40.
51. M. Khoobi, M. Khalilvand-Sedagheh, A. Ramazani, Z. Asadgol, H. Forootanfar and M. A. Faramarzi, J. Chem. Technol. Biotechnol., 2015 DOI:10.1002/jctb.4579.
52. A. Safavi and S. J. Momeni, J. Hazard. Mater., 2012, 201, 125.
53. G. Bayramoglu and M. Y. Arica, Chem. Eng. J., 2008, 139, 20.
54. M. M. Yallapu, Int. J. Nanomed., 2012, 7, 1761.
55. U. Sakulkhu, M. Mahmoudi, L. Maurizi, J. Salaklang and H. Hofmann, Sci. Rep., 2014, 4, 5020.
56. J. Li, L. Zheng, H. Cai, W. Sun, M. Shen, G. Zhang and X. Shi, Biomaterials, 2013, 34, 8382.
57. S. Chaleawlertumpon and N. Pimpha, Mater. Chem. Phys., 2012, 135(2012), 1.
58. H. Cai, X. An, J. Cui, J. Li, S. K. Li, M. Shen, L. Zheng, G. Zhang and X. Shi, ACS Appl. Mater. Interfaces, 2013, 5, 1722.
59. M. Khoobi, T. Modiri Delshad, M. Vosooghi, M. Alipour, H. Hamadi, E. Alipour, M. Pirali Hamedani, S. E. Sadat ebrahimi, Z. Safaei, A. Foroumadi and A. Shafiee, J. Magn. Magn. Mater., 2015, 375, 217.
60. M. Khoobi, S. F. Motevalizadeh, Z. Asadgol, H. Forootanfar, A. Shafiee and M. A. Faramarzi, Mater. Chem. Phys., 2015, 149–150, 77.
61. X. Y. Li, K. Nan, S. Shi and H. Chen, Int. J. Biol. Macromol., 2012, 50, 43.
62. D. Baybas and U. Ulusoy, J. Solid State Chem., 2012, 194, 1.
63. S. Chaleawlertumpon and N. Pimpha, Mater. Chem. Phys., 2012, 135, 1.
64. C. H. Kuo, Y. C. Liu, C. M. J. Chang, J. H. Chen, C. Chang and C. J. S. Chwen-Jen, Carbohydr. Polym., 2012, 87, 2538.
65. H. Cai, X. An, J. Cui, J. Li, S. K. Li, M. Shen, L. Zheng, G. Zhang and X. Shi, ACS Appl. Mater. Interfaces, 2013, 5, 1722.
66. L. Fan, Y. Zhang, C. Luo, F. Lu, H. Qiu and M. Sun, Int. J. Biol. Macromol., 2012, 50, 444.
67. X. Y. Li, K. Nan, S. Shi and H. Chen, Int. J. Biol. Macromol., 2012, 50, 43.
68. X. Yang and J. C. Kim, Int. J. Biol. Macromol., 2011, 48, 661.
69. B. Mendoza-Novelo, J. L. Mata-Mata, A. Vega-Gonzalez, J. V. Cauich-Rodrıguez and A. Marcos-Fernandez, J. Mater. Chem. B, 2014, 2, 2874.
70. T. M. Kolev, E. A. Velcheva, B. A. Stamboliyska and M. Spiteller, Int. J. Quantum Chem., 2005, 102, 1069.
71. V. T. Bich, N. T. Thuy, N. T. Binh, N. T. M. Huong, P. N. D. Yen, T. T. Luong, D. T. Cat, A. Pucci and K. Wandelt, Physics and Engineering of New Materials, Springer publications, 2009, p. 271.
72. K. Watanabe, Y. Nishio, R. Makiura, A. Nakahira and C. Kojim, Int. J. Pharm., 2013, 446, 81.
73. A. J. Hajdu, F. B. Bombelli, M. P. Monopoli, E. Tombácz and K. A. Dawson, Langmuir, 2012, 28, 14983.
74. E. Fröhlich, Int. J. Nanomed., 2012, 7, 5577.
75. M. A. Islama, T. E. Park, B. Singh, S. Maharjan, J. Firdous, M. H. Cho, S. K. Kange, C. H. Yun, Y. J. Choi and C. S. Cho, J. Controlled Release, 2014, 193, 74.
76. A. Zintchenko, A. Philipp, A. Dehshahri and E. Wagner, Bioconjugate Chem., 2008, 19, 1448.
77. S. S. Diebold, M. Kursa, E. Wagner, M. Cotton and M. Zenke, J. Biol. Chem., 1999, 247, 19087.
78. H. Petersen, P. M. Fechner, A. L. Martin, K. Kunath, S. Stolnik, C. J. Roberts, D. Fischer, M. C. Davies and T. Kissel, Bioconjugate Chem., 2002, 13, 845.
79. L. Xu, Y. Chen, J. W. He, Y. Zhonghua and Z. Z. Yi Xue, Chin. Phys. Lett., 2012, 92, 587.
80. C. R. Zhou, Biomateriallogy, Medicine Technology Press of China, 2004.
81. T. Loftsson, M. Masson and M. E. Brewster, J. Pharm. Sci., 2004, 93, 1091.
82. R. V. Yadav, S. Prasad, R. Kannappan, J. Ravindran, M. M. Chaturvedi, L. Vaahtera, J. Parkkinen and B. B. Aggarwal, Biochem. Pharmacol., 2010, 80, 1021.
83. A. Koch, F. Reynolds, M. Kircher, H. Merkle, R. Weissleder and L. Josephson, Bioconjugate Chem., 2003, 14, 1115.
84. R. Campbell, D. Fukumura, E. Brown, L. Mazzola, Y. Izumi and R. Jain, Cancer Res., 2002, 62, 6831.
85. H. H. Tonnesen, M. Masson and T. Loftsson, Int. J. Pharm., 2002, 244, 127.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra13838h

---