Influence of oleic acid coating on the magnetic susceptibility and Fenton reaction-mediated ROS generation by the iron oxide nanoparticles

Fenton reaction-mediated reactive oxygen species (ROS) generation by the iron oxide nanoparticles (IONPs) is responsible for its antibacterial activity. In general, IONPs are surface-coated to facilitate stability, control over size, biocompatibility, solubility, etc. We hypothesize that the extent of surface coating onto the IONPs might affect Fenton reaction-mediated ROS generation, which would eventually impact its antibacterial activity. In the present study, IONPs were prepared using the co-precipitation method, and different weights of oleic acid (OA) were loaded onto the IONPs. Pristine IONPs and oleic acid-coated iron oxide nanoparticles (OA-IONPs) were characterized using Fourier transform-infrared spectroscopy, dynamic light scattering, transmission electron microscopy, X-ray diffraction, vibrating sample magnetometry, goniometer, and thermogravimetric analysis. We found that magnetic susceptibilities of the IONPs were significantly enhanced with an increase in OA loading on the IONPs. The antibacterial study showed that the percentage inhibition was inversely related to the extent of oleic acid coating on the IONPs. The dependency of ROS generation on the extent of surface coating over IONPs was demonstrated using the 2’,7’-dichlorodihydrofluorescein diacetate (DCFDA) assay. Although pristine IONPs showed the least ROS generation, they exhibited maximum percentage inhibition of bacteria. This might be due to mechanical damage to the bacterial cells because of their crystalline nature. In vitro biocompatibility study conducted on L929 fibroblast cell lines indicated that all the nanoparticle preparations were cytocompatible. This study concluded that the extent of surface coating influences the Fenton reaction-mediated ROS generation and also the magnetic susceptibilities of the IONPs.


Introduction
Biomaterials are intended to restore or improve tissue function and often face a serious threat of bacterial infection post-implantation [1].The discovery of antibiotics in 1928 emerged as a potential candidate to treat bacterial infection.Penicillin, methicillin, oxacillin, and vancomycin are a few of the antibiotics developed to treat bacterial infections [2].However, bacterial species started to develop resistance to these antibiotics over a period resulting in multidrug resistance (MDR) [3].Staphylococcus aureus is one of the 'ESKAPE' pathogens, which develops resistance towards antibiotic treatment and hence escapes from the antibiotic action [3].Staphylococcus aureus is also a major causative agent of orthopedic-related infections such as osteomyelitis.Its biofilm-forming ability imposes further complications [4].It is estimated that approximately 75% of the cases of osteomyelitis are due to the Staphylococci species [5,6].To combat such serious issues caused by Staphylococcus aureus, novel approaches are indeed necessary.Metallic nanoparticles have gained a lot of attention due to their antibacterial activity against a wide spectrum of pathogenic bacteria and multimodalities in treating bacterial infections [4,7].
Among several metallic nanoparticles exploited in the biomedical field, IONPs are of special interest due to their multipotency.IONPs mediate several functions in the biomedical field, including potent antibacterial activity, magnetically guided drug targeting, and as a contrasting agent in magnetic resonance imaging [8][9][10].The antibacterial function of the IONPs is dependent on the generation of the reactive oxygen species (ROS) mediated by the Fenton reaction [11].Release of these ROS into the surrounding medium establishes reactions that damage the cellular components, majorly lipids [11].In general, the size, shape, surface charge, and hydrophobicity/hydrophilicity of the nanoparticles influence the ROS generation by the nanoparticles [12,13].Physiological pH greatly defines the amount of ROS generated by nanoparticles.In addition to these parameters, the chemical and crystalline nature of IONPs also influences ROS generation [12].External magnetic fields and radiofrequency can be used to control ROS generation from IONPs [12,13].
Generally, surface modification is carried out on the nanoparticles with either organic or inorganic materials to (1) control the crystal growth of the nanoparticles, (2) improve biocompatibility, (3) improve colloidal stability, and (4) impart functionalization [14].Surface modification has proven to improve the activity of the nanoparticles (table 1).The size of the nanoparticle plays a significant role in biomedical applications such as antibacterial activity, cellular uptake, magnetic targeting, and hyperthermic treatment [15].The smaller the size of the nanoparticle, the higher would be its surface-to-volume ratio, which means a higher number of functional sites [16].Various surface coating agents have been explored for IONPs.Natural polymers (dextran, chitosan, alginate, oleic acid), synthetic polymers (polyethylene glycol (PEG), polyvinyl alcohol (PVA), polydopamine (PDA), polyvinylpyrrolidone (PVP), and other inorganic compounds (silica, carbon-based materials, graphene oxide) are some of the surface coating agents that have been used for IONPs [14,[17][18][19].Natural polymers offer the advantage of biocompatibility.However, these natural polymers are hydrophilic in nature and lack mechanical strength.Sometimes, these natural polymers adsorb non-specifically onto the IONPs [20].On the other hand, synthetic polymers are comparatively better in terms of mechanical strength.The selection of synthetic polymer depends on the required surface properties for the final application.However, the functionalization of synthetic polymers coated IONPs is highly limited compared to the natural polymers [20].Inorganic compounds offer high mechanical strength among all types of coatings.A synergistic effect can be achieved by coating inorganic compounds over IONPs.However, most of these inorganic coatings lack biocompatibility [20] Oleic acid (OA) is a well-known surface coating agent used for the preparation of IONPs.Several authors have reported high colloidal stability, excellent biocompatibility, and firm surface adsorption after coating OA over IONPs [14].Among several surface coating agents explored for IONPs, OA-coated IONPs showed better antibacterial properties [4].It was shown that hydrophobicity and surface charge imparted by the surface coating agent play a significant role in the antibacterial activity of the nanoparticles [4].ROS generation by the Fenton reaction is mediated by two types of iron ions based on the pH of the medium, (1) iron ions present on the surface of the nanoparticle, and (2) iron ions leached out from the surface into the surrounding medium [17].It is observed that surface adsorption of biomolecules on the nanoparticles influences ROS generation.For example, surface adsorption of lysozyme on the silver nanoparticles hindered the release of the metal ions into the media.Similarly, adsorption of bovine serum albumin (BSA) on the surface of iron metal is shown to block the dissolution of metal ions into the surrounding media, thereby altering the ROS generation by the metal nanoparticles [18].However, these biomolecule adsorption-mediated release of metal ions is dependent on the type of biomolecule and metal nanoparticles [18].OA-coating over IONPs has proven to have a better effect on the antibacterial activity compared to other coatings [4].Hence, we have studied the effect of different amounts of OA coating on ROS generation.It is hypothesized that the increased addition of oleic acid might restrict ROS generation, which would impact the biomedical applications of IONPs.This work aims to gain a better understanding of how varying the amounts of OA coating on IONPs impacts ROS generation and its antibacterial activity.
In the present study, we have synthesized IONPs using the co-precipitation method, loaded with varying concentrations of OA, and characterized them thoroughly.We have tested the possible role of the extent of OA coating on ROS generation using DCFDA fluorescence assay.We have also studied the effect of OA coating on Upconversion nanoparticles@AgBiS 2 Synergistic photothermal-photodynamic mediated ROS generation leads to effective antibacterial efficacy Antibacterial activity [20] Titanium-doped barium ferrite Titanium dopant enhanced photocatalysis and thereby inhibited bacterial growth Antibacterial activity [21] the antibacterial activity of the IONPs.Further, the biocompatibility of the pristine and OA-IONPs has been investigated using the L929 fibroblast cell lines.To the best of our knowledge, this is the first report elucidating the influence of the extent of surface coating agent over IONPs on the ROS generation, thereby, its potential antibacterial applications.

Materials
Ferrous chloride (FeCl 2 ), ferric chloride (FeCl 3 ), Luria-Bertani (LB) agar, LB broth, and AlamarBlue were purchased from Himedia, India.Oleic acid and DCFDA were purchased from Sigma-Aldrich, India.Sodium hydroxide was purchased from Sisco Research Laboratories, India.Fetal bovine serum (FBS), antibioticantimycotic solution, and Dulbecco's Modified Eagle Medium (DMEM) were purchased from ThermoFisher Scientific, India.All reagents used in this study were of analytical grade.All the experiments were carried out at least in triplicates.

Preparation of OA-IONPs
Pristine IONPs were prepared by co-precipitation method [22][23][24][25].Briefly, sodium hydroxide (6% w/v) solution was kept under stirring at 400 rpm at 60 °C.Iron salt solutions (FeCl 3 and FeCl 2 ) in a molar ratio of 2:1 were added to the sodium hydroxide solution and left for 5 min for the reaction to take place.IONPs solution was then washed twice with water and finally re-suspended in the ethanol.Different weights of OA (2.5, 5.0, 7.5, 10.0, and 12.5 g) were added to the IONPs solution and kept under stirring for 24 h at 400 rpm at 60 °C.After successful coating, excess OA was removed by washing with ethanol twice.Finally, the pellet was stored at room temperature until further use.Physico-chemical characterization of the pristine and OA-IONPs was carried out.

Physico-chemical characterization of OA-IONPs
Pristine and OA-IONPs were diluted sufficiently (water or chloroform), and then 2 μl of the nanoparticle solution was placed on the Transmission Electron Microscopy (TEM) grid.Subsequently, TEM grids were airdried, and images were taken using a TEM instrument (Tecnai G2 T20) to understand the morphology of these nanoparticles.The sizes of these nanoparticles were measured using dynamic light scattering (DLS) technique.Pristine and OA-IONPs were dispersed in a suitable solvent (ethanol in PBS) using a probe sonicator.The size distribution was measured using a DLS instrument (Malvern zetasizer nano ZS90).Pristine, OA-IONPs, and oleic acid samples were scanned in the range of 4000-500 cm −1 using Fourier transform-infrared spectroscopy (FT-IR) in ATR mode (Bruker Alpha) to understand the interaction between oleic acid and IONPs.Thermogravimetric analysis (TGA) was performed to study the thermal degradation behavior of the pristine and OA-IONPs (NETZSCH DSC 204F1 Phoenix).Approximately 10 mg of the pristine and OA-IONPs were weighed in the alumina pan and heated at the rate of 10 °C min −1 from room temperature to 1000 °C.An inert environment was maintained during the experiment by purging argon gas.The crystalline phase of pristine and OA-IONPs was identified by taking 2Θ x-ray diffraction (XRD) spectrum (Rigaku XRD) in the range of 20-80.Powder samples were spread onto the glass slide evenly and scanned using Cu K-alpha 1 radiation (λ = 0.154056 nm).The magnetic susceptibilities of the samples were measured using a vibrating sample magnetometer (Lakeshore 7410 S).Approximately 10 mg of the samples were weighed, and the magnetic moment of the samples was measured by applying a magnetic field in the range of ± 15000 G at room temperature.The contact angles of the pristine and OA-IONPs were measured by using a goniometer.Pristine and OA-IONPs solution (water or chloroform) was placed on the glass slide, and the solution was air dried.After that, a drop of water was placed using a micro syringe.Images were then captured at time point zero and 5 min.

Bacterial studies
S. aureus wild-type strain MTCC 1144 was chosen for bacterial studies, which is a well-known clinical strain [26,27].A single colony was picked from the overnight grown culture of S. aureus and inoculated into 5 ml of LB broth.The bacterial culture was transferred to the incubator at 37 °C and grown until mid-log phase [28].
IONPs (pristine and OA-IONPs) stock solution (100 mg ml −1 in ethanol or PBS) was diluted in LB broth to reach the working concentration of 250 μg ml −1 .Subsequently, nanoparticle solutions were sterilized by placing them under UV light for 30 min before use.

ROS generation by DCFDA assay
Mid-log phase bacterial culture was diluted 100 times.100 μl of the diluted culture was transferred to the 96-well plate.To each well containing bacterial culture, 100 μl of working concentration of IONPs (pristine and OA-IONPs) solution was added.Further, to each well, 20 μl of DCFDA (0.5 mM) was added and incubated at 37 °C under moderate shaking conditions.Without nanoparticles were used as a control.After certain time intervals (3 and 6 h), reading was taken using a fluorescence plate reader at excitation and emission wavelengths of 485 nm and 538 nm, respectively [29].

Antibacterial activity
Mid-log phase bacterial culture was diluted 100 times.100 μl of the diluted culture was transferred to the 96-well plate.To the wells containing bacterial culture, 100 μl of working concentration of IONPs (pristine and OA-IONPs) solution was added.The plate was transferred to the incubator at 37 °C for 24 h under moderate shaking conditions.Without nanoparticles were used as a control.Optical density (OD) at 600 nm was taken after 24 h incubation.Similarly, OD values of pristine and OA-IONPs without bacterial culture were taken and subtracted from the OD values of the nanoparticles-treated bacterial cultures.The bacterial survival percentage was calculated by comparing the OD values of the control wells [28].

Biocompatibility studies
A cell viability study was conducted on L929 fibroblast cell lines using AlamarBlue assay.Fibroblast cells were seeded in 96 well plates at a density of 1×10 4 cells/well in DMEM medium, 10% FBS and 1% antibioticantimycotic solution.Media was changed once every 24 h.IONPs (pristine and OA-IONPs) stock solution (100 mg ml −1 in ethanol or PBS) was diluted in DMEM medium.100 μl of the diluted solution was added to each well at the concentration of 250 μg ml −1 .The plate was transferred to the CO 2 incubator at 37 °C for 24 h.Without nanoparticles were used as a control.After 24 h, the media was removed, washed thrice with PBS to remove traces of nanoparticles, and then 200 μl of fresh media containing AlamarBlue (0.1 mg ml −1 ) was added to each well.Cell viability was measured as per the manufacturer's protocol [30].

Statistical analysis
The experiments were carried out in triplicates, and data were presented as mean ± standard deviation.Statistical analysis was determined by using one-way analysis of variance (ANOVA) and pairwise Tukey's test at a confidence level of 95% using GraphPad Prism version 5.04.

Results and discussion
3.1.Preparation and characterization of OA-IONPs TEM and DLS results of pristine and oleic acid-loaded IONPs preparations are presented in figure 1 and table 2, respectively.It is evident from the TEM micrographs (figure 1) that the sizes of all the OA-coated IONPs preparations were below 20 nm and quasi-spherical in shape.
It is clear from the TEM images that there was no impact on the shapes or sizes of the nanoparticles with the amount of OA loading.The TEM images show that the size of the pristine IONPs was drastically larger than the oleic acid-coated IONPs.This is due to the agglomeration of nanoparticles in the absence of a coating agent.However, average hydrodynamic sizes of pristine and OA-IONPs determined by DLS (table 2) were found to be drastically larger than the actual size, which is possibly due to the agglomeration of the nanoparticles in the solvent [31] Pristine nanoparticles tend to agglomerate and form larger clusters due to van der Waals interaction which is evident in DLS results [32].Hydrodynamic sizes of the nanoparticles reduced with the increase in the OA loading, which indicates OA prevented nanoparticles from getting agglomerated.
Pristine IONPs are hydrophilic in nature, and OA loading around IONPs induces hydrophobicity by forming an organic layer around the nanoparticles.Sessile drop contact angle measurement was carried out using water as a liquid, and the images are presented in figure 2. Pristine IONPs were hydrophilic in nature as there was no contact angle established, which is typical of pristine IONPs.On the other hand, OA-IONPs showed a contact angle of more than 100°at zero time points, referring to the hydrophobic nature of the nanoparticles.After 5 min, a gradual drop in the contact angle was observed against (2.5 OA)-IONPs, (5.0 OA)-IONPs, and (7.5 OA)-IONPs.However, (10.0 OA)-IONPs and (12.5 OA)-IONPs did not show much drop in the contact angle.This observation infers hydrophobicity of the nanoparticles increased with an increase in the OA concentrations.
Figure 3(a) depicts the FT-IR spectrum of pristine and OA-IONPs.In the pristine nanoparticles, Fe-O vibrational band was observed at 552 cm −1 , whereas in the case of OA-IONPs there was a blue shift observed in the range of 580-600 cm −1 .This shift was mainly due to the movement of iron ions into the tetrahedral sites [33].The small shoulder band at around 630 cm −1 that is visible in the FT-IR spectrum of pristine IONPs indicates the presence of maghemite [34].However, this particular band was either very weak or not visible in the case of the FT-IR spectra of OA-IONPs.This might be because the dense OA coating around the nanoparticles might have hindered the oxidation of magnetite to maghemite around the nanoparticles.In the case of oleic acid and OA-coated IONPs, bands at 2849 and 2923 cm −1 were attributed to asymmetric and symmetric CH 2 stretch of OA, respectively.In the OA spectrum, the band at 1710 cm −1 corresponds to the stretching vibrations of the C=O bond.As the C=O group of OA interacts with the iron atom, it should be absent in the OA-IONPs spectrum.However, this particular band was observed in all the OA-IONPs.This was mainly due to the formation of the OA bilayer around the IONPs.The first layer of OA was chemically interacting with the iron atoms, whereas the second layer of OA was adsorbed onto the first layer of OA through hydrophobic interaction.In addition, two bands were observed in the OA-IONPs spectrum at ∼1530 cm −1 and ∼1450 cm −1 , which belong to symmetric and asymmetric (COO -) stretching vibrations.The difference between these two bands (Δv COO -) indicates the type of interaction between the iron atom and the carboxylate group of OA.All the OA-IONPs showed (Δv COO -) less than 100 cm −1 , which is representative of the bidentate type of coordination between iron ions and the carboxylate group of OA.
Crystal planes (220), (311), (400), (511), and (440) were noticed in the pristine and OA-IONPs.These represent characteristic XRD spectra of magnetite (figure 3(b)) and are in the cubic phase of magnetite [24,25,35].TGA analysis of the pristine IONPs showed only one weight loss up to 100 °C, which is due to the loss of water that was adsorbed onto the hydrophilic pristine IONPs.OA-IONPs showed three weight losses between 200 °C-800 °C (figure 3(c)), matching with previous reports published.Unlike pristine IONPs, OA-IONPs did not show any weight loss till 100 °C, which again implies that OA coating imparted hydrophobicity onto the nanoparticles.The two weight losses observed at 215 °C and 324 °C correspond to the degradation of OA around the IONPs surface.Such two-step weight loss is attributed to two different types of interaction  between OA and magnetite.The first weight loss observed at ∼215 °C is attributed to the loss of the physisorbed secondary layer of OA around the primary layer of OA.The second weight loss at ∼324 °C is due to the degradation of the chemisorbed primary layer around IONPs.A third weight loss was observed above 600 °C in OA-IONPs, which was due to phase transformation from magnetite to hematite, as the latter is more stable at temperatures above 570 °C [36].The total weight loss of different samples based on TGA curves is tabulated in table 3.As explained earlier, weight loss in the case of pristine IONPs was due to loss of water content as these particles are hydrophilic in nature.In the case of OA-IONPs, total weight loss gives information on the amount of OA coated onto the nanoparticles [37,38].As seen from the table, % weight loss increased with the increase in OA-loading onto the nanoparticles.This indicates that there was increased content of OA onto the nanoparticles with an increase in OA loading during the experiment.Magnetic susceptibilities of pristine and OA-IONPs determined at RT using VSM are presented in figure 3(d).The absence of a hysteresis loop in all the preparations indicates superparamagnetic behavior.The saturation magnetization (M s ), coercivity (H ci ), and retentivity (M r ) of pristine and all the OA-IONPs are presented in table 4. M s values of pristine and OA-IONPs were provided by the software that was used for VSM analysis.The values were later normalized with respect to the sample weight.The saturation magnetization (M s ) of pristine IONPs showed a very low value of 23.6 emu g −1 compared to the data published before [30].With the loading of OA, M s of OA-IONPs increased, which was mainly due to the size of the nanoparticles.The size of the nanoparticles plays a critical role in the magnetic susceptibility of the nanoparticles.In contrast to our results, few authors have noticed a decrease in the M s value due to the non-magnetic organic coating (OA) over IONPs [30,36].Coercivity (H ci ) and retentivity (M r ) of all the nanoparticle preparations were insignificant, indicating that the nanoparticles were superparamagnetic in nature.

Bacterial studies 3.2.1. ROS estimation using DCFDA assay
IONPs generate ROS through Fenton-mediated reactions.It is well-established that iron ions generate superoxide and hydroxy radicals through the Fenton reactions [12,39].DCFDA assay was performed to estimate the ROS generation by pristine and OA-IONPs quantitatively.DCFDA, in the presence of free radicals, converts   with respect to the control (without nanoparticles).However, we noticed that ROS generation by IONPs was inversely proportional to the increasing concentration of OA in IONPs.On the other hand, pristine IONPs showed comparatively less ROS generation than OA-IONPs.These observations can be explained by two possible reasons.Firstly, the highly packed OA layer around the IONPs might restrict the release of iron ions into the surrounding medium to initiate the Fenton reaction or hinder the interaction between core surface iron ions with the biological macromolecules.Thereby oleic acid layer around IONPs acts as a shield against the Fenton reaction mediated by IONPs.Secondly, the high rate of agglomeration by pristine IONPs (DLS results and TEM images) reduces active surface iron ions for the Fenton reaction.This result indicates that ROS generation by IONPs-mediated Fenton reaction is inversely dependent on the amount of oleic acid coated onto the IONPs.

Antibacterial activity
Fenton reaction-mediated ROS generation and mechanical rupture by IONPs are responsible for its antibacterial activity [29].The antibacterial test was conducted on the S. aureus MTCC 1144 strain, which is a clinical pathogen [26,27].The bacterial strain was incubated with pristine and different OA-IONPs for 24 h, and the percentage survival of the bacteria with each treatment is shown in figure 5(a).As we see from the figure, all the nanoparticle preparations (pristine and OA-IONPs) showed significant bacterial inhibition.Among all the preparations, pristine IONPs showed maximum inhibition.In the case of OA-IONPs, bacterial inhibition kept reducing with the increase in oleic acid loading over iron oxide nanoparticles.This antibacterial activity by OA-IONPs was in agreement with the ROS generation by the nanoparticles.ROS generation by pristine IONPs was much lesser than the OA-IONPs.
However, pristine IONPs still showed higher bacterial inhibition.This was most likely due to the mechanical rupture of the bacterial cells by the crystalline nature of pristine IONPs.Pristine IONPs were able to kill the bacterial cells through mechanical rupture in addition to ROS-mediated killing because of the rough edges [29].Oleic acid coating over IONPs makes nanoparticles less crystalline.Hence, the mechanical rupture by these OA-IONPs can be minimal based on the extent of coating.DCFDA assay together with the antibacterial test, conclude that OA loading significantly influences the Fenton reaction catalyzed by active iron ions, thereby antibacterial potential of IONPs.

Biocompatibility study
L929 fibroblast cell lines were used for studying the biocompatibility of pristine and different OA-IONPs for in vivo application.Cell viability was assessed using AlamarBlue assay.It was found that pristine and all the OA-IONPs treated cells were more than 95% viable after 24 h incubation (figure 5(b)).A similar observation of no or minimal toxicity toward fibroblast cells was noticed by Rania Ibrahim Shebl et al [4].The selective toxicity of IONPs towards bacterial cells but not against mammalian cells can be explained by the ability of mammalian cells to engulf nanoparticles through the endocytosis process [41].Such intracellular trapped nanoparticles in the form of endocytic vesicles are less harmful.Secondly, if the ROS generated by nanoparticles is surface- bound, then it may preferably kill bacterial cells over mammalian cells [41].The biocompatibility study conducted here using fibroblast cells demonstrates that the OA-IONPs are non-toxic in nature and can be used for biomedical applications.

Conclusion
In summary, IONPs were loaded with different amounts of OA and characterized using various techniques.In our case, with the increase in OA loading, magnetic susceptibilities of the IONPs increased mainly due to control over the size.In addition, OA loading onto the IONPs significantly altered the Fenton reaction-mediated ROS generation by the IONPs, thereby antibacterial activity.We hypothesized that a dense layer of OA coating over IONPs hindered the leaching of either iron ions into the surrounding medium to carry out the Fenton reaction or ROS generated by the surface-bound iron ions.Pristine IONPs showed the least ROS generation among all the preparations.However, they exhibited maximum antibacterial activity.This is possibly due to other mechanisms of antibacterial activity other than ROS.The crystalline nature of pristine IONPs might have caused a physical rupture to the bacterial cells, which probably contributed to the high antibacterial activity.This study helps in better understanding the role of surface coating agents on the physical properties of the IONPs, thereby, magnetic and biological functions in vivo.

Figure 2 .
Figure 2. Water contact angle of pristine and OA-IONPs, measured using a goniometer.

Table 1 .
Role of surface coating agent on the biofunctionalities of the nanoparticles.

Table 2 .
Z-average size of pristine and OA-IONPs with PDI measured by DLS method (n = 3).

Table 3 .
Total weight loss in the nanoparticles based on TGA curves.

Table 4 .
Magnetic properties of pristine and OA-IONPs measured using VSM at RT.