Magnetic transfection with superparamagnetic chitosan-loaded IGFBP5 nanoparticles and their in vitro biosafety

We prepared the superparamagnetic chitosan nanoparticles (SPCIONPs) to study the application of them as gene vectors using a magnetic transfection system for the targeted treatment of lung metastasis of osteosarcoma. The SPCIONPs were characterized by transmission electron microscopy, Fourier transform infrared spectrometry, superconducting quantum interference device and atomic force microscopy. Their biosafety was determined by cell counting kit-8 (CCK8) and live–dead staining assays. The transfection in vitro was detected by laser confocal microscopy. SPCIONPs, which can bind closely to plasmids and protect them from DNA enzyme degradation, were prepared with an average particle size of approximately 22 nm and zeta potential of 11.3 mV. The results of the CCK8 and live–dead staining assays showed that superparamagnetic chitosan nanoparticles loaded with insulin-like growth factor-binding protein 5 (SPCIONPs/pIGFBP5) induced no significant cytotoxicity compared to the control group. The result of transfection in vitro suggested that pIGFBP5 emitted a greater amount of red fluorescence in the SPCIONPs/pIGFBP5 group than that in the chitosan-loaded IGFBP5 (CS/pIGFBP5) group. In conclusion, the prepared SPCIONPs had good biosafety and could be effectively used to transfer pIGFBP5 into 143B cells, and they thus have good application prospects for the treatment of lung metastasis of osteosarcoma.

YT, 0000-0003-1470-3877; DJ, 0000-0001-9935-7066 We prepared the superparamagnetic chitosan nanoparticles (SPCIONPs) to study the application of them as gene vectors using a magnetic transfection system for the targeted treatment of lung metastasis of osteosarcoma. The SPCIONPs were characterized by transmission electron microscopy, Fourier transform infrared spectrometry, superconducting quantum interference device and atomic force microscopy. Their biosafety was determined by cell counting kit-8 (CCK8) and live-dead staining assays. The transfection in vitro was detected by laser confocal microscopy. SPCIONPs, which can bind closely to plasmids and protect them from DNA enzyme degradation, were prepared with an average particle size of approximately 22 nm and zeta potential of 11.3 mV. The results of the CCK8 and live-dead staining assays showed that superparamagnetic chitosan nanoparticles loaded with insulin-like growth factorbinding protein 5 (SPCIONPs/pIGFBP 5 ) induced no significant cytotoxicity compared to the control group. The result of transfection in vitro suggested that pIGFBP 5 emitted a greater amount of red fluorescence in the SPCIONPs/pIGFBP 5 group than that in the chitosan-loaded IGFBP 5 (CS/pIGFBP 5 ) group. In conclusion, the prepared SPCIONPs had good biosafety and could be effectively used to transfer pIGFBP 5 into 143B cells, and they thus have good application prospects for the treatment of lung metastasis of osteosarcoma.

Introduction
Osteosarcoma is the most common primary bone malignancy in children and young people, with lung metastases occurring at the time of diagnosis in approximately 15-20% of patients [1]. At present, a multimodal treatment approach has been adopted for osteosarcoma treatment, with a methotrexate, doxorubicin and cisplatin (MAP) regimen as the first choice. The treatment is mainly based on chemotherapy combined with limb salvage or amputation, but the clinical therapeutic effect is still unsatisfactory. The 5-year survival rate for patients with osteosarcoma is approximately 70%, while the survival rate for patients with metastatic or recurrent disease is extremely low, with an overall survival rate of less than 20% [2]. Therefore, gene therapy has become a trend in osteosarcoma treatment [3].
Insulin-like growth factor-binding protein 5 (IGFBP 5 ) is a member of the insulin-like growth factor (IGF) system, which plays an important role in cell growth, differentiation and apoptosis [4]. IGFBP 5 primarily binds to insulin growth factors to inhibit the IGF1 and IGF2 signalling pathways, thereby inhibiting the proliferation, migration and invasion of osteosarcoma, and is a key inhibitor of in situ osteosarcoma growth and lung metastasis [5]. Our group previously used adenovirus to induce IGFBP 5 overexpression in osteosarcoma cells and confirmed that IGFBP 5 can induce apoptosis and inhibit invasion, migration and proliferation of osteosarcoma cells [6].
Non-viral gene vectors, which are currently a focus of research, will not cause immune responses, potential immunogenicity and lethal inflammation, in contrast to adenovirus vectors [7,8]. Non-viral vectors commonly include liposomes, polyethyleneimine (PEI) and chitosan (CS). Cationic liposomes, which have higher transfection rates but are more toxic and expensive, are the most commonly used liposomes. Cationic liposomes also have no sustained-release effect and thus require repeated administration during experiments in vivo, and they are greatly affected by serum [9]. PEI is currently the most commonly used and most effective non-viral vector due to its strong proton-buffering capacity. PEI is the gold standard for polymeric gene delivery vectors. The proton sponge effect can cause endosomes to swell and release the plasmid into the cytoplasm, which allows the plasmid to enter the nucleus and complete gene expression [10,11]. Therefore, PEI was used as a positive control group in this study. However, due to its severe toxicity, including mitochondrial damage and apoptosis, the in vivo applications of PEI are limited [12,13]. CS is currently used in tissue engineering, drug and gene delivery, wound healing, and antibacterial and anti-tumour applications because of its physiological and biological activity; biocompatibility with a variety of organs, tissues and cells; low toxicity, with generally recognized as safe (GRAS) approval by the Food and Drug Administration (FDA) [14][15][16] and compatibility with chemical or enzymatic modifications. In addition, CS can be dissolved in acetic acid to condense the plasmid and prevent degradation by DNase, and it is a good substitute for viral vectors, but unmodified CS has a low transfection rate as a gene vector [17].
Superparamagnetic iron oxide nanoparticles (SPIONPs) can act as a reagent with both active and passive targeting effects and can concentrate the carried drugs in the target tissue under the effect of an external magnetic field. Because of their advantages, such as biocompatibility, stability, environmental safety and low price, SPIONPs are widely used in biomedicine, including for targeted drug delivery [18][19][20][21], bioimaging, thermotherapy, photoablation therapy, biosensing and thermal olfaction [22][23][24][25]. At present, there are many studies to modify the surface of SPIONPs so that its surface is distributed in different groups, such as polymers, biomolecules, silica and metals which can provide the ensemble functional reactive group, e.g. aldehyde groups, hydroxyl groups, carboxyl groups and amino groups. Their groups can be linked to antibodies, proteins, DNA, enzymes and other bioactive substances for further application [26]. Therefore, in this study, we prepared SPIONPs coated with CS to generate superparamagnetic chitosan-coated iron oxide nanoparticles (SPCIONPs) and conducted experimental studies to investigate their in vitro biosafety and use for cell transfection in the presence of magnetic transfection system.

Preparation of SPCIONPs
Ferric chloride hexahydrate (2 g) and ferrous sulfate heptahydrate (1.4 g) were dissolved in 5 ml of water, followed by ultrasonication until the chemicals were completely dissolved. The solution was then transferred to a three-necked flask and stirred under nitrogen. Subsequently, 20 ml of 0.1% citric acid solution was added to the solution, which was then stirred for 10 min. The three-necked flask was then placed in a water bath at 85°C. A total of 20 ml ammonium hydroxide was added with a dropping funnel; the addition was completed within 10 min, and stirring was continued for another 30 min [27,28]. The resulting suspension was centrifuged at 8500 r.p.m. for 10 min, the sediment was discarded, and the supernatant was removed to obtain the ferrofluid of citric acid-coated magnetite particles (SPIONPs). Finally, 5 ml of the ferrofluid of citric acid-coated magnetite particles was added to MES buffer solution with a pH of 5, and EDC and NHS were subsequently added under stirring to activate the carboxyl group on the surface of the ferrofluid of citric acid-coated magnetite particles, followed by the addition of 5 ml of 20 mg ml −1 50 kDa CS solution in acetic acid and stirring at 300 r.p.m. at room temperature for 6 h [29]. The liquid was collected after 48 h of dialysis with a dialysis bag (retained molecular weight over 50 000). A portion of the liquid was lyophilized and stored at 4°C for later use.

Characterization
The size and morphology of the prepared SPCIONPs were observed with a transmission electron microscope (TEM, HITACHI, USA). The size distribution was determined by measuring diameters of 100 NPs randomly selected on the TEM micrographs. A Zetasizer Nano ZS (Malvern, UK) was used to determine the zeta potential of the CS, SPIONPs, SPCIONPs and SPCIONPs/pIGFBP 5 . The magnetic induction of SPIONPs and SPCIONPs was measured at 300 K using a superconducting quantum interference device (SQUID, a magnetic property measurement system (MPMS), Quantum Design, USA). The structural characterization of CS, SPIONPs and SPCIONPs was performed with a Fourier transform infrared (FT-IR) spectrometer using the KBr compression method. The molecular structure of prepared SPCIONPs were loaded onto the XY scanner of the AFM (IPC-208B, Chongqing University, China) [30][31][32]; the area to be scanned was localized on the monitor, and then images generated at room temperature under ambient conditions using the non-contact mode. A 100 mm scanner and a STM probe were used in the study, with a customized tungsten filament as the microcantilever to detect an area of 12.01 × 12.01 nm.

Protection assay
After incubation of SPCIONPs with the plasmids at different N/P ratios, 1 µl of DNase I (corresponding to 1 µg of plasmid) was added to 10× reaction buffer [100 mM Tris-HCl (pH = 7.5 at 25 mM), 25 mM MgCl 2 , 1 mM CaCl 2 )]. After 15 min of incubation in a 37°C water bath, 1 µl of ethylenediaminetetraacetic acid (EDTA) (25 mM, pH = 8) was added, followed by incubation at 65°C for 15 min to inactivate DNase I. The naked plasmid was used as the control. A 0.8% agarose gel was prepared with 1% TAE, and the samples were electrophoresed at 85 mV for 30 min.
royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 8: 201331 3 2.6. Prussian blue staining Prussian blue can stain iron atoms in cells. First, 143B cells were seeded in 12-well plates at 1 × 10 5 cells/well. After the cells became adherent, SPCIONPs/pIGFBP 5 were added, and a magnetic field was applied for 30 min. After 4 h, the cells were washed three times with phosphate-buffered saline (PBS) and fixed in paraformaldehyde for 20 min. Perls solution was prepared at a 1 : 1 ratio, and 200 µl of the solution was added to each well, followed by incubation at 37°C for 40 min. Cells were washed with ddH 2 O, and 200 µl of neutral red solution was added and incubated with the cells at room temperature for 2 min. After the staining solution was rinsed with ddH 2 O, the cells were observed under a light microscope.

In vitro toxicity assay
The CCK8 reagent was used to calculate the cell survival rate in order to detect cytotoxicity. First, 143B cells were seeded in 96-well plates at 4 × 10 3 cells/well with five duplicated wells in each group. After the cells became adherent, the reagents were added to each of the following groups: the control group, the PUC19 group (the PUC19 plasmid was used to clone Escherichia coli without the target gene), the CS/PUC19 group, the SPCIONPs/PUC19 group and the PEI/PUC19 group, and a magnetic field was applied for 30 min. A total of 10 µl CCK8 reagent was added at 24 and 48 h, followed by incubation at 37°C for 3 h. The optical density (OD) at 450 nm was detected using a microplate reader, and the cell survival rate was calculated as follows: cell survival rate = (experimental group − blank group)/ (control group − blank group) * 100%.

Live-dead staining
The 143B cells were seeded in a 24-well plate at 5 × 10 4 cells/well. After the cells became adherent, reagents were added to each of the following groups: the control group, the PUC19 group, the CS/PUC19 group, the SPCIONPs/PUC19 group and the PEI/PUC19 group. The magnetic field was applied for 30 min. The cells were collected by centrifugation and resuspended in a mixture of 1 µl A (1 mM Live-Dye) + 1 µl B (1 mg ml −1 PI) + 1 ml staining buffer at 24 h and 48 h. After incubation at 37°C for 15 min in the dark, the cells were observed under a fluorescence microscope. The following method was used to calculate the cell survival rate: number of red cells/(number of red cells + number of green cells) * 100%.

Cell uptake
The 143B cells were seeded in a 34 mm confocal dish at 1 × 10 5 cells/well, and FITC-CS/pIGFBP 5 and FITC-SPCIONPs/pIGFBP 5 were added after the cells became adherent. pIGFBP 5 is an expression plasmid with red fluorescence protein (RFP). After a magnetic field was applied for 30 min, the nonendocytosed particles were washed away with PBS, and cell culture was continued for 6 or 24 h. At each time point, the cells were washed with PBS and stained with Hoechst for 20 min. After washing with PBS, the medium was added, and the cells were observed under a confocal microscope.

In vitro transfection assay
The 143B cells were seeded in a 12-well plate, and when the confluency of cells reached 80-90%, the medium was aspirated, and the cells were washed three times with PBS. A total of 800 µl of supplement-free medium was added to each well, and 4 nM HCl was used to prepare serum-free medium at a pH of 5.5. The CS/pIGFBP5 and SPCIONPs/pIGFBP5 was used serum-free medium to dilute to 200 µl per well, and the mixtures were incubated at 50°C for 20 min and used to transfect cells; 200 µl PEI was added to the PEI group (positive control group), and a magnetic field was applied for 30 min. After 4 h of transfection, the cells were washed three times with PBS, cultured for 24 h, and observed under an inverted fluorescence microscope.

Statistical analysis
All analyses were performed using IBM SPSS 20.0 (software version 20.0, SPSS Inc., Chicago, IL, USA) or GraphPad Prism (version 7.00, GraphPad Software, Inc., San Diego, CA, USA). Multigroup comparisons were performed using ANOVA. A value of p < 0.05 was considered statistically significant.

Characterization
SPCIONPs were successfully prepared by the dehydration condensation method, and their physicochemical characterizations were shown in figure 1. The morphology of SPCIONPs was observed by TEM, and the average particle size of a single particle was about 22 nm ( figure 1a,b). Figure 1c shows that the zeta potentials of CS, SPIONPs and SPCIONPs were 42.9, −27.9 and 27.1 mV, respectively. The negatively charged plasmids were combined through electrostatic adsorption to form SPCIONPs/ pIGFBP 5 , which were positively charged with a zeta potential of 11.3 mV.
A comparison of transverse magnetic induction results (figure 1d ) indicated that the superparamagnetic behaviour of SPCIONPs was weaker than that of SPIONPs. Both curves intersected the origin, and the hysteresis was zero. Therefore, both were superparamagnetic, which was consistent with previous research results [29,33].
The AFM also verified that CS was bound to SPIONPs by amide bonds. White dots formed a sixmembered ring representing the basic structure of CS; black dots represented Fe atoms; and red and white dots bound together represented N-H and C=O (figure 1f ).

Plasmid binding assay
After the addition of samples at N/P ratios of 10/1, 5/1, 2.5/1, 2/1, 1/1, 1/2, 1/2.5, 1/5 and 1/10, we found that when the N/P was 2/1, the plasmids did not run out of the electrophoresis lanes, which confirmed that SPCIONPs could fully bind to the plasmids and that the transfection ratio should be greater than or equal to 2/1 ( figure 2a).
The enzyme protection assay showed that the naked plasmid was completely degraded by DNase I. Moreover, at an N/P ≥ 2 and after DNase I addition, the plasmid was not degraded and still gathered in the hole. Figure 2b shows that SPCIONPs also protected the plasmids from degradation by DNase. Figure 3 shows Prussian blue staining after 4 h of incubation with SPCIONPs/pIGFBP 5 . In the SPCIONPs/pIGFBP 5 group, most of the blue particles were concentrated on the cell membrane, and a small number of blue particles were visible in the cells. No significant difference in cell morphology was observed compared to the control group, indicating that SPCIONPs had no obvious cytotoxicity within 4 h.

CCK8 assay
A CCK8 assay was used to detect the toxicity of PUC19, CS/PUC19, SPCIONPs/PUC19, and PEI/ PUC19. Figure 4 shows that the growth of the cells in the PUC19, CS/PUC19, and SPCIONPs/PUC19 groups at 24 h was not inhibited compared to the control group, and the cell survival rate in the SPCIONPs/PUC19 group was even greater than that in the control group, but the difference was not statistically significant. The cell survival rate in the PEI/PUC19 group was 86.59 ± 0.05%, which was significantly different from that in the SPCIONPs/PUC19 group.
After 48 h of the incubation with PUC19, CS/PUC19 or SPCIONPs/PUC19, the cell survival rate was not significantly different from that in the control group, whereas the cell survival rate in the PEI/PUC19 group was 73.37 ± 0.06%, which was also significantly different from that in the SPCIONPs/PUC19 group. No significant difference was observed between the CS/PUC19 and SPCIONPs/PUC19 groups at 24 or 48 h. The overall trend of the live-dead staining assay was generally consistent with that observed in the CCK8 assay. As shown in figure 5, red fluorescence represented dead cells, and green royalsocietypublishing.org/journal/rsos R. Soc. Open Sci.   royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 8: 201331 quantities around the nucleus. The FITC-SPCIONPs/pIGFBP 5 group showed more green fluorescent particles around the nucleus. At 24 h, both groups had red fluorescence expression of the plasmids, while the red fluorescence expression in the FITC-SPCIONPs/pIGFBP 5 group was greater than that in the FITC-CS/pIGFBP 5 group. As shown in figure 7, the red fluorescence expression in the PEI/pIGFBP 5 group ( positive control group) was the highest of all groups. The red fluorescence expression in the SPCIONPs/pIGFBP 5 group at 24 h after transfection was more than that in the CS/pIGFBP 5 group, which was more than that in the control group. This result was consistent with the laser confocal microscopy result.

Discussion
In this study, SPCIONPs were prepared to load the IGFBP 5 plasmids, which could inhibit lung metastasis of osteosarcoma. The gene carrier was composed of carboxylated SPIONPs and CS. The surface of carboxylated modified SPIONPs was rich in carboxyl groups and can be recombined with amino-rich chitosan in the form of covalent bond. In many studies, chlorosulfonic acid supported piperidine-4carboxylic acid (PPCA) [34], citric acid [27,28], PEG [35], oleic acid [36] and polyamide acid [37] were used to modify SPIONPs to make its surface touch carboxyl groups, which can be combined with   amide bond (figure 1e,f ). CS is a polycation due to its amino groups, which are ionized in weakly acidic environments, allowing the polymer to interact with negatively charged surfaces, such as cell membranes [38,39]. After CS was successfully grafted onto SPIONPs, they could bind a large number of concentrated plasmids and effectively protect them from DNA enzyme degradation (figure 2), which was one of the necessary conditions for cell transfection [40]. Because the zeta potential of SPCIONPs/pIGFBP 5 was 11.3 mV, the positively charged SPCIONPs/pIGFBP 5 easily bound to the negatively charged cell membrane (figure 3), as shown by Prussian blue staining, which was the initial condition of cell transfection [41].
The toxicity of gene vectors is an important parameter of gene release systems [38]. Reduced toxicity is a significant advantage of CS-modified carriers, and CS has been shown to have low toxicity in many in vitro experiments [33,42,43]. Unsoy et al. [29] showed that CS-encapsulated superparamagnetic iron oxide nanoparticles (CS-MNPs) had no obvious cytotoxicity to cells, and when CS-MNPs were applied at high doses (1000 µg ml −1 ) to HeLa cells, cell proliferation was reduced by only 2-5%. This observation is consistent with our research results. In this study, CCK8 and live-dead staining assays showed no significant difference between the SPCIONPs/PUC19 group and control group at 24 and 48 h (figures 4 and 5). Therefore, the SPCIONPs prepared in this study can serve as a non-viral vector with good biocompatibility and biosafety.
It is crucial to improve the transfection efficiency of CS as a gene vector system. CS is widely used in various fields due to its good biosafety. However, as a non-viral vector, chitosan does not have a strong proton-buffering effect, making it unable to escape from endosomes in a timely manner, leading to a lower transfection rate [40]. To increase the transfection rate of CS, Xu et al. [17] generated Arg-CS/ pBMP-2 nanoparticles with arginine-linked CS to transfect preosteoblasts, Nam & Nah. [44] used CSlinked PEI to transfect colon cancer (HCT119) cells, and Wang et al. [45] prepared FA-PEG-CS to transfect liver cancer (HepG2) cells. However, magnetic transfection was used in this study, which forced the transfection vector and the target cell to stay in contact. In addition, the oscillation induced by magnetic transfection on the target cell surface can cause the nanoparticle complex to move, which helps increase the transfection rate by mechanically stimulating endocytosis [30,46,47]. Other researchers have shown that magnetic transfection can effectively increase the transfection rate of gene vectors [47,48]. Our in vitro transfection result showed that at 24 h, red fluorescence expression was greater in the SPCIONPs/pIGFBP 5 group than that in the CS/pIGFBP 5 group (figure 7). Laser confocal microscopy result showed (figure 6) that the FITC-SPCIONPs/pIGFBP 5 group exhibited greater red fluorescence expression with magnetic transfection because more nanoparticles accumulated around the nucleus. These findings confirm that magnetic transfection can promote transfection through targeted physical adsorption.

Conclusion
We successfully prepared carboxylated SPIONPs with a negative surface charge, which can combine with CS to form a non-viral vector called SPCIONPs. SPCIONPs have good biosafety, biocompatibility and can be tightly loaded with IGFBP 5 plasmids. Magnetic transfection can effectively promote control CS/pIGFBP 5 SPCIONPs/pIGFBP 5 PEI/pIGFBP 5 royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 8: 201331