RETRACTED ARTICLE: A Comparative Study of the Toxicity of Polyethylene Glycol–Coated Cobalt Ferrite Nanospheres and Nanoparticles

We present a comparative study of the toxicity of polyethylene glycol (PEG)–coated cobalt ferrite nanoparticles and nanospheres. Nanoparticles were prepared by hydrothermal method while nanospheres were prepared by solvothermal technique. The surface of nanomaterials was successfully modified with polyethylene glycol. To investigate the morphology of the prepared samples, X-ray diffraction (XRD), Fourier transform infrared (FTIR) spectroscopy, Raman spectroscopy, thermogravimetric analysis (TGA), and electron microscopy techniques were employed. Structural analyses confirmed the formation of polycrystalline cobalt ferrite nanoparticles with diameters in the range 20–25 nm and nanospheres in the range 80–100 nm, respectively. Kunming SPF mice (female, 6–8 weeks old) were used to investigate the toxicity induced by cobalt ferrite nanoparticles and nanospheres in different organs of the mice. Biodistribution studies, biochemical indices, histopathological assessments, inflammatory factors, oxidation and antioxidant levels, and cytotoxicity tests were performed to assess the toxicity induced by cobalt ferrite nanoparticles and nanospheres in mice. Cobalt ferrite nanospheres were found to be more toxic than the nanoparticles and curcumin was proved to be a good healing agent for the toxicity induced by PEG-coated cobalt ferrite nanomaterials in mice.


Introduction
In recent years, magnetic nanomaterials have received immense interest both in fundamental research and technological applications. These applications include, but are not limited to, drug delivery vehicles [1][2][3], magnetic resonance imaging (MRI) [4][5][6], hyperthermia [7][8][9], biosensors [10], cell separation [11], protein separations [11,12], gene magnetofection [13][14][15], and environmental pollution and remediation [16,17]. Cobalt ferrite, as hard magnetic material, is used as the contrast agent for MRI, targeted drug delivery, and heating mediator in hyperthermia [18][19][20][21][22][23]. Although cobalt ferrite is used in biomedical applications, however it has certain restrictions such as its high toxicity due to remarkable amount of cobalt that releases in the solution, aggregation in solution, and poor accessibility of the surface when surfactants are used. Therefore, this problem was overcome by the use of surface modification with certain biocompatible, nontoxic, and water-stable and dispersing materials [24][25][26][27][28]. Moreover, the fabrication of cobalt ferrite is easy and cost effective with tailored compositions, shapes, and sizes for any particular application. There are variety of techniques adopted for synthesis of nanosized cobalt ferrite, including mechanochemical [29], sonochemical [30], co-precipitation [31,32], micro-emulsion [33], and others [34][35][36][37][38]. Similarly, other techniques including single-step eco-friendly method was adopted for fabrication of tailored fluorescent copper nanoclusters using curcumin as the template [39]. A major drawback of most of these techniques is the low crystallinity of the prepared material, which in turn leads to the significant deterioration of the magnetic characteristics. In this regard, hydrothermal [40] and solvothermal [41] techniques are the most effective and efficient techniques to synthesize cobalt ferrite with controlled morphologies and crystallinities.
In literature, various nanomaterials such as silver nanoparticles (Ag NPs) have been reported to be used for antimicrobial treatment and associated infectious diseases as well as they are used as the nanovehicles for drug delivery and treatment of different diseases [42]. In another review article, ferrates have been reported to be used for elimination of diverse range of chemical and biological species from the wastewater [43]. In biomedical application of cobalt ferrite nanomaterials, the main issue is the accumulation of cobalt ferrite in organs, resulting in the toxicity in the body that requires urgent removal of the collected nanomaterials from the organs and needs healing of the damages induced by cobalt ferrite. Several researchers have studied the antiinflammatory drugs and found that these drugs can reduce the toxicity induced by nanomaterials [44,45]. Curcumin with antioxidant, antimutation, anti-tumor, and carcinogenic characteristics can be used as the healing agent for the toxicity induced by cobalt ferrite nanomaterials [46][47][48]. It has the ability to be used as the TNF blocker from in vitro and in vivo by making bonding to the TNF directly [49].
The objective of this work was to fabricate polyethylene glycol (PEG)-coated cobalt ferrite nanoparticles and nanospheres in the labs with controlled morphologies. Different doses of nanomaterials were intravenously injected into the mice and blood analysis, biodistribution, HE staining, and cell viability studied were performed to assess the toxicity of these nanomaterials. Comparison of the toxicity of cobalt ferrite nanoparticles and nanospheres was made and curcumin was used as the healing agent for the toxicity induced by cobalt ferrite nanospheres in mice. It was shown that cobalt ferrite nanospheres are more toxic than the nanoparticles due to their enlarged surface areas, which make them more toxic and more reactive than the nanoparticles. To the best of our knowledge, this is the first detailed study of this kind that has not been carried our earlier.

Preparation of Nanomaterials
For preparation of PEG-coated cobalt ferrite nanoparticles, we adopted hydrothermal technique [40,47]. For this purpose, solutions of cobalt chloride (0.2 M) and ferric nitrate (0.4 M) were prepared separately in 25 mL deionized (DI) water each and then these solutions were mixed with 25 mL aqueous solutions of polyethylene glycol (2.5 mM) and sodium hydroxide (3 M), respectively. The mixture was then stirred for 20 min and poured into the stainless steel (SS) autoclave, which was heated at 180°C for 6 h. When the process was completed, the mixture was cooled to room temperature and then the solution was washed 2-3 times using DI water and ethanol to remove any unwanted impurities from the mixture. The mixture was dried at about 80°C overnight in the oven and then ground into the fine powders to get the desired cobalt ferrite nanoparticles.
For preparation of PEG-coated cobalt ferrite nanospheres, solvothermal technique was used. For this purpose, cobalt chloride hexahydrate was dissolved in 40 mL ethylene glycol (2.5 mM) which was followed by the addition of 1.35 g of iron chloride hexahydrate and 1 g of polyethylene glycol (PEG). The mixture was then stirred for about 30 min and then sealed in Teflon lined SS autoclave. The autoclave was then heated at 200°C for 8 h and after finishing the reaction, it was then cooled to room temperature. The mixture was washed with deionized water and ethanol and then dried at 80°C overnight in the oven. Finally, the mixture was ground into the fine powders to get PEG-coated cobalt ferrite nanospheres with diameters in the range 80-100 nm. Morphology of the prepared nanomaterials was investigated by X-ray diffraction (XRD) following the method used in Ref. [50], scanning and transmission electron microscopy (SEM and TEM) as used in Ref. [50,51], room temperature Fourier Transform Infrared (FTIR) spectroscopy for determination of functional groups in cobalt ferrite similar to Ref. [51], Raman spectroscopy, and Thermogravimetric (TGA) analysis as used in Ref. [52].

Radioactive Labeling of Nanomaterials
The radiolabeling of PEG-coated cobalt ferrite nanoparticles and nanospheres was performed with 99m Tc using stannous chloride as the reducing agent [53][54][55]. For this purpose, fresh 99m TcO 4 generator eluate (50 μL with activity of ∼ 4 mCi) was prepared by adding it to 30 μL SnCl 2 suspension (1 mg/mL in 0.5 N HCl). With the help of NaHCO 3 solution (1 M), the pH of the suspension was adjusted in the range 8-10. Solutions of nanoparticles and nanospheres (40 μL each) containing~0.4%wt in cobalt ferrite were mixed with suspensions of stannous chloride (50 μg), ascorbic acid (10 mg/mL), and 99m TcO 4 . The mixture was then stirred at 10,000 rpm for 25 min at 80°C. For the accurate measurements, the radioactive counts were recorded within 24 h due to the short lifetime of 99m Tc (~6 h). The supernatant was then decanted after the centrifugation and the remaining material was identified to be 99m Tc-PEG-cobalt ferrite nanoparticles and nanospheres. Paper chromatogram was used to measure the radioactive yields of the labelled compounds, which were more than 65% that reflected the real biodistribution of nanomaterials in the mice in vivo.

Biodistribution of Nanomaterials
As indicated in Fig. 1, Kunming SPF mice (female, 6-8 weeks old, weight 18-20 g) were obtained from the Laboratory Centre for Medical Science, Lanzhou University, China. All mice were kept in cages under the temperature controlled system maintained at 21-22°C and lights were switched-on from 08:00 to 20:00 h. Free access to food and tap water was given to the mice and they were handled following the protocols of the Laboratory Animal Care Formulated by the National Society of Medical Research and the guidelines of the US National Institutes of Health. The mice were randomly divided into several groups with each group containing 5 mice and then they were injected intravenously with 99m Tc-PEG-cobalt ferrite solutions of nanoparticles and nanospheres and killed after 1 h, 6 h, 16 h, and 24 h, respectively. Tissues from the heart, lung, liver, spleen, and kidney were immediately dissected, wrapped in foil, weighed and then the radioactivity of 99m Tc in each tissue was measured using gamma counter detector. The biodistribution of nanomaterials in different organs of mice was presented in percent injected dose per gram of the wet tissue (i.e., % ID/g).

Hematoxylin and Eosin Staining
For hematoxylin and eosin (HE) staining, the paraffin wax was sliced into xylene for dewaxing and the process was repeated twice for about 10 min each. Hydration of the sample was carried by transferring the slides through different ethanol solutions with concentrations of 100% ethanol, 95% ethanol, and 70% ethanol each for 2 min. Rinsed the slides in running tap water at room temperature for about 2 min and when the process was finished, the nuclei were stained in hematoxylin staining solution at 60°C for 10 s and then at room temperature for 1 min and the slides were then placed under the running tap water at room temperature for about 5 min.
Stained the samples in working eosin Y solution for 2 min and then dehydrated the samples first by dipping in 95% ethanol and then in 100% ethanol each for 2 min. The cytoplasm was stained for 7 s by immersing in eosin staining solution for 15 s. After the removal, the cytoplasm was washed and dehydrated with absolute ethanol two times for 1 min each. The tissue was then made transparent with Xylene for 15 s and the cytoplasm was examined and then photographed using neutral gum seals. Microscopic examination of the tissues was performed using Olympus Microphot-CX41 microscope coupled with digital camera.

Biochemical Indices and Inflammatory Factors
Two hundred fifty micrograms of PEG-coated cobalt ferrite nanoparticles and nanospheres was intravenously injected into the mice of the exposure group while the control group was treated with normal saline of 0.9% and all mice were then killed after 24 h. Blood was collected from the mice and centrifuged for about 10 min to obtain the blood serum. The serum contents of TB, ALT, AST, BUN, CREA, and Cys-C were measured by the enzyme-linked immunosorbent assay (ELISA) and western blot. Enzymes linked to the liver, IL-6, IL-8, and TNF-α, play key role in inflammatory response induced by necrosis. Usually high levels of these expressions occur when an organ responds to the inflammation.

MTT Cell Viability Assay
The cytotoxic potentials of PEG-coated cobalt ferrite nanoparticles and nanospheres were determined by MTT, a colorimetric assay for evaluating the cell

R E T R A C T E D
A R T I C L E metabolic activities. Human epithelium cells L-132 and human monocytes THP-1 purchased from Shanghai, China, were exposed to different concentrations of nanoparticles in the range 30-125 μg/mL and nanospheres in the range 50-250 μg/mL and the optical density was measured at 590 nm for different assays using microplate spectrophotometer system (UNICO WFZ UV-2000, Shanghai, China). L-132 cells were selected as the inhalation is a major route for exposure of nanomaterials and THP-1 cells were used due to their role in clearing foreign materials. In each assay, the untreated cells were assessed as the negative control. The inhibition of enzyme activity was observed in the cells, which was compared with untreated (negative control) cells and the values were derived in the form of ratio of the negative control and plotted against the concentration of nanoparticles and nanospheres.

Statistical Analysis
Each data point was reported as the mean value (±sem) of the experiments performed in triplicate. Significance of differences was evaluated using analysis of the variance and statistical charts were drawn with the help of Origin and Microsoft Excel software.

Structural Analysis
Structural analyses (XRD, FTIR, Raman, and TGA) of the prepared nanomaterials are shown in Fig. 2. The XRD results in Fig. 2a represent the coated and uncoated cobalt ferrite at nanoscale, which confirms that cobalt ferrite was successfully fabricated. The positions and relative intensities of all the observed peaks in XRD data confirm the crystalline nature of cobalt ferrite. No extra peaks were observed, which indicates the purity of

R E T R A C T E D
the prepared cobalt ferrite. The mean crystallite size of cobalt ferrite was determined by using Scherrer equation [56], which was found to be~24 nm. Fourier transform infrared (FTIR) spectroscopy was conducted to investigate the cation distribution (of nickel, cobalt, and iron) in cobalt ferrite. Figure 2b indicates the FTIR data collected at room temperature. Theoretically, cobalt ferrite has two strong absorption bands (ʋ 1 and ʋ 2 ) along with few others appearing in the range 400-600 cm −1 . All these peaks are clearly indicated in our data shown in Fig. 2b. In FTIR data, ʋ 1 corresponds to the intrinsic stretching vibrations of the metal at tetrahedral sites, whereas ʋ 2 corresponds to the stretching vibrations of the metal ions at octahedral sites [57][58][59]. The peak appearing in FTIR at 3421 cm −1 corresponds to polyethylene glycol (PEG) which indicates its successful bonding on the surface of

R E T R A C T E D
cobalt ferrite. Raman analysis of cobalt ferrite collected at room temperature is shown in Fig. 2c, which indicates 5 different peaks that can be seen in the data. The peak appearing at below 700 cm −1 is the main characteristics peak (A 1g mode) of cobalt ferrite that corresponds to the stretching of oxygen ions along the Fe-O bonds at tetrahedral sites [60], whereas the other peaks appearing in the data also belong to cobalt ferrite. This confirms the successful fabrication of PEG-cobalt ferrite in our experiment. Figure 2d shows the TGA results of the samples collected in the temperature range 50-380°C which indicate that cobalt ferrite losses its weight at different temperatures. It is also evident in TGA analysis that the thermal stability of PEG is relatively low whereas that of PEG-cobalt ferrite is high. Electron microscopy analyses of the samples are shown in Fig. 3. Figure 3(a) and (b) indicate the SEM images of PEG-coated cobalt ferrite nanoparticles and nanospheres, respectively, whereas Fig. 3(c) and (d) indicate the TEM analyses of nanospheres and nanoparticles, respectively. These results show that the average size of nanoparticles is around 25 nm and that of nanosphere is 80-100 nm. From TEM images of nanospheres, it is obvious that nanospheres are composed for a large number of smaller nanoparticles with large surface areas, thereby making them mesoporous which are highly desirable for medical applications of nanomaterials as the drug-carrying vehicles. All these structural analyses confirm the successful formation of pure phase PEG-coated cobalt ferrite nanoparticles and nanospheres.

Biodistribution Studies
Quantitatively, the biodistribution of PE-coated cobalt ferrite nanoparticles and nanospheres in blood, heart, liver,

R E T R A C T E D
A R T I C L E spleen, lung, and kidney after different intervals of time (1, 6, 16, and 24 h) is shown in Fig. 4. The presence of cobalt ferrite in blood and other organs was assessed within 24 h after the intravenous injection of 99m Tc-PEG-cobalt ferrite solution (nanoparticles and nanospheres). In case of nanospheres shown in Fig. 4(a), the blood retention of cobalt ferrite was found to be 6.5 ± 0.33% ID/g after 1 h of the exposure and then it was gradually decreased over the next time intervals (i.e., 6, 16, and 24 h). It was seen that nanospheres were mainly distributed in the heart, liver, spleen, lung, and kidney; however, most of them were primarily accumulated in the spleen. Moreover, the biodistribution of nanospheres in various organs was found to be highest after the first hour and then decreased gradually and remained less than 30% after 6 h. In case of cobalt ferrite nanoparticles, the blood retention of nanoparticles was about 2.8 ± 0.14% ID/g after 1 h of the exposure, indicating a relatively fast clearance of radioactive material from the body's blood pool and then it was decreased with passage of time as shown in Fig. 4(b). The nanoparticles were distributed in the heart, liver, spleen, lung, and kidney with maximum concentrations in the spleen and liver. It is clear from the figure that the biodistribution of nanoparticles in blood and other organs was highest after the first hour and then gradually decreased after 6 h and finally reached to the lowest values after 24 h. If we compare the biodistribution results of nanospheres and nanoparticles, it is seen that the accumulation/presence of PEG-coated cobalt ferrite nanospheres in blood and other organs of the mice was more as compared with the nanoparticles. This might be associated with large surface area and high porosity of nanospheres as compared with the nanoparticles, which is one of the critical factors to determine the reactivity of nanomaterials interactions with biological systems. In case of nanoparticles, their nonmesoporous nature with low specific surface area made them less reactive than the nanospheres under the same conditions. These features might have reduced the prolonged resistance of PEG-coated cobalt ferrite nanoparticles in blood and other organs of mice. In addition, the nanospheres are causing complex formation with biomolecules and resulting in the increased level of radical species, increasing the level of oxidative stress, damaging the cellular DNA, and resulting in the oxidative stress by lipid peroxidation.

Biochemical Indices
To study the toxicity effect of PEG-cobalt ferrite nanoparticles and nanospheres in mice, biochemical indices were measured and the results are presented in Fig. 5. Various parameters including ALT, AST, BUN, CREA, TB, and Cys-C were measured for the control and exposure group mice. SPSS software was used for data extraction with *P < 0.05 that represents significant

R E T R A C T E D A R T I C L E
changes during the measurements. In both nanospheres and nanoparticles, it is seen that all biochemical indices show significant changes when compared with control group mice (*P < 0.05). In case of cobalt ferrite nanosphere exposure group, the levels of ALT, AST, and BUN demonstrate significant differences (*P < 0.05) compared with control group mice, whereas in case of nanoparticles exposure group, only Cys-C exhibits significant difference compared with control group mice (*P < 0.05). It is seen that TB and Cys-C which are mainly responsible for the biomarker of kidney function were decreased significantly in case of nanospheres. This suggests that kidney is affected more by the exposure of PEG-cobalt ferrite nanospheres as compared with nanoparticles. AST, as the biomarker for the liver, was affected more by the exposure of both nanoparticles and nanospheres. This suggests that the exposure of cobalt ferrite can adversely affect the liver function. From all these results, it is clear that PEG-cobalt ferrite nanospheres are creating more damages in mice in vivo as compared with cobalt ferrite nanoparticles.

Histopathological Study
We have presented histopathology analysis of the control, nanoparticle, nanosphere, and treatment group mice as shown in Fig. 6. If we compare the results of nanospheres and nanoparticles exposure groups with control group mice, it is seen that PEG-cobalt ferrite

R E T R A C T E D A R T I C L E
nanospheres are producing more damages in different organs (liver, spleen, kidney, and lung) of mice as compared with the nanoparticle exposure group. In the kidney, glomerular congestion occurred along-with mild edema and interstitial inflammation cells are seen in case of nanospheres intake when compared with nanoparticle exposure and control group mice. It is also seen that nanoparticles show less inflammation than the nanospheres. In case of nanoparticle exposure, it was found that lungs are relatively less affected whereas in case of nanospheres, the alveolar wall was found to be thickened and mild fibrosis were seen. Additionally, for the nanosphere exposure group, the hepatocytes show swelling and edema was occurred, whereas relatively less inflammation was found in case of nanoparticle exposure group mice.

Inflammatory Factors and Oxidation/Antioxidant Level
The expression levels of IL-6, IL-8, TNF-α, MDA, and T-AOC were measured and the results are shown in Fig. 7. Figure 7a represents the western blot bands of IL-6, IL-8, and β-actin for the control, nanoparticle, and nanosphere exposure groups. The relative protein level of IL-6 and IL-8 for the control, nanoparticle, and nanosphere exposure groups is shown in Fig. 7b, whereas the contents of TNFα, MDA, and T-AOC are shown in Fig. 7c-e with *P < 0.05 for the exposure group versus control group ± sem. The results revealed that the levels of IL-6, IL-8, TNF-α, and MDA for cobalt ferrite nanosphere exposure group mice are higher than that of the nanoparticle group and both these levels are higher than the control group mice.
In case of T-AOC, the level of nanospheres was lower than that of nanoparticle exposure and control group mice. All these results indicate that nanoparticles and nanospheres are causing inflammation in mice, especially in the liver. However, nanospheres are affecting the organs more than the nanoparticles. It well understood that nanomaterials in the body generate oxygen free radicals (ROS), which cause series of qualitative reductions of the antioxidants, resulting in the oxidation damages of biological tissues that adversely affect the cellular organisms [61,62]. Moreover, when the levels of IL-6, IL-8, TNF-α, MDA, and T-AOC for the mice exposed to nanoparticles were compared with those exposed to nanospheres, it was found that cobalt ferrite nanospheres resulted in the more inflammation as compared with the nanoparticle exposure group mice.

Cytotoxicity Assessment
The cytotoxicity studies for different concentrations of PEG-coated cobalt ferrite nanospheres and nanoparticles were carried out and the results are presented in Fig. 8. The percentage survival of L-132 cells is shown in Fig.  8(a), whereas Fig. 8(b) represents the percentage survival of THP-1 cells. It is seen that for concentrations above 100 μg/mL, there are significant changes in cell the viability observed for both the cells, and it is seen that the results are more pronounced in case of PEG nanospheres. This confirms that cobalt ferrite nanospheres are producing more damages as compared with the nanoparticles. Moreover, the cell viability decreases with increasing concentration of both nanoparticles and nanospheres, which indicates that PEG-coated cobalt ferrite in both forms produces more toxicity in mice with increasing concentration. Due to the two different cellular targets (L-132 and THP-1), one can expect the cellular response not to be identical, depending on the cell death mechanism [63]. Possible reason to explain the cellular target specificities even for similar sizes of the particles may be attributed to the function of

R E T R A C T E D
A R T I C L E phagocytosis, which characterizes monocytes (THP-1 cells), but not the lung epithelial cells [64]. It is well understood that single nanosphere is composed of a large number of small nanoparticles. Thus, it possesses large surface area as compared with the nanoparticles and therefore, it has more reactivity and more chances of interaction with biological systems (tissues) as compared with the nanoparticles. Moreover, due to the larger size of nanospheres, they are not able to get secreted-out easily via blood or urine circulation once they enter into the organ. Therefore, they remain in the body (organs) for relatively longer time as compared with the nanoparticles, which in turn affect the tissues adversely. Moreover, the nanospheres cause reduced function of macrophages, reduced phagocytosis of the nanospheres themselves, and reduced macrophage mobility and cytoskeletal dysfunction.

Curcumin Effect on Toxicity
Biochemical indexes in blood serum were studied for nanosphere exposure group and curcumin-treated group and the results were compared with control group mice, which are shown in Fig. 9. It was found that all these indices in the treatment group mice showed significant improvements after the administration of curcumin when compared their values with nanosphere exposure and control group mice. In the figure, it is seen that the expression levels of ALT, AST, BUN, CREA, CYS-C, and TB were approached towards the normal values after the administration of curcumin. This can be attributed to the fact that curcumin has strong antioxidant characteristics which reduces the oxidative stress produced as a result of the toxicity induced by cobalt ferrite [47]. It has also been reported that TNF-α and IL-1 play important role in the induction of hepatic necrosis and curcumin reduces the effect of toxicity by inhibiting the secretion of TNF-α and

R E T R A C T E D A R T I C L E
IL-1 by macrophages [48], similar to the work reported earlier in Ref. [65].

Conclusion
In this work, we successfully fabricated PEG-coated cobalt ferrite nanoparticles and nanospheres via hydrothermal and solvothermal methods, respectively. From structural analyses, it was found that the prepared nanomaterials are highly pure, crystalline, and biocompatible in nature resulting from the successful attachment of PEG. It was found that both nanospheres and nanoparticles of cobalt ferrite are toxic to biological systems. Furthermore, it was shown that nanospheres of cobalt ferrite are more toxic than the nanoparticles due to their large surface area and more reactivity with biological tissues. Positive changes were monitored in biochemical indexes after the administration of curcumin which is a natural chemical possessing no side effects, thus confirming it can be used as the healing agent for the toxicity induced by cobalt ferrite nanospheres.