Studies on Preparation of Photosensitizer Loaded Magnetic Silica Nanoparticles and Their Anti-Tumor Effects for Targeting Photodynamic Therapy

As a fast developing alternative of traditional therapeutics, photodynamic therapy (PDT) is an effective, noninvasive, nontoxic therapeutics for cancer, senile macular degeneration, and so on. But the efficacy of PDT was compromised by insufficient selectivity and low solubility. In this study, novel multifunctional silica-based magnetic nanoparticles (SMNPs) were strategically designed and prepared as targeting drug delivery system to achieve higher specificity and better solubility. 2,7,12,18-Tetramethyl-3,8-di-(1-propoxyethyl)-13,17-bis-(3-hydroxypropyl) porphyrin, shorted as PHPP, was used as photosensitizer, which was first synthesized by our lab with good PDT effects. Magnetite nanoparticles (Fe3O4) and PHPP were incorporated into silica nanoparticles by microemulsion and sol–gel methods. The prepared nanoparticles were characterized by transmission electron microscopy, X-ray diffraction, Fourier transform infrared spectroscopy and fluorescence spectroscopy. The nanoparticles were approximately spherical with 20–30 nm diameter. Intense fluorescence of PHPP was monitored in the cytoplasm of SW480 cells. The nanoparticles possessed good biocompatibility and could generate singlet oxygen to cause remarkable photodynamic anti-tumor effects. These suggested that PHPP-SMNPs had great potential as effective drug delivery system in targeting photodynamic therapy, diagnostic magnetic resonance imaging and magnetic hyperthermia therapy.


Introduction
Photodynamic therapy (PDT) is an effective, noninvasive and nontoxic therapeutics for cancer, senile macular degeneration, actinic keratosis, port-wine stains, rheumatoid arthritis, and so on [1,2]. After bio-distribution, photosensitizer (PS) administered systemically or topically is activated by light of appropriate wavelength and dosage. The activated PS transfers its excited-state energy to nearby oxygen molecular to generate reactive oxygen species, such as singlet oxygen ( 1 O 2 ) or peroxides inducing oxidative damage to target tissue and blood vessels that feed them [1][2][3][4]. Due to minimal invasion and nontoxicity, PDT provides patients, weak or failed in traditional therapy, opportunities to be treated painlessly and repeatedly.
However, the PDT efficacy is compromised by insufficient selectivity and low solubility. Although several methods including drug delivery systems were investigated [3][4][5][6][7][8][9], developing a PS delivery system for higher selectivity and less dark toxicity is still a challenge [5,6]. Magnetic drug delivery system is a promising drug delivery system, which can be steered to the target tissue simply by an external magnetic field [10,11]. Silica nanoparticles, easily prepared with desired size, shape and porosity, are water-soluble, stable and biocompatible. More importantly, silica nanoparticles are permeable to small molecular such as singlet oxygen [4,5], which is the key effector of PDT. Therefore, photosensitizer loaded silica nanoparticles are different from conventional delivery systems which need releasing of the loaded drug [9].
Previous investigations of fluorescent-magnetic nanoparticles mainly focused on the MRI imaging and fluorescence imaging for diagnosis; however, there are few studies on the multifunctional magnetic targeting drug delivery system for diagnosis and therapy [12,13]. In the earliest study of magnetic targeting, a magnetic fluid was developed to which epirubicin was chemically bound to enable those agents to be directed within an organism by high-energy magnetic fields. In vitro and in vivo study of the epirubicin-magnetic fluid indicated biosafety and complete tumor response [10,11], demonstrating the potential of magnetic targeting. Recently, the investigation of PS encapsulated magnetic silica nanoparticles (SMNPs) showed efficient cellular uptake [14] and obvious generation of singlet oxygen in vitro [15,16], which indicated the potential of SMNPs as targeting drug delivery system.

Preparation of Fe 3 O 4 Nanoparticles
In total, 2.51 g (9 mmol) FeCl 3 Á 6H 2 O and 1.25 g (4.5 mmol) FeSO 4 Á 7H 2 O were dissolved in 20 mL water. The solution was vigorously stirred, followed by adding 10 mL 1.5 mol L -1 NH 3 Á H 2 O. The color of the solution was changed into black and the black solid produced was precipitated to the bottom. The Fe 3 O 4 nanoparticles were obtained after the precipitants were washed for five times with 20 mL distilled water and 20 mL ethanol alternatively to remove unreacted chemicals.

Preparation of Silica-Based Fe 3 O 4 Nanocarriers
In total, 1 g Fe 3 O 4 nanoparticles and 10 g oleic acid were mixed with 10 mL ethanol. The suspension was refluxed for 30 min. Fe 3 O 4 /OA nanoparticles were obtained after the excess oleic acid was scoured off with ethanol by the magnetic decantation.
Micelles were prepared by dissolving 0.90 g AOT and 1600 lL 1-butanol in 40 mL doubly distilled water by vigorous magnetic stirring. A solution of 60 lL PHPP (15 mmol L -1 ) in 1-butanol and 0.003 g Fe 3 O 4 /OA nanoparticles were added to above micellar system. After was treated by magnetic separation and washed with ethanol until no PS could be detected in the supernatant by UV-Vis spectroscopy. All the above-mentioned experiments were conducted at room temperature. The products were dried at 60°C for 3 h in vacuum oven.

Characterization
The X-ray diffraction pattern of silica-based magnetic nanocarriers powders was obtained using D/max-2550PC (Geigerflex, Rigaku, Japan) with monochromated CuKa radiation operated at 40 kV and 100 mA. Transmission electron microscopy (TEM) was employed to determine the morphology and size of the aqueous dispersion of nanocarriers, using a HITACHIH-800 electron microscope, operating at an accelerating voltage of 200 kV. UV-Vis absorption spectra were recorded using a Jasco V-530 spectrophotometer, in a quartz cuvette with 1 cm path length. Fluorescence spectra were recorded on a HIT-ACHIH FL-4500 spectrofluorimeter.

Detection of Singlet Oxygen
The PHPP-SMNPs in phosphate buffer (pH = 7.4) were irradiated in the presence of imidazole (10 mmol L -1 ) and RNO (50 mmol L -1 ). The RNO bleaching by 1 O 2 was followed spectrophotometrically with observing the decrease in the 440 nm absorption peak of RNO as a function of irradiation time. The reaction mixture in a 1 cm spectrometric cuvette, placed at a distance of 12 cm, was continuously irradiated using 632.8 nm laser.

Biosafety Assessment
SW480 carcinoma cells (3 9 10 3 cells per well) were seeded in 96-well plates and incubated overnight at 37°C in a humidified 5% CO 2 atmosphere. After being rinsed with PBS (pH 7.4), the cells were incubated with 100 lL varied concentration of PHPP-SMNPs prepared above for 24 h at 37°C in the dark under the same conditions. Rinsed with PBS, the cells were incubated another 48 h. Cell viability was determined by the colorimetric 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Cells were rinsed with PBS and then incubated with culture medium containing 0.5 mg mL -1 MTT reagent for 3 h. The medium was then removed and the formazan crystals formed were dissolved in 100 lL DMSO. The absorbance at 492 nm for each well was recorded by a microplate reader.

Photodynamic Activity Assay
Two plates were set up as dark control and experimental group for the MTT assay and these plates were seeded, exposed identically to the plates prepared for the biosafety assessment. The cells of experimental group were then rinsed again with PBS and immersed in 100 lL of fresh culture medium before being illuminated using a 488 nm argon-ion laser with energy density of 4.35 J/cm 2 from the underside of the culture plate. After 10 min illumination, cells were incubated 48 h in a 5% CO 2 , 95% air humidified incubator at 37°C. Dark control group keep identical to experimental group except illumination. Photodynamic activity assay was also determined by MTT assay as mentioned above.

Statistical Analysis
Cell viability was calculated using the following formula: average A value of experimental group/average A value of control group 9 100%. Results were expressed as means ± SD. Comparisons between two groups were made by unpaired two-tailed Student's t-test using SPSS 15.0 software. P-value of less than 0.05 was taken to indicate statistical significance.

Preparation of Fe 3 O 4 Nanoparticles
Fe 3 O 4 nanoparticles, prepared by co-precipitation method [19,20], were 10 ± 2 nm in diameter by measuring 200 randomly selected particles in enlarged TEM images, which agreed well with the value calculated from Scherrer Equation. Figure 1 shows the spherical morphology (Fig. 1a) and the characteristic peaks (Fig. 1b), which was compatible with the values of standard pattern. In addition to the dispersed and well-separated features, the formed Fe 3 O 4 nanoparticles also exhibited some degree of aggregated morphology. The particles show an attraction to magnetic field, demonstrating the magnetic responsibility.
The Fe 3 O 4 nanoparticles were coated with oleic acid according to the procedure described by Reimers [21,22]. By adding oleic acid at melting point, the Fe 3 O 4 nanoparticles were hydrophobized as illustrated in Fig. 2. The precipitate could be readily redispersed in the solvents such as cyclohexane, chloroform, or 1-butanol.

Preparation of PHPP-SMNPs
Considering the size of the emulsion droplet is directly related to the final nanoparticle size, the formation of the emulsion is the key aspect. Emulsions can be classified in macroemulsions, microemulsions and miniemulsions (or nanoemulsions) The o/w microemulsion method is easy to scale up, it does not need high shear stress, and it is transparent and thermodynamically stable, with droplets mean sizes from 20 to 50 nm [23,24]. So it is widely used for entrapment of hydrophobic compounds. Here, the o/w microemulsion method was combined with sol-gel method which was the classical approach to synthesize SiO 2 nanoparticles [25,26]. The PHPP-SMNPs were prepared in the nonpolar core of AOT/1-butanol/water micelles, as shown schematically in a pictorial representation (Fig. 3). PHPP-absorbed Fe 3 O 4 /OA had priority to disperse in 1-butanol droplets redounding to form the nanocarriers.

Characterization of PHPP-SMNPs
TEM image (Fig. 4a) of PHPP-SMNPs showed that the PHPP-SMNPs were approximately spherical, sized in the range of 20-30 nm, but some agglomeration could be observed. Dynamic light scattering measurements were performed to study the behavior of a suspension containing PHPP-SMNPs. The hydrodynamic diameter of the nanocarriers was about 126 nm and the polydispersion was about 0.116, which verified the aggregates in the  suspension of particles. These aggregates were not stable and could be easily redispersed by simply shake or sonication. Figure 4b illustrates the XRD pattern of PHPP-SMNPs. The (111) peak is derived from the amorphous mesoporous silica spheres and the characteristic (311), and (440) peaks are typical of a cubic structure. The result showed that the crystallinity has not changed after encapsulation. Figure 5 shows the typical FT-IR spectra of (a) PHPP, (b) SMNPs, and (c) PHPP-SMNPs. The bands C-H at 740 cm -1 (Fig. 5a) and C-O/C-N at 1079 cm -1 (Fig. 5c) indicated the encapsulation of the drug in the SMNPs. According to (b) and (c), it was found that the Si-O vibration absorption peak of the 1044 cm -1 shifted to 1079 cm -1 , which might be contributed to the integration of Si-O vibration absorption and C-O/C-N vibration absorption peak. The above facts suggested that PHPP was successfully wrapped in the SMNPs.
The photoluminescence peaks of PHPP were at 625 and 690 nm, as shown in Fig. 6. Figure 7 represents the fluorescence emission spectra of mother liquor and aqueous dispersion of the PHPP-SMNPs at same concentration. The emission signal from the PHPP in the PHPP-SMNPs was almost 20% of the all solution. Fluorescence intensity decreased significantly from (a) to (b), illustrating that PHPP was not completely encapsulated in the SMNPs.  Figure 8 shows a concentration-dependent increase in photoluminescence signal, demonstrating the encapsulation of the PHPP again.

Encapsulation Efficiency Measurements
The amount of drug entrapped within PHPP-SMNPs was determined by dissolving PHPP-SMNPs into hydrochloric acid to destroy the Fe 3 O 4 cores of PHPP-SMNPs for the release of PHPP. After ethanol addition, the absorbance of PHPP was detected and was performed by ultraviolet spectrophotometer at 409 nm. Figure 9a shows the UV spectra of different substances in the 200-700 nm wavelength range, using ethanol as the solvent. The spectrum of PHPP (a) showed the special absorption peaks of PHPP at 399 nm. The increased absorption of PHPP and SMNPs mixture was caused by the absorption of SMNPs. The reaction of PHPP with HCl caused a red shift of the special absorption peak of PHPP from 399 to 409 nm (c). Spectrum (d) indicated absorptions of SMNPs and HCl at 240, 314, and 364 nm, demonstrating that the 240, 314, and 364 nm peak of the spectrum (e) attributed to SMNPs and HCl. Therefore, the 409 nm was the characteristic absorption peaks of PHPP with or without SMNPs.
The PHPP-SMNPs treated with HCl were detected by UV-Vis spectrophotometer and the spectrums are displayed in Fig. 10.
The standard curve had a good linear relation (r = 0.99905) within the range of 7.65 9 10 -7 to 1.02 9 10 -5 mol L -1 , described by the following typical equation: Y = 0.04447 ? 0.02571x (see Fig. 11). PHPP encapsulation efficiency was 20.8%, estimated by the typical equation.   The N,N-dimethyl-4-nitrosoaniline (RNO) was used as an indicator for photo-induced singlet oxygen with imidazole as a chemical trap for singlet oxygen [27][28][29]. The principle of this method is shown in the following formula: The bleaching of RNO by 1 O 2 was followed spectrophotometrically with observing the decrease in the 440 nm absorption peak of RNO as a function of irradiation time.
A decrease in the 440 nm absorption peak of RNO was caused by the Fe 3 O 4 nanoparticles without irradiation. So a 24 h aging of the PHPP-SMNPs and RNO system was necessary prior to detecting the 1 O 2 productivity, until the absorption of RNO at 440 nm did not decline. The system was irradiated. The continued decrease at 440 nm absorption peak was caused by the significant generation of 1 O 2 released from the PHPP-SMNPs (Fig. 12), indicating the potential for efficient PDT.
In Vitro Studies with Tumor Cells: Cellular Uptake, Biosafety Assessment and Photodynamic Activity Assay

Intracellular Uptake
As an essential tool in material science and biology, fluorescence microscopy demonstrates the ability to monitor the precise location of intracellular fluorescence materials excited by light of specific wavelengths, as well as their associated diffusion coefficients, transport characteristics, and interactions with other biomolecules. To test the intracellular uptake of PHPP-SMNPs, fluorescence imaging was performed on human SW480 colon carcinoma cells after incubation with 50 lmol L -1 PHPP-SMNPs for 4 h in cell culture incubator. As shown in Fig. 13, PHPP-SMNPs were taken up by SW480 cells and showed significant intracellular fluorescence (extra nuclear) compared to unincubated cell. As the PDT effects depend on the uptake of PS by tumor cells, the intense fluorescence of intracellular PHPP-SMNPs, related to the PS concentration, predicted an available obvious PDT effects. Subcellular distribution of PHPP-SMNPs in the cytoplasm primarily demonstrated slight effects on DNA.

Biosafety Assessment
Biosafety assessment was essential to evaluate the potential application of silica nanoparticle in clinics. Here, MTT assay was performed to detect the dark toxicity of PHPP-SMNPs. No obvious dark toxicity of PHPP-SMNPs on SW480 carcinoma cells was detected within 0.03-80 lmol L -1 concentration range in comparison with control (Fig. 14a). In addition, negligible cell death and physiological state changes of SW480 cells treated with highest dosage of PHPP-SMNPs were observed in Fig. 14b. It could be predicted that the PHPP-SMNPs had minimal, if any, impact on cellular functions, which indicated the low dark toxicity and good biocompatibility.

In Vitro Photodynamic Efficacy
Likewise, the MTT assay was performed to examine the phototoxicity of PHPP-SMNPs to SW480 colon carcinoma cell lines, which indicated the PDT efficacy in vitro [30]. Cell viability was normalized to control cells (no drug and unirradiated) in Fig. 14. The combination of 24 h exposure of tumor cells to PHPP-SMNPs and 4.35 J/cm 2 irradiation induced a drug concentration-dependent cytotoxicity to SW480 tumor cells, which was significantly different from unirradiated control in statistics, as shown in Fig. 15. With a 10 min light exposure, 80 lmol L -1 PHPP-SMNPs in the safety range measured as above caused approximately 40% cell viability lost, demonstrating obvious photodynamic activity. The group treated with the drug without light exposure showed that the drug alone had no effects on tumor cells which coincided with the result of biosafety assay. In this case, the cell viability at the maximum of concentration (80 lmol L -1 ), slightly lower than the control, was caused by the natural light during the execution of experiments.

Conclusions
Novel multifunctional silica-based magnetic nanoparticles containing photosensitizer PHPP were prepared. The PHPP-SMNPs were approximately spherical and 20-30 nm in diameter, achieving 20.8% encapsulation efficiency of PHPP. They showed no obvious toxicity without irradiation, but significant generation of singlet oxygen and remarkable photodynamic efficacy after irradiation. The PHPP-SMNPs were primarily distributed in the cytoplasm.
It can be concluded that the silica-based magnetic nanoparticles are of great value as effective drug delivery system in targeting photodynamic therapy. The potential of the magnetic core for magnetic resonance imaging and magnetic hyperthermia therapy could also be expected.