D, L-Sulforaphane Loaded Fe3O4@ Gold Core Shell Nanoparticles: A Potential Sulforaphane Delivery System

A novel design of gold-coated iron oxide nanoparticles was fabricated as a potential delivery system to improve the efficiency and stability of d, l-sulforaphane as an anticancer drug. To this purpose, the surface of gold-coated iron oxide nanoparticles was modified for sulforaphane delivery via furnishing its surface with thiolated polyethylene glycol-folic acid and thiolated polyethylene glycol-FITC. The synthesized nanoparticles were characterized by different techniques such as FTIR, energy dispersive X-ray spectroscopy, UV-visible spectroscopy, scanning and transmission electron microscopy. The average diameters of the synthesized nanoparticles before and after sulforaphane loading were obtained ∼ 33 nm and ∼ 38 nm, respectively, when ∼ 2.8 mmol/g of sulforaphane was loaded. The result of cell viability assay which was confirmed by apoptosis assay on the human breast cancer cells (MCF-7 line) as a model of in vitro-cancerous cells, proved that the bare nanoparticles showed little inherent cytotoxicity, whereas the sulforaphane-loaded nanoparticles were cytotoxic. The expression rate of the anti-apoptotic genes (bcl-2 and bcl-xL), and the pro-apoptotic genes (bax and bak) were quantified, and it was found that the expression rate of bcl-2 and bcl-xL genes significantly were decreased when MCF-7 cells were incubated by sulforaphane-loaded nanoparticles. The sulforaphane-loaded into the designed gold-coated iron oxide nanoparticles, acceptably induced apoptosis in MCF-7 cells.


Introduction
Many studies confirmed that sulforaphane (SF) which was found in broccoli, cauliflower, kale and other cruciferous vegetables, and chemically was named 1-isothiocyanato-4-(methylsulfinyl)-butane, acts as an efficient incidence reducer in various types of tumors [1][2][3]. It has been revealed that SF has a promising and powerful anti-carcinogen effect in various cancers such as penicillin G, phosphate buffer (20 mM, pH = 7.8) and corresponding salts which were used throughout this research were obtained from Sigma-Aldrich or Merck. Fetal bovine serum (FBS) and Dulbecco's Modified Eagle's Medium (DMEM) as a culture medium and other supplements were obtained from Gibco (Germany). MCF-7 cells were grown in DMEM medium supplemented with 10% (v/v) heat-inactivated FBS, 2% l-glutamine, 2.7% sodium bicarbonate, 1% Hepes buffer, and 1% penicillin-streptomycin solution (GPS, Sigma) at 37°C in humidified atmosphere with 5% CO 2 . Cells were trypsinized in the solution of 0.05% trypsin and seeded into 96-well micro-plates at the density of 1 × 10 5 cells/well. MCF-7 (human breast carcinoma) cell line was obtained from national cell bank of Iran (Pasteur institute, Iran) and cultured in DMEM medium containing 2 mM L-glutamine and 10% FBS, penicillin (50 IU/ml) and streptomycin (50 μg/ml). Cells were separated using 0.1% trypsin and 10 μM EDTA in phosphate-buffered solution.
The powder XRD spectrum was recorded at room temperature (RT) with a Philips X'pert 1710 diffractometer using Cu Kα (α = 1.54056 A). The morphologies of the NPs were observed using SEM (Hitachi S-4800 II, Japan) equipped with energy dispersive X-ray spectroscopy (EDX). The TEM analysis was performed on a Hitachi H-7650 (Japan) operating at an acceleration voltage of 80 kv. The IR spectra were taken using Nicolet FT-IR Magna 550 spectrographs applying spectroscopic grade KBr. The size of NPs was assessed by DLS (Nano-ZS 90, Malvern Instrument, United Kingdom). The temperature was kept at 25°C during the measuring process and measurements were recorded as the average of three test runs. The zeta potential of the NPs was measured in folded capillary cells using Nano sizer (Zeta sizer Nano ZS90, Malvern Instruments Ltd., Malvern, UK).
Synthesis of α-azide-ω-hydroxyl PEG (II). A mixture of I (3.0 mmol) and NaN 3 (4 mmol) in 150 ml of dry DMF was stirred overnight at 90°C under argon atmosphere. The mixture was cooled down, filtered, and DMF was removed under vacuum. The product was dissolved in CH 2 Cl 2 and washed twice with brine and water. The organic phase was dried over sodium sulfate and reduced to a small volume, and finally precipitated by dropping into Et 2 O. Then II was filtered. 1 H-NMR (DMSO): 4.56 (1H, OH), 3.6 (2H, OCH 2 ), 3.5 (bs, PEG backbone), 3.4 (2H, CH 2 -N 3 ) [36].
Synthesis of α-azide-ω-thioacetate PEG (III). To a mixture of II (3.0 mmol) in 100 ml CH 2 Cl 2 were added NEt 3 (10.0 mmol) and TsCl (10.0 mmol) and stirred overnight at RT. The solution was then filtered, and the filtrate was washed twice with saturated NH 4 Cl solution and water. Next, the organic phase was dried over sodium sulfate and filtered. The filtrate was concentrated by solvent evaporation, and then added drop-wise into dry Et 2 O and the product (αazide-ω-tosyl PEG) was collected by filtration. Afterwards, freshly prepared sodium thioacetate (15.5 mmol) was added to a solution of α-azide-ω-tosyl PEG (3 mmol) in 100 ml dry DMF under inert atmosphere, and the mixture was stirred overnight at RT. After solvent evaporation, the residue was dissolved in CH 2 Cl 2 and treated with active charcoal for 2 h. The mixture was filtered, and the filtrate was concentrated to small volume by rotary evaporation, and was added into dry Et 2 O. Then III was filtered. 1

Synthesis of HS-PEG-NH-FA (V)
Folic acid (0.1 mmol) was added to a mixture of dicyclohexyl carbodiimide (DCC) (0.15 mmol) and N-hydroxy succinimide (NHS) (0.12 mmol) in 3 ml of dry DMSO, the mixture was left to stir at RT overnight in the dark. After the removal of the byproduct, dicyclohexyl urea, the filtrate was mixed with IV (0.11 mmol) and stirred at RT for 24 h. Then V was precipitated by adding dry Et 2 O, dried in vacuum, and was stored at -20°C. 1

Synthesis of HS-PEG-FITC (VI)
FITC (10.0 μmol) and polymer IV (11.0 μmol) was dissolved in 2 ml EtOH and 2 ml CHCl 3 and stirred in the dark, under argon atmosphere. After 4 h, the product was precipitated by adding Et 2 O and dried. Then VI was filtered and stored at -20°C. 1

Synthesis of Magnetic Nanocarrier
Synthesis of Fe 3 O 4 -Gold Core Shell Magnetic NPs. FeCl 2 •4H 2 O (2 ml, 2M) and FeCl 3 •6H 2 O (4 ml, 2M) were stirred at RT for 1 h, under argon atmosphere. Then, a concentrated solution of NaOH was drop-wise added under the inert atmosphere and the pH of the solution was carefully adjusted between 11 to 12. The solution was stirred at 20°C for 1 h and then heated to 90°C for another 1 h. Afterwards the system was kept at the same temperature for 1 h. Subsequently, Fe 3 O 4 NPs were separated by using an external magnetic field and washed with of deionized water (DW) to achieve a neutral Fe 3  @Au] NPs (5 mg/ml) was suspended in DMEM media at RT and placed inside the dialysis bag with a molecular cut-off of 14 kDa and then placed into DMEM cell culture media. The particles were stirred with sampling at 1, 2, 4, 6, 8, 12, 16 and 24 h. The samples were analyzed for FITC content using fluorescence spectrometer (perkin Elmer) at 488 nm. Increasing in FITC content outside the dialysis chamber indicates particle degradation. Therefore, at each time point 100 μl of supernatant was picked up as a sample and then, 100 μl of fresh medium was added to remain the volume at the previous scale. The absorbance of FITC was recorded, and the total release of FITC was calculated. To study physical stability of the synthesized NPs against aggregation, the particle sizes of FITC/FA@[Fe 3 O 4 @Au] NPs were assayed by DLS in PBS buffer. PBS represents the typical pH of physiological medium and is a very common biological buffer. The average hydrodynamic diameters of FITC/FA@[Fe 3 O 4 @Au] NPs was measured as a function of time up to 5 days.

SF Loading and Releasing
FITC/FA@[Fe 3 O 4 @Au] NPs with the final concentration of 0.150 mg/ml was first sonicated with SF with an initial concentration of 0.300 mg/ml for 30 min and then stirred overnight at RT in the dark. The samples were separated applying an external magnet device, and washed with dry EtOH. The SF concentration in the liquid layer was measured using a standard SF concentration curve generated with an UV-visible spectrophotometer at 235 nm from a series of SF solutions with different concentrations. The drug loading efficiency (DLE) of FITC/FA@ [Fe 3 O 4 @Au] NPs can be calculated by measuring differences between initial and residual SF concentrations as: @Au] NPs with the final concentration of 3.5 μmol/l. The cells were washed twice with PBS, before the digital-camera-microscopy. Observation was done under an inverted microscope (Olympus, IX81). The U-MWIB3 (BP-460-495 nm) filter was used to perform fluorescence microscopy. The digital images were captured with a DP72 digital camera. In order to indicate the intensity of FITC in the fluorescence images, images were analyzed using Matlab (The Math Works, Natick, MA). A colorbar was utilized to represent different intensity values.

Cell Apoptosis by Flow Cytometry
Apoptosis of MCF-7 cells was studied by flow cytometry using Annexin-V/PI staining because phosphatidyl serine is unprotected and detectable by Annexin-V at the external surface of the cells at initial stages of apoptosis. Briefly, cells were seeded into 6 well plates (10 5 cells/well) and were treated for 48 h with free SF, FITC/FA@[Fe 3 O 4 @Au] NPs, and SF-loaded FITC/FA@ [Fe 3 O 4 @Au] NPs at IC 50 concentration (24 μmol/l). After incubation, the cells were washed twice with cold PBS and then re-suspended in 1× binding buffer (10 mM Hepes, 140 mM NaCl, and 2.5 mM CaCl 2 , pH = 7.2). 100 μl of the solution transferred to 5 ml culture tube, and 5 μl of Annexin V-FITC, and 5 μl PI were added and incubated in dark situation for 15 min at RT. Then, 400 μl of 1× binding buffer was added to each tube and analyzed by Flow Cytometry (Becton Dickinson FACS Canto II) within 1 h. The thermal cycling conditions were as follows: initial denaturation at 95°C for 10 min, followed by 45 cycles of 94°C for 15 s, and 60°C for 60 s. This program was followed by a melting curve program (60-98°C with a heating rate of 0.3°C/s and continuous fluorescence measurement).

Statistical Analysis
Matlab was used for all statistical analysis. Data are expressed as mean ± SEM. Statistical analysis to detect differences between groups was done by using one-way analysis of variance (ANOVA) with p-value <0.05 to indicate statistically significant differences. Post-hoc pairwise comparisons of Tukey(-Kramer) and Bonferroni tests were preformed and the smaller of the two intervals was taken. TEM and SEM were applied to recognize the structural order and morphology of FITC/ FA@[Fe 3 O 4 @Au] NPs (see Fig 1B and 1C). The analysis of SEM and TEM images showed that the average size of the synthesized NPs is less than 40 nm. The particle size distribution function confirmed that the size of NPs was 37±2.1 nm, which indicates a small heterogeneity in the size of the particles.

Preparation and Characterization of Nanocarrier
Energy dispersive X-ray (EDX) spectroscopy was done to investigate the presence of Au and Fe in the NPs. EDX spectrum of the NPs is presented in Fig 2, which confirms the existence of Au and Fe in the [Fe 3 O 4 @Au] NPs.
To confirm the existence of iron oxide-gold core shell, the powder X-Ray diffraction (XRD) experiment was used as an applicable technique. There were three bands in the XRD pattern which attributable to the corresponding reflections of the solid gold at 2θ = 44.67°, 51.72°, and 76.71°(JCPDS#89-3697). But, the characteristic bands of magnetic core no longer appeared in the XRD spectrum. The absence is due to the fact that all iron oxide NPs are coated with 2 nm thickness of gold [25,38].
In  Fig 2). In the spectrum of SF-loaded FITC/FA@[Fe 3 O 4 @Au] NPs, there are two strong absorbance picks at 2182 and 2109 cm −1 . These absorbance picks are belonged to the stretching vibration of -N=C=S. Besides that, two other absorption peaks at 1452 and 1350 cm −1 are belonged to the deformation vibration of CH 3 .
The thermal gravimetric analysis (TGA) of FITC/FA@[Fe 3 O 4 @Au] NPs showed the first peak at 95°C and the second peak at 395°C which correspond to desorption of water and loss of the organic spacer group, respectively. Therefore, the loading of FITC/FA was about 20%. Following that, the TGA analysis of SF-loaded FITC/FA@[Fe 3 O 4 @Au] NPs showed the weight loss about 70%, which is attributed to removing SF and FITC/FA groups. Hence, in average 2.8 mmol/g of SF was loaded onto the surface of NPs.
Since the existence of the magnetic NP in the designed NP would be a good candidate for in vivo therapy in animal model (using MRI), then it is important to synthesize the magneticcore-shell with sufficient superparamagnetic property. To investigate it, the magnetic (M) hysteresis loop of the same mass of the [Fe 3 O 4 @Au] and FITC/FA@[Fe 3 O 4 @Au] NPs were investigated while magnetic field (H) at RT was applied. As shown in Fig 3,  @Au] NPs were 7.24 and 2.07 emu/g at RT, respectively. The lower magnetic saturation of later NPs could be due to the influence of the PEG group. Furthermore, the hysteresis loop for the samples were completely reversible. The reversibility in hysteresis loop confirms that no aggregation imposes to the NPs in the external magnetic field and the NPs exhibit superparamagnetic property.  nm, respectively. The average size of both are less than 50 nm. Based on our measurement, NPs with and without drug are almost with small standard deviation in size (narrow particle size distribution function).
Another physical quantity which affects on cellular up take is the surface charge of synthesized NPs. The NPs with positive charge (size smaller than 100 nm) cross the cell membrane and readily taken up by cells [39]. The zeta potential of SF-loaded FITC/FA@[Fe 3 O 4 @Au] NP was measured and 9.47±4.6 mv was obtained which is suitable for cell membrane penetration.

Chemical and Physical Stability of FITC/FA@[Fe 3 O 4 @Au] NPs
To investigate the chemical stability of FITC/FA@[Fe 3 O 4 @Au] NPs, release of FITC from NPs was studied by UV-visible spectroscopy at 488 nm. The release of FITC could occurred due to the ligand exchange in the presence of a reductive media. The release profile was measured in two mediums including DMEM ( Fig 4A)   The release of FITC moiety from the NPs in 5% GSH at 24 h showed no significant change in compare to DMEM. To the best of our knowledge, the disulfide bonds are cleaved by reducing agents such as GSH. Our synthesized NPs contains FITC moiety without any linkage S-S bond. Therefore, it is clear why there was not observed a significant change in the release profile of FITC from FITC/FA@[Fe 3 O 4 @Au] NPs.
To elaborate on the physical stability of FITC/FA@[Fe 3 O 4 @Au] NPs, the change in size of the NPs was studied. Change in size may occur during in vitro experiment. Therefore, the physical stability of the FITC/FA@[Fe 3 O 4 @Au] NPs at physiological pH was measured during 120 h. As the graph shows (see Fig 4B) Besides the SF loading, the release profile of SF from the surface of SF-loaded FITC/FA@ [Fe 3 O 4 @Au] NPs was evaluated in two mediums including PBS and citrate buffer ( Fig 5B). As it is shown in Fig 5B,

Cell Viability
To study the cytotoxicity of SF-loaded FITC/FA@[Fe 3 O 4 @Au] NPs, MCF-7 cells were incubated with different concentrations of SF-loaded FITC/FA@[Fe 3 O 4 @Au] NPs and the percentage of live cells after 24, 48, and 72 h of treatments were measured (Fig 6A). There is a   viability at concentrations less than 0.37 μmol/l. Therefore, at concentrations lower than 1.5 μmol/l, the effect of SF-loaded FITC/FA@[Fe 3 O 4 @Au] NPs is not considerable. Besides that, at the concentrations more than 6 μmol/l of SF-loaded FITC/FA@[Fe 3 O 4 @Au] NPs, the cell viability did not change by time. Hence, it seems that the most effective concentration of SFloaded FITC/FA@[Fe 3 O 4 @Au] NPs on cell viability is between 1.5 to 6 μmol/l.
To verify the usefulness of the designed NPs as a potential drug carrier, at the concentration of SF-loaded FITC/FA@[Fe 3 O 4 @Au] NPs, which its efficiency on the cell viability was obtained before, the cytotoxicity of free SF, FITC/FA@[Fe 3 O 4 @Au] NPs, and SF-loaded FITC@ [Fe 3 O 4 @Au] NPs were studied with the same procedure. As Fig 6B shows, at concentrations more than 12 μmol/l, the SF can induce more cell death but inducing cell death with a high concentration of a drug (which is an unstable drug in our case) is not the aim of

Cellular Uptake
In order to investigate the cellular uptake, images of control cells (Fig 7A-7C), incubated cells with free SF (Fig 7D-7F), and incubated cells with FITC/FA@[Fe 3 O 4 @Au] NPs (Fig 7G-7I) were prepared. The emerged images of DAPI with bright field images (Fig 7B and 7E), and FITC-channel ( Fig 7I) (Fig 8A and 8D), fluorescence image of FITC detection (Fig 8B and 8E), and emerged images of FITC with bright field images (Fig 8C and 8F The same method of microscopy was used for 72 h incubation of SF-loaded FITC/FA@ [Fe 3 O 4 @Au] NPs with MCF-7 cells. As is shown in Fig 9A-9L, there is a one-to-one correspondence between live cells (DAPI staining) and nanoparticle/drug uptake by MCF-7 cells. Hence, the FITC/FA@[Fe 3 O 4 @Au] NPs were up taken by all (all in the field of view of the image) MCF-7 cells at 24 h of incubation (Fig 9A-9F). Based on the same colorbar corresponding the fluorescence intensity of FITC-emission, after 72 h, cells uptake more SF-loaded FITC/FA@ [Fe 3 O 4 @Au] NPs. From the images it is possible to evaluate whether the drug release occurs intracellularly following NPs uptake or the drug release occurs extracellularly prior to NPs uptake/delivery. The dynamics of the drug release at two different pH were studied to estimate the percentage of releasing at different environmental conditions. Since the pH of the release medium is merely close to the pH of PBS at 24

Cell Apoptosis
A double staining flow cytometric assay using Annexin V/FITC (A), which reacts with viable cells, and propidium iodide (PI) was performed to quantify the apoptotic MCF-7 cells accurately. Fig 10A shows

Discussion and Conclusions
We have designed, synthesized and characterized a novel and robust nano-magnetic gold DDS to improve the efficiency of SF delivery. The synthesized nano-magnetic vehicle was characterized by a number of techniques such as EDX, XRD, VSM, FT-IR, TGA, DLS, TEM, and SEM.
Owing to the hydrophilicity behavior of PEG function as surface modifier, the aforementioned surface manipulations provide the convenient surface to encompass SF via the electrostatic interaction between the S = O bond from SF and the oxygen groups from hydrophilic  PEG linker. Hence, in order to 1) overcome the SF-instability, 2) improve SF-delivery, and 3) control SF-release, the surface of NPs was manipulated with functionalized PEG.
In vitro cytotoxicity study (MTT assay) of MCF-cells showed that SF-loaded FITC/FA@ [Fe 3 O 4 @Au] NPs are more cytotoxic than free SF. The mechanism of induced appotosis in MCF-7 cells was revealed by double staining flow cytometric assay. Based on the statistical analysis of the obtained data, the rates of total apoptosis significantly were increased after loading onto the synthesized nano-magnetic vehicles in comparison with free SF. Following detection of apoptosis mechanism in cytotoxicity of SF-loaded FITC/FA@[Fe 3 O 4 @Au] NPs for MCF-7 cells, the expression rate of anti-apoptotic genes (bcl-2 and bcl-x L ) and pro-apoptotic genes (bax and bak) were quantified by Real-Time PCR technique. The expression rate of bcl-2 and bcl-x L genes significantly were decreased in comparison with untreated cells.
Notably, in vitro breast cancer study showed that FITC/FA@[Fe 3 O 4 @Au] NP would be a proper candidate nanocarrier for SF-delivery and provides a suitable and appropriate system for pH-dependent delivery of SF. These results would open new horizons to develop promising nano-DDSs to design further nanoparticle systems to improve the therapeutic effect of SF in the future.