Heparin and Carboxymethylchitosan Metal Nanoparticles: An Evaluation of Their Cytotoxicity

In the search for noninvasive diagnostic techniques and new therapies, “nanosystems”, which are capable of binding and targeting bioactive molecules, are becoming increasingly important. In this context, biocompatible coatings are gaining interest, not only for their biological effects but also because they are considered capable to mask nanoparticle toxicity. In this work, we have compared the toxicity of nanoparticles coated with heparin and carboxymethylchitosan in the SKOV-3 cell line. Our results indicate that heparin and carboxymethylchitosan coatings do not guarantee the decrease of nanoparticle intrinsic toxicity which is often envisaged. Nonetheless, these coatings provide the opportunity for further functionalization with a variety of biomolecules for their use in theranostics.


Introduction
Nanomedicine, the application of nanotechnology in healthcare, offers numerous and promising possibilities to significantly improve medical diagnosis and therapy. New sensitive diagnostic devices, in fact, will permit very early personal risk assessment, and the abatement of costs for the disease treatment is a must for healthcare. Due to its high potential, nanomedicine holds the promise to greatly improve the efficacy of pharmaceutical therapy, reduce side effects, and make drug administration more convenient [1].
Among the biological molecules used for NP coating, chitosan, particularly carboxymethylchitosan (CMCS), and heparin appear very interesting also because they are considered capable to mask NP toxicity [11,12]. We should not, in fact, oversee the toxicity of cobalt and nickel oxide NPs [13][14][15][16] nor their potential effect on the environment [17]. Even though heparin is predominantly used as anticoagulant, its ability to interact with proteins makes it very attractive. NPs coated with heparin (NP@heparin) are extensively studied because of their several biomedical applications ranging from tissue engineering to biosensors passing for its use in cancer therapy [3]. As well as heparin, also chitosan NPs have demonstrated anticancer activity in vitro as well as in vivo even though the mechanisms remain to be elucidated [18].
In this paper, we have reported cytotoxicity and uptake of some transition metal oxide NPs (Co 3   NiO) coated with heparin and of Fe 3 O 4 NPs coated with CMCS (Fe 3 O 4 @CMCS) in SKOV-3 cell. Transition metal NPs are especially used to enhance surface electrochemical reactivity to further improve the performance of lithium-ion batteries [19] as well as in catalysis [20,21]. Nevertheless, the therapeutic use of transition metal conjugates was already known in the sixteenth century because of their different oxidation states and ability to interact with negatively charged molecules forming chelation complexes [22]. The results here reported indicate that heparin and CMCS alone did not show any cytotoxicity effect at the concentration used in the experiments. Unfortunately, they did not seem to be able to drastically reduce NP toxicity.
2.2. Nanoparticles Characterization. The particle size distribution was studied by transmission electron microscopy (TEM) using a 90 keV JEOL-1010 electron microscope (Tokyo, Japan). TEM samples were prepared by placing 10 L of a dilute suspension of Fe 3 O 4 nanoparticles in ethanol on a carbon-coated copper grid and allowing the solvent to evaporate at room temperature. The average particle size ( TEM ) and distribution were evaluated by measuring the largest internal dimension of 100 particles.

Synthesis of Carboxymethylchitosan (CMCS)
. CMCS was prepared as reported by Zhu et al. [23]. The chemical structures of chitosan and CMCS are reported in Figure 1. Afterwards, NP systems were separated from the supernatant by a neodymium magnet, centrifuged twice at 15000 ×g for 15 min at 4 ∘ C, then ultracentrifuged at 300000 ×g for 2 h at 4 ∘ C. After centrifugation, supernatants were filtered using a 0.22 m pore size membrane. The amount of Fe (II), eventually released in solution, was determined by complexometric analysis with the o-phenanthroline [25]. The Fe (II), in the presence of o-phenanthroline, form the stable red-orange complex [(C 12 H 18 N 2 ) 3 Fe] 2+ . The intensity of the color does not vary in the range of pH between 3 and 9. The maximum absorption wavelength occurs at 510 nm. The possible Fe (III) is reduced to Fe (II) by treatment with hydroxylamine hydrochloride.

FT-IR Spectra Analysis.
Characterization of the samples was performed using the solid phase Fourier transform infrared spectroscopy (FT-IR). Spectra were obtained using a BioMed Research International 3 Nicolet, Avatar 360. Samples were mixed with infrared grade KBr in a proportion of 2 : 100 (w/w).

Cell
Culture. SKOV-3 cell line was maintained as adherent cells in RPMI 1640 medium, at 37 ∘ C in a humidified 5% CO 2 atmosphere. Medium was supplemented with 10% fetal bovine serum and 2 mM L-glutamine. Cells were passaged as needed using 0.5% trypsin EDTA.

Cell Viability.
Cell viability was determined measuring ATP content by the CellTiter-Glo Assay according to the manufacturer's instructions. In details, 200 L of cell suspension (containing 2 × 10 4 , 1 × 10 4 , 5 × 10 3 , 25 × 10 2 cells depending of the exposure time) were seeded into 96-well assay plates and cultivated for 24 h at 37 ∘ C in 5% CO 2 to equilibrate and become attached prior to the treatment. Then, cells were exposed to 100 L of increasing concentrations of heparin, NP@heparin, CMCS, Fe 3 O 4 @CMCS electrostatic, and Fe 3 O 4 @CMCS covalent for 0.5, 1, 2, 24, 48, and 72 h. After the treatment, plates were equilibrated for 30 min at room temperature and then 100 L of CellTiter-Glo reagent was added to each well. Plates were shaken for 2 min and left at room temperature for 10 min prior recording luminescent signals using the Infinite F200 plate reader (Tecan Group, Switzerland). Cell viability, expressed as ATP content and normalized against control values, was recorded. All the experiments were performed in triplicate.

Cellular
Uptake. 10 4 cells were seeded on a coverslip (12 mm Ø) into 12-well assay plate and cultivated for 24 h at 37 ∘ C in 5% CO 2 to equilibrate and become attached before treatment. Cells were then incubated for 4 or 24 h with 25, 50, and 100 g/mL Fe 3 O 4 @heparin, Fe 3 O 4 @CMCS electrostatic, and Fe 3 O 4 @CMCS covalent and visualized by Prussian blue staining for iron detection. For this microscopic technique, the cells were fixed in ice-cold ethanol for 5 min, stained with an equal volume of 2% hydrochloric acid and 2% potassium ferrocyanide trihydrate for 15 min, and counterstained with 0.5% neutral red for 3 min. The preparations were then washed with distilled water and dried by increasing concentrations of ethanol, than mounted in DePeX (Serva, Germany). Observations were performed by a Zeiss Axiophot microscope under bright light illumination and photographs were acquired by a Zeiss AxioCam ERc5s camera.
Furthermore, for TEM studies, 10 6 cells, seeded in a 10 cm Petri dish, are exposed to 40 g/mL of NP@heparin, Fe 3 O 4 @CMCS electrostatic, and Fe 3 O 4 @CMCS covalent, for 30 min or 3 h. Then cells were harvested, fixed in 2% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.2) for 10 min on ice and for 30 min at room temperature, washed in the same buffer, and postfixed in dark for 1 h with 1% osmium tetroxide in 0.1 M sodium-cacodylate buffer (pH 7.2) at room temperature. After dehydration standard steps with a series ethyl alcohol, samples were embedded in an Epon-Araldite 812 1 : 1 mixture. Thin sections (90 nm), obtained with a ReichertUltracut S Ultratome (Leica, Nussloch, Germany), were stained with uranyl acetate and lead citrate according to the standard methods and observed with a Jeol 1010 electron microscope (Jeol, Tokyo, Japan) operated at 90 keV.

Statistical Analysis.
Cell viability values were expressed as mean ± standard error (SE). Analysis of variance (two-way ANOVA), for balanced mixed-effect experiments (uncoated NPs, coated NPs, and exposure times), was performed using KaleidaGraph 4.0 (Synergy Software). Statistical significant differences were fixed at ≤ 0.05 ( * ), ≤ 0.01 ( * * ), and ≤ 0.005 ( * * * ).    Regarding the comparison between uncoated and coated NPs, our data indicate that the coating did not decrease the NPs toxicity. As demonstrated in Figure 6, Co 3 O 4 NPs were less toxic than Co 3 O 4 @heparin for all the examined concentrations and time of treatment. The differences were less indicative for Fe 3 O 4 NPs and NiO NPs (Figures 7 and 8 Figure 9. The percentage of CMCS bound to NPs was less than 4% of the total weight; therefore, it was reasonable to compare the amount of coated and uncoated Fe 3 O 4 NPs neglecting the weight of CMCS bound. As previously reported ( Figure 5(b)), CMCS itself did not show cytotoxicity at the tested concentrations. On the contrary, Fe 3 O 4 @CMCS covalent, and electrostatic, caused a dose-dependent reduction of ATP (Figures 9(a), 9(b) and 9(c)) more pronounced compared to the bare Fe 3 O 4 NPs. For further details see Supplementary Material Table 4.  (Figure 10(d)) are more internalized compared to Fe 3 O 4 @CMCS covalent (Figure 10(b)) and electrostatic (Figure 10(c)). Apparently, no differences are observed between the two chitosan systems. Coated Fe 3 O 4 NPs are readily incorporated into the cells already after 4 h; therefore, it is not possible to assert a time and dose dependence. In addition, for all the NP systems, it is observed that internalized NP did not interfere with mitosis process (Figures 10(b)-10(d)).

FT-IR Spectra
TEM images ( Figure 11) confirmed that NP@heparin are readily internalized; in fact, already after 30 min of incubation NPs appeared inside the cells. Once entered most of the NPs remained in the cytoplasm, free or inside vesicles ( Figures  11(a)-11(c)). As already highlighted by optical microscope, besides being rapid, internalization of the nanoparticles was aspecific. In these pictures, NP@heparin are identified as high electron density objects since NPs maintained the morphology observed in cell-free environment (Figure 11(d)). Worth to note is that, also after 3 h of exposure, no NP@heparin was observed in the nuclei even though the massive internalization of NPs can modify nucleus shape ( Figure 12).
From our observations, the internalization did not seem to be influenced by the coating. Our hypothesis is confirmed by TEM picture (Figure 13) that did not show appreciable differences in cellular localization between Fe 3 O 4 @CMCS electrostatically or covalently bound and NP@heparin (Figures 11 and 12).

Discussion
In recent years, the use of NPs, particularly MNPs, has expanded into biomedical research. Due to their unique properties such as small size, large surface area, and high reactivity, they are particularly suitable for diagnosis and therapy [1,[26][27][28][29]. Often, NPs have to be covered with molecules to get a core@shell system capable to bind bioactive      molecules, stable in physiological fluids and possibly not toxic to the body. Among the innumerable coating materials, polymers such as heparin, dextran, carboxydextran, chitosan, and polyethylene glycol are considered more advantageous to satisfy the above-mentioned characteristics [30][31][32][33].
In particular, the literature reports several applications of NPs covered with heparin, ranging from use as imaging agent to apoptosis-induced agent in cancer cell, as well as components of nanodevices [34][35][36]. Unfortunately, this wide number of publications does not include toxicity studies of the synthesized systems. In particular, the literature lacks data on the comparison between the toxicity of core and core@shell. To try to fill this gap, in our laboratory, we have studied the characteristics and behavior of Co 3 O 4 , Fe 3 O 4 , and NiO NPs covered with heparin.
From our experiments resulted that the coating had significantly increased the colloid stability and hydrophilic property of metal NPs. In fact, the systems NP@heparin did not agglomerated thanks to the presence of negatively charged groups around the metallic core. The experiments on cytotoxicity, performed on SKOV-3 cells, have shown that heparin itself was not toxic within the range of the examined concentrations (see Figure 5(a)). Furthermore, as expected, Fe 3 O 4 @heparin was the less toxic system, while NiO@heparin was the most toxic one. Contrary to what one would expect, NP@heparin had not been found less toxic compared to the naked NPs for all the examined metals (see Figures 6,7,and 8). Depletion of ATP content, observed in these experiments, could be due to the massive internalization of NP@heparin by the cells, phenomenon substantiated by Prussian blue staining for iron detection. Nevertheless, at the concentrations used in these experiments, internalized Fe 3 O 4 @heparin did not arrest mitosis process and nanoparticles were shared between the daughter cells. Further analysis by TEM have demonstrated that NP@heparin were already present inside the cell after 30 min of exposure (Figures 11(a), 11(b), and 11(c)). In this work, we have not investigated the mechanisms of internalization even though, as shown in Figures 11(a), 11(b), and 11(c) and as reported by the literature [37][38][39], endocytosis is certainly a possible way. Notwithstanding in our previous work we had observed the presence of NPs also in the mitochondria and in the nuclei [40], in these experiments NPs were confined only in cytoplasmic vesicles, even though, sometimes, the vesicle size was so enormous to modify nuclear shape and/or cause mechanical damages to the cell (see Figure 12). When the endocitotic vesicles had sizes that did not justify the mechanical damage, we could assume that cell toxicity could be due to the release of metal ions by the NP system; this hypothesis was supported by the data of cell viability in which Fe 3 O 4 NPs resulted the least toxic metal.
Chitosan, but even better CMCS, preferred because the carboxymethylation increases the chitosan solubility in physiological fluids, is widely studied for theranostic applications [41,42]. Despite the Prussian blue staining indicated that Fe 3 O 4 @chitosan uptake was less efficient compared to that of Fe 3 O 4 @heparin, TEM analysis showed that no differences were noticeable between the two NP systems. Furthermore, as previously reported for heparin, the presence of negative charges on NP surface enhances interactions with the cell membrane facilitating cellular uptake [6,38]. Thanks to its biocompatibility and the presence of active functional groups (amino, carboxyl, and hydroxyl), CMCS is a valid instrument to design novel biocompatible materials with tailored chemical and biophysical properties [43][44][45][46]. Despite the wide use of CMCS little or nothing is known about its behavior when it is associated with metal NPs. This lack of data suggested us to evaluate the properties and the potential toxicity of the system Our studies on cell viability confirmed the biocompatibility of free CMCS at the tested conditions. When cells are exposed to Fe 3 O 4 @CMCS electrostatic, viability decreases with the same trend of Fe 3 O 4 NPs treatment (Figures 9(a)  and 9(b)). The higher toxicity observed for Fe 3 O 4 @CMCScovalent bond (Figure 9(c)) suggested that the method of preparation of the NPs could influence the cellular response.
Also in this case, uptake by SKOV-3 cells was relevant showing massive internalization already after 30 min with NPs stored in cytoplasmic vesicles ( Figure 13) with no detectable difference between NP@heparin and Fe 3 O 4 @CMCS.
Our results have confirmed the data present in the literature about the biocompatibility of heparin and CMCS and their capability to get stable suspensions in hydrophilic fluids when conjugated to metal NPs, but not the ability to reduce the cytotoxicity of metal NPs coated with these polymers. Nevertheless, it is difficult to compare data derived from different experimental conditions such as different concentration ranges rather than diverse cell types which can give diverse responses to the same treatment [2,47]. Moreover, the published data are often related to the whole system prepared and not to the single component, as we did, then the comparison is very difficult if not impossible.
In conclusion, the reactive groups, present on the surface of core@shell systems that we have synthesized, provide the opportunity for further functionalization so that a variety of biomolecules may be immobilized to enhance specific cell recognition for their use in targeting studies. Moreover, as regards Fe 3 O 4 NPs, even though the coating does not reduce their toxicity, the amount of NPs present in the systems is usually so low to render their toxicity negligible. Furthermore, due to their magnetic properties, Fe 3 O 4 NPs can be directed to the site of interest thanks to an external magnet. From this point of view, they could be promising tools as drug carrier for diagnosis and therapy.

Conflict of Interests
No conflict of interests is present. The authors have no financial involvement or interest with any organization or company about subjects or materials discussed in the paper.