Large scale production of superparamagnetic iron oxide nanoparticles by the haloarchaeon Halobiforma sp. N1 and their potential in localized hyperthermia cancer therapy

The Haloarchaeon Halobiforma sp. N1 isolated from Wadi Al-Natrun, Egypt was used throughout this study. It was cultivated on Horikoshi-I medium [16] composed of (g l⁻¹): Glucose 10; peptone 5; di-potassium hydrogen phosphate 1; magnesium sulfate 0.2; sodium carbonate 10, casein hydrolysate 1, and sodium chloride 175, pH was adjusted to 9. No sterilization was required.

Halobiforma sp. N1 was grown in liquid Horikoshi medium for 7 days under shacked condition at 37 °C and 150 rpm. Culture was centrifuged for 10 min at 6000 rpm to separate cellular biomass from the culture filtrate. Both cell mass (CM) and filtrate were examined for the biosynthesis of MIONPs by challenging 10 ml bacterial supernatant or 0.13 g cellular biomass with 10 ml ferrous sulfate (FeSO₄) of 1 mM concentration and incubated at shaker incubator for 48 h at 37 °C, using a modified method of Fatemi et al [17]. Liquid Cultures of Haloarchaeon Halobiforma sp.N1 were prepared using Horikoshi-I medium at pH = 9 and the bacterial inoculum was always set to be 10% (V/V) of the experimental volume. The halo-archaeal morphological structure was observed by using phase contrast microscopy (figure S1(A), available online at stacks.iop.org/NANO/32/09LT01/mmedia), scanning electron microscopy (SEM) (figure S1(B)) and Gram staining technique (figure S1(C)). The cells appeared to be Gram-negative cocci of 0.25 microns in liquid medium. We observed positive results by the formation of particles that show high dispersion in water solution and a magnetic attractive force towards a permanent magnet.

A complete factorial design that is useful with a small number of factors was applied. It reveals the interaction between physical and chemical parameters while the effect it implies on the biosynthesized MIONPs is recognized. The factors studied were the concentration of precursor (FeSO₄), FeSO₄ volume, bacterial biomass, and pH. Table 1 illustrates the abbreviation and concentration of each factor under study at basal (0), low (−), and high levels (+). Accordingly, the experiment included sixteen possible treatment combinations corresponding to 2⁴.

Several characterization techniques have been carried out for a better understanding of the comprehensive properties of the produced MNPs. These include x-ray diffraction (XRD) using Bruker XRD-D2 Phaser, high resolution transmission electron microscopy (TEM) using JEM-2100UHR, and energy dispersive x-ray (EDX) analysis using a SEM Jeol JSM- 5300 equipped with an at EDX Unit. Magnetic properties was measured using Lakeshore 7410 vibrating sample magnetometer (VSM) and functionalization groups were confirmed using Fourier transform infrared spectrometry (FTIR) by Bruker Tensor 37. These techniques allow us to study crystal phase structure, surface morphology, size distribution, chemical composition, the spatial distribution of the functional group, and magnetic properties of the synthesized MIONPs, respectively. Feasibility for hyperthermia of 5 mg MIONPs dispersed in 1 ml of water was measured by using a G2-D5 Series Multi-mode 1500 W Driver from Nanoscale Biomagnetics. The sample solution was placed into a G2-D5 coil, where the alternating magnetic field and frequencies were applied. The induced temperature of the solution was measured and recorded using a fiber-optic sensor. The feasibility of MIONPs for magnetic hyperthermia treatment depends considerably on the SAR of these particles dispersed in water solution under applied AC magnetic field. The SAR expresses the heating capabilities of MIONPs in water solution. The applied AC magnetic field was in the range of 100–500 Oe and at frequencies of 144, 164, and 304 kHz.

The intracellular biosynthesis of the MIONPs from the haloarchaeon Halobiforma sp. N1 is unconventional. No reports are available on MIONPs formation from haloarchaea, while some literatures reported the use of some archaeal species in silver nanoparticles formation such as Halococcus salifodinae BK3, BK18 [14, 18]. We successfully biosynthesized MIONPs by the formation of nanoparticles when bacterial biomass was challenged against a 1 mM FeSO₄ substrate solution. S-layer (surface layer) proteins/glycoproteins commonly found in the cell envelope of archaea are capable of self-assembling into monomolecular layer on suitable surfaces or suspension [19]. Results were recorded visually by observing the magnetic attractive force of MIONPs towards amagnet. First experiments showed formation of only a small amount of particles, but by the aid of 2⁴ factorial experimental design, a full conversion occurred. Conditions employed in trial #15 appeared to provide successful biosynthesis of MIONPs (table 2). The conditions applied in trial #15 are CM 2 mg/100 ml, FeSO₄ concentration (FC) 50 mM, FeSO4 volume (FV) 100 ml, and pH 9. The solution turned to a dark brown color with a high attraction of particles towards the permanent magnet leaving a clear solution (figure 1). Moreover, under this condition the highest yield (∼15 g l⁻¹) was obtained showing a high attraction force towards the magnet.

Table 2. Design matrix and results of the 2⁴ factorial experimental design.

Trial	Factors under study	Variables
		CM	FC	FV	pH	Attraction to magnet
1	Base	−	−	−	−	−
2	CM	+	−	−	−	−
3	FC	−	+	−	−	−
4	FV	−	−	+	−	−
5	pH	−	−	−	+	−
6	CM, FC	+	+	−	−	−
7	CM, FV	+	−	+	−	−
8	CM, pH	+	−	−	+	−
9	FC, FV	−	+	+	−	−
10	FC, pH	−	+	−	+
11	FV, pH	−	−	+	+	−
12	CM, FC, FV	+	+	+	−	−
13	CM, FC, pH	+	+	−	+	−
14	FC, FV, pH	−	+	+	+	−
15	CM, FV, pH	+	−	+	+
16	CM, FC, FV, pH	+	+	+	+	−

Zoom In Zoom Out Reset image size

This experiment aimed to reach higher yield than that of the primary experiments, but the results were fascinating and surprisingly turned to a large-scale production of functionalized MIONPs. It is clear too from figure 1 that the formed MIONPs are highly dispersive in solution, which is an important property for our future in vitro/in vivo studies of hyperthermia. However, to confirm the optimization of these biosynthesized MIONPs, the physical properties such as crystallography, size, homogeneity, magnetic, and hyperthermia must be characterized.

The crystal phase structure of the optimized MIONPs was characterized using XRD [20, 21] as shown in figure 2. Figure 2 shows XRD patterns of the synthesized MIONPs. Typically, the 2θ peaks found at 30.1, 35.5, 42.6, 53.6, 57.0, and 62.8° are assigned to diffraction of the (220), (311), (400), (422), (511), and (440) planes, respectively. Those peaks are characteristic for inverse spinel, with O²⁻ ions forming a face-centered cubic magnetite (Fe₃O₄) lattice and iron cations occupying interstitial sites as referenced by Shukla et al [22] and Rasouli et al [23]. In addition, from the full width at half maximum of the peaks, we determined the average crystalline size using Scherer equation [24] to be 16 nm.

Zoom In Zoom Out Reset image size

The morphology including the particle size of the formed Fe₃O₄ MIONPs was determined using TEM as shown in figure 3. The TEM image (figure 3(A)) illustrates that the particles have spherical like shape. High-resolution TEM images showed polycrystalline features and d spacing (lattice fringes) of 0.25 nm, which is consistent with the XRD data of 311 planes (figure 3(B)) of magnetite (Fe₃O₄) nanoparticles. In addition, the particle size distribution shown in figure 3(C) was determined using more than 250 particles and was fitted using Gaussian distribution to reveal average size of 25 ± 9 nm. The result from TEM was in reasonable agreement with XRD where the calculated crystalline size using Scherer equation lies in the range of size distribution determined from TEM.

Zoom In Zoom Out Reset image size

The obtained results from TEM analysis are comparable with other biologically synthesized MIONPs by Fatemi et al [17] who reported spherical IONPs from Bacillus cereus with size ranges from 18.8 to 28.3 nm, while Sundaram et al [12] reported the spherical biosynthesized IONPS by Bacillus subtilis with size range between 60 and 80 nm. Spherical iron oxide nanoparticles of 20 nm were observed by high resolution of TEM in green synthesized iron oxide by Amutha and Sridhar [24]. VG and Prem [25] reported the green synthesis of nearly square iron oxide nanoparticles using Phyllanthus niruri extract with size of 10 nm. Polycrystalline character of iron oxide nanoparticles synthesized by Pyrodictium delaneyi and Pyrobaculum islandicum was reported by Kayshap et al [26]. The biosynthesized Fe₃O₄ MIONPs in our study were of higher uniformity and smaller size compared to other microbial synthesized MIONPs; however, our study can be extended to present particle size control in the future.

The purity of our MIONPs was determined by the EDX analysis as shown in figure 4. EDX spectrum confirms the formation of pure iron oxide nanoparticles. It displays 94.71% of iron oxide purity. The EDX confirmed the presence of both iron and oxygen in NPs. The composition was 70.45% and 24.26% of oxygen. Since our nanoparticles were biosynthesized from biological source, it confers some impurities that explains the low intensities in XRD. The impurities were displayed in the EDX as peaks for C, Na and S elements as seen in figure 4. By comparing our results to the published literatures, we found that Jagathesan and Rajiv [27] produced iron oxide nanoparticles biosynthesized from Eichhornia composed of 77.08% of iron and 22.97% of oxygen. Rasouli et al [23] reported that the EDX spectra of the Fe₃O₄-NPs synthesized by co-precipitation in presence of natural honey confirm the presence of elemental Fe without any impurity peaks. Therefore, the MIONPs biologically synthesized using natural honey show a very high purity as well as the NPs biosynthesized in this study. However, the production of large quantities of NPs from natural honey will be of a high cost, in contradiction with our method where the NPs were biosynthesized in an eco-friendly and cost-effective manner in large quantities (15 g l⁻¹).

Zoom In Zoom Out Reset image size

In order to investigate the magnetic properties of our biosynthesized MIONPs, we measured the magnetization dependence on magnetic field up to 2 T using VSM. Figure 5 represents the closed hysteresis loops revealing superparamagnetic-like behavior of our biosynthesized Fe₃O₄ MIONPs with high saturation magnetization (M_S) of 62 emu g⁻¹, small coercivity (H_C) and small spontaneous magnetization. Based on these results, the magnetization studies can provide another alternative to obtain more information on the magnetic domain size (D_mag) from the initial slope of the closed hysteresis loop [28, 29]. Therefore, one can estimate an upper boundary for the magnetic domain size to be 8 nm which confirms the superparamagnetic behavior of our Fe₃O₄ MIONPs.

Zoom In Zoom Out Reset image size

In comparison to the published work by others, the magnetic properties of the iron oxide nanoparticles of Krishnan et al [3] showed a magnetization of 50 emu g⁻¹, and that of Walujkar et al [30] showed magnetization value of 28 emu g⁻¹. Fatemi et al [17] showed IONPs with M_S of 59 emu g⁻¹. Darroudi, et al [31] showed IONPs with M_S ranging 17.5–42.1 emu g⁻¹. Kandpal, et al [32] reported Fe₃O₄ with M_S of 66 emu g⁻¹. In comparison with the previously mentioned literatures, the MIONPs produced throughout this study are of very high magnetization compared to that of NPs produced by biological methods, and comparable to those NPs produced by chemical methods such as co-precipitation with less cost-effective microbial method using Halobiforma sp.N1.

In order to investigate the functionalization groups that are responsible for the high dispersion in water solution and formed on the surface of the MIONPs, FTIR have been used. In the IR spectrum of biosynthesized nanoparticle (figure 6), absorption bands around 3189.53 cm⁻¹ are characteristic stretching vibration of the hydroxyl functional group (O–H) on the surface of nanoparticles. The stretching vibration of the carboxylate group (C=O) is observed around 1404.52 cm⁻¹. The stretching vibration of the C=C group was also localized at 1635.46 cm⁻¹. The absorption band around 1110.77 cm⁻¹ is assigned to the stretching vibration of the C–O group. Multiple peaks appearing at 439.67 and 542.42 cm⁻¹ are due to Fe–O in conformity with the literature Kumar et al [33]; Srivastava et al [14], and Mirza et al [34]. The formation of Fe₃O₄ NPs was confirmed by characteristic peaks laying in the region between 400 and 600 cm⁻¹ corresponding to Fe₃O₄ as mentioned by VG and Prem [25] in which the IR spectrum also showed a peak at 583.02 cm⁻¹ corresponding to the Fe-O stretching band of magnetite (Fe₃O₄). Also the FTIR analysis and presence of organic groups (C=O) and C=C group observed around 1404.52 cm⁻¹ and 1635.46 cm⁻¹ respectively may act as a natural coating for the particles which explain the reduced agglomeration and the high dispersion of the highly magnetic Fe₃O₄ nanoparticles. As previously reported, in FTIR spectrum the protein functional groups may confer the effect of a capping agent for nanoparticles due to sorption of proteins on surface of nanoparticles from biological routes [17]. Comparison of the different iron oxide nanoparticles from different routes of synthesis are illustrated in table 3.

Zoom In Zoom Out Reset image size

We have found that our nanoparticles biosynthesized by Halobiforma sp.N1 have a significant M_S; also, the obtained size is very good in such a one-step approach, which is cost-effective, easy, and not time consuming. The biological synthesis offered our MIONPs a protein cap from cellular material leading to a very high dispersion, solving the agglomeration problem of highly magnetic IONPs.

The SAR expresses the heating ability of a magnetic material and, therefore, the feasibility of a material for application in magnetic hyperthermia [37]. The value of SAR increases as the frequency of AC magnetic field increases [38]. The SAR value is calculated from the initial slope of the T versus t curves (figure S2, supplementary information) using the following equation:

$$\begin{eqnarray*}&&{\rm{S}}{\rm{A}}{\rm{R}}=\displaystyle \frac{C}{{m}_{{\rm{a}}{\rm{c}}{\rm{t}}}}{\left|\displaystyle \frac{{\boldsymbol{dT}}}{{\boldsymbol{dt}}}\right|}_{t=0},\end{eqnarray*}$$

Here, m_act is the mass of the magnetically active material and C the heat capacity of water (C = 4.18 J g⁻¹ K⁻¹). As shown in (figure 7), at f = 304 kHz SAR values are 50, 155, and 230 W g⁻¹ corresponding to the magnetic field applied (H) of 200, 300, and 400 Oe, respectively. At f = 164 kHz, SAR values are 25, 50, 75, and 77 W g⁻¹ corresponding to applied magnetic field (H) of 200, 300, 400, and 500 Oe, respectively. At f = 144 kHz, SAR values are 25, 45, 75, and 80 W g⁻¹ corresponding to applied magnetic field (H) of 200, 300, 400, and 500 Oe, respectively. It is clear that SAR increases when applied field and frequencies are increased.

Zoom In Zoom Out Reset image size

In addition, intrinsic loss power known as normalized SAR, was calculated as shown in FIGURE S3, confirming that SAR of Halobiforma sp. N1 MIONPs induced a very good heat at the low tested frequencies in comparison to all the published values for commercial materials prepared by physical and chemical methods [39]. Therefore, the biosynthesized MIONPs in this study have very promising properties to be used efficiently for the magnetic hyperthermia in vitro/in vivo and clinical approach in the future.